Table of Contents
16.1 Factors that Promote Finishing Problems and Damages
Finishing processes can be affected in a damaging manner by influences and decisions that occur before and after finishing (Fig. "Detecting weak points from finishing"). Increased operating loads demand higher-strength materials. The resulting more difficult machinability of modern engine components seems to increase the damage rates in rotors considerably (Fig. "Tracing back finishing cracks in rotors")
Design construction is of considerable importance. It must be done in accordance with material specifications (such as design lines) and regulations. The use of ever harder materials for parts under correspondingly greater loads presents tremendous demands on the base materials and undamaged, probably even specifically treated surfaces. These treatments include hardening and residual stresses from chip-removing machining. The condition of an undamaged surface is referred to as surface integrity (Chapter 16.2.1). The task of ensuring surface integrity makes chip-removing machining more difficult and is often counteracted by the increased sensitivity of high-strength materials (to comma cracks, etc.; Ill. 126.96.36.199-2). For example, reaming (Ills.188.8.131.52-6 and 184.108.40.206-2) blade slots in disks with a high material strength, which are usually under high dynamic loads (HCF, LCF), becomes considerably more difficult. This means that the decision between a cast or forged part version will have serious consequences for finishing. Even conducting surface treatments of cast parts will remove the casting skin, which may have had significant favorable properties. If cavities from the casting process are opened, they may pose a problem with regard to their influence on strength properties, and their acceptability may need to be reviewed. Subsequent quality improvement through cavity removal in an HIP process (Fig. "Properties of cast parts influenced by HIP") is not possible if cavities have been opened. If the finishing process includes grinding and/or welding, the tendency to cracking will depend largely on the structure; i.e. properties such as grain size, grain orientation, and the hardening state (Fig. "Damage potential of a coating").
Many parts consist of material combinations. They may be related materials in cast and forged versions that are then connected together. However, very different material combinations may be found in parts. This is the case in coated parts. Non-metallic materials can broaden the spectrum considerably. Metallic materials are frequently combined with ceramic or synthetic coatings. This type of combination complicates the coating process (e.g. temperature control in the case of thermal spray coatings), heat treatments, etching, and cleaning, to name but a few.
The drive to lower costs leads to a minimization of part numbers (Fig. "HC Hub ratio as design characteristic"). A lower number of compressor stages requires wider blades with higher rpm and, therefore, greater loads. This means that, for example, a typically highly-stressed zone such as the contact surface of the blade root will become a special problem zone. These zones demand the highest possible finishing precision (Fig. "Limits of demads on blade roots") with additional work such as special shaping of the contact surface (3D curved) and strictly specified coatings. This increases the demands on the finishing techniques and also the probability of potential finishing problems.
New design principles, such as the transition from inserted blades to blisks, demand new or adapted finishing processes (e.g. linear friction welding). This often results in new requirements for the finishing process with new typical weak points. For example, bonding surfaces in diffusion, laser, EB, and friction welding are almost impossible to safely detect with serially implementable non-destructive testing methods. The trend towards integral designs such as blisks and stators is tending to increase the finishing risks. If these parts are assembled from many individual parts through welding or soldering, the probability of single unallowable (at best, they can be subsequently machined) flaws is very high. The low natural damping of these parts promotes higher dynamic loads during operation. This increases requirements with regard to surface quality. Through the effect of “mistuning” (Fig. "Safe operating due to dimensionalaccuracy"), tiny scattered dimensional/geometric deviations (e.g. in blades), often within the usual tolerances, can increase the dynamic loads on single part zones to levels several times greater than the average in the surrounding parts. This further increases the relative importance of surface integrity in these parts.
Material combinations that are heat-treated in their combined state may be problematic. A typical phenomenon is the properties of one material worsening (e.g. strength losses through solution annealing or reduced ductility) while those of the other improve. High-temperature soldering or diffusion welding may require temperatures in the range of the solidus. Even minor deviations in the alloy composition and/or temperature can lead to damages. For example, slightly excessive temperatures may cause melting, while low temperatures may prevent optimal high-temperature soldering from being successful (e.g. bonding flaws, embrittlement).
If residual stresses are induced in raw parts, they can complicate the subsequent finishing process. In this way, material removal may lead to warping and etching can cause cracking.
Unfavorable designs may make it difficult to clamp parts in order to machine them. This indirectly prescribes clamping methods such as “hard clamping” (in a clamping device) and “soft clamping” (through pouring), as well as their respective specific risks.
Although process engineering is part of the finishing process, its influence on finishing problems, damages, and quality is often not appreciated as highly as it should be. Process engineering is based primarily on experiences from serial finishing and finishing development. In this case, specialized experience of the technical personnel is of decisive importance for risk minimization. For example, knowledge of particularities and the specific behavior of the finishing facilities is a prerequisite for the highest quality demanded by engine finishing. Examples include the stiffness of machining equipment, temperature distribution and the temporal temperature changes in the available ovens, and design characteristics of tools used in electro-chemical machining. The realization of suitable equipment must minimize influences such as removed material, fouling of the part (Fig. "Damaging by foreign materials on parts"), as well as deformations (Fig. "Fastening of parts during finishing"). In the case of chipping processes, chip transport (Fig. "Dynamic fatigue strength influenced by fused chips") and the necessary coolant flow must be guaranteed. In coating processes, dust deposits that may reduce bond strength must be prevented.
The especially problem-relevant influences include work sequences. Suitable chronological order of the finishing steps minimizes the probability of damage. This includes the order of solderings, diffusion coatings, and heat treatments. The arrangement and methods of quality controls in the work sequence has a large influence on damage risks and finishing costs. Compromises are unavoidable here, as well. A thorough inspection after many finishing steps may minimize the inspection work and accelerate the production process. However, if flaws occurred in an early finishing step, they will only be detected after a considerable investment has already been made. In addition to the high costs, this situation is also likely to result in many more affected parts in the finishing process ahead of the inspection step.
Figure "Tracing back finishing cracks in rotors" (Ref. 16.1-5): According to FAA data from 1999, in the years 1990-1999 25% of cracks and fractures and rotors could be traced back to anomalies in the finishing process. In contrast, only 9% were due to material production and 5% were related to the forging process. It is surprising that there is a clear percentual increase in finishing flaws since 1970. This could well be due to the higher-strength materials with correspondingly higher operating loads. Higher operating loads cause cracks at smaller flaws to be capable of growth. At the same time, the high-strength materials are more difficult to machine, increasing the damage risk during the finishing process.
Where should a finishing test be placed, both organizationally and physically?
The testing of serial production and the process engineering rely on known previously implemented production processes and/or new production technologies from production development. They serve especially to minimize the production work/costs and the scrap rate of the serial parts. In large batches, for example, these concern changes in dimensional accuracy, wear, and replacement of tools and optimized equipment.
Production development determines/minimizes later problems in the series to a significant degree. This process develops new production technologies and adapts existing ones for new parts. This can occur simultaneously during the production of parts in the framework of development and prototype production. A task of production development that is often underestimated is the identification of process-specific problems and flaws, as well as their effect on the part behavior. Understanding the processes by which they originate is a prerequisite for targeted solutions and the verification of serially-implementable, sufficiently safe non-destructive tests (Fig. "Tracing back finishing cracks in rotors"). The first question to be answered is where the production testing should take place. Is a separate independent development better than one that is integrated into serial production? Both possibilities have advantages and disadvantages:
Independent production development enables rapid action to be taken, free of series priorities. It can use more flexibly usable equipment with parameters that can be adjusted across a wide range. They do not have to be optimized for series-typical demands such as low production costs and large production volumes, allowing more flexible development approaches. The personnel may be selected with an eye to the demands of development. However, experience has shown that a considerable problem may occur when transferring the developed production technology to the series. It is entirely possible that developments were conducted on equipment that does not sufficiently correspond to that used in serial production. For example, the chamber size of an electron beam welder may be a factor. It may influence the vacuum that can be reached in an acceptable time. It may also affect the cleanliness of the parts, since the quality of the weld seams is influenced by the quality of the vacuum. If the electron beam guns are different, variations in beam energy and focusing properties can influence the welding rate and therefore affect the creation of flaws (Fig. "Electron beam welding (EB) process"). In this case, the transition to serial production equipment and processes is unavoidable. This relativizes the applicability of steps of the process engineering and process sequences. Different personnel that is not as familiar with the developed procedural steps will increase the potential for problems. This makes a certain additional testing of the production process unavoidable. Experience has shown that this can be exacerbated by the almost certain desire engineers have for independent solutions, or at least improvements and optimizations. It is rare that the developed process will be adopted without corrections in this context. However, this means that there is a considerable risk of disimprovement occurring.
Production development in series has the advantage that process engineering, production personnel, equipment, and procedures do not require drastic transitions with “transfer losses”. Experience has shown, however, that problems arise with regard to decisions regarding priorities betwen the series and development. Equipment suitable for series use will have a less flexible range of process parameters. For personnel in serial production, the strict observance of prescribed parameters and procedures is a requirement. However, development work requires changes. Replacing this personnel is therefore made more difficult by the respective necessary attitudes, which cannot be combined when attempting to attain optimal results.
Figure "Detecting weak points from finishing": Finishing flaws are different from weak points, which are by definition accounted for in the design. The size of the weak points and the corresponding tolerable part loads depend on the sensitivity of serially-implementable non-destructive testing methods (Fig. "Probability of detection of non destructive tests"). If a new technology or process development may result in new flaw types or cause known flaw types to change (location, etc.), suitable testing processes must be developed within the framework of the production development process. One example is the typical macro-cracking found in some nickel alloys (Ills.220.127.116.11-14 and 18.104.22.168-25) around electron beam welds (Ref. 16.1-3). For some procedures, there is no known testing method that is sufficiently sensitive and safe as required for the utilization of high strength. These procedures include contact points in shear-stressed diffusion welds (Fig. "Diffusion weld of a ‘dual property’ part").
A rather “second-class” alternative to non-destructive testing methods is ensuring flawlessness through process monitoring alone, despite the fundamental rule that it is easier to avoid flaws than to find them through later testing.
Figure "Problems before the finishing process": As is the case with all damages, finishing damages are usually affected by several different causal factors, including some that can not necessarily be attributed to the finishing process itself. One example of this is a design that does not permit sufficiently safe finishing due to poor accessibility, poor testibility, and/or limited reworking potential (Fig. "Failed detection of large cracks"). Further examples include welding problems that are caused by material conditions that are less than optimal, but that cannot be changed for the finishing process.
Specifications and regulations that cannot be satisfied with the desired low scrap rate using the available methods can also be partially responsible for any damages.
The work environment in which the finishing process takes place can also have a considerable influence on the development and resolution of damages. For example, experience has shown that forcing lower scrap rates through personnel-related repressive measures will be less effective in the long term than an objective approach based on the problems and issues (Fig. "Negative motivation"). In the context of “peer pressure”, doing things such as distributing charts showing the temporal increase in scrap rates can be problematic.
Along with increasing operating demands, such as aerodynamic properties (dimensional tolerances of the blade profiles) or strength (e.g. rotor components), the tolerances for deviations become ever tighter. This also applies to any necessary reworking (Chapter 17.5).
16.1-1.1 “ASM Handbook” Volume 4 (“Heat treating”)
16.1-1.2 “ASM Handbook” Volume 5 (“Surface Cleaning, Finishing, and Coating”)
16.1-1.3 “ASM Handbook” Volume 6 (“Welding, Brazing, and Soldering”)
16.1-1.4 “ASM Handbook” Volume 7 (“Powder Metallurgy”)
16.1-1.5 “ASM Handbook” Volume 14 (Forming and Forging“)
16.1-1.6 “ASM Handbook” Volume 15 (“Casting”)
16.1-1.7 “ASM Handbook” Volume 16 (“Machining”)
16.1-2 Peter Adam, “Fertigungsverfahren von Turbotriebwerken” Birkhäuser Verlag,, 1998, ISBN 3-7643-5971-4.
16.1-3 M. Field, J.F. Kahles, “Übersicht über die Oberflächenbeschaffenheit bearbeiteter Werkstücke `Surface Integrity'”, periodical “Fertigung” 5,72, pages 145-156.
16.1-4 A. Schäffler, presentation at the TU München
16.1-5 AIAA Rotor Manufacturing Subcommittee, “Management Summary”, March 29, 1999.