Table of Contents
5.4.1.2 Damage due to Corrosion without Mechanical Loading
Extensive corrosion damage not caused by mechanical loads has been reported especially in older engine types that still used low- and high-alloy steels for housings, stators, and rotor parts. For this reason, large corrosion studies were conducted, especially in the 1950s. In all cases the corrosive medium was moisture in the air, and frequently involved ocean environments. The studies also showed that the salt content of the air above oceans/the sea was largely dependent on the weather conditions, the height of the waves, and the distance between the aircraft and the surface of the water. Moisture in the air is carried into the engines of standing aircraft both by wind (dynamic soak) and by an airflow caused by the engine cooling off after shut-down (static soak). In the worst cases, 200 ppb (10-9) salt concentration was measured in the air inside an engine (Ref. 5.4.1.2-1).
The damage process during moisture corrosion usually involves the corrosion occurring primarily during standstill. When the corrosion has sufficiently weakened supporting cross-sections of engine parts, it can cause dynamic/fatigue cracks or spontaneous part failures.
A special corrosion danger with watery media is present when, on parts made from corrosion-sensitive materials, mating surfaces (such as at holes for clamping bolts/reamed bolts or mating zones of disks and spacer rings) cannot be coated with sufficient corrosion protection due to extremely close tolerances. This are for example tight fit tension bolts through the holes of rotor disks.
“Dry corrosion” caused by fused salt deposits from the air flow can occur in the hotter engine areas during operation. This requires sufficiently high operating temperatures (for sulfidation on hot parts made from Ni-alloys, see Chapter 5.5; for pitting caused by salt on titanium alloys, also see Chapter 5.4.2).
Figure "Typically affected components": The components of engine compressors typically affected by corrosive attack are the rotor blading, the stator vanes, and the inner housing walls, especially around abradable coatings.
Rotor blading (“A”): Blades made from Al- or steel alloys (such as typical 12% Cr steel) without corrosion protection are prone to pitting (Fig. "Behavior of a corrosions system"). This causes considerable local decreases in dynamic strength and increases the danger of dynamic fatigue fractures in the blading.
Abradable and rub-tolerant coatings (“B”): These relatively soft and therefore porous coatings can absorb condensation water and corrode, and/or carry electrolytes to the base material and cause it to corrode. The polyester spray coatings filled with Al powder that are used in the front compressor area at operating temperatures up to about 100 °C have shown themselves to be especially vulnerable. The corrosion occurs in layers and can be recognized by blistering. The most sensitive area is the transition of the coating edge to the housing, since electrolytes can easily enter into the gaps.
Soldered compressor stator vanes (“C”): If cell action occurs due to the metallic contact of different metals that are far apart on the electrochemical series (Figs. "Marine atmosphere" and "Types of watery corrosion"), corrosion will be especially pronounced. A typical example is compressor stator blades/vanes made from Cr-steel sheeting and joined with silver or copper solder. Chains of corrosion pittings occur in the root area on the edge of the braze parallel to the chord. This is also where the largest dynamic loads are to be expected in case of a flexural mode in the blade. Therefore, this type of corrosion is especially dangerous with regard to dynamic fatigue fractures in vanes.
Due to the corrosion danger, today compressor blades made from Cr-steels are generally coated with inorganic lacquers filled with Al-powder, which can be used at operating temperatures up to about 500 °C. Coatings that are sintered or shot-peened with glass beads become electrically conductive and provide the base material with cathodic corrosion protection. This means that, around small scratches, the base material is protected by the greater allowable dissolving of Al-coating components.
Figure "Corrosion in the area of the compressor": In the compressor areas of older engine types there are still various corrosion-sensitive materials in use. These are primarily:
- low-alloy steels (boiler steels) in housings and disks
- high-alloy steels (Cr-steels) in bladings, disks, control systems (such as spherical bearings) and threaded connections. In the middle compressor area of modern engines, disks and housing rings made from steels with low thermal-expansion coefficients are frequently placed between parts made from Ti-alloys and parts made from Nickel alloys. These parts must be equipped with corrosion protection that is suitable for the operating temperatures.
- Al-alloys in rotor and stator blades, as well as gearing and regulator housings
- Mg-alloys in compressor and gearing housings/casings
Today, these materials are all coated with corrosion protection layers, resulting in a considerable improvement. However, experience has shown that corrosion is to be expected under influence of condensation water in ocean/sea environments, especially when various materials are combined (cell action, top left diagram) and at separable mating surfaces (crack corrosion, e.g. on flange surfaces). This corrosive attack, which usually takes the form of pitting, often originates in damaged corners/edges or worn areas of seating surfaces. A common area is the seating surfaces of dovetail blade roots (top right diagram).
Compressor stator vanes made from steel sheeting that have been joined by brazing are subject to increased corrosion along the braze seam due to cell action (silver or Cu solder).
With adjustable stator vanes, the conditions around the adjustment mechanisms promote crack corrosion in the magnesium housings/casings (top diagram).
Example "Material alteration" (Ref. 5.4.1.2-11):
Excerpt: “Other changes include removing magnesium from the engine, which reduces the powerplants susceptibility to saltwater corrosion and erosion. ..the engine's magnesium inlet and gearbox housing were changed to aluminium, and the magnesium compressor housing was changed to less expensive stainless steel.”
Comments: This case concerns an increased-performance version of an engine in a large, twin-engine military helicopter that has been in use for a considerable time. Evidently the material changes were necessary in the light of the bad experiences with the earlier versions.
Example "Proper corrosion test" (Ref. 5.4.1.2-1 and Ref. 5.4.1.2-3):
Excerpt: “ …Corrosion- prone hardware identified as a result of the …test included the blade-tip rub seals employed in the first three stages of the high-pressure compressor. Corrosion within the bond coating of the seals resulted in widespread delamination or the seals partway through the program. The resulting loss of the seal material decreased engine performance by more than five percent and produced impact damage to over one hundred compressor airfoils.
Comments: These findings are from a very elaborate testing program in which the entire engines were subjected to proper load cycles in a simulated ocean atmosphere (a sea salt solution was sprayed into the compressor). The test duration was up to 1200 hours with longer periods in which the engines were running. This became necessary after it was realized, that the testing conditions of the 1960s were insufficient. In the earlier tests, the salt was inserted into the engines as they ran at a mere few hundred RPM, which did not sufficiently reproduce the corrosion processes (e.g. sulfidation) in the hot parts.
Figure "Corrosion of rubbing coatings during marine operation": Rub coatings designed to minimize tip clearances in compressor housings are usually made from porous spray coatings, which can let through watery corrosive media or soak them up like a sponge. These coatings are attached to bond layers. A typical composition of an bond layer is 5% aluminum and 95% nickel. This bond layer can be damaged by corrosion, causing parts of the rub coating to delaminate (middle diagram). Consequential damages in the compressor cause the compressor efficiency to unallowably worsen.
This sensitivity was discovered in very extensive and realistic corrosion tests (Ref. 5.4.1.2-1), and countermeasures were implemented (Example "Proper corrosion test").
The bottom diagram shows typical corrosion symptoms on sprayed coatings. Sprayed coatings in housings/casings can be made from an organic matrix (polyester, etc.) with metallic filling (Al-powder, etc.) and have a relatively dense structure, or they can be made largely from metal with a porous structure, such as is typical in Ni/graphite spray coatings. These coatings are usually attached to the base material by an bond layer.
Spray coatings on rotor spacer rings can be made from relatively dense ceramic coatings on an bond layer. The most common failure mechanisms are:
- Corrosion of the abradable coating itself:
The corrosive medium is carried into gaps and cracks by the capillary effect and leads to a layered attack along corrosion-sensitive filler materials. This is followed by blistering and layered breakouts within the coating. This type of corrosion is most common in abradable coatings with a synthetic matrix and filler materials.
- Corrosion of the bond layer:
The corrosive media travels to the bond layer through porosity, cracks, and gaps in the abradable coating. It then damages the bond layer and causes the abradable coating to delaminate. This corrosion form is more common in porous metallic and non-metallic spray coatings.
Example "Low utilization" (Ref. 5.4.1.2-3, compare with Fig. "Corrosion as cause of a disk burst"):
Excerpt: “The U.S. National transportation Safety Board is urging the FAA to mandate more frequent inspections of certain high-pressure compressor discs in …. engines to preclude uncontained failures induced by corrosion pitting. …the action which centers on …engines with low-utilization HPC disks, is prompted by the rupture and subsequent disintegration of the 9th stage HPC disk …The aircraft had departed…when the No.1 engine's 9th-stage HPC disk rim separated and shrapnel exited the high pressure compressor case…As a result, fragments from the disk struck the left wing leading edge forward of the main spar, puncturing the “A” system hydraulic lines, which led to a complete loss of pressure in that system…
Metallurgical analysis of the fractured disk in… laboratories revealed a fatigue crack that had originated in the rim and migrated inward until reaching a critical length… In addition there was `extensive corrosion pitting in multiple sites that were concentrated on the outer web and rim areas of the disk'… Further examination indicated that the failure originated in a dovetail slot. `The large number of pits present suggests that corrosion pitting was the most probable cause of the failure!….
(the engine manufacturer) defines an HPC disk as `low utilization' if it accumulates fewer than 1,300 flight hours or 900 cycles per year, and it remains in that classification until it is recoated or replated, according to the company. A service bulletin currently requires that these disks be inspected every 10 years after being replated with nickel-cadmium. But NTSB… said that investigation of the…failure indicates that the 10-year inspection interval `is too long' and should be reduced to significantly less than eight calendar years…
Both the BEA and the NTSB have `found no evidence to suggest that the failed…9th-stage disk had been improperly inspected or replated eight years before the accident'…
The NTSB suspects that low-utilization HPC disks are more vulnerable to corrosive attack chiefly because they are not operated at a minimum number of hours or cycles per year. Corrosion pits eventually penetrate into the disk material and develop fatigue cracks….the safety board said corrosion pitting has been a factor in at least one other uncontained failure or (engine of the same type), low utilization HPC disk….(nearly 5 years before an aircraft of an other type) was destroyed by fire after the crew aborted takeoff …
Analysis of the disk showed that it had ruptured from a fatigue crack on the disk rim that had originated at a corrosion pit measuring 0.010 in. (0.25 mm) deep. The rim also exhibited heavy surface corrosion and pitting…“
Comments: At least a few compressor disks of this older engine type are still typically made from a pitting-sensitive steel.
The different assessment of the safe inspection intervals is interesting. There is also a remaining question as to why no sufficient measures to prevent future damage were taken after the first damage, which occurred years before.
The ultimate moment of damage, when the blade rim separated, typically occurred during start-up or shortly afterwards.
Figure "Corrosion as cause of a disk burst": This damage to a compressor disk made from a corrosion-sensitive steel was evidently caused by pitting resulting from condensation water that attacked the engine during relatively long standstill times. The damage process is described in Example "Low utilization". A similar failure in the turbine region describes Ill. 12.6.3.3.
Figure "Fracture of the spacers" (Refs. 5.4.1.2-4 and 5.4.1.2-5): These corrosion damages (see Example "Failing spacers") can be traced back to an unsuitable design combined with corrosion-sensitive materials and maintenance problems (Example "Corrosion in bores" and Example "Rotor tension bolts"). It is not entirely clear, to what degree the corrosion was caused by condensation water during operation or through cleaning solutions that had entered during maintenance work.
rulemaking
Example "Failing spacers" (Ref. 5.4.1.2-4):
Excerpt: ”…investigators discovered pieces of a spacer lying alongside the runway. Its location indicated it the left engine at approximate aircraft liftoff, penetrating engine case, fan duct and cowling.
…According to an FAA notice of proposed rulemaking…compliance with the inspection order will be required within two years, or 4,000 engine cycles…For certain high-time engines, inspection is sought within 1,000 cycles.
The notice also calls for the replacement of the suspect spacers, located between the seventh and eighth and also the ninth and tenth compressor discs in the engine's high-pressure section, as well as four other spacers in the same area, with an existing version of a more recent design. In most cases the substitution is to be made during the engine's next scheduled overhaul. The spacer problem affects 1,840 engines on domestic airliners and another 1,838 engines worldwide.
…(according to a representative of the FAA) the spacer problem is `not viewed as an immediate, safety of flight' matter. He said (the engine manufacturer's) procedures require removal and inspection of older version spacer tie-rod tubes during overhaul, though the FAA uncovered some instances where this was not done. The engine manufacturer also published service bulletins in 1972 and 1981, both of which emphasized proper inspection and, where necessary, repair of…high-pressure section spacers.
As used in the (engine type), a spacer is a 3.in.-long (7.5 cm, see arrow 5.4.1.2-5) metal alloy sleeve that slips over the engine's rotating central shaft and separates turbine discs from each other. There are six spacers in the high-pressure section…A `stack' of discs and spacers is bolted together using 12 tie-rods, which run through the discs and spacers parallel to the engine shaft.
Older (engine type's) high pressure section spacers incorporate 12 tubes, all mechanically pressed into place, to accommodate tie-rod passage. Undetected corrosion between and around these tubes has resulted in eventual cracking and failure of the spacer.
(the engine manufacturer) began exclusively shipping a new design…spacer with corrosion-defeating integral tie-rod passages over two years ago. Many of these are already operating in aircraft in the field. However, as replacement of the old-design spacer was not mandatory in the past, both new- and old-design spacers may be mixed on an individual engine….A total of 4,950 of the older- version spacers have yet to be 'attritioned out' domestically, while another 4,939 are still flying…
Since the (related engine type) was introduced into commercial airline service in the middle 1960s, 46 spacer-related engine failures have been reported. Thirty resulted in excessive vibration or in-flight engine shut-downs. The remaining 16 spacers separated from the engine shaft with eight having enough energy to penetrate the engine case, fan duct and cowling, in some instances causing airframe damage. Each replacement spacer costs approximately $3,500.”
Comments: The large number of parallel damages (?) and the long time period over which the problem developed are astounding. Evidently even several uncontained failures were not considered sufficiently dangerous to mandate an earlier removal of the potentially dangerous parts.
Example "Corrosion in bores" (Ref. 5.4.1.2-9):
Excerpt: “The FAA is calling for detailed inspections to detect cracks in engine support fittings… that could result in separation of the Nos. 1 and 3 engines from the aircraft…Two of the reports noted that the cracks originated from the large fastener holes next to the fuselage, and a third report described a crack that had propagated almost completely through the fitting…The cracks were caused by fatigue induced by corrosion pitting on the surfaces of the fastener holes in the fittings.”
Comments: Engine suspensions are often made from high-tensile steels that are sensitive to pitting corrosion. In the inside of bores, especially, capillary action can cause corrosive media (salt water) to collect and, over time, attack fresh metal surfaces that have been created by friction wear/fretting. A further problem is the fact that it is usually very difficult to visually inspect bolt connections from outside without disassembling the engine (Ill. 12.6.1-12).
Example "Rotor tension bolts" (Ref. 5.4.1.2-6, see Fig. "Fracture of corroded clamping bolt"):
Excerpt: “The (NTSB) board's recommendation follows discovery of broken tie rods on an engine…The aircraft experienced an uncontained failure of the low-pressure turbine…during takeoff roll…
The board recommended…that all …engines be disassembled at next shop visit and that the low-pressure compressor rear tie rods be removed, cleaned and inspected, recoated with antigallant and reinstalled or replaced. The rear tie rods are long bolts that hold together the four stages of the …low pressure compressor (LPC). One of the tie rods…was fractured and had surface corrosion, and because of overstress, the 11 remaining tie rods also were fractured.
Maintenance records showed the engine, including the LPC, had been repaired …(about 12 and 8 Years before)…the board said. NTSB said it concluded that an antigalling coating was not applied to new LPC rear tie rods..(by the OEM) at the time of manufacture.
The board said that while… (the OEM) stated that …(this) engine failure has been the only one that involved the fracture of an LPC tie rod that was supposed to have been coated with the antigallant, the lack of the antigalling coating on all 12 of the LPC rear tie rods… suggests the problem may be widespread.”
Comments: This does not necessarily mean that the failure of a single tie rod will cause the entire rotor connection to fail, before vibrations caused by imbalances indicate damage early enough to avoid a catastrophic failure.
Because this engine type is very widely used, the quoted reference literature estimates the potential costs for necessary preventive measures to be about $109.
Figure "Fracture of corroded clamping bolt" (Ref. 5.4.1.2-6): Surface corrosion caused a clamping/tensioning bolt in the low-pressure compressor to fail (black arrows in bottom diagram, Example "Rotor tension bolts"). After this, the remaining clamping bolts failed due to overloading and the shaft separated. This resulted in the now free-running low-pressure turbine reaching overspeed and failing and creating uncontained fragments.
Example "Compressor spacer rings" (Fig. "Fracture of the spacers", Ref. 5.4.1.2-7):
Excerpt: ”…However, the core problem seemed to be the compressor spacer rings…which have to be replaced in 1840 engines worldwide. The technical problem is very complicated in its details, but its fundamental principle is relatively simple: the seal system of an…engine (of the affected type) is held together by long bolts which pass through pipes (“1”) in the spacer rings (“2”). During inspections, these pipes must be removed in for cleaning and overhaul. However, this was often neglected in order to lower costs. As a result, cleaning solution collected between the pipe and spacer ring, causing rust to form. …The US company (engine manufacturer) maintains that it had insisted that all airlines inspect the spacer rings for cracks and replace them by early 1988 at the latest.“
Comments: The above explanation, in which the corrosive media only entered into the damaged area in the form of cleaning solution during overhauls, cannot satisfactorily explain the life span-related inspection intervals in Example "Failing spacers". Also, the involved media is a watery electrolyte which will evaporate after short run times in this rear area of the compressor. Therefore, in order for dangerous corrosion to take place, new electrolyte must always form during standstill. This can only be plausibly explained by condensation water, which dissolves dried cleaning solution remnants and makes them corrosive again. This probability seems rather remote when one considers that formation of condensation water requires an air exchange, making elements of the marine atmosphere far more likely as a corrosive media.
Example "Rear compressor variable vane" (Fig. "Critical areas of a variable vane system",Ref. 5.4.1.2-8):
Excerpt: ”…Both training accidents…(of an military aircraft type with one engine) resulted in a successful crew ejection following engine stagnations at low altitude. The …crash brought on worldwide inspections of the engine rear compressor variable vane synchronizing arm bearing and control cable for freedom of movement and the rear compressor push-pull control cable. Out of 271 powerplants checked, 12 were rejected. Six more have been taken out in inspections being conducted 10 hr. after the first check.
Sixteen days before the crash, it (the engine) had undergone a 50 hr. phase inspection at 84.9 total operating hours…
Inspection of the engine showed that:
The rear compressor variable vane (RCVV) feedback cable to the unified fuel control had a broken link at the RCVV bell crank. The cause was fatigue failure….
(the accident investigation on the crashed engine showed)…The RCVV feedback push-pull cable was found broken at the bell crank end because of fatigue failure.
The report said inflight failure of the cable would cause engine operation problems because of misscheduling of the vanes. Experiences with (an other military aircraft type with two engines of the same type) have shown six instances of cable failure, five of which resulted in stagnation - two at part power and three in augmentor operation. Three of those stagnations had successful airstarts.“
Comments: The repeated fracture of the replaced feedback cable after a short run time indicates that overload conditions that were not eliminated by the replacement were already present. This may have been, for example, some stiffness. The literature does not indicate how much of a role corrosion played in this case.
Figure "Critical areas of a variable vane system": Compressor stator vane adjustment systems usually contain steels, which tend to corrode in marine environments, in bearings “1”, seals “2”, (also see Fig. "Corrosion in the area of the compressor") hinge eyes “3” and cable connections “4”. It has been discovered that lubricants containing MoS2 can even promote the sticking of plain bearings in marine environments. This can result in overloading of the actuator components.
Other possibilities include pitting corrosion, resulting in dynamic fatigue fractures. Additionally, the high strength of the steels used in these parts makes them susceptible to stress corrosion cracking (see Chapter 5.4.2).
Failure of the adjustment system or its control and regulator systems can cause compressor surges/stalls, resulting in engine failure (Example "Rear compressor variable vane").
Figure "Stator vane adjuster" (Ref. 5.4.1.2-8): The failure or malfunction of the stator vane adjuster in the compressor is an extremely dangerous situation. It can lead to immediate compressor stalls and engine failure (Example "Compressor spacer rings"). If only one vane is affected, it can lead to dynamic fatigue fractures in the neighboring stages after a reasonably long run time. This dynamic overstress is due to the turbulent flow caused by the improperly positioned vane.
Typical damages and damage causes in variable stator vanes are:
- Complete or partial shutting of the variable vanes of the affected stage
- Fractured transmission links
- Sticking and/or fracture of bearings and joints
- Mechanical or electrical damage to the regulator/feedback cables
- Fracture of the vane bearings (shafts/axle lugs)
Example "Fuel leakage" (Ref. 5.4.1.2-10):
At the endcap of the fuel distributor a considerable fuel leak developed. An investigation showed, that the producer of this component did not met the specified production process after the heat treatment. This had as result a susceptibility for intercrystalline corrosion. An operation failure was the consequence, which can lead to a fire and the shut down of the aeroengine with a emergency landing.
Comment: Crack formation at the coupling nuts of pipe line screw connections can be traced back predominantly at a material structure, susceptible for a crack corrosion. Usually the cracks run axial, correspondent the structure orientation of the rod material from which the parts are produced. At steels an unusual high hardness (above 32 HRc) is a certain feature for an unsufficient heat treatment during production. It is not clear how far the intercrystalline path of crack is also influenced by operation caused tensile stresses and/or residual/internal stresses from the heat treatment process (Fig. "Coating damage I"). In such a case the corrosion can be rather assigned a stress corrosion cracking (Fig. "Vulnerable components" and Fig. "Sensitization during operation"). The sketch was reconstructed from the scarce informations of the literature on hand. Deviations are possible.
Figure "Corrosive attack on Mg alloy": Corrosive attack on the edge of a lacquered/coated gear casing made from a cast Mg alloy. The pitting corrosion under the influence of marine atmosphere is typical (Cl ions). Repair is possible in areas subject to low mechanical loads by filling them with synthetic material.
Proper corrosion protection in accordance with regulations on magnesium parts has been proven to ensure long damage-free operating times. Therefore, if corrosion occurs, the cause is most likely defects or damage in the coating.
Corrosion of edges usually occurs at mechanical damage to the lacquer coating, insufficient coating thickness, or coating defects. Corrosion in the flange area is promoted by corrosion cracking conditions and cell action in contact with the mating surface.
Corrosion pittings on lacquered/coated surfaces are often related to improper preparation of the surface to be lacquered, such as corrosive remnants of cleaning baths and shot-peening materials.
This results in typical blistering due to corrosion between the coating and the base material.
Figure "Corrosive attack underneath coatings": Corrosion of lacquered/coated parts has typical damage symptoms that can indicate the cause of the damage. The final stage of the corrosion is usually blistering due to the increased volume of the corrosion products.
Figure "Corrosion dangers at nacelle": The experience shows, that especially corrosion must be expected if splash water can get into the nacelles of the aeroengines. Such situations develop during rain at the landing approach with extended landing gear or at at puddles on the runway. Prone are from experience airplane types with unfavorable aeroengine position in relation to the nose landing gear (sketch below, example 10-10.1). That is also true for turboprops when the main landing gear is swung out of the aeroengine nacelle (sketch above).
Especially endangered by corrosion are components at and around the aeroengine as
- mountings of the aeroengine and the pylon. Here the shear bolts (attachment bolts), mounting bolts from high strength steel tend to the formation of corrosion pittings. The result is a dangerous drop of the fatigue strength. Fatigue cracks cause the separation of the aeroengine (Ill. 10-10.1) with potential catastrophic consequences.
- Electrical lines, contacts, connectors (Ill. 19.2.1-1.1 and Ill. 19.2.1-1.2) lo0se by corrosion through breaks, slack joints and shorts temporary or permanent its function. Consequences are malfunctions or the outage of components like control units as well as dropped out or false alarm signals. With this serious incorrect decisions of the pilot are pre-programmed. Typical are erroneous fire alarms with unnecessary in flight shut down (IFSD) of the aeroengine. Even worse may be, if fire is not signaled.
- The variable guide vane actuation is positioned at the compressor outside. So humidity from rain and splash water can enter. With this, corrosion (Fig. "Critical areas of a variable vane system" and Fig. "Stator vane adjuster") can trigger fracture (Fig. "Failure of variable vane adjuster"), jamming or unacceptable wear. In an extreme case this causes surging of the compressor with thrust loss of the aeroengine during a delicate situation (start).
References
5.4.1.2-1 F.T. Carroll, D.R. Parish, “Testing Considerations for Military Aircraft Engines in Corrosive Environments (a Navy Perspective)”, AGARD-CP-558, proceedings of the conference “Erosion, Corrosion and Foreign Object Damage Effects in Gas Turbines”,Rotterdam, The Netherlands, 25.28 April 1994, Chapter 37, page 37-10.
5.4.1.2-2 J.Broichhausen, “Schadenskunde- Analyse und Vermeidung von Schäden in Konstruktion, Fertigung und Betrieb”, pages 344-368.
5.4.1.2-3 E.H.Phillips, “NTSB Targets JT8D Cracks”, periodical “Aviation Week & Space Technology, February 3, 1997, page 47.
5.4.1.2-4 “FAA Proposes Inspection of JT8D Spacers After Crash Investigation”, periodical “Aviation Week & Space Technology”, January 1986.
5.4.1.2-5 “FAA acts on JT8D”, periodical “Flight International”, 11. January 1986, page 5.
5.4.1.2-6 “NTSB Seeks JT8D Inspections”, periodical “Overhaul & Maintenance”, April 1999, page 73.
5.4.1.2-7 “Rost an den Flügeln”, periodical “Wirtschaftswoche Nr. 9, 21.2.1986, pages 92 and 94.
5.4.1.2-8 D.R. Griffiths, “Crash Investigations Spur Engine Recommendations”, periodical “Aviation Week & Space Technology”, September 15, 1980, pages 86 and 87.
5.4.1.2-9 “FAA Orders Inspections of Boeing 727 Fittings”, periodical “Aviation Week & Space Technology”, March 17, 1997, page 101.
5.4.1.2-10 “Blow-dry your jet”, periodical “Flight International”, 18 May 1985, page 49.
5.4.1.2-11 “Stanely W.Kandebo, Helo Gains With T55, Aviation Week + Space Technology, 5.17.1999, Vol. 15, Issue 20 P 67”