Table of Contents

16.2.2.1 Surface Topography, Roughness

The three-dimensional surface structure is referred to as topography. In this chapter, the topography includes characteristics with height differences of tenths of a millimeter. In a cross-section, i.e. two-dimensional view, the measured data refer to the roughness (Fig. "Roughness data tell not enough"). In this case, the technically relevant height differences (roughness depth) are usually less than 0.1 mm. Finely worked surface roughnesses are less than 0.001 mm. Many of the problems dealt with in the following text are therefore included in the concept of topography. The term texture is used when elementary structures, similar to roofing tiles, repeat themselves on a larger surface. The surface topography has a strong determining effect on the operating characteristics of the parts, which can sometimes be unexpected and surprising (Fig. "Topography and roughness influencing loads"). This extends beyond just the notch effect of grooves and their influence on dynamic fatigue strength (Ills. 16.2.2.1-3 and 16.2.2.1-4).
The dynamic fatigue strength of a surface depends on any hardening and residual stresses in the groove with the greatest notch effect (deepest sharp groove) that runs perpendicular to the stress direction (Fig. "Estimating fatigue strength by fracture mechanics" ).
The aerodynamic quality of a surface is also influenced by the orientation of the roughness, in this case perpendicular to the flow. However, in this case the highest peaks are the decisive factors (Fig. "Differences between turning grooves" ). In both cases, the roughness depth R (sometimes also labeled k or Rk) is most important, not the average roughness depth Ra or mean roughness depth Rz.
However, with regard to the carrying capacity of contact surfaces, the topography corresponding to the carrying length of the tips may be most important.
In many cases, the orientation of the roughness has a considerable influence on important characteristics of the surface. These include sealing effects under relative movements (e.g. floating ring seals), solder flow in gaps (Fig. "Bond face of braze joints"), and the tribological behavior of abradable and rubbing systems (Volume 2, Ills. 7.1.4-15 and 7.2.3-13).

Figure "Roughness data tell not enough" (Ref. 16.2.2.1-5): Roughness is an important characteristic of surfaces in engineering and is described using various defined values (Ref. 16.2.2.1-6).
The measurement results depend decisively on the type and parameters of roughness measurement. Basically, the measured roughness pattern, which depends on the limited length of the measuring sensor or measuring beam wavelength, can never exactly reflect the actual roughness pattern. This is especially true if the only roughness is sharp, crack-like grooves (bottom right diagram).

The measurement will provide a standard for comparison, such as for quality assurance purposes. However, unique characteristics of the roughness pattern, and their influence on physical and chemical processes, should be viewed critically. Classic roughness measurements are two-
dimensional. They determine the roughness in a line. Depending on the orientation of this line and its location on the part, different measurement values may result. The design engineer must consider this in his drawing specifications. He must specify the part zone in which the measurement must be done, as well as the measurement direction and also the measuring method. This is important because there are many different measurement systems with different sensitivity and relevance to the specific case (Ref. 16.2.2.1-5).

The roughness depth “R” (in the past, also known as “K”) corresponds to twice the “Rm” and only describes the distance between the highest peak and the deepest valley on the defined measurement line (hill curves of the profile). Therefore, it can indicate the maximum expected groove depth. As mentioned above, depending on the size of the sensor tip, crack-like notches are not included in the measurement. Therefore, R does not explain the shape of the roughness profile, and has no relevance to influences that react to the profile shape. In the field of finishing, R is above 0.025 mm in rough-machined surfaces, below 0.01 mm in ground and finely worked surfaces, and well below 0.001 mm in very finely ground, lapped, or polished surfaces.
Roughness requirements of the type required in drawing specifications, must be selected by the design engineer on the basis of their relevance to the desired operating behavior and their ability to ensure it (Ills. 16.2.2.1-2 and 16.2.2.1-3).

In order to better define the roughness profile, the average roughness value “Ra” was introduced. It defines the distance between the base of the profile (inner envelope) and the line that averages the surface proportions of the profile. In order to make the measurement more representative, 5 Ra measurements are combined in a mean value “Rz”. “Ra” represents properties of the surface such as carrying capacity, wear, and reflection more accurately than R does (bottom diagram).
All of these roughness values fail to account for very narrow, crack-like notches which cannot be registered by the sensor or sensing beam. Therefore, this type of roughness does not permit conclusions to be drawn regarding dynamic fatigue strength (Fig. "Roughness influencing dynamic fatigue"), and thus the degree of danger it represents (Fig. "Differences between turning grooves").

Figure "Topography and roughness influencing loads": The surface of a part is its connection with the environment. It is subjected to many different influences. The topography of the surface determines how these influences affect operating behavior.
In highly-stressed parts in modern engines, this occurs in two ways: by determining life span and/or functions. For this reason, it may be necessary to go beyond the usual demands that have been placed on load-specific topographies up to now. This increases the importance of surface geometry.
Dynamic fatigue strength is affected by notches and fretting (Fig. "Fretting damage loweing by shot peening", Volume 2, Ill. 6.1-19), such as when grooves act as weak points and/or micro-movements occur.
On the other hand, surface topography can also have a decisive influence on the acting loads.
For example, if the roughness of a blade is the cause of a rotating stall, this means that it will increase the loads on the blading. In extreme cases, it may result in a surge process with at least temporary loss of engine power.
The following questions are critical: When are the usual roughness requirements in drawing specifications no longer sufficient? Which effects in which part zones must be taken into consideration? When must “new” measuring methods be used that can depict surface topography in its 3-D structure and meaningfully analyze it? High-resolution, contact-free sensors (beams, fields) for surface evaluation, combined with computer-aided analysis of the findings, can potentially capture and specify surface structures.
Two roughness characteristics are especially important: orientation and shape.

The orientation of the topography/roughness is understood to be the primary direction of topographical features. Usually, these are grooves from chip-removing machining processes such as turning, milling, clearing, and grinding. The causes of oriented roughness can also be the flow direction of a machining fluid (e.g. ECM or abrasive flow machining) or the direction of an angled abrasive particle jet.
If currents occur in baths (circulating, thermal) during chemical processes and carry bubbles (reaction, steam, air, Ref. 16.2.2.1-2), they can create corresponding tracks.
Even without directional material removal, a primary direction can be created in the topography. One example of this is chemical material removal (etching), when grain boundaries and/or structural components (e.g. carbides) of a directionally solidified structure (e.g. grain direction or single crystal) are revealed (Ref. 16.2.2.1-2).

Shape of topography/roughness: The three-dimensional shape of undirected roughness is also important for the operating behavior of parts. It is possible that in some cases, special surface structures are being unconsciously utilized. The typical calottes of shot-peened surfaces (Ref. 16.2.2-10, Fig. "Fretting damage loweing by shot peening") are one example of this. They are dependent on peening parameters such as shot size, peening angle, and coverage. Usually, the parameters are selected according to the desired peening intensity. This also influences residual stresses and hardening in the surface. The geometric function of the surface, such as its micro-contact pattern and reservoir effect are secondary factors, at best. Depending on the measuring system, if non-directional roughness with dangerous notch effects, such as grain boundary corrosion, pitting corrosion, sharp imprints from abrasive shot, or micro-burrs (Ref. 16.2.2-10) do not occur in roughness documentation, it indicates the limited relevance of these measurements to part behavior.

The following text discusses typical examples of the influence of topography on part loads: Aerodynamic behavior: The resistance of a surface in a flow, typically a blade, depends highly on its roughness (Fig. "Aerodynamic demands finishing process"). Grooves in the direction of flow have less of an effect than those that are at an angle of more than +/- 10° to the flow (Fig. "Aerodynamic demands finishing process").

Excessively high roughness peaks on the blading influence the efficiency and operating behavior (surge limit) of the compressor, and therefore the entire engine. If roughness is a causal factor for a stall, then it will endanger the operation of the engine, as well as the integrity of the individual parts.

The roughness of new parts can promote the adhering of deposits (fouling) and thus have an indirect effect. Similar indirect effects of new part roughness include the seal effectiveness at the tip clearance gap or erosion behavior.
In cooled hot parts such as turbine blades, the roughness influences the heat transition and the flow resistance on the hot gas side and cooling gas side. It is desirable to minimize the heat flow from the hot gas into the part. This requires the smoothest possible surface. On the other hand, the heat transition on the cooling air side should be as intensive as possible. On this side, it may be beneficial to have directional roughness that is suitably oriented to the cooling air flow (Ref. 16.2.2.1-2). This type of beneficial roughness can be created by structural effects (e.g. ECM boring in single crystals) and process characteristics (gas bubbles from chemical reactions, Fig. "Boring methods for cooling air"). If the process parameters change so that the unconsciously required rough surface does not form, and this is not recognized, it can significantly shorten the part life. This example shows the risks of changing processes.

Behavior of contact surfaces: This primarily concerns the contact surfaces of blade roots, flange surfaces, fitting surfaces, and socket connections. The contact surfaces of blades and disks made from titanium alloys are especially affected by the high surface pressures and static and dynamic loads, as well as damage through micro-movements (fretting). This can cause the dynamic fatigue strength of untreated surfaces to fall by up to about 70%. An additional important factor for the critical stress in the contact zone is the combination of bending stress and shear stress (Volume 2, Ill. 6.1-11) which is dependent on the coefficient of friction. For this reason, for this type of contact surface, a low coefficient of friction is desirable.

Nowadays, contact surfaces under fretting stress are shot peened. In addition to the hardening effect and protective compressive stresses, there are advantages to the non-directional topography. A low coefficient of friction over long operating times is ensured by embedded lubricants (e.g. graphite lacquer) and the absorption of wear products (top diagram). The relaxing effect of the micro-contact surfaces on the roughness peaks should also have a positive effect on the dynamic fatigue strength.
Directional topography on the contact surfaces, such as lateral ridges (bottom diagram), could have a beneficial effect by separating fretting-damaged zones from the highly-stressed notch base. Naturally, the notch effect of the ridges must not exceed the positive effect. There have not been any reports of specific research into the application of this type of treatment to blade roots.
Roughness also influences the seal effect between the contact surfaces. This can be important in flanges that seal off high gas pressure differences (e.g. housing flanges or flanges in the oil and fuel systems). Radially oriented roughness that does not seal properly can cause leakages.
An additional problem can result from the machining grooves in bores, centering collars, and fitting diameters, which are usually oriented along the circumference. If the resistance is too great when they are pressed together (Fig. "‘Plucking’ damages by a machining process"), and this is not noticed, it can cause assembly flaws. Possible consequences include imbalances and vibrations. In extreme cases, cold welding (galling, Fig. "Verification of local damaging deformations") may occur, which will cause a dangerous loss of dynamic fatigue strength.

Dynamic fatigue: Fundamentally, notches in the shape of grooves and non-directional damages (e.g. corrosion, impressions) are considered to reduce dynamic fatigue strength (Fig. "Roughness influencing dynamic fatigue"). Therefore, production processes that create grooves (machining processes such as grinding, sanding, turning) should be optimized accordingly (Fig. "Influence of surface topography and finishing"). Of course, these influences are overlayed by other factors, such as hardening and residual stresses, which often mitigates their effect. Unfortunately, it is not possible to rely on these rather incidental effects. The indirect increasing of dynamic stress on a part due to unfavorable surface topography has already been discussed in the context of aerodynamic behavior and friction behavior of contact surfaces.

Sliding and rubbing surfaces: These are usually sealing surfaces. In tribological systems, such as the tip gaps of blades or labyrinths, the roughness of the new part is only significant for as long as it takes for rubbing to change the surface. In order to optimize the rubbing behavior of coated surfaces, it is possible to create topographies that result in improved chip removal from the opposing surface and/or lower friction forces and heat development (Ref. 16.2.2.1-7). In any case, the chip-removing surface should have sharp cutting edges (Fig. "Examples of high temperature brazings") and sufficient volume for accepting the wear products/chips.
Function-specific roughness requirements exist for sliding surfaces such as those found in variable vane actuators. In this case, play-free, smoothly moving control and effective sealing under long-term environmental influences are required.

Floating ring seals should maintain their production-specific roughness over longer operating times at the low expected wear rates. In seal rings made from elastomers, the profiling of the seal surfaces is used to minimize the leakage flow. Even a conveying effect against the pressure gradients can be made use of. On the other hand, unfavorably oriented grooves on the shaft increase the wear from the seal lip and therefore also the leakage flow. The roughness of the contact surfaces also influences the friction forces, friction heat, and wear rate, and therefore also the dry-running operating properties and life span.

Erosion behavior: Gas and fluid flows that carry particles or gas bubbles have an erosive effect. Directionally structured surfaces can have various erosion sensitivities, depending on the direction of flow and roughness. In general, smooth surfaces are less sensitive.

Radiation: The behavior of a part surface under radiation (reflection, absorption), is dependent on the type and size of non-directional roughness, as can be seen in the diagram. In addition, the angle at which the radiation strikes the surface is important. This effect is increased by directional structuring.
The roughness of surface exposed to radiation is important if the heating of the part occurs primarily through heat radiation. Typical examples include combustion chamber inner walls and turbine inlet stator vanes exposed to radiation from the flame.

Reactivity: Chemical processes such as corrosion and oxidation are determined by the reactivity of the new part surface. In addition to the “activated” metallic surface, its size is an important factor. Special structures such as a large number of fine ridges and peaks (e.g. micro-burrs on ground surfaces, Ref. 16.2.2.1-10) and/or the structure of etched surfaces increase the size of the effective surface considerably relative to the size of the macro-geometric surface. Corrosive media and fouling can more easily adhere to dissolved topographies and are more difficult to remove through cleaning. This affects the fouling behavior of surfaces in a flow (oil, air).

Adhesive strength (Fig. "Finishing influencing thermal spray coating") of coatings and lacquers is naturally connected to the topography of the bonding surface. This influences both mechanical interlocking and chemical effects. Thermal spray coatings such as thermal barriers are exposed to high thermal stresses during operation due to the differences in heat strain between the ceramic coating and the metallic substrate. The bond surface must therefore be sufficiently roughened and structured for good mechanical interlocking to occur.
Specially defined surface structures have been tested for optimal adhesive strength in thermal spray coatings (top right diagram, Ref. 16.2.2.1-8). However, serial application has not been reported. The adhesive strength of synthetic resin- and elastomer-based lacquers and coatings is more dependent on chemical bonding forces. In this case, the reactive surface is the most important factor.

Figure "Operation loads demand surface features": The new part roughness influences the operating behavior of the parts (also see Fig. "Topography and roughness influencing loads"). At the same time, different zones of a part will be subjected to different operating influences. This can be compensated by specifically designing topography/roughness.

Compressor blades: The blade tip is subjected to rubbing stress and has to maintain an optimal sealing effect. In this case, the new part roughness should only apply during the first occurrences of rubbing.
The blade faces are subjected to erosion, especially on the pressure side. In this case, it can be expected that the new part roughness will become less significant after a certain operating time. On the suction side, the highest aerodynamic quality, i.e. low roughness, is required. Grooves that are oriented along the direction of flow have less of a negative effect (Fig. "Differences between turning grooves"). This can be made use of during chip-removing machining by selecting a favorable machining direction (milling).
The high-frequency vibrational stress increases towards the blade platform (fundamental flexural modes). Therefore, axial grooves should be avoided in this zone.

Contact surfaces of the blade root are subjected to fretting, high tension and shear stresses, and high contact pressure. In addition to high dynamic fatigue strength (minimal damage), the surface should have the lowest possible coefficient of friction over the operating interval. Edge loads should be avoided, placing special demands on dimensional accuracy (Fig. "Limits of demads on blade roots").

Roller bearings: In order to guarantee the fatigue life that is determined by the rolling processes, a high level of smoothness of the races needs to be ensured.

Combustion chambers: The wall around the flame is subjected to high cyclical thermal stresses and temperatures in oxidizing conditions. Heating is primarily the result of heat radiation from the flame (Volume 3, Ill. 11.2.2.1-7). The cooling is dependent on heat transferral to the cooling air film. This means that roughness characteristics are important for the heat removal through the cooling air and/or for the reflection and absorption. The roughness of a new part has temporally limited effectiveness, however, due to the typical erosion of chemical thermal barriers and oxidation of uncoated metal surfaces.

Fretting wear under intensive oxidation conditions with relatively low contact pressure has an important influence on the plug and socket connections of the combustion chamber. It can be assumed that the roughness will affect this wear process.

Turbine blading: The tips of the rotor blades must maintain high seal effectiveness while experiencing rubbing. This is achieved through wear-resistant coatings with hard particles (Fig. "Examples of high temperature brazings"). The topography of the new part must be optimized for these requirements, even if it wears down during the first rubbing occurrence.
The blade face should absorb as little heat as possible from the gas flow, and the radiation loads on the first guide vane stage should especially be minimized. The surfaces of the cooling air bores within the part and to the blade surface (cooling air film) must be optimized for a maximum cooling effect. This demands a compromise between heat transferral into the cooling air and an acceptable flow resistance. At the same time, good oxidation protection must be ensured.
The roots of the rotor blades must tolerate the high cyclical loads and fretting stresses, especially in the contact surfaces. The plug-and-socket connections of the stator vanes and the transitions to the combustion chamber, housing, and sealing plates are also subject to fretting wear and influenced by their roughness.

Figure "Fatigue strength and work or machining direction": The work direction can affect the dynamic fatigue strength of a part surface in several ways:

  • De-hardening (removing a hardened zone)
  • Tension residual stresses
  • Machining notches

In this context, the tools (geometry, stiffness, cutting material), the machining direction, and the intensity (infeed rate, cutting depth, cutting rate) are all important factors.
The following discussion is limited to the parameters of the topography.

Directional grooves of the type that are created in the machining direction influence the dynamic fatigue strength most strongly if they run across the direction of the stress. For example, in compressor blades this applies to cyclical centrifugal forces (LCF) and the fundamental flexural modes (LCF). In higher order vibrational modes, such as the lyre mode of blade edges (Volume 3, Ill. 12.6.3.1-6), radial grooves can also cause dynamic fatigue cracks. The blade transition into the root platform, the shaft to the root, and the root profile (dovetail) are subjected to especially high dynamic loads. The dynamic fatigue strength in and around the contact surfaces can be additionally reduced through fretting wear. This effect is also influenced by the topography.
In rotor disks, there is a pronounced two-dimensional stress distribution in the surface area. Tangential and radial stresses are dependent on the diameter. Around the hub bore, the tangential stresses are usually especially high, while the radial stresses are zero. Therefore, axially oriented grooves are dangerous in this area. Fortunately, they are not usually expected from the typical machining processes (turning). The radial stress increases towards the outside, making circumferential grooves more important (also see Fig. "Disk fracture caused by smoothing procedure"). In the annulus, the radial stress in the blade slots results from the centrifugal force of the blades and disk cams.
Unfortunately, typical machining processes use machining directions that create grooves that run across the main stress direction (bottom diagrams).

Rounding blade edges through abrasive blasting or sanding creates grooves that run across the primary dynamic stress. This may also apply to blade milling. Milling or grinding the blade root as shown in the diagram creates axial grooves that run across the loads resulting from blade vibrations (HCF range) and cyclical changes in centrifugal force (LCF range).
Fir tree and dovetail slots are usually reamed rather than milled. In this case, the axial machining direction runs perpendicular to the cyclical tangential stresses from centrifugal force and thermal strain. The stresses are increased by the notch effect resulting from the slot geometry. In the disk membrane, the grooves from the typical turning processes run perpendicular to the radial stresses.

Turbine and compressor rotors, especially in older engine types and smaller gas turbines, are affixed using a central tie rod (clamp bolt) or multiple tie rods distributed around the circumference. These tie rods are subjected to high tension/compression cyclical loads due to the cyclical camber (thermal strain between the rotor and tie rod) and possibly due to flexural vibrations of the rotor and/or the tie rod (Volume 3, Ill. 12.6.3.3-15). These loads act perpendicular to the turning and grinding grooves on the highly-stressed shaft (Fig. "Roughness influencing dynamic fatigue").
The bottom frame shows an example that demonstrates that even apparently secondary machining processes must be exactly specified. The mass of a turbine disk is primarily subjected to tangential stresses. When some of this mass is removed (milling, grinding), it results in the situation shown at left. However, this results in unfavorable radial machining grooves. Therefore, a machining direction that corresponds to the right diagram should be prescribed, so that only tangential machining grooves are created.

Note:
The machining direction in highly-stressed parts must take into consideration the load directions and must be specified by the design engineer. This must be strictly observed unless there are specific instructions to do otherwise.

Figure "Estimating fatigue strength by fracture mechanics": The stress concentration can be compared with a load, while the fracture toughness (with the same dimension) corresponds to a material property such as the fracture stress. In the Paris diagram, the crack growth is plotted over the stress concentration “k” as the oscillation amplitude “D k” for a cyclical load. “k” represents the loads at the crack tip under external stress. If the depth of the crack or a sharp notch is sufficiently shallow, i.e. the stress low enough, crack growth will not occur. The threshold value “D kth” is not exceeded. In special cases, as can occur with titanium alloys, for example, even very small cracks and notches below the threshold value may be capable of growth.

Disregarding effects such as hardening and residual stresses, it is possible to use fracture mechanics to estimate the depth “a” (or roughness depth “R”) of grooves that could cause cracking under a known dynamic load “Ds”. At the same time, the crack toughness “D k” must be known as a material property (Volume 3, Ills. 12.2-4 and 12.2-12 ).

Figure "Influence of surface topography and finishing": Finishing processes create the surface topography, but are also influenced by existing topography. This interaction is described using several typical examples.

Soldering: Increased roughness increases the surface that is wetted by solder, but does not have a positive effect on strength (Fig. "Influences at the thrength of brazings", Ref. 16.2.2.1-11). The orientation of the roughness of machining grooves influences the flow of solder into the gap. Therefore, pronounced grooves perpendicular to the desired solder flow direction should be avoided (Fig. "Bond face of braze joints", Ref. 16.2.2.1-12). Calotte-like topographies (Ref. 16.2.2.1-13) on shot peened surfaces can evidently also hinder the solder flow. In contrast, machining grooves in the direction of the desired flow can have a beneficial effect. It is possible to create this type of desirable topography with the aid of a specific process. The influence of the roughness on the soldered joint should be especially noticeable in cases with thin gaps and viscous melts. High-temperature Ni-based solders used in hot parts are especially sensitive. In order to avoid damaging the base material as much as possible (diffusion, structural changes), the high soldering temperatures selected are as low in the application range as possible. This results in higher viscosity of the molten solder with correspondingly poor gap-filling behavior. The longer the gap, and therefore the solder flow, the more pronounced the effects and the influence of the roughness become when diffusion processes between the melt and base material additionally increase viscosity.

Thermal spraying: In order to attain the desired high adhesive strength in metallic or ceramic coatings with the aid of mechanical interlocking with the surface (Fig. "Bond strength of thermal spray coating structures"), the topography of the surface should be optimized. This can usually be done by specifying the machining process and the roughness. However, it must be mentioned that the roughness requirements alone will not guarantee desired topography such as open porosity and sufficient jaggedness (Fig. "Roughness data tell not enough"). Therefore, the process that ensures the structure of the bonding surface is created cannot be changed (for reasons of cost, availability, etc.) without sufficiently reliable verification. Abrasive blasting is usual for the base material surface. If this is followed by the application of a bond layer, this must also have a sufficient topography for proper bonding with the “functional coating” to occur.

Processes in a vacuum (electron beam welding = EB): During EB and diffusion welding, the surfaces being joined are usually pressed together. If these surfaces have grinding or turning grooves in their usual direction of circumference, it can result in fouling, etching and cleaning agent residue, or air being trapped (Ills. 16.2.1.3-26 and 16.2.1.3-27). If this prevents sufficient degasing in the vacuum, the gases may be released during the welding process and create flaws such as oxidation and porosity. Gases and fouling trapped in the roughness can also complicate evacuation during heat treatments. Depending on the heating rate, this can result in undesirable reactions (oxidation) on the surface.

Cleaning and rinsing: The effectiveness of these processes also depends on the roughness of the surfaces. Fouling that has settled in indentations is understandably difficult to remove. Even if cleaning is successful, there is still a higher probability of baths and fouling being transferred to other areas. It is possible that a film of fouling can appear on the bath surface and contaminate other parts (Fig. "Difficult crack detection by fouling on baths"). Insufficient cleaning can also affect subsequent testing processes, including the sensitivity of penetrant testing and the result analysis of macro-etching on forged parts.

Non-destructive testing (Ref. 16.2.2.1-2): The topography and roughness of a surface can have a decisive effect on some non-destructive testing methods. High roughness is unsuitable for attaching ultrasonic sensors. This especially complicates the detection of flaws in the surface region.
High roughness and topographies that trap fouling (micro-burrs, Ill. 17.3.1-4; galling, structural components, open porosity) lead to background fluorescense during penetrant testing. In addition, they significantly complicate the identification of fine cracks from parallel grooves. These unfavorable surfaces are also created by typically intensive abrasive blasting of fine cast surfaces after the casting process and removal of the casting molds. A complication is the hammering-shut and blocking of flaws (porosity, micro-cracks) through blasting media. Eddy current testing reacts especially sensitively to very minor inhomogeneities and roughnesses near the surface that mask flaws. It is easy to recognize that the high demands for sensitivity and reliable detection in the named testing processes (Ills. 17.3.1-2, 17.3.1-3) require correspondingly low roughness.

Galvanic processes: The quality of precipitated metal layers, primarily Cr and Ni coatings, is also dependent on the roughness profile (improved precipitation at the peaks) and the roughness depth (Fig. "Typical flaws on galvanic coatinga"). Larger flaws such as etching pits and micro-burrs can positively (e.g. bond strength) or negatively (e.g. thickness variations, structure) influence the coating. The reactivity of a large surface can affect first the precipitation rate, and then the coating properties (e.g. residual stresses, cracking), especially smoothness/shininess. Therefore, it can also influence the results of subsequent work such as grinding (cracking).

Reforming processes: Processes such as deep drawing and pressing, in which the material is locally plastically deformed through contact with the tool and without material removal, can affect surface roughness in two different ways:

  • Notches such as grooves and etching pits, especially those perpendicular to the primary deformation direction, can lead to cracking (crack fields) and even fracturing of the part.
  • The roughness may result in greater friction forces during reforming, causing the material to be overstressed (cracking). The thickness of the metal sheet may also have undesirable variations and dangerously worsen the later operating behavior of the part (e.g. vibrations, Fig. "Problems by deviation of wall thickness").

Chemical processes: These include etching and burnishing. The topography created by etching can vary greatly, depending on the material, bath conditions, and process parameters. Typical characteristics include etching pittings, grain boundary corrosion, or grain surface corrosion (Fig. "Damages by not approved processing baths", Ref. 16.2.2.1-10). The resulting roughness influences processes such as penetrant testing and the analysis of micro-etching used to detect segregations (segregation etching).
In engine production, burnishing is used on steels for toothed gears and shafts (Fig. "Cracking in steels processed in burnishing baths"). If the tensile stresses at the part surface are sufficiently high, it can result in cracking (SCC) around small notches such as machining grooves.

Mechanical smoothing processes: These understandably become more difficult with increasing roughness.

Diffusion coating: Roughnesses and any fouling trapped in them can compromise the diffusion process and cause coating flaws.

Diffusion welding: This process can react very sensitively to excessive roughness. If the contact pressure is insufficient, bonding may only occur at the roughness peaks. This results in kissing bonds, which are virtually impossible to detect (16.2.1.3-37).

Machining to remove material: Machining grooves can affect many different subsequent production steps, and have already been discussed in the relevant sections.
One roughness effect that can influence the machining process itself is cooling lubricant sticking to the surface to be machined. This can reduce tool wear and cutting forces, thus affecting the topography of the new work surface.

Figure "Roughness influencing dynamic fatigue" (Ref. 16.2.2.1-16): This older literature contains interesting data regarding the effects of roughness on dynamic fatigue strength in metallic materials, although the tests were primarily done on steels.

With regard to the roughness depth, the geometric shape of the individual grooves was not taken into consideration. However, it can be assumed that process parameters such as infeed rate and turning tool shape also have an effect on the dynamic strength (Fig. "Differences between turning grooves").
In the top diagrams, it can clearly be seen that the reduction in dynamic fatigue strength is related to the roughness; i.e. the production-specific surface quality (top left diagram). However, this reduction, which corresponds to an exponential function, only occurs beyond a threshold roughness depth (top right diagram). It is evidently independent of the load type (flexing cycles, tension/compression, cyclic tension).

Today, the threshold roughness depth can be explained using fracture mechanics. It is related to the threshold value of the stress intensity (Kth ; Fig. "Estimating fatigue strength by fracture mechanics"). Below this material-specific value, which corresponds to a dynamic load on a specific flaw size, crack growth can only be expected in special cases. Like Kth , the threshold roughness depth is material-dependent. It is suspected that the threshold roughness depth is influenced by the grain size. This indicates a connection with the growth of small cracks below Kth , which is known today.
The influence of the roughness on the fatigue resistance is evidently independent of the stress distribution and the stress gradients into the part. This can be explained by the fact that the roughness is limited to a very thin surface zone. Therefore, the part size should also not have an influence on the effects of the roughness. A noticeable influence can be expected when the crack growth phase is dominant.

The process-specific influence of the surface quality (polished, ground, rough-machined) on the fatigue resistance depends on the roughness depth, as shown in the bottom right diagram. This is also true for single grooves. However, it must be noted that grooves that create significant additional tensile residual stresses (e.g. grinding) have a much more damaging effect on the dynamic fatigue strength than a rough-machined state with comparable roughness depth.
The influence of the roughness on the dynamic fatigue strength is influenced by the pre-stressing and mean stress (middle right diagram). At low mean stress levels, the rough surface leads to faster dynamic fatigue strength losses than smoother surfaces do. However, this effect becomes less pronounced at higher mean stress levels.

If the dynamic fatigue strength of a coating is greater than that of the base material, it is beneficial for the dynamic fatigue strength of the system. The bottom right diagram shows that these coatings can also reduce the influence of the roughness on the dynamic fatigue strength. The prerequisite for this is sufficient coating thickness. Nitrided coatings that are thinner than 0.1 mm increase the dynamic fatigue strength relative to the uncoated state considerably less in rough-machined surfaces than in polished surfaces. The dynamic fatigue fractures in all specimens originated below the coating (Fig. "Flexual load and flat stress gradient"). Therefore, they are not related to the roughness. In the case of nitrided coatings with sufficient thickness (above 0.2 mm), fine finishing may not be required.

Figure "Differences between turning grooves": This LCF fracture occurred on a clamp bolt for the axial fastening of a rotor. The material is a forged Ni-based alloy. One can clearly see the cracks in the circumferential turning groove (top left diagram). The top right diagram shows a metallographic cross-section across the turning grooves. The sharp-edged groove base is cracked in several places. The bottom diagrams show the various damaging effects of different groove shapes (Fig. "Roughness data tell not enough").
It must be noted that hardening and residual stresses in the groove base will vary considerably in the two groove types. In case of a rounded groove base, protective compressive residual stresses and hardenings can be expected. In contrast, the sharp groove base will result in dynamic fatigue strength-reducing tensile residual stresses, in addition to the notch effect. The hardenings should also be less pronounced in this case.

Figure "Loads and contact surfaces affect operaring" (Ref. 16.2.2.1-15): The importance of optimal surface structures and treatments, and therefore the topography, to the loadability and safety of a part can be seen in the root connection of a compressor blade to a disk.

Cracking in the contact surfaces of compressor blade roots, especially those made from high-strength titanium alloys, is a well-known problem (Volume 2, Ill. 6.2-3). In the depicted case involving a large fan engine (middle diagram), a fan blade fractured at the root and several cracked blade roots were found in other cases. The cause was the failure of the solid film lubricant between the contact surfaces. In addition to high shear stress and bending stress, the root surface experiences fretting due to rubbing movements that damage the material (especially titanium alloys) and drastically reduce the dynamic fatigue strength (Fig. "Fretting damage loweing by shot peening").

The lower the coefficient of friction, the lower the shear stresses in the surface of the blade root, and therefore the crack-inducing loads, are. This necessitates a solid film lubricant that remains effective over the entire operating period. The solid film lubricant is contained in the calotte structure of a shot-peened surface (bottom right detail). The calottes can also store wear products. In addition, the shot peening process induces compressive residual stresses that lower the tensile operating stresses, and increase the dynamic fatigue strength through hardening. It is also assumed that the special surface structure has a beneficial influence on the micro-scale stress distribution (Fig. "Topography and roughness influencing loads"). The fretting-threatened contact points at the tips are relieved by the operating stresses. The more highly stressed calotte base remains undamaged. In some cases, an additional protective loose sheet metal insert is used to further minimize fretting damage (bottom left diagram).

Figure "Aerodynamic demands finishing process" (Ref. 16.2.2.1-14): The aerodynamic requirements on blade faces vary across the profile (top diagram). In Zone 1, the highest quality is required to prevent increased friction losses and flow stalls. Grooves that run at an angle of more than 10° relative to the flow direction have a similarly strong effect as perpendicular grooves. On the other hand, grooves that run along the direction of flow are usually tolerable. However, the reduction of dynamic fatigue strength must be assessed.

The diagram shows allowable roughnesses for engines with typical applications in fighter aircraft and civilian aircraft. The necessary roughnesses decrease due to the laminar boundary layer, which becomes thinner towards the back of the compressor (bottom diagram, influence of flow speed, temperature, and pressure).

One can see that, depending on the various flight missions, military applications require roughness qualities in the low-pressure compressor (LPC) as high as those in civilian high-pressure compressors. Due to the low flight speeds at high altitude, compressor blades in civilian engines can have somewhat greater roughness than comparable blades in military engines. Fan blades permit the greatest new part roughnesses.

References

16.2.2.1-1 “ASM Handbook”, Volume 4 (“Heat treating”), Volume 5 (“Surface Cleaning, Finishing, and Coating”), Volume 6 (“Welding, Brazing, and Soldering”), Volume 7 (“Powder Metallurgy”), Volume 14 (Forming and Forging“), Volume 15 (“C asting”), Volume 16 (“Machining”).

16.2.2.1-2 Peter Adam, “Fertigungsverfahren von Turbotriebwerken” Birkhäuser Verlag,, 1998, ISBN 3-7643-5971-4, pages 90 and 129-212.

16.2.2.1-3 M. Field, J.F. Kahles, “Übersicht über die Oberflächenbeschaffenheit bearbeiteter Werkstücke `Surface Integrity'”, periodical “Fertigung” 5,72, pages 145-156.

16.2.2.1-4 Aircraft Accident Report, NTSB/AAR-98/01 (PB98-910401, DC A96MA068), “Uncontained Engine Failure Delta Air Lines Flight 1288”, Accident July 6,1996, Report from 1998, pages 1-129. (3273)

16.2.2.1-5 F. Klocke, “Measuring and Testing in Production”, “Manufacturing Technology I, Lecture 2”, WZL, RWTH Aachen.

16.2.2.1-6 “Dubbels Taschenbuch für den Maschinenbau I” , Twelfth Edition, 1961, Springer-Verlag, pages 525 and 526.

16.2.2.1-7 Gerd Lütjering, J.C. Williams , “Titanium”, SpringerVerlag Berlin, 2003, ISBN 3-540-42990-5, pages 197-223.

16.2.2.1-8 T.E. Strangman, “Ceramic lined turbine shroud and method of its manufacture”, patent specification EP 0256700B1, 1987.

16.2.2.1-9 F.Bordeaux, R.G.Saint-Jacques, C.Moreau, S.Dallaire, J.Lu, “Thermal Shock Resistance of TiC Coatings Plasma-Sprayed on Macroroughened Substrates”, Proceedings of the “Fourth National Spray Conference”, Pittsburgh, PA, USA, 4-10 May 1991, pages 127-134.

16.2.2.1-10 I.Engel, H. Klingele, “Rasterelektronenmikroskopische Untersuchungen von Metallschäden”, published by the Gerling Institut für Schadensforschung und Schadensverhütung GmbH Köln, ISBN 3-9800043-0-9, 1974, pages 37,38, 66 and 202.

16.2.2.1-11 J. Colbus, W. Hauch, “Beitrag zum Verhalten von Lötverbindungen aus Hochtemperaturloten auf Edelmetallbasis an hochwarmfesten Werkstoffen bei Zeitstands-Belastungen bis 1000 Stunden und Temperaturen bis zu 800 Grad C”, periodical “Metall” , Volume 23, October 1969, Issue 10, pages 994 -1002.

16.2.2.1-12 W.Wuich, “Rationalisieren des Hartlötens durch Verwendung von Formteilen”, periodical “Metall” , Volume 24, April 1970, Issue 4, pages 371-374.

16.2.2.1-13 “Das Löten, Überblick und Anwendungsstand”, “Mitteilungen der BEFA” , Nr. 11, Volume 14, 1963, pages 1-16.

16.2.2.1-14 A.Schäffler, “Experimental and Analytical Investigation of the Effects of Reynolds Number and Blade Surface Roughness on Multistage Axial Flow Compressors”, ASME Paper No. 79-GT-2 of the “Gas Turbine Conference”, San Diego, California, March 12-15, 1979, pages 1-8.

16.2.2.1-15 ATSB, “Examination of a Failed Fan Blade, Rolls-Royce RB222 Trent 892 turbofan engine, Boeing 777-300,A6-EMM”, Technical Analysis Report No: 8/01, Occurrence No: 200100445, Reference: BE/200100004, Oct. 31. Jan. 2001, pages 1-21.

16.2.2.1-16 E.Siebel, M.Gaier, “Untersuchungen über den Einfluss der Oberflächenbeschaffenheit auf die Dauerschwingfestigkeit metallischer Bauteile”, periodical “VDI-Z”. 1956, Nr. 30, October 21, pages 1715-1723.