A continuing challenge

Citation

Ford, T. (2004), "A continuing challenge", Aircraft Engineering and Aerospace Technology, Vol. 76 No. 5. https://doi.org/10.1108/aeat.2004.12776eaf.002

A continuing challenge

A continuing challenge

Keywords: Aircraft engineering, Corrosion

Measures to limit the effects of corrosion have occupied engineers for many years. However, in aerospace during the 1950 and 1960s corrosion was not viewed as a serious life-limiting factor. Most aircraft in this era were replaced due to technical obsolescence and only a few types had to be prematurely retired on account of the occurrence of unexpected fatigue problems. In the 1970s it became apparent that commercial transports would remain in operation well beyond their original projected service life of 20 years.

This has to be viewed against the background of the evolution of means for guarding against the potentially catastrophic effects of fatigue failure. The three main stages of dealing with this are: safe life; fail safe; and damage tolerant design. The first has seldom been used for large transport category aircraft since the late 1950s (except in areas where other approaches have proved impractical, such as landing gear). The strength of much of such a structure is concentrated in a few major load carrying elements with little or no effort put into designing a structure or in choosing materials to prevent rapid propagation of fatigue cracks. During the 1950s also, accidents caused by fatigue failure led to an increasing realisation that dependence on safe life principals alone also means susceptibility to errors in engineering design and construction. This resulted in a move by manufacturers towards fail safe design which set out to provide structures with an acceptable level of strength in the presence of failure or partial failure of each principal structural element.

More recently, the term Damage Tolerant design has been used to denote a structure which has been subjected to modern fatigue or damage tolerance requirements. These call for a rigorous evaluation of potential fatigue cracking scenarios with due account being taken of multi-site crack origins. The evaluation has to cover crack growth rates as well as residual strength. Crack growth rate data are then used to determine maximum permissible inspection rates for critical structure.

In the 1970s it became apparent that commercial transports would remain in service well beyond their original projected service life of 20 years and the attention of constructors was drawn towards the possibility of certain age-related defects other than fatigue. Concern was directed towards more extensive damage than might have been considered in the original fall-safe design of the aircraft. This referred to a number of small adjacent cracks developing suddenly into longer cracks; failures or partial failures in other locations following an initial failure; and concurrent or partial failure of multiple load path elements.

Events and consequences

The various considerations relating to ageing aircraft caused the creation of supplementary inspection documents (SIDs) which addressed damage tolerance and multiple element damage concerns. Despite the work that followed the issue of these documents, an incident occurred in 1988 that was to have a profound effect on the aviation community. This was the near disaster involving a Boeing 737-200 of Aloha Airlines. It was realised that what had occurred was a significant failure of the total airworthiness system.

This event prompted considerable activities relating to ageing airframes, among them being the formation of the airworthiness assurance task force (AATF), the immediate objectives of this being for each aircraft model, to reconsider all fatigue related documentation and service/test experience; to create model specific corrosion control programmes (CCPs); to readdress the supplementary structural inspection programmes (SSIPs); and to consider repair quality issues.

The most probable cause of the Aloha 737 accident was linked up multi-site-damage (MSD) in the fuselage skin at the upper row of countersunk rivets of the lap splice at stringer 10 on the left side of the aircraft. Cold cured adhesive, between the skin panels, became delaminated due to the corrosion of the metal in the faying surface of the joint. The purpose of the adhesive in the design of the joint was to remove load from the rivets which were installed in knife edged countersunk holes. After delamination, the rivets started to react hoop tension loads due to cabin differential pressure. Knife edged rivet holes have very poor fatigue life. This caused multi-site- damage due to fatigue. Cracks linked up into a critical crack which was not arrested by the crack stopping system. The bonded crack stoppers were also delaminated.

Several factors became apparent, the first being that the manufacturer did not action a known problem but relegated the required corrective action to repetitive visual inspection and damage repair. This was despite the fact that there was sufficient information available to the operator to alert him to the problem of cracking in the lap joints associated with corrosion based deterioration. In addition, the FAA airworthiness directive (AD) which caused the SIDs to be issued, should have addressed all the lap joints and not simply that at the stringer where the problem had been first identified. There was no evidence that the operator had performed a satisfactory nondestructive inspection. Also, the operator had not instigated sufficient training to promote awareness of fatigue problems amongst the mechanics and to focus their attention on corrosion control and crack detection.

Corrosion control

The AATF made sure that corrosion is recognised as a significant worldwide problem. It is not just an ageing aircraft concern since in many cases, if proper precautions are not taken, corrosion can cause significant structural degradation in a reasonably short time. It becomes of even greater importance with ageing aircraft with widespread corrosion as well as fatigue cracking occurring and a combination of the two becomes more likely. It was decided that baseline corrosion prevention and control programmes (CPCP), to use the US designation, should be established for each aircraft model. The object was to define the minimum requirements that might affect the continued airworthiness of the worldwide fleet. Three broad levels are defined: Level 1 – within structural repair manual (SRM); service bulletin limits; Level 2 – outside normal repair limits – principle structural element replacement; Level 3 – urgent action outsides SRM. An effective programme is one which controls corrosion of all primary structure to Level 1 or better. An FAA AD was issued for each model, as for the modification programmes, that for Boeing aircraft being issued in 1990 followed by those for other large transports and commuter aircraft.

The Boeing approach is quoted as typical of a CPCP. The company formed a task force in 1988 to develop a proposal for CPCPs for its ageing 707/7207 727/737 and 747 aircraft. To be sure that the programmes would be effective and representative of real-world conditions, manufacturers, operators and regulatory authorities participated in their development. It was necessary to define the structure as primary or secondary and the level of corrosion control required. Primary structures are required to carry fully factored design flight and ground loads and consist of all the main load carrying structures of the aircraft such as frames, skins, stringers, spars, ribs, ailerons and elevators. Secondary structures include aerodynamic fairings, and non-load- carrying doors and panels. Total elimination of corrosion is impossible and therefore, it is necessary to establish the level at which corrosion has little effect on aircraft damage tolerance. This assessment must also consider the potential for combined forms of damage. CPCP procedures are needed for ail current in-production aircraft and for later models, as well as the 757 and 767. A CPC will be included in the basic maintenance program for the 777.

Corrosion can be classified into different types and knowledge of these can aid in selecting the best corrective action and preventive treatment. One type is stress corrosion cracking which occurs when structure is subjected to a moist environment and a sustained stress. It tends to follow a single plane or path, usually related to the grain flow formed by rolling, extruding and forging of parts during manufacture. Stress corrosion is minimized during the design process through material selection, recognition of secondary clamp-up stresses and proper corrosion measures.

Exfoliation corrosion is similar to stress corrosion cracking in that it follows grain boundaries created during manufacture. However, exfoliation corrosion attacks many grain boundaries, resulting in a leafing or delammation effect. The volume of corrosion product can exceed ten times the original-material volume. In fastened joints, the pressures created can cause noticeable bulging or deformation of the structural members.

Galvanic corrosion results when two dissimilar metals are in contact or are otherwise connected in the presence of a corrosive medium. The most reactive metal corrodes after the protective finish system is damaged. In design, the use of dissimilar metals is minimized. When dissimilar metals cannot be avoided, they are isolated from one another by protective sealants. Maintenance of the sealants is essential to promote durability.

Concentration cell, or crevice corrosion is a result of differences in the environment at a metal surface. It typically occurs in a crevice or stagnant area. A commonly encountered form is oxygen differential cell corrosion where the entrapped moisture in the crevice has a lower oxygen content than at the open surface. Additionally, when moisture and salt are present, chloride ions migrate to the oxygen-depleted zone (anode) inside the joint creating an acidic and corrosive condition.

Filiform corrosion occurs as a network of threadlike filaments of corrosion products on the surface of a metal coated with a paint film. A form of crevice corrosion, begins at a break in the paint film, typically around a fastener head. The development of more flexible paint systems such as polyurethane enamels has minimized the extent of this type of damage.

Pitting corrosion is a localised corrosion that begins on a metallic surface by galvanic or concentration cell mechanisms, the corrosion penetrates into the metal and forms a pit. The developing corrosion pits can act as stress concentrations that can evolve into fatigue or stress corrosion cracking.

Proper consideration of corrosion control must include material selection with aluminium and low-alloy steel susceptible to corrosion. Corrosion resistant aluminium alloys and tempers are used to increase resistance to exfoliation corrosion and stress corrosion cracking, an example being the replacement of 7150-T651 aluminium plate on upper wing skins with the less susceptible 7055-T7751. Major structural forgings are shot peened to improve the fatigue life of aluminium and steel parts and reduce susceptibility to stress corrosion cracking. Corrosion-resistant titanium alloys are used in severe corrosion environments and corrosion-resistant steels are used whenever possible. Fibre-reinforced plastics are of course, corrosion resistant but plastics reinforced with carbon fibres can induce galvanic corrosion in attached aluminium structure.

The most practical and effective means of protecting against corrosion involves finishing surfaces with an appropriate protective coating. For aluminium alloy, the coating system usually consists of a surface to which a corrosion-inhibiting primer is applied. It has become a common practice not to seal the anodised layer. Although this reduces the corrosion resistance of the anodised layer, the primer adheres better to the unsealed surface. As a result, it is less likely to chip off during manufacture and service, producing improved performance. For low-alloy steel parts, the coating system consists of cadmium plating to which a corrosion-inhibiting primer is applied. Stainless steel parts are cadmium plated and primed if they are attached to aluminium alloy or steel parts. This is to prevent the stainless steel from galvanically corroding the aluminium or steel. For the same reason, titanium parts are primed if they are attached to aluminium or steel parts. The corrosion inhibiting-primers used are Skydrol-resistant epoxies formulated for general use.

Effective drainage of all structure is essential to prevent fluids from becoming trapped in crevices. The Boeing approach is that the entire lower pressurized fuselage – is drained by a system of valved drain holes. Fluids are directed to these drain holes by a system of longitudinal and cross-drain paths through the stringers and frame shear clips.

The company effectively eliminates the potential for joint crevice corrosion by sealing the fay surfaces with a polysulfide. The polysulfide sealant is typically applied to areas such as the skin-to-stringer and skin-to-shear tie joints in the lower lobe of the fuselage, longitudinal and circumferential skin splices, skin doublers, the spar-to-web chord and chord-to-skin joints of the wing and empennage, wheel well structure, and pressure bulkheads. Earlier and improved designs in the region of the outboard wing front spar are illustrated.

An example of avoiding coupling materials of different galvanic properties unless required by economic and weight conditions can be found in the Boeing 777 carbon fibre reinforced plastic (CFRP) floor beam design incorporating corrosion protection measures. The fibres are good electrical conductors and they produce a large galvanic potential with the aluminium alloy used. The only effective way to prevent corrosion by keeping moisture from simultaneously contacting aluminium structure and carbon fibres is by finishing sealing using durable isolating materials such as fibreglass, and providing drainage. A comprehensive corrosion prevention and control programme including the features described is necessary to ensure continued structural integrity and minimize the need for corrosion-related maintenance.

Longitudinal lap splices

A particular concern shown by the Aloha incident was the possibility of interactions between corrosion and fatigue, especially as aircraft age. Corrosion (disbanding of a lap splice and associated fail-safe tear straps) in combination with MSD along the lap splice can lead to a sudden loss of structural integrity. Work has been undertaken on the disassembly and investigation of longitudinal lap splices on several types of older aircraft. This has concentrated on: the characteristics of corrosion and MSD fatigue crack initiation and early crack growth; any association between corrosion and MSD; and suitable MSD fatigue modelling and lap splice fatigue analysis methods.

One was a lap splice sample taken from a Boeing 727-100 in which the corrosion appeared to be concentrated along the upper side of the lap splice/stiffener connection, but the D Sight aircraft inspection system (DAIS) showed that corrosion-induced “pillowing” was present throughout the lap splice. The configuration is shown in the figure. After sample removal, eddy current and X-ray non-destructive inspections were undertaken before sending it to the NLR (Netherlands) laboratory where it was again inspected. Severe corrosion was found to be in a well-defined area covering all three rivet rows of the lap splice. However, the crack indications were almost all confined to the upper rivet row; only two indications were outside the severely corroded area.

Detailed examination showed that most of the cracking, in terms of crack locations, was intergranular owing to exfoliation corrosion and stress corrosion, the latter especially as corrosion products built up between the sheets to cause locally high stresses. Some of the larger intergranular cracks were associated with small, secondary fatigue cracks. These usually initiated from intergranular cracking, but sometimes occurred directly from rivet holes, without any evidence of fatigue initiation due to local corrosion (pitting). DAIS appears to be more sensitive to detecting severe corrosion than eddy current.

Four types of lap splice from different aircraft were analysed and in the 727-100 sample, severe internal corrosion did not lead to MSD fatigue crack initiation. However, this lap splice had small, secondary fatigue cracks. These usually initiated from intergranular cracks due to corrosion and stress corrosion, but sometimes occurred directly from rivet holes, without any evidence of fatigue initiation due to local corrosion (pitting). A number of the latter were found without any corrosion-induced cracking. These cracks most probably initiated owing to a combination of normal in-service stresses and additional stresses caused by corrosion product build-up (pillowing) between the outer and inner sheets of the lap splice.

Corrosion fatigue

A complex problem which has occupied researchers for some time is the two very different loading systems acting in a combination of fatigue loading and corrosive environment, which can occur is most alloys used in aerospace structures. Added to this is the difficulty of accurate simulation of this condition. Whereas correct simulation of fatigue alone is possible in the laboratory, a similar procedure for the combined condition has many difficulties. It is necessary to test components with their protection systems under the variable amplitude stress-time history and to simulate a typical sequence (corrosion-fatigue-corrosion, etc.) in order to obtain a result which corresponds to the real result in service. One method which has been attempted is to corrosion-fatigue test components which have suffered corrosion damage in service.

Differences between corrosion fatigue tests in the laboratory and corrosion fatigue life in service have been the subject of considerable investigations in Europe and worldwide. Of particular concern can be the corrosion and fatigue loads on a tactical aircraft which may undertake 100 to 100 flight of about 1h each per year and spend the rest of the year on the ground corroding away, if the environment is corrosive enough or if the corrosive products of previous service are present on the inside and outside of the structure. It is this prolonged period, in civil or military aircraft, particularly in the latter, which is most difficult to simulate.

This work involved produced some conclusive guidelines. These included the only question of engineering significance: How long is the structure going to last in its corrosion and fatigue environment in service and what can be done to improve that performance ? The conjoint action of corrosion and fatigue during the (usually short) fatigue periods and of corrosion alone during the usually long) corrosion periods in typical aircraft service are not easy to simulate, such tests also have to feature the slow deterioration of the corrosion protection system. Notched fatigue specimens may give qualitative answers as well as crack propagation tests. It was also suggested that all corrosion fatigue tests from notched or cracked specimens to complex components should be carried out under the relevant stress-time history. It is a consideration that “corrosion boxes” be fitted over typical details of a structure, the results being the nearest approach to service possible.

At several locations, research into the combination of fatigue loading and a corrosive environment have continued with the understanding that although protective schemes are generally effective, paint films can be damaged and allow ingress of aggressive environments such as bilge fluids, acid rain, or salt water. DRA Farnborough has been particularly active in this area with investigations showing the effects of corrosion fatigue leading to the reduction of fatigue life by premature initiation of fatigue cracks and increasing the fatigue crack growth rate. One typical activity compared the corrosion fatigue crack growth rate properties in the aluminium-lithium alloy 8090 with those of conventional 2000 and 7000 series plate alloys. The results showed corrosion fatigue crack growth rates for 8090 in the medium strength condition to be similar to those for the damage tolerant 2024 alloy; these were lower than for the high strength 7000 series alloy. At this facility, flight simulation loading and constant amplitude fatigue tests are performed using computer-controlled machines with automatic data acquisition and load monitoring.

Detection and avoidance

Until effective “shot peening” was introduced for particular areas, the only effective method was to machine out the corrosion, which was by no means always successful and often resulted in the scrapping of the part. The Metal Improvement Company developed controlled search peening where the induced compressive stresses stretch and yield the outer material surface and induce visible blistering and flaking at the surface indicating the existence of exfoliation corrosion that is selectively attacking the aluminium substrate via the grain boundaries. It is notable that stress corrosion cracking cannot occur in an area of compressive stress. As an example, a component after controlled search peening may show signs of edge exfoliation corrosion. An aluminium oxide blast of the surface is made to clear debris and remove surface oxides for further inspection and dressing and another pass with the controlled search peening equipment may be required. The inspection and process will continue until no further blistering occurs. More recent developments in this area include laser peening which employs short bursts of intense light that create a high pressure plasma on the surface of the material and the residual compressive stress being deeper, provides greater resistance to fatigue and corrosion.

Nondestructive inspection (NDI) techniques for the detection of corrosion and cracks have progressed considerably in recent years and include sophisticated X-ray systems such as those of the German Seifert company which can deal with specific requirements for the inspection of aircraft sections made of fibre-reinforced composite materials, aluminium and titanium. One example has an overhead X-ray tube suspension system with isocentric swivel device for fluoroscopy, programmed and manual operation of the four motion axes. It is equipped with ISOVOLT X-ray tube 160 kV and low inherent filtration input window image intensifier.

Another example of equipment available is that supplied by the UK Panametrics NDT company of which the Model 500PR is typical. This is a pulse-receiver which is used in conjunction with transducers and an oscilloscope, and is the most basic unit in any ultrasonic system for NDT testing or materials research. It can be used for flaw detection and thickness gauging in a wide variety of metals, plastics, ceramics and composites.

The US Zetec organization provides a range of multi-frequency eddy current instruments to perform a variety of tasks on aircraft structures, engine components and wheels. It can detect surface and subsurface fatigue cracking, corrosion and heat damage or determine material type. One application detects corrosion in a multi-layer structure with varying air gaps caused by lack of adhesive. Typically, the signal from the gap variance moves in the same direction as the corrosion (material thinning) signal. The dual-frequency mixing incorporated allows the inspector to almost completely reduce the unwanted gap signal, while enhancing the corrosion signal. The few examples mentioned illustrate the advanced techniques now available for dealing with corrosion-related events.

Terry Ford