10. Engine Mounts
The alternately influence of mounts and operating behaviour of the engine is discussed under different aspects in other volumes respectively chapters.
- Deformation of casings (Fig. "Clearance changes by flight loads") and its influence at the engine operation (Fig. "Housing deformations near engine suspension"). Especially concerned is the consistent drop of the efficiency (deterioration) respectively an increase of the fuel consumption and the gas temperatures due to the abrasion/rubbing of the labyrinth seals (Fig. "Seals and engine performance"). This is affected by aerodynamic loads of the engine nacelle ( ), inertia forces by maneuvers (Fig. "Deformations by 'G-forces'") and gyro forces (Fig. "Gyroscopic forces at fan and low pressure shaft"). Thereby the influence of casing stiffness and with this of the mounts play a role.
- Introducing forces and moments (Fig. "External loads at the engine mounts" and Fig. "Design of an engine mount") into the nacelle respectively the wing (Fig. "Load transfer from the engine into the wing").
- Assembly problems: See German edition, Volume 5 Ills. 19.1-8, 20.2-2 an 20.2-3.
- Safety relevant position of the engine mount:
- At the engine: Volume 3, Ill. 188.8.131.52-9.
- At the aircraft regarding safety aspects in case of uncontained fragments: Fig. "Rotor failures risk tree" , Fig. "Containment with 'nesting' in the fan area" and Fig. "Destroyed engine mount by uncontained fragments". Thereby of cause, the naturally special features of the mount at the nacelle have to be considered.
To come up to the multifaceted demands, mounts of engines can be surprisingly complex (Fig. "Design of an engine mount" and Fig. "Engine mounts specific problems of elements"). Because they must be adjusted at special features (e.g. stiffness/rigidness, planes of containment/fragment trajectories) of the particular engine type, aircrafts will be only offered in combination with an approved engine type.
Problems with engine mounts which even lead to the separation of the engine are seldom (Fig. "Load transfer from the engine into the wing" and Fig. "Fracture of the engine connection at the pylon"). Usually first signs will be identified in time by maintenance or overhaul before a catastrophic failure may occur (Example "Connecting bolt failing by corrosion" and Example "Connection rods showing cracks"). In such cases instructions like AD-notes serve the risk minimisation. These are addressed at the operator to check and possibly change the affected parts/components.
Figure "Requirements for engine mounts": The requirements for the engine mounts, which have to be considered by the designer of the engine and the nacelle, surpass by far the bearing and the transfer of the already complex forces from the normal operation (Fig. "External loads at the engine mounts"). So the configuration of the connection bolts (fuse pins) at the joints of the mounts is something to worry about (Fig. "Engine mounts specific problems of elements").
Of great importance for the safety requirements for the operation is an optimal maintainability and an easy control/inspection for damages respectively failures like wear/fretting, corrosion or in an extreme case cracks..
Figure "External loads at the engine mounts": This sketch shows the conditions at a larger fan engine with high bypass ratio.The mounts for the loads into the casings are a matter of strength and stiffness and must be optimised. For this purpose the flanges and connections of the flaps and maintenance doors from the nacelle/cowl as well as the installations (e.g. thrust reverser at the fan) are used. Big nacelles have a high aerodynamic lift (Fig. "Backbone bending' by aerodynamic forces"). Especially its moment on the engine must be adjusted without unacceptable deformations, e.g backbone bending, (Fig. "Housing deformations near engine suspension" and Fig. "Problems by deformation of compressor housing").
Figure "Aircraft specific mounts" (Ref. 10-20 and Ref. 10-22): This picture shows examples of engine mount attachments. They depend on design features of the engine like:
- Bypass ratio,-Number of shafts (two, three) and axial position of the containment relevant rotors.
- Bearing inner structure of the engine: Casing struts (Fig. "Engine mount of an airliner"), position of the main bearings.
- Fan engine, turboprop engine or turbo shaft engine (helicopter propulsion).
Figure "Turboprop engine mounts / support" (Ref. 10-20, Ref. 10-21 and Ref. 10-22): The mount and suspension of turboprop engines, appropriate of external propeller gears (sketch below) is demanding. In this cases have additional high torques to the propeller and relative low frequent vibrations (possibility resonances) to be considered . For those loads stiffness/rigidity of the thrust frame have to be adjusted (upper sketch).
Figure "Engine mount of an airliner": A suitable load bearing inner structure (sketch below) allows, like here at a three shaft fan engine, a front mount on the fan casing (sketch in the middle). This requires naturally suitable rigidness e.g. design features. Frequently in other engine types the front mount is adapted to the core engine (Fig. "Design of an engine mount").
Figure "Engine mount of a fighter": In case of fighter aircrafts the engine is positioned in the fuselage. This can reduce the risk that the engine separates in a catastrophic manner when a failure of the not thrust loaded mount occurs. This showed a case where the rear mount failed. The engine lowered only till the the distance of few centimeters to the surrounding wall of the engine bay. The slight deviation of the of the exhaust direction hardly could be observed. So the failure was only identified during inspection on the ground.
Normally the thrust supporting mount uses a thrust pin (sketch lower left). The other mount is circumferential orientated and connected with hinges, using a ball joint (sketch lower right). Naturally, because of the cramped circumstances the maintenance and exchange of the engine have to be considered especially by design.
Figure "Design of an engine mount": This picture shows an example with front and rear mount direct connected with the core engine (Fig. "Aircraft specific mounts"). They must transfer forces and moments from the engine to the nacelle (Fig. "External loads at the engine mounts"). There are additional demands for OEMs which use typical design elements/features.
- Minimal elastic deformation of the engine casings in the region of the load initiation.
Used are circumferential lands.
- control of vibrations.
- fail Safe-feature by redundancy.
- suitable behaviour during overload (Fig. "Engine mounts specific problems of elements").
- screw joints/bolts at the support area to the pylon.
- joining rods (Fig. "Load transfer from the engine into the wing") with
- ball joints (Fig. "Engine mount of a fighter") and
- connection pins/shear pins (Fig. "Engine mounts specific problems of elements") as well as protection against slipping out.
Example "Engine mount maintenance problem" (Ref. 10-5): The problem developed within a maintenance task, ordered by the OEM (service bulletin). It goes back to an other airworthiness directive. Thereby a too low torque at the connection bolt of the front engine mount, differing from the instructions, was applied. This resulted in a too low initial tension which means a higher dynamic load (German edition, Volume 5 Ill. 184.108.40.206-2) and increases the risk of a fatigue failure.
Concerned are components of the „secondary thrust load path“. This shall guarantee the fail safe-behaviour ('only active upon failure') should the primary thrust load path fail.
In an “emergency airworthiness directive” (EAD) more then 200 concerned aircrafts are addressed. These are different types of a twin engined aircraft. The concerned connection bolts must be assembled again with the right torque.
Comment: Just because the importance of the concerned component for the fail safe-behaviour tis feature is not second-rank. Because if the primary thrust load path fails and also the second it's to reckon with the separation of the engine (Example "Engine mount maintenance deficit"). In this case a catastrophic accident becomes probable.
Figure "Engine mounts specific problems of elements" (Ref. 10-6):
Overload during operation: This develops by unbalances or the failure of a mount element (Example "Engine mount maintenance deficit"). Not to endanger the structure of the wing or the nacelle (Fig. "Engine mounts specific problems of elements"), certain connection bolts (fuse pins) may only fail in a defined sequence (Fig. "Separating an aeroengine for safety"). This is also required to avoid the fracturing of the wing tanks (Fig. "Fracture of the engine connection at the pylon"). With this feature the design will be a difficult task. In one respect, the connection bolt must meet the high safety requirements against a failure under the normal long time operation loads. To those belong especially dynamic loads.These are apparently not always sufficient known (Fig. "Design of an engine mount"). Thereby the integrity of those highly loaded components must not be affected by wear and corrosion. On the other hand the bolt must breake as planed early enough when extremely overloaded. This performance will be achieved by mainly shearing the bolt at overload, in contrast to bending by the normal operation loads.
To meet the demands, it is possible to use hollow connection bolts with special cross section features (Fig. "Fracture of the engine connection at the pylon").
Wear: Especially two kinds of relative movements can occur at the connection bolts and hinges.
- Siding wear by the planned macro movements.
- Fretting as result of micro movements by elastic strain differences. Wear can be reinforced by corrosion.
As a consequence of an unsuitable lubricant e.g. containing MoS2 (see German edition Volume 5 Ill. 22.4.1-1) in combination with humidity and/or accumulated wear/abrasion paricles this can lead to jam of hinges and so to overload.
Corrosion represents especially for connection bolts a great danger (examples 10-10.1/.2/.3). The bolts consist of high strength steels. Those are corrosion resistant against water (e.g. condensation /atmospheric moisture) and especially sea atmosphere not. Therefore a suitable coating as corrosion protection is used Is this damaged, consumed or worn by fretting over long operation times typical corrosion pits can develop. Those notches lower the fatigue strength and promote fatigue cracks/ fractures (Example "Cracks and corrosion in engine mounts").
Damages during maintenance, respectively assembly form notches that trigger fatigue failures. Plastic deformations (burrs, bulgings) can be unfavourably influence contact conditions as cause for an overload (Example "Fatigue cracks at the connection bolt of an engine mount"). Also with galling/seizing of the bolts must be reckoned (volume 4 Ill. 220.127.116.11-10).
Assembly problems always must be seen in connection with deficiencies of „human factors”. First there is the accessibility by hand and visual (Fig. "Demands on the engine mounts of an agricultural aircraft" and German edition Volume 5 Ill. 19.1-8).
A typical problem are unsufficient fastening torques of the nuts (Fig. "Demands on the engine mounts of an agricultural aircraft"). They increase the vibration stresses of the bolt. From this problem suffer also other bolts like those so fix lockings that serve against slipping out. (Example "Fatigue fracture of fastening bolt securing connection"). Also possible is the loosening of an unsufficient tightened nut (Example "Engine separates by turbine dusk failure").
Forgotten securing elements also have been the cause of the failing of a connection because the bolt slipped out (German edition Volume 5 Ill. 19.1-8).
Lay out design of engine mounts is very demanding. It is possible that not all operation loads are known, e.g. in a special application (cargo use, agricultural use) respectively operator specific (Fig. "Demands on the engine mounts of an agricultural aircraft"). Thereby the utilizable material strength, depending from long time influences like corrosion or wear, have to be regarded.
A further problem is the controlled failing at extreme overloads (Example "Engine separates by turbulence" and Example "Separating an engine purposeful at dangerous unbalances?"). To this belong external loads like from accelerations (inertia forces) by turbulences(Example "Engine separates by turbulence") and the impact of an emergency landing (e.g. on water, Fig. "Separating an aeroengine for safety"). Also other forces must be considered like extreme unbalances (Example "Separating an engine purposeful at dangerous unbalances?") as a result of failing components (fracture) of a suspension/mount (Example "Engine mount maintenance deficit" and Example "Separated engine after fatigue fracture of an 'anchor rod'") must be considered. .
Dangerous influences like fire (titanium fire, hotgas) or fragments (in case of containment) must also be considered (Fig. "Stresses during blade fragment containment"). If there are new findings, retrofits can be necessary (Example "Destroyed engine mount by uncontained fragments").
We should always see material strength in connection with the dimensioning/design. The strength must be with sufficient safety higher than all the design loads in the normal operation. This is also true over the whole scheduled operation time under estimated influences like corrosion and wear. Here problems with later material changes can emerge. An example is, when a part made of wrought/forged material is changed to a casting (Example "Cast materials requiring high caution").
Also repairs (e.g. weldings) in highly stressed zones of a mount must be considered very critic.
A defined failing under overload must consider effects like embrittlement during shock/strike (Fig. "Brittle material behaviour at impact").
Example "Engine mount maintenance deficit" (Ref. 10-12): After the start the right engine separated about 60 meters over ground. However the climb was continued and an emergency landing at the same airport was carried out without problems (volume 1 example 3-5).
The following investigation showed, that the conical connection bolt of the secondary thrust load path (secondary support assembly) of the rear mount had failed by a fatigue crack. After that both front connection bolts were also overloaded by the forward swinging engine and failed.
Obviously the fatigue crack existed already before the accident and was finally backtracked to maintenance deficits.
Comment: Unfortunately the the literature at hand gives no information about the maintenance problem which was responsible for the fatigue crack.
Does the rear engine mount break, the engine rotates around the front mount which is thereby overloaded. Probably also a knee lever effect occurred.
Figure "Engine separation risks" (Ref. 10-6, Ref. 10-24): The separation of an engine under low thrust/ power or standing will be considered by the designer for the case of overload (Example "Separating an engine purposeful at dangerous unbalances?", Ref. 10-25). This situation can be expected if the engine failure is primary. Thereby a damage of the wing, the nacelle and/or a parallel engine must be avoided (upper sketch). For this an optimal failing sequence of the connection bolts shear from front and rear mount as scheduled (shear bolts, Fig. "Fracture of the engine connection at the pylon").
In case of high engine power/thrust (start, climb, upper sketch) the risk of a catastrophic secondary failure is far higher (Fig. "Fracture of the engine connection at the pylon" and Example "Separated engine after fatigue fracture of an 'anchor rod'"). When the rear mount fails (Example "Fatigue cracks at the connection bolt of an engine mount") the engine first tilts upwards (phase „1a“). Thereby the leading edge of the wing with flaps and installations can be dangerous damaged. 'Shoots' the separated engine forward in front of the wing, accelerated by the short time still acting thrust (phase „1b”), it will be caught up in the next moment by the airplane. Thereby the danger of a collision with the parallel engine exists. This can be torn out or catastrophic damaged.
Figure "Load transfer from the engine into the wing" (Ref. 10-24): Looking at engine mounts (sketch at the left side) we are bound to include also the structure of the pylon (lower sketch at the left side) to the wing, respectively when required to the nacelle. Of influence are stiffness/rigidness/elasticities and damping as well as failure limits and failure sequences under overload. Thereto belongs, that safety important structures and other engines will not be destroyed. Also the dimensioning of the engine mounts is affected. In the pylon we find also connections with typical elements, similar to the mounts at the engine. To these belong deflection rods (lugs), connection bolts (fusion pins) and ball joints. They all are exposed to similar failure mechanisms (Fig. "Fracture of the engine connection at the pylon").
The shown configuration suffered an catastrophic accident when during the start an engine with the pylon separated. Thereby the flap control of the wing was badly damaged.n and the crash unavoidable (German edition, Volume 5 Ill. 20.2-2 and Ill. 20.2-3). The cause were during installation of the engine overloaded structures at the mount to to the pylon of the wing.
Figure "Fracture of the engine connection at the pylon" (Ref. 10-6):Concerned is an airliner in cargo version. The engine 3 (at the airplane inner side right, upper sketch) broke and separated with the pylon during climb flight in about.2000 m above ground. Thereby a loud bang occurred. It swayed to the outer side and backward and damaged the wing leading edge with the flaps. Then it hit the engine 4. This separated also with it's pylon from the wing.
Because the engines of this type of aircraft are only difficult (outer engine) or nor at all (inner engine) to see from the cockpit, the crew believed that only the function of the engines is concerned. Because only the flaps at one wing were damaged the maneuverability of the airplane was limited. It crashed in a big dwelling house.
The connection of the pylon to the wing (upper sketch) is so designed, that in case of a separation due to extreme overloads, the damaging of a wing tank does not occur (Fig. "Engine mounts specific problems of elements"). A `clean' separation of the engine nacelle and/or of the pylon shall be guaranteed with the shear of the connection bolts (shear bolts/pins).
The pylon is designed as a torsion box with 4 compartments (bulkheads). Those in front provide the connection to the engine, the rear (upper sketch) to the wing. The associated connection bolts (fuse pins) are called primary. The forces from the engine are absorbed by the front bulkhead. The next bulkhead adopts vertical and lateral reaction forces, drag of the engine nacelle as well as torsion. The four attachments (mounts) to the wing are loaded over beams of the pylon. Primary, drag forces are guided by a compression strut into the diagonal connection (diagonal brace). An additional strut (side brace) takes the side thrust.
Hydraulic systems which can be damaged by the failing of the pylon can be designed multi redundant. In this case there are four independent systems.
After this flight accident the inspection intervals of the connection bolts (fuse pins) of the concerned aircraft type were reduced. The result was a multitude of failure notes from the operators. The majority dealt with cracked bolts in different mounts/connections from the pylon to the wing.
Figure "Destroyed engine mount by uncontained fragments": Merely by chance a spectator made at the same day before the previous landing pictures of the airplane. On those photos can be seen, that the engin 3, compared with the other engines was slightly angular tilted upward (Example "Connecting bolt failing by corrosion"). With this the question arises after an already existed failure at the engine suspension which with enough attention during routine inspections could have been identified.
Example "Cracks and corrosion in engine mounts" (Ref. 10-14): There were found cracks and corrosion at the mounts of a big fan engine.
In an airworthiness directive rework (refurbish) of the engine mounts was instructed. This is carried out with a visit in the shop after a scheduled number of operation cycles, corresponding instructions (bulletin) of the OEM.
Concerned are engines with more than 5000 cycles since `new'. Than in intervals of 36 000 operation hours respectively 12 000 cycles:
Comment: Corresponding with the scheduled inspection intervals there is obviously a long time problem. The relatively short period of time after the new condition hints at a problem during lubrication and the corrosion protection. It can be supposed, that the development of the crack stands in connection with advanced corrosion. Possibly triggered by the notch effect of the corrosion pits. Those could be in a condition that can be refurbished by polishing. It seems, that a crack-mechanism like stress corrosion can be ruled out. This could hardly be intercepted by a rework, because typical, relatively fast crack propagation.
Example "Destroyed engine mount by uncontained fragments" (Ref. 10-15): In four cases of the concerned engine type (big fan engine), fracture fragments were hurled into the gas channel which than rub/grind through the casing of the low pressure turbine.
Tis airworthiness directive should prevent that the casing of the low pressure turbine will be penetrated (uncontained) and the rear engine mount with all attachments is destroyed (Fig. "Load transfer from the engine into the wing"). In such a case the breaking off of the engine must be expected (Example "Engine mount maintenance deficit" and German edition Volume 5 Ill. 20.2-2 and 20.2-3).
As measure skin doubler pads for strengthening and deflectors on the casing wall are attached at the casing wall or the exchange with an already equivalent prepared casing was assigned.
Comment: There are no informations about the cause of the primary failures, especially about the development of the fracture fragments. Because there is spoken about a grinding through (wore through) it's obvious to suppose that a whole ring of turbine guide vanes came loose. This can rotate and mill through the casing wall (Fig. "Failing of retaining bolts at contact areas"). That would also explain, why the normally by design warranted containment of the turbine casing was not sufficient.
Figure "Demands on the engine mounts of an agricultural aircraft" (Ref.10-18): The engine mount, respectively the thrust frame (sketch in the middle) seems for agricultural aircrafts subject of especially high, unexpected loads. The diagram shows events/incidents related to the mounts of three different engine types (2 piston engines, 1 turboprop) over a period of about 15 years. It concerns fractures of the connection bolts/fuse pins and engine components like tubes and angle sections. The upswing of the curve is primarily dictated by turboprop drive refitted aircrafts. A faulty design as cause seems unlikely. More likely are regarded:
- The authorized operation data don't match sufficient the real ones.
- Deficits with the assembly of the engine. To this hint fatigue fractures of connection bolts which are connected with differences to the assembly instructions. Thereby an unsufficient accessibility seems to play a role (human factors). So it came to a too low torque at the bolting (German edition Volume 5 Ill. 18.104.22.168-2).As remedy the repeatedly inspection of the engine mount, respectively the trust frame for cracks was instructed by an airworthiness directive.
The picture shows the position of a typical failure: In a welding zone of the lower tube from the engine frame, near the engine mount (sketch in the middle), a crack was found. If it comes to a fracture the engine may separate from the aircraft.
Example "Fatigue cracks at the connection bolt of an engine mount" (Ref. 10-17): During climb at about 1300 Meter over ground, the airplane rolled (Fig. "Engine affecting by inertia forces / acceleration") to the right. At the same time the pilot noticed, that the lever for acceleration (throttle slam and lock) was at idle and the airplane shuddered. A short time later, the right engine separated and dropped on an open field. Thereby the hydraulic actuation of the flaps on the wing was damaged that they were only limited movable. However a save emergency landing could be carried out.
The investigation showed that the conical connection bolt (fuse pin) from the rear mount of the right engine was broken. A fatigue fracture which propagated from the undercut radius at the end of the thread was identified. The cracks started from opposite sides. Likely cause for the cracks was a damage of the bolt. This lead to bad contact conditions and unusual loss of tightening torque (pre-stress) under cyclic loads. This triggered vibration fatigue.
After the rear mount failed, the thrust tilted the engine and overloaded the front mount (Example "Engine mount maintenance deficit").
Comment: The fracture appearance points to a reverse bending load. Especially the undercut radius to the thread (detail right, German edition Volume 5 Ill. 22.214.171.124-4) as a sharp change in stiffness than can be dynamic overloaded. This example shows the significance of expertise and attention of the engine mechanics. Did they have a chance to identify the flaw early enough?
Example "Wing specific loading of connection bolt of an engine mount" (Ref. 10-9): During the start, the crew heard in the moment of the start rotation a loud sound. At once the aircraft rotated to the right. Passengers and observers from outside saw, that the right engine separated. The aircraft continued the start and returned to the airport where it landed.
In a first investigation fatigue cracks in the rear mount of the engine are found. After that, the operation of the airplanes from the concerned airline had been forbidden (grounded).
Fatigue fractures of the conical connection bolt from the rear mount are seen in connection with several similar cases of the concerned airplane type (Example "Engine mount maintenance deficit" and Example "Fatigue cracks at the connection bolt of an engine mount"). It is interesting that in all cases it was the right engine that failed.
At the beginning it was supposed that the breaking loose of the engine stands in causal connection with an FOD. In this case a forced fracture would be expected. Under extreme overload the connection bolts are so designed, that they fail in a controllable manner (e.g. as shear-bolt, Fig. "Engine mounts specific problems of elements") to protect the aircraft nacelle/wing against catastrophic damage (Fig. "Engine mounts specific problems of elements").
However when it was found that the cause is a fatigue fracture, doubts arose about the thesis of overload..
Comment: Unfortunately withe available documents many questions can not be answered. Not clarified is, if we deal with short time fatigue (LCF) as result of dangerous high unbalances. But this however would not explain the parallel cases. Why obviously in the majority cases the right engine is concerned? A chance seems can be ruled out. Thinkable is that also gyroscopic forces play a role. These load the rear mount of both engines in different tangential directions. Gyroscopic forces depend from the sense of rotation of the rotors. They point at both engines in the same direction. But at the stiffness of the wing the moments act different. So gyroscopic forces can be `the last drop that brings the keg to overflow'. The rotation of the aircraft to the right occurred in the moment of the incident.
These cases show clearly who complex the load cases and operation loads can be for an engine mount. This requires very experienced designers and exhaustive tests. Precondition for a safe design is the knowledge of all relevant conditions. That especially is true for the connection bolts which during normal operation have also to be safe against vibration forces under the influence of wear ands corrosion, but must purposeful fail under extreme overload.
Example "Cast materials requiring high caution" (Ref. 10-7): On elements of the front mount at the side of the casing of this big fan engine, cracks were found in the contact area. (Fig. "Load transfer from the engine into the wing"). The same applies for linkage of the stiffening rod (yoke, Fig. "Engine mount of an airliner") against axial bending of the engine. Here also cracks at the casing side were discovered. Both areas consist of titanium castings. Does the mount fail, exists the danger of a separation of the engine with catastrophic consequences (German edition Volume 5 Ill. 20.2-3).
Apparently the weak points stand in connection with weld repairs. The serial numbers of such suspect components which lie in a non redundant zone are refered. Concerned parts have to be exchanged.
The manufacturer (OEM) of the airplane obviously for this reason, carried out analysis of the loads on the engine mount. It emerged that the titanium casting mounts may only be loaded with lower cyclic stresses.
The inspection interval for the assigned penetrant inspection equals 4000 start-stop cycles.
Comment: Cast material generally has, depending from its structure (coarse, inhomogeneities), a lower fatigue strength than the wrought condition. This is also true for titanium alloys. Are there in highly loaded zones additional welding repairs, even more caution is required. Such welding repairs frequently are carried out during the production on the raw part in allowed/specified limits. It's rather not to assume that cracks which developed during operation were re-welded.
Also the OEM should have known about the lower fatigue strength. Obviously the operation loads of the mounts were not sufficiently known, so the danger was not identified.
Example "Fatigue fracture of fastening bolt securing connection" (Ref. 10-10): In an airworthiness directive the visual inspection of the front engine mount thrust connection was instructed. Thereby every deviation must be corrected before the next flight.
Problematic is the securing end cap respectively its bolts (`end cap bolts') to prevent the slipping out of the connection bolt. Because there are no detailed informations these have been interpreted, corresponding with the sketched detail.
The directive is a reaction at the fatigue fracture of an end caps fastening bolt, securing the connection bolt. Does this securing exist no more, the engine can separate when the primary load bearing element fails (slipping out of the bolt?).
Comment: The dynamic overload of the securing end cap bolts hints at relevant axial forces on the connection bolt. With this, the direct danger exists, that the unsecured bolt moves out. Obviously a visual inspection relates to the fracture of a bolt, respectively the lack of the bolt head.
Example "Separated engine after fatigue fracture of an 'anchor rod'" (Ref. 10-6): This cargo version of an elder airliner type lost during climb an, inside on the wing positioned engine, that after separating hit the outer engine and tore it off. The cause was a fatigue fracture of an 'anchor rod' to the wings middle bearing structure (midspan fitting lug).
Comment: A similar case occurred only one month later. However here the outside positioned engine was not torn off. It's eye-catching that obviously cargo versions are concerned (Fig. "Fracture of the engine connection at the pylon"). In both cases the engines had been retrofitted with so called 'hush kits'. Which role the life of the engines from those normally sorted out passenger airliners and/or the maintenance played is not clear.
Example "Connecting bolt failing by corrosion" (Case 1, Ref. 10-6): After about 42 000 flight hours with about 9 000 starts when the airplane stood on the airfield, it was noticed that the engine 3 together with the pylon was slightly twisted to the fuselage. A two centimeter wide gap to the wing had been formed. An investigation showed, that the connection bolt to the outside supporting beam was broken and at three positions heavy corroded.
Example "Crack forming near corrosion pits" (Case 2, Ref. 10-6): The airplane had about 52 000 flight hours with about 11 000 start cycles. During a visual inspection of the pylon/wing connection a crack was discovered in the hollow rod which is connected to the inside bearing strut at engine 3. The crack was located in a zone of distinct corrosion pits.
Example "Connection rods showing cracks" (Case 3, Ref. 10-6): The airplane had about 50 000 flight hours with about 10 500 start cycles. During a corrosion check in line with a C-inspektion a crack in the connection rod to the inside bearing strut at engine 3 was detected. A second crack was found in the diagonal connection strut.
Comment: Unfortunately, from the available descriptions doesn't arise how far the corrosion was causative for the fracture of the connection bolts. But this can be assumed. The high strength steels of such components are from experience susceptible for deterioration by corrosion under the influence of sea atmosphere, respectively condensate with sea salt Insufficient corrosion protection is especially alarming. This can be due to the production. But this also can occur over longer operation times (in all 3 cases several 10 000 flight hours!) by wear or consumption of the protection coating. Especially dangerous is the notch effect of the corrosion pits (Fig. "Types of watery corrosion"), an embrittlement as result of hydrogen charging (Fig. "Stress corrosion cracking" and Fig. "Role of hydrogen in corrosion processes") and drop of the fatigue strength simultaneously influence of the corrosion medium during vibration load (corrosion fatigue, Chapter 5.4.3).
Example "Engine separates by turbulence" (Ref. 10-6): This cargo version came into heavy turbulences during climb in about 600 meters over ground. Thereby the engine 2 separated with the pylon from the wing. The leading edge of the wing to the outer engine was badly damaged. The pilot succeeded an emergency landing at the start airport although the airplane was difficult to control. During the following investigation a crack about 75 mm long was found in the thin walled structure of the pylon midspan attachment near the linkage. The OEM found in this crack sealing compound. This indicates, that the crack existed already some time. Anyway as main cause the overload due to the turbulence was supposed.
Comment: From the available informations can not be concluded if the fracture runs through the found crack. The location where the turbulence occurred is known for such extreme events (Fig. "Heavy turbulence").
Figure "Separating an aeroengine for safety":With the size if the engines, especially of the fan the dynamic loads in case of a rotor blade/fan blade fracture increase. Typical are unbalances and high torsion moments as result of a sudden retarding of the rotor by seizure. Thereby the structure of the airplane /wing, pylon will be endangered by a catastrophic failure/fracture. The question arises if this can be practicable avoided by a controlled failure of shareable connection bolts/ fusion pins which fail at extreme overloads. So the risk can be minimised, that components (e.g. tanks, actuation of flaps) get damaged with catastrophic results. But those systems have also problems. They must not fail by the normal high dynamic operation loads (see Example "Dealing with catastrophic engine seizure").
A further situation for a purposeful separation of the engines by an controlled explosion blow is an emergency ditching into water. The high water drag of the engines can force the airplane to tilt forward even at optimal horizontal flight attitude when both engines immerse at exactly the same moment.
However it must be reckoned due to a sloping position. In such a case the airplane will be tilted in a manner that even the fuselage will break. To avoid this situation both engines should be controlled separated before the water will be touched. This must occur not until the connection bolt at the engine mount fails uncontrolled during touching the water. Otherwise extreme loads act at the airplane.
Therefore obviously a technology from the aeronautics is discussed. There explosive predetermined studs are attached, which when a dangerous overload occurs, or possibly automatic by sensors, can be triggered by the pilot before touching the water. A benefit would be that the predetermined explosive stud could be sufficient over-dimensioned to rule out a failing by normal operation loads. A grave problem may exist in the possibility of an unintended ignition. Therefore the reliable triggering must be guaranteed under all situations.
Example "Separating an engine purposeful at dangerous unbalances?" (Ref. 10-3): This cargo airplane lost the left outer engine in about 5300 meters during climb.The pilot could carry out the emergency landing. Later the engine was recovered from a lake.
Cause was a dynamic over load of the front and rear engine mount as result of unbalances. Those developed, because the disk rim (outer portion) of the second high pressure turbine stage fractured with blades (German edition Volume 5 Example 126.96.36.199-1). The disk failure was due to a maintenance fault at the second stage turbine nozzle. So it could tilt backwards and touch/rub the disk. Since years exists an explicit warning in the engine handbook about such a possibility.
Comment: Because it's a failure of the engine mounts it can be assumed that the engine separated with the pylon.
Example "Engine separates by turbine dusk failure" (Ref. 10-4): After a former lubrication of the thread a slipped out connection bolt from the upper cross link of the pylon was found. This can cause a separation of the engine from the wing. As cause the possibility exists, that the nut after a remove was no more sufficient tightened.
In an airworthiness directive adequate inspections for a high number of airplanes of the concerned type are required.
Comment: The loosening of the bolts during operation is discussed in the German edition Volume 5, Ill. 188.8.131.52-5.
10-1 „767-400 engine mounts strengthened“, Zeitschrift „Flight International”, 2.8 June 1999, Page 8.
10-2 G. Norris, „Ultimate Power“, Zeitschrift „Flight International”, 9-15 June 1999, Page 163 bis 166.
10-3 Special Airworthiness Information Bulletin SAIB: NE-07-4, „Subj: Pratt & Whitney (P&W) JT9D HPT2nd Stage Vane/Disk Failures“,August 3, 2007, entspricht ESA Safety Information Notice No.: 2007-20 vom 08 August 2007, Page 1-3.
10-4 Civil Aviation Safety Authority (Australia), „P&W JT9D Aft Engine Mount Tangential Link”, AD/B747/157 / 96 Amdt1, Seite 1.
10-5 Emergency Airworthiness Directive (EAD) by Direction Generale de L'Aviation Civile (France), GSAC/T 32/05 „Subject: Engine-Forward mount bolt (ATA 71)“ Airbus A319/A320/A321, Page 1 und 2.
10-6 Nederlands Aviation Safety Board, Aircraft Accident Report 92-11, „EL AL Flight 1862”, Boeing 747-258F 4X-AXG, Bijmermeer, Amsterdam, October 4, 1992, Page 1 and 47.
10-7 Civil Aviation Safety Authority (Australia), „General Electric Turbine engines- CF6 Series, Forward Engine Mount Platform“, AD/CF6/53 , 13/2003, Page 1.
10-8 Airworthiness Directive FAA-2006-23871, FAA Aircraft Certification Service 2007-05-14, „General Electric Company, Turbofan engines with cast titanium assembly engine mount platforms…”, Page 1-9.
10-9 D. Kaminski-Morrow, „Nationwide (Airlines) 737 engine-loss inquiry finds fatigue crack in mount“, Source Flightglobal.com, 04/12/2007, Page 1.
10-10 Civil Aviation Safety Authority (Australia), „Engine Thrust Links PW 4000 Powered Aircraft”, AD/B747/215, Page 1 and 2.
10-11 Civil Aviation Safety Authority (Australia), „Agricultural Aircraft Safety Review“, December 2008, Page 38 and 51.
10-12 Aviation Safety Net, (aviation-safety.net), „ASN Aircraft accident Boeing 737-232 N322DL Dallas-Fort Worth, TX”, (Engine separated from wing), 7. Januar 1992, Page 1 and 2.
10-13 Federal Aviation Administration (FAA), Docket No. FAA-2008-0410, „A Viking Air Limited Model (Caribou) DHC-4 and (Caribou) DHC-4A Airplanes“, (Engine separated from airplane), Januar 23, 2008, 1992, Page 1-6.
10-14 Civil Aviation Safety Authority (Australia), „Engine Mounts (RB211-535E4, cracks and corrosion)”, AD/RB211/28, 6/2001, Seite 1.
10-14 Civil Aviation Safety Authority (Australia), „General Electric Turbine engines- CF6 Series, Forward Engine Mount Platform“, AD/CF6/53 , 13/2003, Page 1.
10-15 Federal Aviation Administration (FAA) Docket No. FAA-2007-28367, AD 2008-21-11, “General Electric Company CF6-80C2 Series and CF6-80E1 Series Turbofan engines (Loss of the mount system structural integrity by fragments of the LPT-casing could result in the engine separating,) Fragments of LPT-Casing ”, AD/CF6/53 , 13/2003, Page 1-9.
10-16 Federal Aviation Administration (FAA) Docket No. 99-NE-45-AD, AD 2000-12-08,”General Electric Company (prevent failure of left hand engine mount link assemblies) Fragments of LPT-Casing “, August 28, 2000, Page 1-5.
10-17 NTSB Identification :NYC88FA050, „Engine separation Boeing 737-2B7, December 05, 1987” (mounting bolt fatigue due to engine installation ), Page 1.
10-18 Civil Aviation Safety Authority (Australia), „Airtractor AT-300, 400 and 500 Series Aeroplanes (cracks in engine mounting tube) “, AD/AT/29, 17/2008, Page 1-3.
10-19 Federal Aviation Administration (FAA) Docket No. FAA-2007-27360, AD 2007-06-05, “Airbus Model A318, A319, A320 and A321 (incorrect torque to the forward engine mount bolts during maintenance) ”, March 30, 2007, Page 1-7.
10-20 „The Jet Engine” , Rolls-Royce plc, 3rd edition Seite 159 von 1969 und Fifth edition ( ISBN 0 902121 2 35) von 1996, Page 248 and 257.
10-21 I.E.Traeger, „Aircraft Gas Turbine Technology, Second Edition“ , Glencoe, ISBN 0-07-065158-2, 1994, Seite 437.
10-22 Z.S.Palley, I.M.Korolev, E.V.Rovinskiy, „Structure and Strength of Aircraft Gas-Turbine Engines”, Foreign Technology Division, FTD-HT-23-903-68, Übersetztung aus dem Russischen, 1994, Page 265.
10-23 „Trent 800, Anatomie eines Turbofans“, Zeitschrift „Flug Revue”, April 1996, Page 58 and 59.
10-24 National Transportation Safety Board (NTSB), Aircraft Accident Report NTSB-AAR-79-17, „American Airlines, INC., DC-10-10, N110AA, Chicago-O-Hare International Airport, Chicago, Illinois, May 25, 1979“, Page 62.
10-25 „Engine Structural Integrity Program (ENSIP)”, Department of Defense Handbook, MIL-HDBK-1783A, 22. March 1999, Page 110 up to 112.