16:162:1622:16225:16225

16.2.2.5 Mechanical Overstress and Damages

In the following, finishing-related mechanical overstresses and damages are understood to be influences in the course of a finishing process that use force to have an unallowable effect on the operating behavior of the parts. Typical examples of damage due to dynamic force application include dynamic fatigue cracking due to excessive grinding of blade tips and ultrasonic baths that are operated using the wrong parameters (Fig. "Damage at turbine blades by ultrasonoc cleaning").

Even fastening parts for working can overstress them (Fig. "Fastening of parts during finishing"). Machining processes such as reaming, milling, and turning can tear part surfaces open. Typical mechanisms are chattering (Fig. "‘Plucking’ damages by a machining process") and comma cracks (Ill. 16.2.1.1-2).

New and alternative procedures are always accompanied by new dangers and challenges. One example is tumbling in vibrating containers, during which the parts can damage one another (Fig. "Damage risk during ‘secondary tasks’").
If shavings are not sufficiently removed, they can fuse with or be pressed into the part surface, or cause grooves and notches (Fig. "Dynamic fatigue strength influenced by fused chips").

Tool breakages (Fig. "Dangers of tool-fractures") and intensive friction contact between tools and parts (Fig. "Friction contact during clamping") are especially damaging.

Dynamic overstressing of the parts, resulting in macro-cracking, can occur during many finishing processes (Fig. "Dynamic fatigue by finishing processes i"). If damaged parts are installed in engines, they can cause extensive safety-relevant operating damages (Fig. "Dynamic fatigue cracks by ultrasonoc cleaning").

Parts can be damaged when, for example, they are torn out of their fastening during machining, have undefined contact with tools, or strike something during handling. Plastic deformations such as the bending of blades are typical results (Fig. "Cracking of brittle coatings by deformation"). In this case, especially with the expensive parts used in engine construction, there are questions concerning straightening processes and their risks (Ills. 16.2.2.5-13 and 16.2.2.5-14).

In addition to the actual deformation, in parts with brittle coatings there is a danger of cracking, which is difficult to detect (Fig. "Cracking of brittle coatings by deformation").

Sufficiently safe and serially-implementable prevention of potential damages is a special challenge. First, they must be recognized. Next, specific solutions require exact identification of the damage causes (Ills. 16.2.2.5-10 and 17-11). If reworking is considered, the risks related to this must be assessed (page 16.2.2.5-22). In the case of large deformations, straightening processes may be necessary, which also have certain related risks (Ills. 16.2.2.5-12 and 16.2.2.5-13).

Figure "Fastening of parts during finishing": Fastening parts on machine tools and equipment can plastically deform parts and cause dangerous damage. Filigreed parts such as rotor components (rings, disks, and blisks) are especially sensitive to this. They are subject to the highest safety and strength demands, and extremely valuable. Geometric deviations from the design drawings can be detected with sufficient accuracy. In this case, cost is the primary concern. The expenses can be significantly greater if safety-relevant damages remain undetected and are installed in engines. Typical scenarios for the damaging of parts during fastening are:

Mechanical fastening with the aid of adjustable equipment: A typical example is a three-jaw chuck (top left diagram). Excessively high fastening forces in the contact areas can elastically or plastically deform the part. This can be seen in the form of impressions on unmachined fastening surfaces. Residual stresses from the machining process can create additional tension between the part and fastening equipment. The part will then warp as it springs back following release from the equipment.
If highly filigreed parts cannot be fastened tightly enough, there is a danger that they will be partially or completely torn out by the tool. Experience has shown that this type of self-increasing process will usually result in overstress and destruction of the part.
If contact surfaces of fastening equipment are fouled by shavings or particles, they can be pressed into the part, where they can act as dangerous notches.

Affixing centering surfaces: Large diameters tend to jam and be damaged while being slid on (bottom left diagram). If the contact forces are too large, then galling (cold welding) can occur both when the part is slid on and/or off. Foreign particles promote this type of damage. Galling marks in the shape of grooves are especially dangerous (bottom detail). Especially hard foreign particles such as metal dust or blasting media can create pronounced grooves. Tilting and jamming can leave impressions that act as notches in the contact surfaces. All of these damages can decisively shorten the cyclical part life. Except for those located in small bores, they are generally sufficiently safely detectable. Because the listed damages can also influence the material below the surface, in the life span-determining part areas such as the hub bore (bottom right diagram), the possibility of reworking should be thoroughly analyzed (Chapter 17.5).
An additional problem is the increase in the joining forces in case of galling. For example, a specified amount of force may become insufficient for the axial positioning of a part in a machining device, which is necessary to ensure dimensional accuracy.
A similar problem occurs during shrink-fitting or pressing-on seats. Even minor deviations in tolerances and temperatures necessitate powerful pressing forces. This creates a high potential for damage through galling. The risk is further increased by the lack of lubricant. In this case, galling marks are covered up and no longer visible. This is especially dangerous if these connections are located in highly-stressed part zones such as hub/shaft connections.

Cast workholding: This large-scale fastening method uses low melting alloys or thermoplastic polymers. It is used for parts with complex geometries (e.g. blades, top right diagram). In this case, it must be ensured that no wear products from the casting material remain on the part, where they could cause damages at operating temperatures (Ills. 16.2.2.3-10.1 and 16.2.2.3-14). If, during melting for removal of the part, sufficiently high thermal stress is created between the solid cast block (greater expansion) and a filigreed part area (e.g. blade with internal cooling configuration), it is possible that brittle diffusion layers may crack (Fig. "Cracking of brittle coatings by deformation"). These very fine cracks are difficult to detect and can reduce the oxidation life considerably, as well as compromising the dynamic fatigue strength.

Figure "‘Plucking’ damages by a machining process" (Refs. 16.2.2.5-6 and 16.2.2.5-7): Chatter marks can dangerously lower the dynamic fatigue strength. They are created during machining with a defined cutting edge (turning, milling, reaming, boring). Especially during reaming, chatter marks are often created in the highly-stressed reamed grooves that accept the blades. The cause of this is self-increasing vibration. More energy is imported by the machining process than is used up by damping friction. These processes are also known in other technical fields, such as in the squeaking of brakes. The cause is a stick-slip effect in which a friction force cyclically builds up while braking energy is stored. Chattering is an unstable dynamic process with sudden jumps in amplitude. Chattering can damage machine tools and wear them down prematurely. However, safety concerns mean that the potential direct or indirect (tool wear) damage to parts is more important.
Chatter vibration in machine tools is usually caused by several simultaneously acting phenomena. In addition to the stiffness of the machine tools, the machining parameters, the condition of the tools, and the cooling/lubrication are all important. During optimization of the machining process, the so-called threshold cutting depth is determined. This is the cutting depth below which no chatter occurs. Naturally, it is only valid for the specific tested tool conditions. If wear has changed these, then dangerous chatter can occur. One can see that arbitrary changes to the machining process, including the use of untested but cheaper tools, threatens the safety of the parts being worked on.

There are two major types of chatter vibrations that occur in machine tools:

“Plucking” is a type of chatter with decreasing cutting force: In this case (bottom right diagram), the vibration moves in the same direction as the cutting motion. Experience has shown that this type of chatter will damage the surface area, especially. It will tear open the surface and create micro-cracks. A stick-slip effect occurs because the thermally-dependent friction conditions mean that greater cutting forces act when the tool swings forward than when it swings back. This situation is caused by the lower effective cutting rate.

Regenerative chatter occurs with a vibration that is perpendicular to the suface (direction of thrust; bottom right diagram). The vibration builds up while machining a wavy surface, and has a reciprocal influence on the machining force.
In machining processes without defined cutting edges, such as grinding, vibrations in the direction of the tool infeed will be more damaging. In the phases of the deeper impact, greater heat development can be expected. It is also possible that the supply of cooling lubricant (Chapter 16.2.1.1.1) may be compromised. This explains tarnishing that occurs in a striped pattern.

Figure "Dynamic fatigue strength influenced by fused chips" (Ref. 16.2.2.5-5): The danger that insufficient chip removal during machining poses for the operating safety of parts is often not sufficiently understood. The Pensacola accident (Ills. 16. 2.1.1.-9.1 to -9.5), in which a fan disk burst due to finishing-related cracking, makes this especially clear. The cause for a chip jam during turning machining is usually unfavorable spatial conditions (left diagram). This applies especially to hollow spaces that further complicate subsequent damage detection due to their poor visual accessibility. When boring or milling small openings, chip jams can often be prevented only through periodic removal of the tool (right diagram). In these cases, the poor visual accessibility makes it difficult to evaluate the machined surface. Chip jams can cause damages in various ways:

  • Galling of the chips with the part surface (Ill. 16.2.1.1-3.1).
  • Heat development due to intense friction and/or restricted cooling/lubricating media.
  • Influencing of the cutting process: tool wear due to high cutting forces increases the danger of chatter damage (Fig. "‘Plucking’ damages by a machining process").

Damages caused by chip jams:

  • Cracking
  • Tensile stresses
  • Oxidation, embrittlement
  • Strength losses due to structural changes

Figure "Friction contact during clamping": Similar to chip jams (Fig. "Dynamic fatigue strength influenced by fused chips"), friction contact between non-cutting surfaces presents a danger of fretting (Ill. 16.2.1.1-3.1). The necessary relative movement results from the stationary tool and moving part (e.g. turning), or vice versa (e.g. milling).

The contact surfaces are primarily the blade holders (right diagrams) or fastening devices for the tools (left diagram). They are usually made of steel. This results in the transfer of foreign material (Fig. "Verification of local damaging deformations", Ref. 16.2.2.5-9), which can have additional damaging effects during subsequent finishing processes such as ECM, etching, or welding (Fig. "Fouling of Titanium welds"). Intensive rubbing on parts causes structural changes, plastic deformations, embrittlement, micro-cracking, heat development, and high temperatures.

Figure "Dangers of tool-fractures" (Ref. 16.2.2.5-10): Tool fractures are a relatively common problem. The risk of tool fractures is increased by the tendency towards higher machining speeds and harder materials in the parts being machined. These factors require harder and more temperature-resistant cutting materials, which are usually more brittle.

Insofar as the tools are made from hard metals consisting of tungsten carbide (WC) in a cobalt matrix, there are possibilities for serially applicable verification. However, these are not entirely satisfactory, and improvements are being developed. One of these is a type of eddy current testing that reacts to changes in the electrical and magnetic properties relative to the uninfluenced base material. For non-magnetic materials such as titanium and nickel alloys, the magnetism of hard metals presents a possibility for detection. Intensively and quickly formed structures can create adiabatic shear bands (Fig. "Verification of local damaging deformations"). These can also be detected by eddy current testing.

If the broken particles have already been removed by a subsequent machining process, then problematic remaining areas with plastic deformations and structural changes can still be detected. This is possible, at least in titanium alloys, with the aid of suitable etching techniques (blue etch anodizing = BEA, Fig. "Verification of local damaging deformations").

Figure "Damage risk during ‘secondary tasks’": Even apparently harmless finishing processes should be considered. These include tumbling processes in vibrating containers. In these processes, the parts move freely in part-specific vibrating grinders. This usually occurs in a rotating and rolling motion. Proper loading (e.g. number of parts) should ensure that no damaging contact between parts can occur. Depending on the part geometry and later operating loads, damage is possible. Blades are some of the more sensitive parts. Possible damages include fretting marks (false brinelling), brinelling (Fig. "Dynamic fatigue cracks by ultrasonoc cleaning"), and notches.

The loading of the machine with the parts and grinding media must be done in accordance with regulations and with sufficient care to prevent damage to the parts.

Illustrations 16.2.2.5-7.1 and 16.2.2.5-7.2: At first glance, one would not expect that dynamic fatigue could damage parts even during finishing. There is no satisfactory, serially-implementable, non-destructive method to detect dynamic fatigue before noticeable cracking has occurred. Therefore, it cannot be determined whether considerable dynamic fatigue has already occurred in a crack-free part. Destructive verification of the damage can only be expected with the aid of dynamic tests (e.g. with an electrodynamic shaker), which will show reduced life spans and/or decreased dynamic strength. It is necessary for the test to stress the damaged part zone. In addition, the minimum life span for undamaged parts under the dynamic loads in question must be known with sufficient exactness for purposes of comparison. Therefore, the damaging vibrational mode of the finishing process should be applied if possible. Therefore, the damaged part zone and the damaging vibrational mode must be understood. For this reason, damage verification done with the aid of a serially-implementable non-destructive testing method such as penetrant testing is a necessary condition if a sufficiently large (macro) crack has already formed. The detection of these cracks can be considerably complicated by subsequent finishing processes before crack detection. These influences include processes such as shot peening and abrasive blasting (Fig. "Limiting influences non-destructive testing"), as well as machining processes such as grinding or turning. If the surface is plasticized, cracks can be pressed shut or smeared over. The detectibility of cracks with penetrant testing can also be worsened by oxidation during heat treatment or by the presence of foreign media such as oils. Of course, internal cracks (e.g. in the cooling structure) cannot be detected in many cases, at least without a specific search in a closely limited and defined part area. The effectiveness of X-ray testing, another commonly used non-destructive serial testing method, is no better. Especially if the location and path of the crack on the part are not known in advance, then optimal adjustment of the testing parameters is difficult. If a crack is found during finishing, it must be assumed that it is merely the tip of the iceberg, rather than an isolated case (Volume 1, Ill. 2-3.3). Part lots that meet the requirements for vibration excitement are always suspect.

Therefore, preventing dynamic fatigue in finishing requires a thorough understanding of the affected processes. Strict observance of safe process parameters, machines, and equipment must be emphasized. The following chart contains critical processes and possible causes of vibration:

“1” Ultrasonic cleaning: This intensive cleaning process is used during new part finishing, as well as during repairs. High-frequency vibration of the part in a cleaning fluid dissolves even firmly adhering fouling. Excessive ultrasonic energy and/or resonance of the part can lead to dynamic fatigue. Wear problems on roller bearings (false brinelling, Fig. "Dynamic fatigue by finishing processes i") and cracking in cooled turbine blades (Fig. "Dynamic fatigue cracks by ultrasonoc cleaning") have been reported.
High-frequency vibrations are also considered to be especially problematic because dangerously high vibrational stresses can be created even at very small amplitudes (around 0.1 mm).
A combination of high-frequency vibrations with other machining processes such as grinding and ECM would at least potentially be capable of exciting dangerous vibrations in parts. For this reason, these processes must be thoroughly tested for this damage mechanism before they are implemented.

“2” Electrochemical machining: There are several indicators that electrochemical milling (ECM) can cause dynamic fatigue cracks in parts (Fig. "Problemzones of the ECM process"). The vibration excitement is caused by processes in the gap between the part and the tool (Ref. 16.2.2.5-12). These processes include bubbles, instability in the electrolyte flow (eddys, fluttering of elastic cross-sections), or pressure fluctuations from the supply pump.

Figure "Dynamic fatigue by finishing processes iii" shows a turbine disk used to power the fuel pump of a rocket engine. Dynamic fatigue cracks were found in several blades after finishing. There were two fundamental excitement mechanisms that may have caused the cracking:

  • The machining process during the removal of the ring racks on the sides (“4”) following ECM.
  • The ECM process.

“3” Grinding: A typical example is subsequent grinding of the blade tips of compressor rotors in order to minimize the clearance gaps. In this case, the pressing of the blade roots against the rotor due to centrifugal force usually provides sufficient damping to suppress vibrations. During subsequent grinding, the blades are excited to vibrate by the grinding disk, similar to a bow on a violin string. Depending on the grinding conditions, this process can act similar to vibrations during operation, and excite fundamental flexural modes (cracking in the transition radius to the root) as well as high-frequency vibrations (lyre mode; Volume 2, Ill. 7.1.3-4) with cracking in the edges near the tip. Damping measures such as rubber inserts on every blade have proven effective for preventing these problems.

“4” Machining: Intense, unusual whistling during the machining process can be a warning sign of dangerous vibration of the part (Ill. 17-16). This type of vibration can usually be classified as chatter (Fig. "‘Plucking’ damages by a machining process"). The stick-slip effect is also capable of causing high-frequency vibrations and suddenly overstressing the part. Intermittent machining forces from high-speed tools with multiple cutting edges, such as milling cutters, also act as high-frequency excitements.

“5” Cast part production: Although it is rare, dynamic fatigue cracks occur in cast blanks. In this case, the casting is not the problem, but rather subsequent treatments, including:
Removal of the casting mold: The typical use of vibration hammers with relatively large amounts of energy is suspected of exciting dangerous vibrations. This is also true if adhering residue from the cast mold should have a damping effect.
Separating work to remove the sprue and risers, and to release the cast part from the cluster. The separation is usually done using a high-speed disk with bound polymers. This disk can excite vibrations through a stick-slip effect (“3”). The intensity depends on the separating parameters (infeed force, cooling, sharpness, stresses in the part) and possible jamming of the disk (casting stresses). The loud noise produced by a cutting process can cover up the sound of vibrations of the part, so that the danger of damages is not realized.
Experience has shown that dynamic fatigue cracks in cast parts are difficult to detect. The typical intensive surface treatment with abrasive blasting smears the cracks (Fig. "Limiting influences non-destructive testing"). The rough surface of this type of treatment additionally complicates penetrant testing.

“6” Water jet cutting and other processes that use fluid jets: These are de-coating/stripping processes. Excitement can be caused by the pulsation of the jet, and therefore the supply pump. If a part vibrating in resonance suitably changes the force of the fluid jet, such as making the jet periodically strike in a different manner, it is possible that a self-increasing effect could occur.

“7” Material-removing ultrasonic machining: Especially when countersinking brittle materials such as ceramics with ultrasonic processes, there is a danger of damage occurring. The vibrating tool nozzle uses abrasive media (slurry) to import vibration energy into the part. In special cases, ceramic thermal barriers on metallic parts are machined using this process. In a system with such large differences in material properties (metallic base material, ceramic coating), significant damping can be expected. This makes it very unlikely that vibration excitement will occur and cause cracking in the base material. However, it is possible that the bond strength of the coating could be lowered in areas with high vibrational loads, and that the coating could separate in extreme cases.

Figure "Dynamic fatigue cracks by ultrasonoc cleaning": Ultrasonic cleaning with insufficiently verified process parameters can dangerously damage parts (Fig. "Damage at turbine blades by ultrasonoc cleaning"). In addition to dynamic fatigue and macro-cracking, fretting can occur with loose contact surfaces. For example, false brinelling with the same spacing as that between the rollers has been observed on the races of roller bearings (Ref. 16.2.2.5-2).

Figure "Damage at turbine blades by ultrasonoc cleaning" (Example 16.2.2.5-1,Ref. 16.2.2.5-8): This case is a perfect example of the dangers of improper process parameters. A seemingly harmless cleaning process, ultrasonic cleaning, makes parts vibrate at ultrasonic levels in a tank of fluid. Evidently, if the settings are not correct, this can cause potentially dangerous dynamic fatigue cracks.

Example 16.2.2.5-1 (Ref. 16.2.2.5-8, Fig. "Damage at turbine blades by ultrasonoc cleaning"):

Excerpt: “The FAA and…(the OEM) are investigating the miscalibration of an ultrasonic cleaning machine that induced potential fatigue cracks in nearly 8,200 turbine blades used in four different types of the company's engines….A company-issued alert affects 15 airlines, including two in the U.S.

The problem centers on cracking of the first- and second-stage high-pressure turbine blades installed on ….(four totally different engine types of different power classes) that power a variety of ….(different transports of two big manufacturers). Damage on the blades was caused…by a new machine installed … (some months before). The investigation by …(the OEM) and the FAA determined that only eight engines may have contained suspect blades…There have been no incidents associated with the potentially-flawed blades and no in-flight shutdowns have occurred. The affected blades already had accumulated time in service, and had been returned … for repair and reconditioning….
The ultrasonic process removes foreign material that accumulates within the hollow blades during normal engine operation, but `overaggressive' cleaning made the internal surfaces of the blades vulnerable to high cycle fatigue cracks
According to the alert letter, the `nature of the cracks makes them difficult to detect' and `traditionally nondestructive inspections can not detect all of the cracks'. As a result, all blades processed during the period are suspect. The cracks initiate from the internal trailing edge of the blade airfoil and propagate to a length of up to 0.500 in., with some extending through the wall to the external surface. Similar cracks in the trailing edges of other blades `have led to fractures in service'. Although no fractures are known to have occurred with any of the suspect blades…
The length of time the blades were exposed to the cleaning solution `was not a factor' in the incident…' The (cleaning) process was wrong because some of the blades had cracks', but not because personnel `threw the blades in the tank' and ignored them….a post-cleaning X-ray inspection revealed cracks in some of the blades. After determining the apparent cause, tests were conducted in the same tank using the improper process. The tests resulted in reproduction of the cracking problem on scrap hardware…“

Comments: Because blades from different turbine stages in different engine types are affected, it can be assumed that this is not a part-specific problem, e.g. due to unique design characteristics (thickness, geometry). Evidently, no new parts were affected, but only previously run parts that were ultrasonically cleaned during an overhaul. This process can also be used on new parts, which is why it is suitable as an example. The description does not clearly indicate whether the ultrasonic process caused the cracks or whether it merely caused rapid growth of operating cracks that were already present. It can be assumed, however, that overly aggressive ultrasonic cleaning could also cause dynamic fatigue cracks in uncracked parts. If similar cracks in other blades led to fractures during operation, the question is how they were caused. The record does not indicate whether the crack initiation zone was subjected to especially high operating loads. This would considerably increase the risk of operating damages on pre-cracked blades.

Figure "Verification of local damaging deformations" (Ref. 16.2.2.5-9): Local plastic deformations can incur serious losses in dynamic fatigue strength. The damaging effect is primarily based on a local loss of strength and/or stress increases. The following are influencing factors:

  • Overheating (friction heat)
  • Structural changes (embrittlement, loss of hardness)
  • Cracking
  • Notches, increased by higher notch sensitivity
  • Tensile residual stresses

Therefore, it is very important that this type of flaw is prevented. In addition to preventive measures, it is important that flaws are detected sufficiently early, and that their causes are understood in order to take specific measures (Ill. 17-11).
Examination of the available literature revealed that, at the moment, there are evidently only very few serially-implementable testing methods. These include visual inspection, penetrant testing, and micro-etching (e.g. BEA of titanium alloys). However, these do not ensure sufficiently reliable detection of these flaws, especially if they were consciously or unconsciously reworked (Fig. "Minimizing scrap rates throuch reworking"). For this reason, there are concerted efforts to improve existing processes (e.g. eddy current testing) and develop new ones (magnetometry, automated optical inspection).

Dangerous material deformations can primarily be attributed to two mechanisms in the finishing process:

Impressesions with and without embedded particles, caused by:

Detection of pronounced structural changes (top right diagram) or smeared foreign material with the aid of etching processes can be considered to be non-destructive. This can be aided by polishing, as is typical in metallography (also see Fig. "Non destructive microscopic inspection").
If notches or particle residue are still present, an SEM (Fig. "Scanning electron microscopy (SEM)") can be used to examine typical surface structures, and perhaps an analysis of the particle composition, can be used to determine their cause (e.g. wear products from contact surfaces, impressed particles, bottom left diagram).

Sliding contact with galling (cold welding, Fig. "Friction contact during clamping"): Typical causes are

Galled areas have typical structures (top left diagram). These can be identified in an SEM with the aid of a surface imprint (Fig. "Non destructive microscopic inspection") or a direct inspection of the part (usually a size problem).
Here, as well, etching processes (bottom right diagram) can be used for verification if there are sufficient changes to the surface in the damaged area.

Figure "Cracking of brittle coatings by deformation": For parts with brittle coatings, there is a danger of crinkle-lacquer type cracking occurring even in the case of very small plastic deformations. Typical examples include turbine blades with a diffusion coating that behaves brittly at room temperature (Fig. "Oxidation protecting Al diffusion coating problems").

These deformations can occur in certain zones if the part is clamped in an unfavorable way (Ill. 17-13). A typical occurrence is cracking around dimensional notches and stiffness jumps. One of these zones is the transition of a turbine guide vane to the shroud. This cracking is usually crack fields with concentric or parallel fine cracks that are very difficult to recognize visually. The special problem with these very fine cracks is that they are under compressive stress from the springing-back (Fig. "Straightening difficulties"). This presses the crack edges together and makes penetrant testing difficult. It is entirely possible that the cracks will only be revealed by oxidation during operation (Volume 3, Ill. 12.6.2-15). However, by this point it is almost impossible to determine whether the cracks occurred during operation, or already during finishing, handling, or assembly. One indicator can be the location of the crack field and/or the crack orientation. If these are unusual for operating cracks, but can be easily explained by a fastening process, it indicates that it is a new part problem.

Figure "Straighrening bent parts": Problems such as careless handling or unforeseen effects of a tool (such as a programming mistake with extreme infeed thrust) can cause plastic bending of the parts. Due to delivery agreements and high part costs (e.g. blisks cost as much as several new cars), this can mean a major loss. Therefore, there is always a consideration of possible straightening as a form of reworking. This must be systematically planned and executed (see frame on page 16.2.2.5-25 and Chapter 17.5).
Plastic deformations in a cold state, as are typical for damages, influence the material and part properties in many different ways:

  • Material· Hardening, i.e. increase of the flow limit and therefore the elastic limit ratio (flow limit/fracture strength). Increase in hardness. Increase in dynamic fatigue strength.
  • Embrittlement: reduction of the fracture strain.
  • (Micro-) cracking in brittle coatings such as oxide coatings on Ti and Al alloys. This type of cracking can reduce the dynamic fatigue strength considerably.
  • Breakdown of the desired effect of a surface hardening treatment such as shot peening.
  • Residual stresses are induced by springing-back (Fig. "Straightening difficulties"). Tensile stresses created near the surface reduce the usable dynamic fatigue strength.
  • Changes to the dynamic part behavior: The changes in shape can occur with local stiffening and shifting of mass in the part. This influences the resonant frequencies and vibration forms (nodal lines).
  • Influences on aerodynamic properties: Form changes in aerodynamically effective surfaces such as blades can cause performance losses and even rotating stalls, leading to increased dynamic stress.Possible changes to specified part properties make straightening necessary. In this case, the material is changed relative to the original finishing state. Its state after cold straightening is not exactly definable. For this reason, the allowability of a straightening process must be assessed by the responsible technical departments, and perhaps also the OEM, authorities, and customers.

Figure "Straightening difficulties": Often, cold straightening or straightening in the range of normal operating temperatures are the only possible options. Otherwise, there is a risk that the properties of the material, strictly tolerated mass, or operating behavior will be unallowably influenced.

Before conducting cold straightening, one should remember the difficulty of straightening a paperclip with one`s fingers (top diagram). The paperclip can only be easily bent outside of the bends that are already present. After some effort, the best possible result is a slightly wavy line. This behavior can be attributed to the hardening of metals when they are plastically deformed (top right diagram). During cold forming, regardless of whether it is compression or expansion (Fig. "Increasing fatigue strength by hardening"), the flow limit (elastic limit) increases to the stress that corresponds to the maximum loads in the true stress-strain curve (without considering constriction; top right diagram). During the next forming attempt, plastic deformation only occurs above this increased flow limit. At the same time, the flow limit approaches the fracture strength. This means that the material has less toughness after every new plastic deformation. In other words, the material becomes more brittle, and the danger of cracking during straightening increases (Fig. "Cold straightening of titanium parts").

Straightening as a permanent bending utilizes local plastic deformation. During the bending process, tensile stresses act on one side of the cross-section, while compressive stresses act on the other side. Upon relief, the affected area springs back somewhat. This creates compressive stresses in the plastically deformed area that previously had tensile stress. Accordingly, tensile stresses are induced in the formerly compressed side (bottom left diagram). The spring-back is especially pronounced in titanium alloys (Ill. 16.2.2.5-14, Ref. 16.2.2.5-11).

Negative effects, especially the spring-back, can be minimized by straightening at suitable temperatures. The prerequisite for this is that the part (mass, local heating) and the material (structure, strength properties, embrittlement in air) allow this. The proper technical departments need to be consulted in this regard. In some cases, the plastic deformability can be improved (toughness, necessary force) and risks of damage (cracking, tensile residual stresses) minimized.
During deformation (usually bending), the part changes considerably (strength, residual stresses, toughness), which must be considered during straightening.

  • Hardening usually increases dynamic fatigue strength, which is a positive change. However, hardening makes plastic deformation more difficult and reduces toughness (danger of cracking).
  • After bending, there is a complex residual stress state (bottom left diagram, Ref. 16.2.2.5-11). It does not correspond to the similar external deformation under the influence of the externally acting bending force.Although the part geometry corresponds to the design drawing requirements after straightening, unlike an uninfluenced new part, it will have high tensile residual stresses. During straightening, the stress patterns will have changed relative to the bent conditions (bottom right diagram). Tensile residual stresses mean lower dynamic fatigue strength in the part. They must then be broken down in accordance with the assessment made by the responsible experts, if possible with the aid of suitable methods such as shot peening or heat treatment.

Figure "Cold straightening of titanium parts": Special problems during straightening are spring-back (Fig. "Straightening difficulties") and cracking and fractures of the part in the deformation zone. This fracture risk is not necessarily very high in materials with pronounced hardening properties, such as steels and nickel alloys. Of course, the risk will be greater if the strength of the material has already been attained by strain hardening. Cold-drawn aluminum alloys are an example of this.

Titanium alloys which only exhibit minor strain hardening have pronounced spring-back (Ref. 16.2.2.5-11) and are especially at risk of fracturing during straightening. The spring-back is related to their high basic strength and low modulus of elasticity. Plastic deformations and the accompanying induced residual stresses lead to correspondingly large elastic strain (spring paths).

In a low-hardening material, the plastic deformations should be limited to a relatively narrow area. This can be expected in form notches, such as the transition radius of a blade into the root platform. In contrast, a high-hardening material will encourage the plastic deformation of surrounding areas that have not yet been hardenend.

In the depicted case involving a machined prototype part made from a high-strength titanium alloy (left diagram), one blade fractured and several others were deformed. It was suspected that the primary cause for the fracture of the blade was an unallowable material brittleness. However, a laboratory analysis (SEM, Fig. "Scanning electron microscopy (SEM)") revealed that it was a micro-ductile forced fracture typical of the material. The orientation of the dimples and the shape of the shear surfaces in the surface
area (right diagram) indicated other factors. The forced fracture evidently occurred under a bending force that was in the opposite direction to the forces responsible for the bending of the surrounding blades. An investigation revealed that the fracture only occurred during an attempt at cold straightening. This was done in order to correct deformations that occurred when the blade was ripped out of the lathe. This example shows the problems with insufficiently prepared attempts at straightening.

With heating and combined bending-stretching processes at low deformation speeds, the chances of success of a straightening process increase. It is important that the deformation is distributed as evenly as possible in order to prevent local overstress and cracking. However, it must be ensured that the temperature does not unallowably change the material. This applies to the structure, strength behavior, and creation of brittle coatings in titanium alloys. For titanium alloys, the temperature range for straightening is between 540°C and 815°C, depending on the alloy (see specifics in Ref. 16.2.2.5-2). The straightening process requires a sufficient protective gas film.
Information regarding risks in the operating behavior of straightened parts may be found in the relevant repair manual. This can also be useful for verifications regarding new parts.

Before reworking is done, it is necessary that the responsible specialists conduct the following steps:

  • Determine the type and cause of damage
  • Estimate potential damage (type, depth)
  • Develop and specify a rework process
  • Acceptance of the rework (design construction, strength, licenser, authorities, etc.)
  • Documented execution of the rework
  • Answering important questions:
  • Have dynamic fatigue fractures during operation been reported in the part area in question?
  • Does the maintenance handbook contain any information?
  • Is the straightened zone in an area that is potentially subject to high dynamic stress?
  • What are the vibration forms of the part?
  • If possible, technical tests to verify the safety of the rework.
References

16.2.2.5-1 D.A.Dornfeld, J.S.Kim, H.Dechow,J.Hewson, L.J.Chen, “Drilling Burr Formation in Titanium Alloy, Ti-6Al-4V”, (CIRP Annals 1999, Manufacturing Technology), Volume 48/1/1999, pages 73-76.

16.2.2.5-2 “Metals Handbook Ninth Edition, Volume 11, Failure Analysis and Prevention”, ISBN 0-87170-007-7, American Society for Metals, 1986, page 500.

16.2.2.5-3 “Metals Handbook Ninth Edition, Volume 16, Machining”, ISBN 0-87170-007-7, American Society for Metals, Issue 4, 1999, pages 1-48, pages 838-843.

16.2.2.5-4FAA warns Boeing 717 operators of BR 700 engine component failure”, periodical “Flight International”, December 18, 2000.

16.2.2.5-5 NTSB, “Uncontained Engine Failure, Delta Air Lines Flight 1288, McDonnell Douglas MD-88, N927DA, Pensacola, Florida, July 6,1996”, NTSB/AAR-98/01, PB88-910401, DC96MA068. page 31.

16.2.2.5-6 W.D.Feist, F.Niklasson, K.M.Fox, “The Influence of Manufacturing Anomalies on Fatigue Performance of Critical Rotating Parts in the Aero-engine”, paper from the EC-Program Nr. G4RD-CT2000-00400 5 GROWTH “MANHIRP-Integration of Process Controls with Manufacturing to produce High Integrity Rotating Parts for Modern Gas Turbines”, pages 1-7.

16.2.2.5-7 B.Denkena, “Versuch: Ratterschwingungen an Werkzeugmaschinen”, Universität Hannover, Institut für Fertigungstechnik und Werkzeugmaschinen (IFW), May 2003, pages 1-17.

16.2.2.5-8 E.H.Phillips, “Pratt & Whitney Recalls Defective Turbine Blades”, periodical “Aviation Week & Space Technology”, April 13, 1998, page 63.

16.2.2.5-9 W.D.Feist, G.Mook, J. Hinken, “Zerstörungsfreie Charakterisierung fertigungsinduzierter Anomalien in Flugtriebwerksrotoren”,Paper from the GROWTH-Projekts G4RD-CT2000-00400 MANHIRP of the European Union.

16.2.2.5-10 G.Mook, J.Simonin, W.D.Feist, H. Wrobel, J. Hinken, “Wolframkarbide in Titanlegierungen-Herkunft und Nachweis”,Paper from the GROWTH-Projekts G4RD-CT2000-00400 MANHIRP of the European Union.

16.2.2.5-11 “ASM Handbook, Volume 14, Forming and Forging”, ISBN 0-87170-007-7, American Society for Metals, 1998, pages 841 and 882.

16.2.2.5-12 P.Adam, “Fertigungsverfahren von Turboflugtriebwerken”,Birkhäuser Verlag, 1998, ISBN 3-7643-5971-4, page 74.

© 2024 ITTM & Axel Rossmann
16/162/1622/16225/16225.txt · Last modified: 2022/05/16 19:10 by Valentin Mostetschnig