Table of Contents
21.2.7 Straightening of deformed part zones
Operation influences like creep at hot parts or FOD at fanblades can produce permanent deformations of the part. A danger of such deformation has different reasons (Fig. "Straightening of parts").
Turbine stator vanes are subject to high, buckeling thermal stresses (volume 3, Ill. 12.5-15). Turbine rotorblades which are tensioned with the tip shrouds, will permanently untwist under the centrifugal forces (volume 3, Ill. 11.2.3.2-6). But warpages can also form during a repair step like welding, brazing, heat treatment or handling of the parts (volume 4, Ill. 16.2.2.5-13).
In many cases where deformations will stay in the limits prescribed by manuals and specifications, a defined straightening process is approved. So leading edges of a one piece cast (integral) turbine stator can be suitable bent with an simple tension device (Lit. 21.2.7-2). Also FOD caused deformations at fan blades are straightened (Fig. "Straightening of parts").
Such straightening processes are absolutely not unproblematic and have a damaging potential (volume 4, Ill. 16.2.2.5-12 and Ill. 16.2.2.5-13) . To those count (micro-) cracks (Fig. "Repair at Fanblades"), change of material structures with drop of strength (e.g., from uncontrolled heating), embrittlement effects (e.g., oxidation during heating) and internal tension stresses.
Therefore approved straightenings are detailed specified. They are based on positive part specific operation experience. The straightening process can be carried out, depending from the material in a certain temperature range, with local of overall heating. Preferably it is carried out under cover gas (inert gas, eg., at titanium alloys) and in a defined time lapse (e.g., during forming by creep).
To release for the fatigue strength unsuitable internal tensile stresses caused by the straightening process, a following heat treatment (stress relieve annealing) or shot peening offers itself.
Figure "Straightening of parts": Deformations of parts can become noticeable negative in different ways during operation. That is the reason for approved/certified straightening processes during overhaul. The risks of deformations are highlighted in the following examples.
„S1“ Notches and cross section steps increase locally the operation stresses and lower with this the fatigue strength. So the danger of vibration fatigue with crack formation up to the fracture arises.
„S2” Internal tensile stressses caused by deformation and spring-back (Band 4 Bild 16.2.2.5-13). Tensile stresses add to the operation stresses, This rises the mean stress and lowers the usable fatigue strength (volume 4, Ill. 16.2.2.4-4) .
„S3“ Vibration excitation of a deformed part like a compressor blade. They can be triggered from a local stall at the aerodynamically no more optimal profile. Develops in the compressor a rotating stall (volume 3, Ill. 11.2.1.1-1) a dangerous high frequency vibration load can occur.
Also blades of neighbored rotor stages can be excited from the flow disturbance of a damaged compressor blade to dangerous vibrations (volume 3, Ill. 12.6.3.3-21).
„S4” Resonance as result of the change of the vibration behaviour. A deformation can be adverse change of stiffness and/or mass distribution and so influence unsuitable the vibration behaviour of the part. Concerned is as well the vibration mode (line of nodes) as also the natural frequency. Especially endangered are blades which changed their resonance behaviour in a way, that existing excitations during operation, which normally are harmless, now trigger high vibration loads.
„U1“ Damages of the cooling from hot parts. Typical example is the warpage of combustion chambers (volume 3, Ill. 11.2.2.1-10) and turbine guide vanes (nozzles).
Thereby as well the cooling air supply in the channels/bores can be hindered, as also the formation of a cooling air veil on the hot gas impinged surface. As consequence a local rise of the part temperature with exponentially dropping creep life (volume 3, Ill. 12.5-4). In an extreme case the warpage of a combustion chamber leads to a failure of the casing with hot gas exit (volume 3, Ill. 11.2.2.2-9). So, besides the danger of a safety relevant early failure of the part, the probability of high costs for overhaul /spare parts exists.
“U2” Performance drop and decrease of thrust can be especially influenced by the warpage of the high pressure turbine guide vanes/nozzles (Fig. "Straightening of deformations from turbine vanes"). Also the warpage of the sealing surfaces from labyrinths, respective its supporting casing structures, can be markedly noticeable (volume 2, Ill. 7.0-3 and Ill. 7.0-4). Especially concerned are static sealing segments in the casing above the tips of the high pressure turbine rotor blades. If it is necessary to rise the gas temperature to get a sufficient thrust, this will be at expense of the hot parts lifetime. This drives with the increased fuel consumption the operation costs.
„K1” Deterioration of the specific fuel consumption see „U2“.
„K2” Increased costs of repair and spare parts see „U1“ and „U2”.
Illustrations 21.2.7-2.1 and 21.2.7-2.2 (Lit. 21.2-22): The fracture of a fan blade between clapper and root platform (sketch below right) caused a failure with fragment exit (uncontained) from the aeroenine (upper frame, left side). The laboratory investigation of the part of the blade, which remained in the disk (right side, below left), showed that the fracture is an about 85 mm long fatigue crack (HCF) which started at the trailing edge (right sidee, sketch and detail abobe right). From this crack a forced fracture follows. Concentric to the region of the crack origin traces of a rework can be seen. This rework obviously lead to a shortening of the chord to 181,14 mm. This is below the minimum length of 182,12, specified in the manual.
The depth of the dark coloured flaw is about 0,6 mm from the leading edge (detail in Fig. "Repair at Fanblades", above right). This shows a microscopic zone with features of a ductile forced fracture. Indications of a material flaw, a failure causing influence on the material or FOD are not present. There are traces of oxidation, accordant to the dark discolouration.
This discolouration of the flaw at the fracture origin and the features of a forced fracture can be explained with a straightening process under temperature at the trailing edge. Such straightening procedures at 650 °C have been carried out two times at this blade in the course of time. Thereby the initial crack may have formed unnoticed. It was not identified by the exclusively eddy current test. Obviously also periodic NDT at the operator could not find the flaw.
The potential failure promoting impacts at such a rework contains the Ill. 22.2.7-2.1.
Directly before the assembly, about 320 operation hours before the incident, the leading edge of the blade (not at the trainling edge where the flaw is located!) was tested with eddy-current.
Comment: It may be, that this testing is limited to the trailing edge because there possible FODs can be expected and/or the OEM from experience sees here rather the likelihood of a fatigue crack.
Figure "Straightening of deformations from turbine vanes" (Lit. 21.2.7-24): The failed aeroengine showed, that the 1st stage of the low pressure turbine (LPT) lacked. Only a 120° segment of the turbine wheel lay at the runway. The laboratory investigation suggested four opposed fatigue cracks. This would be typically for a disk vibration mode with two nodal diameters (volume 3, Ill. 12.6.3.1-8).
The aeroengine had since the new assmbly about 8800 operation hours with about 5200 start cycles. About 100 hours and about the same number of cycles before the incident a mainmn inspection (major periodic inspection = MPI) was carried out. This showed, that the 1st stage stator and disk (cross section middle right) must be exchanged. The turbine disk was a new part. However the exchanged turbine stator was a purchased repair part. The list of the OEM with the approved/certified repair shops did not contain the delivering.
The concerned stator (sketch bottom left) with 67 vanes is casted in one piece. During the repair a flow check was carried out. Thereby the stator was fixed in a device and the flow rate evaluated. As measured value a mean value is specified. This shows, that a straightening process at the vanes is necessary. A creep deformation by the operation is typical for such parts (volume 3, Ill. 12.5-15).Nonuniform passage cross sections between neighbored vanes means in the gasflow corresponding pressure differences at the circumference. They represent an intense potential vibration exitation.
A check of the serial number from the failed part showed inconsistencies.
Straightening process during stator repair: For every aeroengine individual rotor speeds and gas temperatures are necessary for a certain flow cross section. A computer program determines the vanes of the turbine stator, which must be suitable formed. Thereby at particular vanes the trailing edge is adjusted with a bending device. The computer program was introduced because of former failure cases. During those NDT disks without flaws broke after few hundred operation hours by vibration fatigue when combined with an overhauled stator (HCF, volume 3, chapter 12.6.3.1).
This behaviour could be explained with a unsuitable distribution of unequal flow cross sections between two neighboured vanes. This triggered a resonance vibration of the 1st stage NDT-Stufe. There was a warning about this possibility by the OEM and the aeronautical authority (FAA) and a service bulletin to the certified repair shops have been informed. Additonally a check was defined, which is not in the manuals. A repair shop that was not certified by the OEM, like in the incident, didn't get this instruction. This was also valid for later instructions of the OEM. To this belongs the application of the computer program to identify the position of vanes which must be straightened. So it is understandable, that this instructions have not been considered for the failure part. Therefore an increased risk of a dangerous vibration exitation of the turbine wheel behind existed.
References
21.2.7-1 Australian Transport Safety Bureau, Technical Analysis Report No. 9/02, „Analysis of a failed Pratt & Whitney JT9D-7R4 turbofan engine, Boeing 767-238, VH-EAQ“, Occurrence 27 November 2001 (No.: 200105627), page 1-14.
21.2.7-2 AAIB Bulletin No.: 6/2004, „Serious Incident, Severe damage to right engine and thrust reverser” , Aircraft HS125-700A, Triebwerke Honeywell TFE731 3R 1H, page 1-5.