Table of Contents
184.108.40.206 Baths for Cleaning, Etching, and Rinsing
Very different types of processing baths are used during finishing (Ref. 220.127.116.11-15). Auxiliary substances from process steps such as milling (e.g. cutting oils) or forming (e.g. lubricants) must be removed in order to prevent them interfering with later finishing steps (e.g. coating application) or testing (e.g. penetrant testing). Tested and approved cleaning baths and cleaning fluids remove fouling without having an undesirable effect on the part. Primarily, this involves watery cleaning baths with a pH value that has been adjusted to the specific application (acidic, neutral, alkaline). These cleaning fluids are mixed with additives (e.g. detergents) containing tensides (anti-foaming agents) and inhibitors (to prevent corrosion). Only in special cases are organic cleaning agents used, since their costs are very high, also because of equipment and environmental measures (Ref. 18.104.22.168-2). A special difficulty with cleaning and rinsing baths is the fact that they are used at the end of the finishing process, when all the production costs have already been invested. This means that the damage potential is especially high.
Etching baths in new part production are primarily used for the following purposes:
- Creating metallic, reactive, easily-bonding surfaces, for example, for coatings and solder.
- Opening flaws near the surface, such as cracks (e.g. from grinding) or cavities (in cast parts), which were pressed shut during a machining process or during shot peening. This is a prerequisite for conducting effective penetrant testing with sufficient flaw detection (Fig. "Opening of cracks before penetrant testing").
- Removal of coatings for re-coating during a reworking process.
- Removal of brittle oxide layers, e.g. before plastic forming such as deep drawing.
- Removing lacquers.
Very aggressive etching baths are required in order to dissolve the corrosion-resistant Ni alloys and/or their chemically highly stable oxides. In case of mistakes, transferrals, or fouling, these media will cause damages. In order to remove any bath residue, subsequent rinsing is especially important.
If hot water is used for rinsing, the heated part will dry completely more quickly after removal from the rinsing bath. As with cleaning baths, there is a risk that rinsing baths could become contaminated and foul other parts (Fig. "Problems by contaminated cleaning baths").
Finally, one must remember that in baths in which abrasive processes (e.g. vibration grinding, tumbling, or ultrasonic baths) continually create fresh metallic surfaces, effects such as corrosion or chemical attack are especially aggressive. These damages will be much more pronounced in these cases.
Figure "Damagings by processing baths": This diagram shows important tasks and special problems related to processing baths for cleaning, etching and rinsing in the context of development, finishing, and repair.
The treatment of parts in processing baths is comparable to machining processes such as chip removal with regard to reproducibility and reliability.
The typically large number of parts that pass through a processing bath means that, if damages are detected late in an advanced stage of the finishing process, there are special risks regarding safety and costs.
Deviating from tested, optimized process parameters and part properties can cause damages such as cracking, embrittlement, and corrosion notches.
Process baths used for degreasing (wetting) and/or etching (opening cracks) are used as preparation for penetrant testing (Fig. "Opening of cracks before penetrant testing") and therefore affect part safety. Naturally, this especially applies to baths that also represent the testing process in the shape of segregation etching. In general, bath treatments affect subsequent finishing steps and can indirectly lead to damages (e.g. bonding problems in coatings).
Technical personnel with suffucient experience are necessary in order to ensure the quality of the parts in cleaning, rinsing, and etching sequences and to prevent costly damages from occurring.
Development: This primarily concerns the selection of the bath and determining optimal process parameters. This must take into account costs and availability of the process bath, as well as part-specific factors such as the base material, any coatings, or subsequent finishing steps.
Naturally, the parts must not be damaged (Ills. 22.214.171.124-6 and 126.96.36.199-7). Effects on the later operating behavior must be given special consideration.
A special problem is the transferral of the process from development into serial production. Important factors such as plants, equipment, and part conditions can be very different during the two stages.
The large number of parts in serial production is capable of altering the process properties (changing the bath, etc.) or importing fouling much more quickly than the far smaller number of parts during the development stage.
If there are any uncertainties regarding the use of processing baths, then suitable, realistic, operation-relevant tests should be undertaken to verify the effectiveness of the process.
Finishing process: Serial finishing requires a stable process with a sufficient certainty that it will not drift outside of the specification limits. This includes monitoring the baths and auxiliary substances.
If the bath properties change, it can lead to less material removal or cause selective attack (e.g. on the grain boundaries).
Auxiliary materials from previous production steps (e.g. cutting oils or separating agents) can be carried into the bath. This creates a risk that these foulings could be transferred to other parts (Ills. 188.8.131.52-3 and 184.108.40.206-4).
This undesirable influencing of subsequent processes such as penetrant testing, segregation etching, or coating must be prevented through bath monitoring and inspection of the parts.
Repair: Testing and selecting suitable processing baths, as well as their correct application within the framework of repairs is a very demanding task. This is due to several factors.
Small, sporadically arriving numbers of parts increase the risk of time-related bath changes. Operation-dependent changes and fouling of the part can cause unexpected behavior. One example is intergranular cracking in sensitized materials. Heavy oxidation on long-running hot parts requires especially intensive etching processes, usually in combination with abrasive blasting. The varying oxidation levels in the different part zones can lead to pronounced local attack, the appearance of which may vary, e.g. as a grain boundary attack or as laminar material removal.
Fouling on parts, such as splashed or smeared iron or nickel resulting from rubbing during operation (blade tips, labyrinth tips) can decisively change etching attacks. In etching baths, titanium alloys that are fouled with iron tend to unusually heavy absorption of hydrogen with hydride formation. The result is pronounced embrittlement (Fig. "Hydrogen embrittlement by etching process", Ref. 220.127.116.11-11). Similar foulings that may occur during the finishing process (smeared material from steel brushes, fusion of machining sparks) must be avoided.
Figure "Problematic changing of long used baths": Unnoticed alteration of baths over longer usage and/or standing times presents a danger. The possible causes for this aging can be very different (top frame) and influence one another reciprocally. These creeping changes to a bath can result in desired part properties not being attained and/or damage to the part (top frame, Fig. "Problems by contaminated cleaning baths"). For this reason, suitable monitoring of the processing baths is absolutely necessary. In this context, clear, testable limits for the changes in the baths must be specified. On the basis of these, the baths must be regenerated or replaced when necessary.
One important possibility for changes in bath properties is bath transferral (bottom frame). This can occur over the course of several sequential baths, making the potential changes very complex and their effects highly varied. In parts with hollow spaces and bores, the danger of transferral is especially high. Typical examples of this include cooled turbine blades (Fig. "Damage risks by etching baths") and complex welded constructions such as housings (Page 18.104.22.168-2).
Figure "Problems by contaminated cleaning baths": Especially cleaning and rinsing baths, which seem harmless at first glance, can be the cause of dangerous substances being transferred. These contaminants can originate in processing baths as well as in the processed parts (Fig. "Corrosion of ‘resistant metals’ by contaminations").
Typical bath contaminants can have specific damaging effects (also see Fig. "Problematic changing of long used baths"). In the following, damage mechanisms are described in greater detail. These effects are not all documented in the available literature regarding actual cases of damage. However, known characteristics of the contaminants and damage mechanisms indicate the potential damages.
Corrosive substances: These are primarily Cl compounds from
- etching baths,
- cutting oils,
- electrolytes (e.g. from electrochemical processes, Ref. 22.214.171.124-2). For example, ECM of titanium alloys is usually done with saline solutions.
Anti-friction coatings and separating agents may contain fluorine compounds. These can have damaging effects in the bath itself, as well as in subsequent finishing steps (e.g. heat treatment).
If titanium alloys are under sufficiently high tensile stress at high temperatures, stress corrosion cracking can be expected, e.g. around welds (Fig. "Stress corrosion cracking by process baths and hand sweat").
In contrast, stress corrosion cracking is not as readily expected in titanium alloys in alcohol baths with minor halogen contamination (Ref. 126.96.36.199-3). The most pronounced effects are from contaminants in methanol and ethanol. Minor amounts of water suppress this effect.
In CrNi steels, sensitization can result in intergranular corrosion. In case of element formation with other surface contaminants such as splashed metal, pitting corrosion can occur (pinhole corrosion).
If residue containing Cl dries in cooling air bores, there is a danger of thick oxide coatings forming during operation, constricting the bore cross sections. In combination with an insulating effect, these oxides are capable of causing overheating and extensive consequential damages (Ills. 188.8.131.52-8 and 184.108.40.206-9). If chlorine residue causes grain boundary corrosion, it will promote thermo-mechanical fatigue cracking in the cooling air bores.
Substances that influence wettability, such as silicon compounds and PTFE: Silicon primarily originates in anti-foaming agents used in cutting fluids. Hand creams can also import silicon into the process. PTFE primarily originates in liquid or sprayed separating agents. If residue of these compounts has a notable effect on wettability, it can have serious consequences:
- The function, i.e. sensitivity, of penetrant testing may be compromised.
- The bond strength of laquers and organic coatings, as well as adhesive connections, may be affected.
- Weakening of the bonds between layers of fiber-reinforced plastics (FRP).
Substances containing sulfur: These compounds occur in lubricating greases and anti-friction coatings containing MoS2 . Sulfur compounds are also found in cutting oils (Ref. 220.127.116.11-6). Sulfur can cause sulfidation in Ni-based materials during subsequent heat treatment, as well as in hot parts at the operating temperatures in the rear compressor area. Sulfides can form on non-ferrous metals and silver. The poor friction properties of silver sulfide affect the tightening force of threaded connections and can cause serious damage to gliding systems (e.g. axial piston pumps).
Ferrous contaminants such as shavings, grinding dust, or rust particles can have negative effects on corrosion behavior (e.g. of titanium alloys) and weldability (cracking, Fig. "Fouling of Titanium welds"). In baths, iron and Ni particles on titanium alloys can lead to intense hydrogen absorption with hydride formation. The result is serious embrittlement (Fig. "Hydrogen embrittlement by etching process", Refs. 18.104.22.168-3 and 22.214.171.124-11). In order to prevent this, oxidizing additives are added to the baths.
Figure "Corrosion of ‘resistant metals’ by contaminations": A typical example is metallic fouling on parts that is transferred to a bath. Baths that are contaminated in this way can then transfer the fouling to other parts. The consequences are described in Fig. "Problems by contaminated cleaning baths". This occurrence can also compromise the effectiveness of the bath.
In the bath, it is possible that reactions between the base material and fouling will occur (Fig. "Damages by not approved processing baths"). This fouling can be transferred to the part in many different ways:
The top diagram shows a rotor disk as an example of a “scooping part”. Liquid that has been contaminated with shavings or grinding dust can collect and, after the fluid has evaporated, the fouling can be firmly bonded to the surface.
Wear products from chipping tools such as drills or turning tools.
Shavings may be smeared on or pressed in (bottom right diagram, Fig. "Dynamic fatigue strength influenced by fused chips").
Material from the machine tool itself may be smeared onto the part (e.g. wire brush or steel wool), or there may be wear products from other parts smeared onto the tools or abrasive media (e.g. abrasive belts, Ills. 126.96.36.199-6, 188.8.131.52-1, and 184.108.40.206-2).
Fastening equipment (e.g. brass, copper) and sprue (e.g. lead, bismuth) can leave potentially damaging wear products on the part surface.
Firmly bonding residue from the “sparks” (bottom right diagram, Chapter 220.127.116.11) of machining processes such as grinding and separating.
Splashes from thermal cutting and welding processes that are created during the melting phase (e.g. laser boring and cutting, torch cutting, laser welding, electron beam welding).
Figure "Bath treatment altered by material deviations": The risk of damage increases if the surface of a part differs from those parts (or specimens) that were used to test and optimize the processing bath. In this case, the consequences may range from disturbing effects (appearance) to dangerous damages (cracking). Typical deviations and their effects include:
Residual coating remnants: This includes insufficiently removed coatings, as well as reaction layers such as oxides. These change the etching process so that, for example, the material removal is heavier around the coating residue, and/or the coating residue remains after the etching process. This makes subsequent processes such as soldering or diffusion-coating much more difficult.
Tensile stresses: If processes (Chapter 18.104.22.168) deviate from tested parameters, unusually high tensile residual stresses can be induced in the surface. These stresses are sufficient to cause stress corrosion cracking in some materials (Ills. 22.214.171.124-6 and 126.96.36.199-8). Plastic deformation of the surface through a machining process can reduce the sensitivity of the part to intergranular corrosion (Fig. "Preventing corrosion cracking by shot peening,"). If this effect was used unconsciously, changes to the work process can suddenly make parts sensitive to corrosion.
Roughness differences: The surface that reacts with a processing bath is influenced by roughness and topography. If, during segregation etching, an etching surface is used to determine quality, changes in the etching pattern unrelated to the material composition can also influence the safety of the part. This is also true for changes in a machining process relative to its process verification testing.
Structural changes may be traced back to a difference in heat treatment conditions, such as solution annealed rather than hardened.
Local structural changes around welds can result in increased grain boundary corrosion (sensitiziation, Volume 1, Ill. 188.8.131.52-4, Ref. 184.108.40.206-12).
If, after a change in supplier, raw parts are introduced that have a different grain size than those previously used, it can affect the etching process. This situation occurs in the case of a combination of coarse- and fine-grained material in a welded joint (Fig. "Influences at micro-cracking").
If carbides or other phases are selectively attacked by an etching agent, their size and distribution, which may depend on the raw part supplier, will become apparent through potential damages.
Depending on the type, surface fouling can cause various damage mechanisms (Fig. "Corrosion of ‘resistant metals’ by contaminations" and Chapter 220.127.116.11).
Figure "Damages by not approved processing baths": Processing baths such as etching baths and coating/lacquer removal baths can cause serious damages to parts if they are not used for the proper applications or with unsuitable process parameters. In the following, examples are used to explain typical damage mechanisms:
“1”, “5”, “6” and “8”: Residue from etching baths such as FeCl3 + H2O on “scooping” parts (Fig. "Corrosion of ‘resistant metals’ by contaminations") can cause intergranular corrosion around pitting corrosion.
During temporary storage, this type of residue can reach high concentration levels due to evaporation, and can unexpectedly cause corrosion.
In the bath, these types of damages are related to deviations such as overtimes and aging/fouling of the bath. The same is true for the surface and material conditions of the part (Fig. "Bath treatment altered by material deviations", Ref. 18.104.22.168-7).
Seemingly harmless baths such as glycolic acid, which are used to remove ceramic coatings, can dissolve carbides out of the base material surface. This results in dangerous micro-notches.
If insufficient rinsing causes etching bath residue to remain in cooling air channels in hot parts, increased oxidation can be expected during later operation. This will worsen the cooling properties and lead to dangerous shortening of part life (Ills. 22.214.171.124-3 and 126.96.36.199-9).
Baths that selectively react with solder materials can cause weakening or even complete dissolving of soldered connections (diagram “8”).
“2” and “4”: Element formation in parts made from different materials (Fig. "Risk of combined materials in etching baths"), such as soldered connections consisting of a solder type that is considerably different from the base material, can cause corrosion even in cleaning and rinsing baths. There is an especially high danger of corrosion in cases in which the process deviates from the tested and proven process parameters. A typical example of element formation is hard steel compressor stator vanes soldered with silver or copper solder, which were common in older engine types (Fig. "Brazing applications in engines"). In this case, there is a danger of pitting corrosion at the transition of the solder to the base material (Volume 1, Ill. 188.8.131.52-1). Similar corrosion can occur at the transition zone of coatings that behave more nobly than the base material. An example of this is wear-resistant WC coatings on blades (Volume 1, Ill. 184.108.40.206-8.1).
Transmission housings and compressor inlet housings in older engine types are often made from cast light metal and are fitted with threaded inserts made from hardened steel (Ref. 220.127.116.11-7). Due to the tightening force, these inserts may be silver-coated. Even stay bolts (martensitic material) or inserts for oil pipes (CrNi steel) may remain in housings during lacquer removal or surface preparation before a coating process. This is usually the case during overhauls, not in new part production. Insufficient masking and/or aggressive baths can cause a corrosive element to form between the housing and the insert. This is also true of bath residue in threaded bores and blind holes.
“3”: Damage to coatings related to processing baths is often reported. These problems are primarily caused by mistakes in bath selection or baths that have not been sufficiently tested. One example is the partial or total removal of wear-resistant coatings made from tungsten carbide (WC) bound in a Co matrix. These coatings are widely used in engines, especially on surfaces that are subjected to fretting wear. These cases have made it especially clear how difficult it is to retroactively accurately determine the process flaws.
“7”, “9”, “10”: Abradable coatings can suffer material-specific damage through baths and bath residue. Adhesive bonds may be dissolved, depending on whether the adhesive is organic or inorganic. For example, inorganic adhesives filled with Al powder (Fig. "Factors influencing adhesive joints") rapidly dissolve in weak acids or bases. If a porous coating becomes soaked full of bath residue that then remains in the coating, it can act over an extended period of time.
Organic adhesives and coating materials such as filled elastomers or synthetic resins can embrittle, shrink, soften, or tear if they are placed in unsuitable baths.
Figure "Etching problems": This compilation of potentially damaging effects of etching baths is intended to help with the identification of probable damage mechanisms in case problems occur. Characteristic indicators are covered in Ills. 18.104.22.168-2 to 22.214.171.124-6.
There is additional information in Volume 1:
Chapter 5.4 deals with corrosion, and therefore also discusses important effects that occur in connection with etching baths.
Chapter 5.4.1 deals with stress-free corrosion: element formation, sensitizing, intergranular corrosion (IGC).
Chapter 5.4.2: stress corrosion cracking (SCC).
Chapter 5.4.4: hydrogen embrittlement (Fig. "Hydrogen embrittlement by etching process").
In the right diagram, damages related to etching baths are shown in accordance with the usual sequence of processing stages.
Raw parts and blanks: Surface changes on these parts can have an undesired effect on the results of processing baths, especially etching. Affected parts are those that already have their final dimensions. These include precision-forged compressor blades (Ti alloys, forged Ni alloys), cast turbine blades (Ni alloys), and cast housings for accessory equipment (Al or Mg alloys). Structural deviations relative to the tested parts may be the result of differences in casting or forging parameters. Deviations in the surface treatment, e.g. a blasting process, can change the reactivity or cleanliness of the surface (loading effect, Fig. "Loading effect by blasting processes"). Changes to the heat treatment, e.g. related to the furnace atmosphere, can affect the oxidation state or lead to depletion of alloy components due to diffusion.
Pre-treatment: The production processes before etching affect the surface properties, and therefore also the behavior of the surface in the etching bath. Influencing factors include plastic deformations and residual stresses resulting from chipping machining processes and/or blasting (Fig. "Preventing corrosion cracking by shot peening,"). Naturally, material combinations with different behavior in etching baths will also have an effect (solder/base material).
It is important that no production steps occur before etching that could be damaged by the etching process, such as coating, adhesive joining, or soldering (Fig. "Damages by not approved processing baths"). It must be ensured that the part is not threatened by unavoidable element formation or structural changes. This may require adjustment of the production sequence. This task is primarily the responsibility of the design engineer and the work preparation.
Etching: Bath testing and process optimization are prerequisites for damage-free etching. Deviations between test specimens and the part or laboratory conditions and serial production can lead to nasty surprises.
Even apparently “harmless” etching baths can cause damages. For example, formic acid can cause hydrogen embrittlement in high-strength steels and titanium alloys (Ref. 126.96.36.199-8). Etching with glycolic acid is used to remove ceramic coatings. However, it can also dissolve material-specific carbides out of the surface, creating micro-notches.
Critical process parameters include bath volume, throughflow relative to the etching surface, contamination possibilities, and equipment such as masking/covers or baskets. Bath monitoring to prevent unallowable changes such as aging and contamination must ensure the stability of the process. Of course, it must be ensured that suitable rinsing baths effectively remove dangerous etching agent residue (Fig. "Damages by not approved processing baths").
Finishing steps following etching: Problems during etching can have effects later in the finishing process. If the desired reactive surface for a coating or soldering process cannot be created, then a stable result cannot be expected. If surface flaws such as cavities and cracks are not opened, then there is a doubt as to whether penetrant testing (Fig. "Opening of cracks before penetrant testing") will be sufficiently safe.
Operating influences: Safe operating behavior can be influenced significantly by etching processes during finishing. Etching agent residue in cooling air ducts can cause overheating of hot parts (Fig. "Damages by not approved processing baths"). Insufficient joining surface preparation can cause separation of coatings and layers during operation.
Figure "Chlorine in process baths causing stress corrosion": Titanium materials can suffer stress corrosion cracking above 150°C under the influence of chlorine compounds with sufficiently powerful tensile stresses. Degreasing baths (top left frame) such as trichloroethane (TRI), tetrachloroethane, or perchlorethane (PER) can release chlorine under certain conditions. Stress corrosion cracking has also been observed in titanium alloys in connection with Cl in water or methanol (Ref. 188.8.131.52-14). Several reactions, which may also influence one another reciprocally, lead to decomposition that releases Cl ions:
- Secretion of hydrochloric acid (HCl) occurs through photolysis (light-influenced), thermolysis (temperature-influenced), hydrolysis (reaction with water), or oxidation (reaction with air).
Metal chlorides such as aluminum chloride, iron chloride, or zinc chloride can promote decomposition. These chlorides can form on the parts or on the equipment itself (clamps, walls, etc.).
If the PER or TRI solution reaches a contamination level of over 50% (water, greases, and oils), then the temperature of the bath in the sump below the degreasing vapors (top left diagram, Ref. 184.108.40.206-1) will have a decomposing effect and release Cl. In order to prevent this, neutralizing and oxidizing inhibitors are added. However, their effectiveness is limited. Therefore, additional measures are required (Ref. 220.127.116.11-11):
- High alloy steel should be used for the equipment and rigs.
- Indirect heating can prevent overheating of the sump.
- Analytic monitoring of the content of oil contamination should keep its level safely below 20% by weight.
- Monitoring and adjusting the pH value to keep it above 7. This ensures that there will be a sufficient amount of inhibitors.
In the following, the highly interesting results from Ref. 18.104.22.168-10 are discussed in abbreviated form:
The tests described in the following investigated cracking that occurred under the influence of bath agents. A tension rig (bottom left diagram) was used for the SCC test. The specimens (bottom right frame) made from Ti6Al4V were placed into both fresh and artificially aged degreasing baths (pH<7). The (fracture-mechanical) specimens used had notches and dynamic fatigue cracks. In PER vapors, SCC crack growth occurred into the centimeter range. In TRI vapors, crack growth was only in the millimeter range. It is suspected that the higher boiling point of PER (121 °C) relative to TRI (87 °C) is significant with regard to the differences in crack growth.
As expected, smooth tensile specimens showed no cracking.
These results can be interpreted to mean that, in contrast to vapors, no SCC can be expected in baths of TRI or PER. For parts with no sharp edges, there should be no danger of SCC even in TRI and PER vapors. In summary, it can be said that:
In the case of high tensile residual stresses and sharp edges (e.g. small grinding cracks or hot cracks on welds), dangerous crack growth cannot be excluded, and PER baths should be classified as especially dangerous.
There seems to be a risk of cracking even after degreasing if the tensile stresses are significantly high and the temperatures are greater than 450°C. Thin reaction coatings containing Cl can form on parts in degreasing baths. The thickness of these chloride films, which form in an insufficiently stabilized degreasing bath, was measured at 100 angstrom (Ref. 22.214.171.124-11). This is several times the value in a bath with optimal composition. At temperatures around 450°C, chlorine can be released and have a crack-inducing effect. A similar effect has been observed in common salt fouling (hand sweat, marine environments) on tensile specimens made from titanium alloys (Ill.126.96.36.199-16).
Penetrant testing of thin, degreased metal sheet specimens from welding tests using WIG (TIG) and electron beam (EB) did not reveal any cracking. However, it must be noted that the flat metal
plates only had a thickness of 1mm (Ref. 188.8.131.52.3-11). This makes it doubtful, whether sufficiently high tensile stresses could have built up for SCC to occur.
Experience seems to show that SCC conditions can evidently occur in production processes such as heat treatments or welding. As a rule, chlorinated solvents should not be used to degrease the welding zone (Ref. 184.108.40.206-13). The case shown at the top right is an electron beam weld on a compressor rotor made from Ti6Al4V. Axially oriented SCC fields were observed next to the weld. Their development during the welding process could plausibly be explained by a reaction layer containing Cl. For this reason, many companies from a very early stage prohibited the use of degreasing baths for parts made from titanium alloys (Ref. 220.127.116.11-11).
Baths containing chlorine compounds or contaminated by chlorine represent a potential cracking danger for titanium alloys. They can cause cracking to occur even during subsequent finishing steps, if temperatures above 450°C are reached and there are sufficiently high tensile stresses.
Figure "Hydrogen embrittlement by etching process": Following cleaning (scale conditioning, Ref. N18.104.22.168-15) with 70% NaOH + 30 % Na2CrO4 at 130°C, extreme embrittlement was confirmed on parts made from various high-strength titanium alloys. This was traced back to hydride formation caused by hydrogen absorption in the cleaning bath (Refs. 22.214.171.124-3 and 126.96.36.199-11). In both cases, Ni particles were stuck to the surfaces of the embrittled areas. This can be explained by rubbing against Ni/graphite abradable coatings during operation. In addition, traces of iron were present in all cases.
It is plausible that this type of splashed material may build up on titanium parts during finishing processes. Grinding and milling create suitable sparks and dust. It is also possible that this fouling could have been created by cutting tools, clamping equipment, and contaminated shot. Even bath contaminants that adhere to the part surface can be problematic in connection with hydrogen absorption.
The left diagram shows a well-run compressor rotor blade with splashed material from a rubbing occurrence adhering to the rear edge. After treatment in the bath, this area was embrittled to the point that it could easily be broken away with one`s fingernail.
The right diagram shows the brittly fractured labyrinth tips of a compressor disk. The embrittlement mechanism can be explained as follows. It is well known that TiFe alloys eagerly absorb hydrogen and store it as metal hydride. Therefore, it is likely that TiFe or TiNi particles that formed during the rubbing process during operation present a weak point in the protective oxide layer. Hydrogen can pass through this weak point to the base material and diffuse into the substrate. It diffuses especially quickly into the b-structure of the titanium material at the relatively high bath temperature of 130°C.
Tests have shown that if the adhering splashed particles are removed in a nitric acid bath (HNO3), the subsequent cleaning bath will not have an embrittling effect.
Figure "Damage risks by etching baths": Extremely aggressive agents are required for the etching of superalloys. Deviating process parameters such as overtimes or fouling of the bath can cause dangerous crack-like corrosion (intergranular corrosion = IGC). This dangerously lowers the dynamic fatigue strength. The top frame shows a case in which it was noticed that parts did not sound metallic when struck, but made a dull sound. A metallographic inspection showed extreme grain boundary corrosion, which had a damping effect.
The middle frame shows a case involving new turbine rotor blades that suffered serious corrosion on the inside. The area around cracks between the impingement air bores was especially affected. It is possible that the cracks were created during heat treatment and destroyed the protective oxide layer. This allowed the etching agent to have an unexpectedly powerful corrosive action.
The bottom frame shows a turbine blade that had an unusually rough surface following an etching process. This was due to grains that had fallen out (grain decay) after their grain boundaries were dissolved by the etching bath.
Solutions for Problems and Damages in Baths used for Cleaning, Etching, and Rinsing
Fundamentally, as with other finishing processes, the use of processing baths requires sufficient application-specific testing, insofar as they are not precisely specified by the licenser. The testing must take into account factors such as possible expected deviations in surface conditions, material (Fig. "Etching problems"), series conditions (facilities, number of parts, etc.). If these optimal, specified process parameters are strictly adhered to during the entire bath life, then the high safety demands of aircraft engine construction can be met.
This includes keeping aging-related changes and fouling in the bath within the tolerable limits.
Special attention must be given to the possibility of fouling being transferred between baths (Ills. 188.8.131.52-2 and 184.108.40.206-3). Fouling can cause problems and damages in subsequent finishing steps.
An important prerequisite for damage-free bath treatments is the experience and technical knowledge of the responsible personnel. This includes being able to register unusual appearances of the parts or changes in the bath as indicators of problems (Fig. "Warning signs of unusable baths"). This is an important factor in being able to detect and prevent damages early and to introduce corrective measures.
The safety of a bath treatment also depends on the design engineer and the work preparation. In this regard, the most important problems of the processes should be known and understood.
If parts are made from multiple materials, e.g. due to coating or soldering, the damage possibilities increase due to element formation or differences in material behavior (Fig. "Risk of combined materials in etching baths"). The application of potentially damaging processes (Fig. "Chlorine in process baths causing stress corrosion") should be avoided if possible, especially on parts that are difficult to mask and/or test, such as the inside of turbine blades (Fig. "Damage risks by etching baths").
Baths that use electrical currents require special attention where they contact the part. The type of contact must conform to the requirements of the specifications in every detail (also see Fig. "Importance of galvanic processes for safe operating"). Changes such as wear or damage to insulation are unallowable. Otherwise, there is a risk of dangerous material damage occuring through spark formation in highly-stressed part zones (Fig. "Danger of contact points during electroplating"). If necessary, suspensions and/or contacts must be developed and their safe application verified in part- and series-relevant tests.
Naturally, the more difficult conditions of later repair processes must be considered. An expensive part that cannot be repaired within acceptable work and cost limits represents a high financial risk for the operator. This can be important with regard to the decision on whether or not to procure an engine.
Figure "Risk of combined materials in etching baths": Parts with material combinations place greater demands on the development and application of processing baths. There are many different types of combinations, which can be created in various ways (top diagram):
Different materials joined by welding (“1”): If, for example, a turbine disk made from a cast Ni alloy is welded to a shaft made from heat-treated steel, then element formation and the differences in chemical resistance will limit the use of etching baths. Masking and covers may be necessary. Even during temporary storage during the finishing process, it is important to closely control (e.g. through suitable transport containers and covers) possible corrosive influences (splashed bath agents, condensation water).
Adhesive joints (“2”) are very different from the metallic base material. If the adhesive bond of a porous metal felt is damaged upon contact with a bath, separation can be expected, at the very latest in operating conditions. Without clear separation occurring, this type of damage is almost impossible to detect in series using non-destructive methods. The result is extensive damages (Fig. "Operating influence on adhesive metals joints").
Brazing on compressor guide vanes made from 13% Cr steel (“3”): The copper- and silver-based hard solders form a galvanic element with the base material. Unsuitable baths or splashed water or bath liquid can lead to increased corrosion at the solder transition (Fig. "Corrosion of brazed joints").
Experience has shown that wear-resistant coatings made from tungsten carbide in a cobalt matrix (“4”), which are used on the contact surfaces of titanium parts, can be damaged by unsuitable baths to the point that they are removed completely. Unfortunately, the exact damage process is evidently not yet completely understood, and specific damaging baths are not yet known.
Housings made from light metal alloys (Al, Mg) are often combined with steel inserts. Heat-treated steels are used for threaded inserts, stay bolts, and bearing carriers, while CrNi steels are used for cast-in oil lines. Threaded inserts may also have a silver coating to improve gliding. These combinations tend to element formation. Primarily, the light metals (usually cast parts) and martensitic steels are at risk. Common solutions are removal of the removable inserts and/or suitable masking or sealing. Naturally, in the case of tight tolerances, material removal must be given extra attention.
Elastomers (“6”) as well as filled abradable coatings made from rubber can be damaged by unsuitable baths and/or excessively high processing temperatures. Depending on the coating and bath, it is possible that swelling, shrinking, embrittling/cracking, or separating may occur. Some damages are not always detectable using non-destructive methods. If in doubt, it is necessary to test the baths before serial implementation.
Figure "Warning signs of unusable baths": Because many damage mechanisms (embrittlement due to aging or hydrogen absorption, grain boundary corrosion, damage to adhesive connections) in processing baths such as etching or rinsing baths cannot be easily detected, if at all, during serial production, experienced personnel is especially important for operating and monitoring the processing baths. However, even when damage is discovered, it is highly likely that the damage to largely completed parts will result in very high costs.
Changes to the part and/or bath can be important indicators for possible unallowable effects of the bath on the part. These changes are deviations from the typical, “normal” appearance of the bath or part. Recognizing these demands sufficient experience, and a prerequisite is knowledge of the “normal” state of the bath and part. This type of indicating characteristic will itself not necessarily have a negative effect on the part. If the effects of these changes are not already categorized as harmless or if their origin is not understood, the appropriate specialized technical departments should be consulted.
Typical changes to the part:
The first important characteristic is the state of the part surface before it is placed into the bath. Signs of fouling, such as discoloration, flow marks, or adhering particles, must be taken seriously. For each bath treatment, the part must show the optimal (specified) surface conditions as a prerequisite for the reproducibility (stability) of the process.
If a part displays unusual wetting behavior, it must be assumed that the effectiveness of the etching bath may be compromised. If wetting problems occur in cleaning and rinsing baths, it may indicate the presence of contaminants such as silicon compounds. In this case, there is a danger of these contaminants being transferred to other parts by the affected bath. This may necessitate extensive reworking, for example, before penetrant testing.
- Discoloration: In titanium parts, as long as the colors are shiny metallic blues and yellows, tarnishing can be seen as merely an optical change, and is no proof for high part temperatures, as would be the case in steels, for example. The causes of this tarnishing are thin, non-damaging surface coatings created by chemical processes. While they are not damaging themselves, they may indicate a change in the reactivity of the surface or the presence of certain contaminants.Unusual tarnishing that is observed during slight heating following a bath treatment may indicate stability problems in the bath treatment process.
- In the area around flow marks, the effectiveness of the bath is probably different than on the rest of the part. Depending on the type of fouling, the effectiveness may be increased or decreased in this area. For example, liquid contaminants (e.g. greases or oils) already on the part can reduce the effectiveness of the bath. Flow marks can also indicate surface areas in which the flow conditions in the bath are different. These changes are especially disruptive during segregation etching, as they change the testing results. Bubble marks can cause similar effects as flow marks. They also locally alter the contact conditions of the bath with the surface.
Surface luster: If the luster of a part surface changes, such as taking on a matt appearance, it is indicative of a change in bath effectiveness. Matt surfaces usually indicate corrosion, although this will not necessarily be damaging. In contrast, if a surface that is expected to be matt is shiny in appearance, it indicates insufficient effectiveness of the bath.
An unusual appearance of the surface may be related to aging or fouling of the bath. However, it may also indicate changes in the base material or even mistakes in material selection. In any case, it must be verified that these changes do not indicate damages such as intergranular corrosion (IGC, Fig. "Bath treatment altered by material deviations").
Changes to plates of fasteners and electrical contacts: If there are even minor unusual changes to these parts, the responsible specialized department should be consulted for an evaluation. Special attention must be given to the insides of bores. These surfaces represent highly stressed notch zones that react very sensitively to small flaws (Fig. "Danger of contact points during electroplating"). The slightest signs of overheating and/or spark formation are alarming (Fig. "Danger of contact points during electroplating"). External indicators of this are tarnishing, small craters (Fig. "Dynamic fatigue lowered by electric arcs"), and visible structural changes (in etching baths).
In parts treated in baths with ultrasonic excitation, one must be aware of possible damages to contact surfaces between the parts or between parts and equipment (e.g. baskets). Of course, the necessary equipment should be designed and properly tested before being implemented in a serial production process.
Characteristic bath properties:
Bubbles and/or foam: Recognizable characteristics include the amount, temporal progression, and distribution on the part.
If confusion in bath or part selection (wrong material) has already been ruled out, unusual appearances could indicate changed activeness of the part surface and/or the bath. Mistakes related to mixing up parts become more likely in cases in which there are different part series with different materials and/or coatings, but an identical appearance.
The increased aggressiveness of an etching bath can lead to unexpectedly heavy material removal and selective corrosion (grain surfaces, grain boundaries). In aged etching baths, especially in cases in which decreasing effectiveness is intended to be balanced by longer exposure times, material removal may be concentrated on the more easily soluble grain boundaries (e.g. in forged Ni alloys), thereby causing intergranular attack. Bubbles (reaction gas, air) can cover the part surface and prevent the desired action from occurring.
Discoloration and/or clouding of the bath fluid are indicators of fouling or aging, and therefore changes to the bath properties. If the bath appearance is unusual, a proper bath analysis should be undertaken in order to prevent it from drifting out of the specification limits.
Reaction times and intensity can sometimes be estimated by the start of bubble formation. It may be possible to detect changes in baths with the aid of continual monitoring of current profiles and/or voltage characteristics. The trends of material removal over several series are another opportunity to detect changes sufficiently early.
Buildup (sludge) on the bath floor is also an indicator of possible bath aging.
If foreign materials are floating on the bath surface, there is a danger that these will be carried to other parts (Ills. 220.127.116.11-2 and 18.104.22.168-3). These contaminants sometimes appear as flow marks on the bath surface (bottom diagram), which may sometimes shine. The fouling should not progress so far as to create “grease drops”, for this could have serious effects on subsequent finishing steps and quality assurance.
Figure "Importance of the drying process for parts": The drying process after cleaning and/or rinsing baths can have a major effect on subsequent finishing steps (see later explanations). Therefore, its importance must not be underestimated. There are usually three different methods to choose from for drying parts during finishing (top frame):
Heating the part in a hot rinsing bath and drying it upon removal. In this case, it must be ensured that the drying time is sufficiently long for drying potential water collection areas such as gaps and hollow spaces (e.g. bores, overlapping on welded parts).
The dwell time to heating should be adjusted to correspond to the part mass and the material cross-sections. Minimum bath temperatures, i.e. part temperatures, must be ensured.
Depending on the specific part, drying processes in furnaces also require minimum dwell times and minimum temperatures. These must naturally also have upper limits in order to prevent damages, for example to elastomers (abradable coatings) or adhesive connections.
If a drying process occurs in a vacuum, the dwell time must be sufficiently long to prevent freezing of the moisture. This effect may seem unexpected at first glance, and is due to the cooling during the pressure decrease, lack of convection, and insignificant heat radiation (at the relatively low temperature).
Drying must occur immediately before the finishing step (middle diagram) that requires the dried part. This should ensure that no further moisture (e.g. condensation water) can settle on the part. The dryness of the part also depends on the surrounding conditions such as temperature, air humidity, and the prevention of splashed liquids.
Water residue can influence finishing processes and cause dangerous damage to parts even in very small amounts, such as a film of condensation water. A primary concern is the danger of hydrogen embrittlement when welding martensitic steels. Examples of these are compressor housings in older engine types and turbine outlet housings in modern fan engines (Fig. "Weld quality by cover gas").
Water in cracks can compromise the sensitivity of penetrant testing.
A film of water can affect the hardening process of synthetic resins and elastomers (adhesive joints, abradable coatings), and can impact their strength properties and bonding strength.
Moist parts can compromise finishing processes that occur in sealed chambers. This applies to processes that require a very high vacuum quality (e.g. electron beam welding, heat treatments) or gas quality (heat treatment, vapor deposition). If this requirement is not met, there are a wide variety of process-specific flaws and damages that can be caused by even a very small amount of moisture (also see Fig. "Weld quality by cover gas").
Figure "Hydrogen absorption during etching process": Martensitic steels (case hardened/heat treated steels) and titanium alloys (Fig. "Hydrogen embrittlement by etching process") can absorb hydrogen in etching baths (Ref. 22.214.171.124-16) and galvanic baths (e.g. Cr and Ni coatings), which will result in embrittlement. In the case of titanium alloys, this concerns the formation of brittle hydrides. The diagrams apply to steels. This embrittlement depends on diffusion processes (Volume 1, Chapter 5.4.4). Embrittlement results in considerably reduced life spans in notched specimens under static loads. A short-time load such as a notch impact test is not suitable for determining hydrogen embrittlement in steels. The danger of hydrogen embrittlement can be avoided through the use of a heat treatment following the bath treatment in question. This process is known as disembrittlement, although it is actually preventive, and must be done before the onset of embrittlement. The literature also uses the confusing term “degasification”. This process is not very plausible, at least in those cases in which a gas-impermeable coating (e.g. Cd) prevents gas from escaping from the base material. The fact that a disembrittling effect occurs during heating can be explained by improved distribution of the hydrogen and so-called irreversible hydrogen traps (Ref. 126.96.36.199-16) in the metallic lattice. These prevent the movement of the hydrogen in the lattice, which prevents the hydrogen from diffusing and recombining to flaws.
The minimum temperature for effective disembrittlement is 170°C (bottom diagram). Higher temperatures are safer and should be used if the heat-treatment state of the part allows. Depending on the temperature level, the time at temperature should be several hours.
The length of time between the bath treatment and disembrittlement is also very important. It should not be longer than a few hours. Otherwise, there is a risk of irreversible damage occurring through hydrogen that is trapped in the atomic lattice and/or recombined. If embrittlement has already occurred, it can no longer be “undone” through heat treatment.
Depending on the stress levels, the top left diagram shows the influence of time at a temperature of 150°C on the life span of tensile notch test specimens made from a very high-strength (i.e. highly-sensitive) steel. The embrittlement effect is still present even after a period of 24 hours. This supports the experience-based assumption that effective temperatures should be considerably higher. The effectiveness of disembrittlement is also highly dependent on the loading level, i.e. the amount of hydrogen that has diffused into the material. This is reflected in the curves for 150°C and 480°C (Ref. 188.8.131.52-17).
The left diagram shows that as hardness/strength and hydrogen content increase, brittleness also increases. Experiences with high temperature-treated or hardened parts such as bolts (Ref. 184.108.40.206-16), threaded inserts, and springs verify this (diagrams, also see Volume 1, Ill. 220.127.116.11-1). High-strength steel parts, which are understandably exposed to correspondingly high load levels during operation, have especially sensitive reactions to hydrogen embrittlement.
The success of a disembrittlement process is decisively dependent on keeping the time period between hydrogen absorption and disembrittlement as short as possible. The specified time frame must be adhered to.
18.104.22.168-1 ASM “Metals Handbook”, “Volume 5 - Surface engineering”, ISBN 0-87170-377-7, 1999, “Alkaline Cleaning”, “Solvent Cold Cleaning and Vapor Degreasing”, “Ultrasonic Cleaning”, “Acid Cleaning”, “Pickling and Descaling”, pages 3-54, 707, 753-757, 835-838.
22.214.171.124-2 P.Adam, “Fertigungsverfahren von Turboflugtriebwerken”,Birkhäuser Publishing, 1998, ISBN 3-7643-5971-4, pages 149,150, 242, 244, 254-260.
126.96.36.199-3 ASM “Metals Handbook Ninth Edition”, “Volume 11 - Failure Analysis and Prevention”, ISBN 0-87170-007-7, 1986; B.E.Wilde“Stress-Corrosion Cracking”, pages 203-224; C.D.Kim, “Hydrogen-Damage Failures”, pages 223, 224, 245-251.
188.8.131.52-4 W.N.Schmitz, “Selecting the `Best Buy' Coating Removal Process”, pages 1-11. (3558)
184.108.40.206-5 S.W.Kandebo, “GE Win Signals Entree Into F-15 Business” , periodical “Aviation Week & Space Technology”, April 29, 2002, pages 27 and 30.
220.127.116.11-6 R.Völler, “Einflüsse des Kühlschmierstoffs beim Normal - und Hochgeschwindigkeitsschleifen”, periodical “Schmiertechnik und Tribologie” Volume 21 Issue 4, 1974, pages 75-79.
18.104.22.168-7 A.Rossmann, “Rissbildung an galvanotechnisch behandelten Bauteilen, Teil A. Entstehung und Vermeidung von Rissen”, periodical “Metalloberfläche” 35 (1981), 10 pages 390 - 396.
22.214.171.124-8 W.N.Schmitz, “Selecting the `Best Buy' Coatings Removal Process”, firstname.lastname@example.org, 1997, pages 1-11.
126.96.36.199-9 H.Simon, M.Thoma, K.Maier, “Dampfentfetten von Bauteilen aus Titanwerkstoffen mittels Trichloräthylen und Perchloräthylen” periodical “Metalloberfläche” 33-1979, 3 pages 95-98.
188.8.131.52-10 H.Simon, “Oberflächenreaktionen an Titanwerkstoffen”, periodical “Metalloberfläche” 5-1982, 10 pages 211 - 217.
184.108.40.206-11 M.Thoma, “Hydrogen embrittlement and Stress Corrosion Cracking of Titanium Alloys Caused by Cleaning Processes”, SAE Technical Paper 850709, Proceedings of the “21st Annual Airline Plating and Metal Finishing Forum”, Atlanta, Georgia, February 18-21, 1985.
220.127.116.11-12 H.Speckhardt, “Beitrag zur interkristallinen Korrosion rost- und säurebeständiger Stähle”, periodical “Metalloberfläche, Angewandte Elektrochemie” 26-1972, Issue 8, pages 851 - 290.
18.104.22.168-13 Titanium Metals Corporation (Timet), “Properties and Processing of TIMET Al6-4”, www.timet.com, pages 1-6.
22.214.171.124-14 J.B. Cotton, “Stress Corrosion of Titanium Alloys”, Proceedings AGARD No. 1 of the conference “Stress Corrosion Cracking in Aircraft Structural Materials”, 1967, pages VI-3 to VI-17.
126.96.36.199-15 H.Simon, “Oberflächenbehandlung im Flugtriebwerkbau”, periodical “Metalloberfläche” 30 (1976) 12, pages 557-565.
188.8.131.52-16 K.H.Kloos, St.Beyer,, R.Landgregbe “Wasserstoffinduzierte Sprödbrüche bei Schraubenverbindungen - Ursachen, Einflüsse, Abhilfemaßnahmen”,“VDI Report No. 766” (1989), pages 211-242.
184.108.40.206-17 P.Bünger, H.Simon, “Elektroplattieren und Stromlos Plattieren von Stählen hoher Festigkeit unter dem Aspekt ihrer Wasserstoffversprödung”, (1979).