Aeroengine Safety

Figure "Brazing applications in engines": Especially in older engine types, brazing was done with unrelated solders. The motivation for the use of this soldering technology was primarily cost savings relative to integral milled constructions, and the feasibility of the latter. Modern digitally controlled chipping machines and increased strength requirements make integral designs such as cast parts and machined forged parts more advantageous. This is also true for the extensive use of highly weldable titanium alloys. The solders have good wettability, but are brittle (Ref. 16.2.1.4-13).
Due to the low heat resistance of unrelated brazing solders relative to the base material, they were primarily used in the compressor and attachments, such as pipes of the oil, air, and fuel systems. The joined materials were low- and high-alloy steels. In newer engine types, soldered constructions are again used in built compressor stators.

In the following, some typical uses of brazing are treated separately:

“A”: This is a front bearing carrier in the inlet housing of a small gas turbine, built from steel sheets, cast, and forged parts joined with copper or silver solder. The bearing struts/inlet guide vanes are hollow and carry de-icing air that comes out of openings on the rear edge when necessary.

“B”: Compressor stators (Ref. 16.2.1.4-9) were usually built from steel sheets joined with copper solder in older engine types. However, the high operating temperatures of modern engines require either very expensive gold-based solders (Ref. 16.2.1.4-16) or high-temperature Ni solders made from related materials (Fig. "Examples of high temperature brazings"). High-temperature solders have poor gap-filling qualities and tend to embrittle when the soldering temperature is kept low. However, the temperature may need to be kept low to avoid exceeding the damage threshold of the forged materials (dynamic fatigue strength, Fig. "Weak points of braze joints").

“C”: These compressor guide vanes are made from Cr steel (13% Cr) joined with copper or silver solder. This blade type has performed very well in the compressors of an older fighter aircraft engine of which many have been built. One problem was a tendency for element formation resulting in increased corrosion at the solder transition to the base material (Fig. "Corrosion of brazed joints"). The solution was cathodic corrosion protection with an inorganic coating (an aluminum powder-filled coating with ceramic bonding). An additional problem was the failure of solders in the root box under dynamic loads. In the “softened” structure, dynamic fatigue cracks could form in the base material.

“D”: The compressor stator of the radial compressor of a small gas turbine made from CrNi steel joined with silver solder. The flat design promoted dynamic fatigue fractures where the guide vanes were soldered to the side walls (Volume 3, Ill. 12.6.3.3-18), and in soldered fastening bolts.

“E”: Air duct made from CrNi steel, joined with silver solder. In this case, the problem was verifying and producing the required solder quality (minimum bond share, Fig. "Bond face of braze joints").

“F”: Soldered wear resistant layer with hard material particles on the tip of a compressor rotor blade. The problem is the desired distribution of the hard material particles and the solder in order to ensure optimal dynamic strength in the blade tip, which is under high dynamic stresses (Volume 2, Ill. 7.1.4-13).

“G”: Distributer for an oil jet made from CrNi steel (CrNi 18/9) joined with silver solder. This construction design is marked by a large stiffness jump at the solder joint, and therefore tends to fracture under increased dynamic loads (Ills. 16.2.1.4-17 and 16.2.1.4-18).

“H”: The carrier (low alloy steel) of the abradable coating on a labyrinth in a small gas turbine. A metal felt is attached with silver solder. In this case, finishing must solve the problem of securely soldering a porous material (Fig. "Brazing porous coatings"). Non-destructive testing of the solder quality is also problematic (Fig. "Methods of qualitative bond strength testing").

“I”: Carrier (low alloy steel) of an abradable coating made of silver solder. This was frequently used in older engine types of all sizes, but is almost never used today. This development is probably due to the potential danger of liquid metal embrittlement (LME, Fig. "Cracking due to metal melt drops"). This occurs at the labyrinth tips in connection with rubbing, and in hot parts as the result of buildup of coating wear product. An additional problem is the coordination of the thermal strains between the solder layer and base material. If this is not satisfactorily done, it can cause the coating to break off and create extensive safety-relevant consequential damages (Fig. "Separation of brazed coatings").

Figure "Examples of high temperature brazings": High-temperature soldering with related soldering materials are indispensible for hot parts, even in modern engines. They are used in areas where welding causes too many problems. This is the case, for example, when the materials are poorly weldable (e.g. Ni-based cast material), the cross-sections are too filigreed, or coatings must be applied (e.g. wear-resistant layers with hard particles). In the following, typical examples of high-temperature soldering are introduced:

“1” Turbine blade segments: In order to improve stiffness and mountability, as well as to make use of the easier production processes for single blades (casting, boring cooling air holes), blades, especially those with complex inner cooling air structures, are soldered into segments on the shrouds. In this case, there can be a problem with reactions of the solder to subsequent diffusion coating (Ills. 16.2.1.4-8 and 16.2.1.4-16). Solder joints must be configured and positioned in such a way that failure of the joint, e.g. due to overheating or thermal fatigue, does not cause the entire part to fail.

“2” Wear-resistant layers on turbine rotor blade tips: In order to minimize tip clearance losses, especially at the tips of blades in the high-pressure stages, hard particles (cubic boron nitride, silicon carbide) are soldered on for the first rubbing incident. They wear down the opposing hard ceramic abradable coating (zirconium oxide; Volume 2, Ill. 7.1.4-14). A special problem in this context is an unallowable reaction of the particles with the solder at the high operating temperatures.

“3” Turbine seal segment with a honeycomb seal: These static parts are located opposite the turbine blade tips or the labyrinth tips of blade shrouds. The seal structure consists of thin metal sheets (honeycomb) and is soldered on the whole surface. There are specific problems such as insufficient bonding due to warping, a risk of embrittlement due to solder diffusion, and the weakening of the metal sheets through solder erosion (Fig. "Embrittling by solder diffusion")

Figure "Influences at the thrength of brazings": The strength and quality of soldered joints are determined by influences related to the materials and processes. The soldering gap (Ref. 16.2.1.4-10, also see Fig. "Bond face of braze joints") and the topography of the joining surfaces (Ref. 16.2.1.4-11, Fig. "Selecting a solder type") are also important.

Formation of the soldering gap:
The width of the soldering gap is especially important for ensuring that a soldered joint is flawless. The soldering gap influences the strength of the base material and the size of the soldered surface, as well as the tensile and shear strength of the soldered joint (top diagrams). To obtain optimal strength, a minimum gap width is necessary, and is usually about 0.1 mm. Gap widths of a few hundredths of a millimeter are usually too narrow, resulting in soldering flaws. Gaps of several tenths of a millimeter do not improve strength, but rather tend to reduce it (slightly). Therefore, it is important to determine the optimal gap width and to maintain it during the soldering process.

Even if the settings are correct before soldering, the soldering gap can change unallowably during the soldering process. These changes may be caused by thermal strain. This results from uneven heating or differences in thermal strain properties, which may be caused by differences between the base materials being soldered, or differences between the part and the soldering jig. Deformations resulting from changes in the residual stress conditions may also unallowably alter the soldering gap. Uneven heating or cooling can also create powerful residual stresses in the solder. Thermal strain differences that are “frozen” by this process are capable of seriously lowering the (dynamic) strength of the joint or causing unallowable deformation during finishing and/or operation.

Too narrow soldering gaps:
These can have several different negative effects.
If the amount of flux reaching the joining surface is insufficient, it will result in insufficient bonding, because the oxides are not completely removed (Ref. 16.2.1.4-18). The dissolved oxides contaminate the flux, increasing its toughness and surface tension. This promotes flux inclusions. For this reason, in the case of narrower soldering gaps, the soldering gap length should also be shorter (Fig. "Bond face of braze joints"). If soldering is done in reducing gases, when the gaps are very narrow, the gas exchange will not be sufficient to reduce and remove the oxides. Changes in the solder composition due to diffusion with the base material become more pronounced, the narrower the gaps are. This causes an increase in solder viscosity, which leads to increased bonding problems at the ends of the gap.

Too wide soldering gaps:
As the gap width increases, the capillary effect decreases sharply (bottom right diagram, Fig. "Bond face of braze joints"). If the soldering gaps are too wide, they prevent the development of the capillary forces required to transport the solder, especially if the solder must rise against gravity. The unfilled soldering gap
creates hollow spaces. An even gap width with no local widening is desirable. If the solder shoots forward into tighter spaces and surrounds the widened areas, it will result in bonding problems and/or gas/flux inclusions (Ill.16.2.1.4-4). With Ni-based high-temperature solders, gap heights greater than 900 mm can be reached with gap widths below 50 mm and soldering temperatures around 1200°C (bottom right diagram, Ref. 16.2.1.4-6).
Topography of the soldering surfaces:
The attainable bonding strength of a soldered joint also depends on the surface structure of the soldering surfaces. The strength increases with the smoothness of the surface. The surface roughness should therefore be as low as possible (bottom left diagram, Fig. "Problems by very high brazing temperature"). There is no “locking” mechanism as occurs with thermal spray coatings. Grooves (e.g. from machining) across the direction of solder spread and intentionally structured surfaces (cross-knurled or non-directed, Ref. 16.2.1.4-10) hinder the solder flow and should be avoided.

Figure "Bond face of braze joints": Important factors for the quality of a soldered joint, related to gap-filling, are the gap width (top right diagram), contact surface topography (Fig. "Influences at the thrength of brazings"), and reactions between the solder and base material. It must be noted that the wetting of the base material with solder is not the same as the capillary effect of the soldering gap.

The wetting effect of the solder is characterized by the wetting angle (Fig. "Inferring from solder seam contour") and the climbing height (bottom left diagram). It is determined especially by the oxides on the joining surfaces and is responsible for the bonding of the solder. For this reason, oxide-removing pretreatments and oxidation prevention before and during soldering are especially important. Oxide removal is done with the aid of fluxes, especially when soldering in air. They contain combinations of boron, fluorine, and chlorine compounds. These aggressive components can have very dangerous effects on part strength if residue remains in following finishing steps and/or during operation (Fig. "Corrosion of brazed joints"). In contrast, furnace soldering usually occurs in a reducing atmosphere, in a vacuum, or in protective gas.
The flux is changed by the dissolved oxides. This usually results in increased viscosity and surface tension. In unfortunate cases, the flux is not displaced by the solder. This results in bonding flaws in the shape of flux clusters (middle right diagram).
A similar effect can also be seen with the solder. The reaction between the solder and the higher-melting base material occurs in the flow direction of the solder and results in an increasing alteration of the solder composition (top right diagram). This raises viscosity and worsens flow properties and wettability, preventing longer soldering gaps from being completely filled. Lack of fusion occurs especially at the end of the solder seam (in the direction of solder flow). Changes in the solder properties as contact time with the base material increases are also a reason for poor soldering results in repeated (repair) soldering.
Often, the working temperatures must be increased significantly with every soldering cycle. This frequently causes melting of the base material during high-temperature soldering. The melting results in solder erosion (Fig. "Embrittling by solder diffusion"), which can dangerously weaken thin cross-sections. The danger of melting can also occur in part zones that are removed from the solder seam. This is caused by solder components that evaporate from the solder and lower the melting point of the base material.
A special type of material-specific soldering flaw occurs when two different base materials are soldered together. The danger is present even if the two materials are well solderable with themselves. These poorly solderable material combinations include steels joined with materials containing Al, B, or Zr (e.g. Al bronzes). Only very low strength levels can be expected from these solder combinations. The solder absorbs aluminum from the base material through diffusion. Through this, aluminum is transported to the steel surface, creating an aluminum oxide coating that prevents fusion. Possible solutionsfor this include coating the steel parts (e.g. galvanic, vapor deposition) foil diffusion barriers that are soldered in. For certain materials, such as titanium alloys, the joining strength is subject to compromises. The pronounced brittleness of soldered joints on titanium is primarily due to the formation of brittle phases.
Influence of the work temperature on the quality of a soldered joint: The bottom left diagram shows the typical behavior of molten solder, with the climbing height as a measure of wettability, plotted over the temperature. The wettability must not be confused with the capillary effect of the soldering gap (Fig. "Inferring from solder seam contour"). The climbing height increases along with the soldering temperature. Soldering temperatures near the upper limit usually make possible the highest flow paths. Both the solder and the base material, insofar as they are not eutectic alloys, have a melting range. This is limited at the lower end, at which the first phases melt, by the solidus temperature. The liquidus temperature signifies the reaching of the complete melted state. The soldering temperature (work temperature) should be roughly in the middle of the melting range (melt interval) of the solder. However, this can lead to problems if, as is the case with many high-temperature solders, the melting ranges of the base material and solder overlap (Fig. "Causes of brazing flaws").

Work temperature too low:
The part temperature must reach the work temperature of the solder. This is in the soldering range (work range) above the liquidus temperature, i.e. the temperature of the fully melted solder (bottom left diagram). If the soldering temperature is below the soldering range, soldering flaws will occur due to high viscosity, surface tension, and poor wetting (Fig. "Weak points of braze joints"). Although this requirement may seem self-evident, in some cases it is very difficult to meet. For example, if a part made from a forged Ni alloy is subjected to high operating temperatures (e.g. a soldered compressor stator in the rear stages), it may be necessary to use a high-temperature solder made from the same material. The soldering temperature of this solder may cause unallowable damage through occurrences such as recrystallization and grain growth, which reduce dynamic strength. Because the thermal strength of alternative, unrelated materials is not sufficiently high, it may be necessary to work at the lower end of the soldering temperature range. This means that the solder will not have optimal flow properties and minor temperature fluctuations will result in defects. If in doubt, the work temperature must be determined using part-specific test conditions before soldering is used in the finishing process. During re-soldering of flawed soldered joints, the work temperature must usually be raised because the melting point of the solder has increased due to diffusion and evaporation during the first soldering process.

Excessively high work temperatures (overheating) and/or overly long soldering times (overtime):
When soldering with flux on corrosion-resistant steels, the flux will react with the air oxygen and the oxygen bound in oxides on the part surface. If the temperatures are too high or act for too long, unsoluble oxides can be created on the base material, preventing wetting by the solder (Fig. "Inferring from solder seam contour"). The solder then balls up, and adding fresh flux will not solve the problem.
Special fluxes can largely eliminate the described effect.
The maximum possible work temperature (annealing temperature) of a subsequent diffusion annealing process is determined by the base material. Annealing is used to break down brittle phases and to even out alloy components in the solder and to the base material, which must not undergo any unallowed changes to its structure or strength.
An additional problem is damage in the solder seam area in the form of excessively deep diffusion and “solder erosion”. Excessively high temperatures lead to significant changes in the solder and base material. The alloy components of the higher-melting base material that are dissolved in the solder raise the melting point of the solder. Large diffusion zones can embrittle the base material (Fig. "Proprerties of high temperature brazings"). This problem increases with the length of the melting intervals. In this context, the heating rate is especially important, because the molten solder will act on the base material for a longer time. During solder erosion, melting leads to material removal and weakens the cross-section of the base material. This type of damage is especially noticeable in thin-walled parts such as cooling air ducts or honeycomb seals made from sheet metal (Fig. "Embrittling by solder diffusion"). Several tenths of a milimeter of removed material are dangerous for metal sheets in honeycomb seals and turbine stator inserts. If the inserts break, it can result in overheating and extensive consequential damages during operation.

If high ductility is required, gold- or palladium-based solders are often the only alternative.
If the work temperature is in the range of the softening temperature of the base material, it will result in erosion-like material removal from the base material (solder erosion). This causes unallowable weakening of thin-walled parts (Fig. "Embrittling by solder diffusion"). Another type of damage is large melted zones on the part due to excessively high local temperatures or local lowering of the melting point. The result is deformation or cracking (Fig. "Problems by very high brazing temperature"). This danger is especially pronounced in cast Ni alloys with high thermal strength, which typically have a relatively low melting point.

Excessively slow heating:
In a solder with multiple structural components, the lowest-melting one is liquified first. Therefore, this component can enter into the soldering gap before the work temperature is reached. This raises the melting temperature and viscosity of the remaining solder. This may cause the work temperature to be insufficient for filling the soldering gap (top right diagram).

Figure "Inferring from solder seam contour" (Refs. 16.2.1.4-10 and 16.2.1.4-19): The top diagrams show the relationship between wetting and bonding. This is characterized by the wetting angle relative to the part temperature. In the top left diagram, the solder drop solidifies before bonding can occur. In the middle, the solder heats the neighboring part surface above the solidus temperature of the solder. This results in bonding. However, the wetting is not sufficient for the drop to spread, since the surrounding surface is still too cold. Therefore, the wetting of the surfaces by the solder is a prerequisite for a metallic bond. The wetting of a surface also depends on its treatment. For example, it has been observed that surfaces blasted with SiC have considerably better wetting and flowability for Ni solders than surfaces blasted with Al₂O₃ . This can be explained by the load of remaining blasting particle residue in the surface. SiC reacts strongly with Ni melts (Volume 2, Ill. 7.1.4-14). It is also understandable that oxidation and fouling can have a considerable effect on wetting, and even prevent it. In order to minimize these effects, fluxes or suitable soldering atmospheres are used (Fig. "Bond face of braze joints"). The cross-sections (bottom row of diagrams) show the influence of the soldering atmosphere on the wetting angle, and therefore on the soldering process. Through targeted adjustment of the atmosphere, a soldering process with a tight specified work temperature can be optimized.

The purity of the protective gas is also important. The purity may be insufficient at the part surface, even if it has the specified value when it enters the furnace. Powerful vacuum pumps can maintain an apparently good vacuum even when there is unallowable leakage. In this case, there is enough oxygen present for unallowable oxidation to occur. Possible causes for contamination and impairment of the cleaning effect of cover gas annealing include:

• Encroachment of outside air, even during cooling.
  
• Using the furnace for different purposes. These include heat treatment (nitriding, case hardening, diffusion coating) with various auxiliary materials (e.g. covering materials) and atmospheres (nitrogen, carbonizing atmospheres).

The wetting temperature is specific to the base material and is reached when the solder drop bonds. It may be below the work temperature, if the solder still sufficiently locally heats the base material. This soldering is also called braze welding and is only used in engine construction for applying soldered coatings (e.g. abradable coatings for labyrinths) and thermal spray coatings. Understandably, the wetting temperature depends on the thermal conductivity and thermal capacity of the joining surfaces. The bonding temperature is the lower limit of the wetting temperature. The part surface needs to be at this temperature in order to bond a drop of solder that strikes it. The top middle diagram shows this case. The drop heats the base material above the solidus temperature of the solder, wets it, and bonds. However, it cannot spread because the surrounding part surface is still too cold.
Using soldering to join parts is known as fusion brazing depending on capillary action. If the base material has reached the work temperature, the solder will shoot into the gap with sufficient intensity (top right diagram).

It must also be noted that the capillary effect of the gap should not be equated with wetting. In addition to the wetting (adhesive effect), it also depends on the surface tension (cohesive effect) of the molten solder. Therefore, for a capillary effect, the adhesive forces of the wetting of the solder on the surface must be greater than the surface tension in the solder created by the cohesive forces. The force with which the solder is drawn into the gap is called the capillary pressure. This is influenced especially by the gap shape and gap width. The capillary pressure decreases sharply with increasing gap width. This means that overly wide gaps promote soldering flaws (Fig. "Bond face of braze joints").

Figure "Selecting a solder type": The selection of a solder type must consider many criteria. These are determined by the requirements and can be categorized into the following main groups:

• Finishing/soldering process type
• Soldering surfaces
• Design/shape of the solder joint and surrounding parts
• Base materials being joined
• Solders, properties, and operating behavior

These influences and requirements act reciprocally upon one another. This makes solder selection a complex task that requires sufficient experience.
Several of these requirements will now be treated in greater detail.

Finishing and surfaces being soldered: The costs of the solder should not be underestimated. Examples of expensive solders include gold or palladium solders, which are superior to Ni solders in their toughness with high thermal strength and low working temperatures. These properties reduce the risk of unallowable changes to the structure of the base material occurring. For this reason, these solders are used on stators in the rear compressor stages, which are assembled using forged materials (Fig. "Brazing applications in engines"). Requirements for the use of furnaces, in some cases with vacuum equipment, are another cost factor. Other factors include tight gap tolerances and high joining surface quality, which are related to the flowability, gap filling, and gap bridging capability of the solder. These properties (Fig. "Bond face of braze joints") have high requirements for fouling- and oxide-free surfaces, as well as favorable topography. For optimal strength, the allowable roughness must not be exceeded.

Design and configuration of the soldering joint and surrounding area: The configuration of the solder deposit (Ills. 16.2.1.4-20 and 16.2.1.4-21) determines the form in which the solder will be used. In addition, the solder must be suitable for the specific soldering surface and the shape of the soldering gap (Fig. "Controlling gap filling of a brazing"). In foils containing a binding agent, it must be ensured that the volatile component is removed from the soldering gap (Ref. 16.2.1.4-1). The volume lost by the removal of the binding agent must be replaced by feeding in solder from outside. All-metal foils also cause soldering defects because, unlike ductile pastes, they cannot smooth out roughness. Another problem is their poor form-fitting ability, especially on spherically curved surfaces, which is caused by their sheet-like stiffness. In this regard, foils made from amorphically solidified solder material perform better. At the same time, a vacuum will enable the complete evaporation of the binding agent better than a protective gas atmosphere.
Stiff constructions using brittle solders with residual stresses can cause cracking during and after the soldering process. This also applies to later operation. Thermal stresses (thermal fatigue) can easily cause cracking in brittle solders (Fig. "Brazing design of thermal stressed parts").
Depending on the assumed operating loads, the strength configuration in the framework of constructive design determines the required minimum strength of soldered joints. This is decisive for the solder type, the shape of the soldering surfaces, as well as the location of the solder joint on the part. Naturally, one should strive to keep the load levels as low as possible and to prevent peeling loads (Ill. 16.2.3.4-18).
If a solder seam cannot be diffusion-coated, its location on the part may be determined by this requirement (masking possibility; Fig. "Braze damaged by diffusion coaating").

Base materials to be joined: The work range of the solder is also dependent on the soldering temperatures that the base material can tolerate without damage, such as recrystallization, grain growth, or stress losses. The alloy composition prescribes the solder type with consideration of the highest possible strength in operating conditions (e.g. thermal fatigue, creep) with diffusion and solder erosion. High base material strength also makes possible high solder joint strength (Fig. "Influences at the thrength of brazings").

Thermal stress and the modulus of elasticity of the soldering surface determine the residual stresses in the solder during the soldering process and later operation. Unfavorable conditions can promote cracking if brittle phases form in overly thick solder seams (Fig. "Proprerties of high temperature brazings"). It may be that the solder selection must take into account ductility.
In cases in which the base material must be protected with a coating that also covers the solder, its compatibility with the solder must be verified (Fig. "Braze damaged by diffusion coaating").

Solder properties and operating behavior: Solders used in engine construction must be approved in the same way as other materials. This already limits the number of available solders.
On the whole, good flowability and wetting characteristics on as many different materials as possible is desirable. Some reasons for this universality are storage, approval, and logistics. This results in limiting solder types to as few as possible. The use of fluxes in unrelated solders creates a risk of corrosion if any flux residue remains (Fig. "Corrosion of brazed joints"). Another problem is bonding flaws resulting from enclosed flux (Fig. "Bond face of braze joints"). Related high-temperature solders, on the other hand, are worked in a vacuum, in protective gas (e.g. argon), or in active gas (e.g. hydrogen). If possible, the corrosion behavior of the part should not be worsened by the solder. Element formation between the solder and base material should be prevented if possible (Fig. "Corrosion of brazed joints").
The oxidation and sulfidation behavior of the solder and/or surrounding zones should be taken into account in order to guarantee the specified operating life spans of hot parts.
Good solders and their joints are expected to have high ductility and strength. These properties are especially advantageous under peeling loads (Fig. "Peeling of solders and brazings") and thermal fatigue. For this reason, especially in high-temperature solders, ductility should remain over the largest possible temperature range (especially operating temperatures) and over a sufficient operating life (formation of brittle phases?). The soldered joint must be largely free of reactions (diffusion) between the solder and the base material, and structural changes in the solder. Another requirement is the least possible reaction between the solder and base material. This applies to the soldering process and to long-term loading at operating temperatures. This includes the penetration of solder into the grain boundaries (liquid metal embrittlement, LME, Fig. "Cracking due to metal melt drops"). The base material should be dissolved as little as possible, i.e. there should be no solder erosion or damaging structural changes such as the formation of brittle phases.
Additional requirements for solders are good gap-filling capability, even with long gap lengths, and low sensitivity to fouling on the soldering surfaces. When joining porous metals (e.g. metal felts), a sufficient viscosity (toughness) of the melt is desirable in order to prevent too much of the solder from being drawn into the porous metal (Fig. "Brazing porous coatings").
High-temperature solders: The thermal strength of the solder must be determined primarily by the expected operating temperatures.
Nickel- and cobalt-based solders: Adding metalloids such as boron, silicon, and phosphorus makes possible technically useful soldering temperatures between 1000 °C and 1200 °C (Ref. 16.2.1.4-6) for Ni- and Co-based solders. These melting point-reducing additives positively influence the wetting and flow behavior (Fig. "Bond face of braze joints"). However, they are also responsible for the brittleness of these solder joints due to hard material phases (Ills. 16.2.1.4-7 and 16.2.1.4-9).
The fracture strain of these solders is usually less than 1 %. This brittleness is due to solder-specific hard material phases. Improved ductility, and also improved bonding strength, can be achieved in conventional high-temperature solders through the use of long soldering times or subsequent heat treatments that break down the hard material phases. The smaller the soldering gap width (< 50 µm), and therefore the smaller the diffusion paths, the more pronounced this effect is.

Figure "Weak points of braze joints": In the following, weak points are considered to be accepted characteristics that are based on technical feasibility and are accounted for in the design.
Like other technologies, soldered joints have characteristic weak points. These cannot be prevented with sufficient safety during the production process and/or through non-destructive testing.

“1”: Unavoidable thermal strain, warping, and dimensional tolerances can result in variations of the soldering gap. This can also affect the gap-dependent minimum seam strength (Fig. "Influences at the thrength of brazings").

“2”: Depending on the use of flux or gas atmosphere, pores or flux inclusions can be expected. These must be taken into account in the drawing requirements and any non-destructive verifications (Fig. "Bond face of braze joints").

“3” and “7”: Solder diffusion and solder erosion, which is a process-specific occurrence in high-temperature soldering, can reduce the dynamic strength significantly relative to the base material. The thermal fatigue behavior of hot parts can be especially affected by this. For this reason, these solder seams should be located outside of part zones under significant loads.

“4”: When examining a solder seam under sufficient magnification, e.g. in an SEM, it is not uncommon to find shrinkage cavities that are usually stretched in the direction of the soldering gap. These weak points are found in solder accumulations, such as where the gap could not be maintained. One example is the edges of soldered blades in compressor stators (bottom frame). If these microscopic discontinuities are unavoidable below the size of detectability of a binocular with typical magnification (10 x 20 x), they must be defined as weak points (see Note on page 16.2.1.4-17).

“5”: Depending on the gap width, high temperature solders can be expected to form brittle phases, primarily in the middle of the seam, due to their containing elements such as boron and silicon, which lower the melting point (Fig. "Proprerties of high temperature brazings"). This means that the weld seam can behave considerably more brittly than the base material. This promotes cracking under loads in the plastic range and under thermal fatigue. If the base material permits diffusion annealing (recrystallization?, strength losses?), the embrittlement can be broken down.

“6”: When using flux during soldering, it is possible that corrosive attack will occur on neighboring surfaces.
“8”: If a soldered joint must be located in part areas that are under high stress, fail safe behavior and the possibility of simple inspections must be ensured. In case of unavoidable cracking, this should occur in a part area under low stress and dwell in this area for a longer period. These properties must be verified in realistic tests that represent operating conditions.

Figure "Causes of brazing flaws": Soldering flaws, like flaws in general, are different from weak points (Fig. "Weak points of braze joints") in that they are not allowable. They are outside of specifications and design requirements.

The mechanisms and types (top diagrams) are similar to, but more pronounced than those of the weak points dealt with in Ill.16.2.1.4-7. Soldering flaws include crack-like longitudinal splits that cannot be sufficiently reliably detected using serially implementable non-destructive testing methods. Especially those visible using a binocular at normal magnification are unallowable.
The seam formation (wetting angle, Fig. "Inferring from solder seam contour") can provide clues regarding suspected abnormal soldering conditions, such as wetting problems or low working temperatures.

Bonding flaws that reduce the percentage of fused area to 70% of the surface should not be acceptable. This also applies to the seal surfaces of pipes (Fig. "Bond face of braze joints") with flux inclusions, even though these only experience minor mechanical loads. Accordingly, soldered joints under significant mechanical loads are subject to more demanding specifications.
In order to prevent bonding flaws when vacuum soldering, it is necessary to ensure that air and bonding agents will be removed from the soldering gap over the necessary time period (Ref. 16.2.1.4-1).

The lower diagram shows problems that are related to soldering, but do not necessarily occur directly in the area of the soldered joint. These include:

Signs of overheating on the base material: These problems occur during high temperature soldering if the work temperature is close to the solidus temperature of a cast base material with segregations. This results in local melting, which also takes the shape of “exudations” (Fig. "Problems by very high brazing temperature") with a recognizable change to the part surface.

Structural changes in the base material: In forged alloys, these usually involve grain growth or recrystallization at high soldering temperatures over long periods.

In hardened materials, if solution annealing effects with strength losses occur, and cannot be reversed through subsequent heat treatment, the strength potential of the base material cannot be fully utilized. This must be taken into account during the design process.
Solder deposits and fouling of the part surface with solder (solder powder, solder spray). The result may be local melting with considerable strength losses. These flaws are especially dangerous in highly-stressed part zones, such as the fir tree roots of turbine rotor blades (Ills. 16.2.2.3-1 and 16.2.2.3-2). If solder drips on the base material while it is under high tensile stress, it may result in liquid metal embrittlement (LME; Ills. 16.2.2.3-5, 16.2.1.4-13, and 16.2.1.4-14).

Flux residue usually contains aggressive halogen compounds (chlorine, fluorine) to remove oxides from the part surface. Independent of the material, if flux residue remains, it can cause intergranular attack and/or stress corrosion cracking (under sufficient tensile stress; Fig. "Corrosion of brazed joints").

Damage related to coatings and thin-walled structures: Some solders can froth up like cauliflower if they are diffusion coated (Fig. "Braze damaged by diffusion coaating"). This situation can occur during reworking or in case of unfavorable processes. If there are already soldered joints on the part (e.g. metal felt coatings), these can melt again and form flaws. Longer annealing times at the high soldering temperatures can damage thin-walled structures such as honeycomb seals through increased solder diffusion.

Figure "Proprerties of high temperature brazings": In the depicted case, soldering is used to join individual turbine blades to a larger segment (top diagram). High seam quality with a minimum of hard material phases depends on optimal gap width (left detail). If the gap is too wide, it will result in brittle phases that can cause premature cracking during operation (top detail). On the other hand, if the gap is too narrow it will result in bonding flaws (bottom detail).
At gap widths around 25 micrometer, nickel-based alloys form hard material phases (Ref. 16.2.1.4-20) that have a negative effect on the ductility and strength of the joint. The fracture toughness of these solders is in the range of 1% or 2% (Ref. 16.2.1.4-6). The hard phases in the soldering gap can cause large fluctuations of the modulus of elasticity and strength. Their distribution in the solder seam has a decisive influence on dynamic strength, especially thermal fatigue.
Subsequent diffusion annealing can break down the embrittlement. During the annealing, the concentrations are balanced with the base material. Melting-point-reducing additives such as B, Si, and P are diffused. The soldering gap should be smaller than 50 mm in order to keep the diffusion paths sufficiently short. In the middle of the seam, a softer, ductile, diffusion-free zone is
created. Micro-ductile forced cracks spread through this zone. However, complete dissolution of the hard phases cannot be expected. Therefore, a soldering joint will always result in a certain strength loss. Because these weak points are found in the entire solder cross-section, dynamic fatigue cracks in soldering joints can also spread from the inside (Ref. 16.2.1.4-4).
The evaporation of the solder additives also raises the melting point of the solder (Fig. "Bond face of braze joints"). This becomes apparent during any repair soldering.

Strength behavior of soldered joints at temperature:

The behavior of solders under long-term heating without significant mechanical loads.
High-temperature soldered joints usually tolerate operating temperatures up to about 800 °C and short-time overtemperatures of up to 1100 °C. The Cr content of the solder is especially important for its strength and oxidation resistance. Phosphorus and manganese solders, on the other hand, only have low oxidation resistance at very high strengths.

The different compositions of the solder and base material result in concentration gradients of the alloy components, which promotes the diffusion of single elements.

Diffusion tends to act in the direction of lower concentrations, but in exceptional cases it can also act in the reverse direction (“up hill”). Diffusion not only changes the structure of the base material in the contact area of the solder, but also changes its strength properties (embrittlement, reduction of creep strength). This means that after longer heating times, entirely different bonding characteristics will be present.
High-temperature strength of soldered joints (short-time loads at temperature): Normally, increasing temperatures will be accompanied by a steady decrease in strength.

This decrease usually begins earlier than in the base material, although the bond strength is related to the strength of the base material. The greater the base material strength, the better the bond strength with the same solder (Fig. "Influences at the thrength of brazings"). However, sometimes a strength maximum can be observed in a limited range of increased temperatures. This is due to the corresponding increase in hardness in the solder caused by the formation of a special structure. In high-temperature solders made from closely related materials, which only contains small amounts of melting-point-lowering additives, it is possible that the diffusion of these additives will cause the strength of the joint to increase as the operating time increases. The creep behavior of soldered joints is similar to that of the base materials (Volume 3, Ill. 12.5-3): initially, relatively fast creep occurs, and its speed decreases. Next, creep continues at a relatively low and even rate, accelerating again towards the end of the part life. Therefore, like the short-term behavior, the creep behavior of soldered joints depends not only on the solder, but also on the strength of the joined base materials. Often, above a certain temperature, an alloy component will diffuse from the solder into the base material/solder contact surface. During this process, pores will be created, especially on the contact surface. It is not clear how important the Kirkendall effect is in addition to the creep pore formation. The growth of the pores can lead to cracking and fractures. Above about 700°C most high-temperature solders only have minor creep resistance due to the diffusion processes.

Figure "Embrittling by solder diffusion": Soldering honeycombs as a rubbing surface for labyrinths (top right diagram, turbine seal segment) and clearance gap seals opposite blade tips usually requires the use of solder foils (Ref. 16.2.1.4-1). The honeycomb strips are either tacked down by welding them to the previously applied solder foil, or solder powder is spread into the tacked honeycomb strips. A special problem is maintaining optimal soldering gaps with the aid of tacking. If the capillary effect along the honeycomb faces is not sufficient, the solder will rise too high on the honeycomb walls. This increases the risk of larger diffusion zones and solder erosion (bottom right detail). If these parts are subsequently machined using chipping processes, it can result in fractures and cracking of the honeycomb walls (top right detail). This weakens the cross-sections of the honeycomb walls in several different ways. Another problem is increased oxidation at the transition between the solder/base material in hot gas.

Embrittlement due to diffusion and the formation of brittle phases can also occur later during operation. Damage is primarily dependent on the material temperature. The thin cross-sections of the honeycombs are understandably especially threatened (Volume 2, Ill. 7.1.3-14).

Figure "Problems by very high brazing temperature": The high work temperature of high-temperature solders means that there is a danger of overheating of the base material. The result of temperatures only slightly below the melting interval of the base material is melting in areas with segregations, or unallowable structural changes and hot cracks may occur. In the melting interval of the base material, the entire part can melt. Rapid cooling can further increase cracking. Unusual diffusion in the soldering zone (top detail) can have an embrittling effect. In forged parts, grain growth and re-crystallization can be expected. The heating and cooling, as well as different thermal strains in different materials being joined, can induce dangerous thermal stresses with the risk of cracking (bottom detail). The depicted example shows the joining of a shaft made from low-alloy steel with a turbine disk made from cast Ni-based alloy.

Thermal strain differences can change the soldering gap in a way that makes it impossible for full-surface bonding to be achieved. The heating rate can be strongly influenced by the reflective behavior of the materials (absorption and radiation). Relatively minor differences in the roughness or tarnishing can result in temperature changes that are unfavorable for the soldering process.

Figure "Brazing porous coatings": Due to their porosity, metal felts tend to have a blotting effect, since they suck up the solder. In order to achieve a good soldered joint with full-surface bonding between metal fibers and the base material, without excessively penetrating the metal felt (top left diagram), the following measures are necessary:

• The full surface of the metal felt must be pressed against the base material during the soldering process. This cannot be ensured if the metal felt yields at the operating temperature due to creep and/or lifts off of the base material due to thermal strain difference between the metal felt and the base material, or due to warping caused by thermal stress. Preventing this from occurring requires jigs with sufficient elasticity, good dimensional stability, and tuned thermal strain (materials, cross-sections), which are configured for each specific task. If the solder only spreads across the base material and does not come into contact with the metal felt, the result is a lack of fusion (bottom right diagram).
  

• Correct amount of solder with suitable viscosity and wetting properties. Excessive infiltration of the metal felt can unallowably worsen its abradable properties. In this case, there is increased danger of hot cracking occurring in the tips during labyrinth rubbing (Volume 2, Ill. 7.2.2-9.1), and it can even result in catastrophic failure due to overheating and smearing (Volume 2, Ill. 7.2.2-4).

• Suitable temperature control in the metal felt and base material. It is problematic if the metal felt is heated too rapidly relative to the base material, e.g. through radiation, and reaches the wetting temperature too early. This creates a danger of the felt absorbing the solder and preventing fusion with the base material from occurring (bottom left diagram).

Figure "Surface fouling by soldering paste": Solder characteristics such as wetting and diffusion mean that solder residue, such as solder powder or solder paste, has material-changing effects beyond the actual soldering zone. Solder residue can result from careless application of solder. Solder powder can be spread onto other parts as dust by air currents, improper removal, or vacuuming (Ill. 16.2.2.3-2).

The danger of local material changes due to solder carries a risk of strength losses and embrittlement. In addition, solder erosion and form notches can develop around these areas (Fig. "Cracking due to metal melt drops"). The melting of the solder residue can occur during the soldering process, or later, such as during diffusion annealing.
A special danger is present when damages occur in highly-stressed, life span-determining part zones. A typical example is shown in the top diagram. In this case, damage occurred in the fir tree root of a turbine rotor blade (Fig. "Operation behaviour of parts by surface fouling"). Due to the tight dimensional tolerances, repair (Fig. "Minimizing scrap rates throuch reworking") of this type of flaw in the machined fir tree roots is usually not possible. It is difficult to estimate the potential depth of the damage. If uncontrolled reworking is attempted, it will result in a flaw that is very difficult to verify using non-destructive methods.
The bottom left diagram shows the back side of a radial turbine disk with raised, pinhead-sized flaws on the machined surface, as well as on the cast surfaces. A metallographic cross-section (bottom right diagram) showed recrystallization that clearly penetrated into the base material in the flaw zone. It can be assumed that highly heated machining chips of the same material caused this effect when they remained on the part surface during high-temperature soldering (Fig. "Problems by very high brazing temperature"). It is possible that melting-point-lowering solder additives may have diffused, lowering the melting point of the thin, rough chips. The fine grain can be expected to have lower thermal strength (creep resistance). The acting depth shows the limited repairability. If uncontrolled reworking is done in this area, it is certain that flaws will remain that are very difficult to detect with non-destructive methods.

Similar effects, although more superficial than deep-acting, can be seen in the region of melted segregations on cast Ni parts, such as this part.

Figure "Cracking due to metal melt drops": Splashes and surface fouling through solder represent a potential danger that must not be underestimated.

A special damage type related to soldering and welding is liquid metal embrittlement (LME). This is a type of material damage in which a wetting metal melt penetrates into the hard (“suitable”) contact material. This occurs very rapidly along the grain boundaries (it shoots in), and in steels along the former austenite grain boundaries (Ills. 16.2.1.4-15, 16.2.2.3-1, and 16.2.2.3-5).
Through this explosive effect, embrittlement, strength losses, and cracking can occur even in surprisingly thick cross-sections (top diagrams). One of the best known occurrences is the penetration of copper, as a pure metal or alloy, into steel at temperatures above the g-a transition. However, high temperatures are not absolutely necessary for liquid metal embrittlement, as can be seen in the action of cadmium on steel (350°C) and mercury on brass (at room temperature).

Liquid metal embrittlement can only occur under certain conditions:
The materials, which are wetted with melted metals, must be subject to tensile stresses greater than a minimum level determined by the material combination. For Cu/steel combinations, tensile stresses greater than 150N/mm²are apparently necessary.
For example, liquid metal embrittlement has been observed in the following metal combinations (Fig. "Damages by metallic surface fouling from finishingfinishing"):

• Ni alloys: silver, gold (Fig. "Liquid metal embrittlement (LME)")
• Steels (ferritic): bronze, Cu, brass, Pb, Cd, Ag, Sn, Zn
• High-alloy austenitic steels: Pb, Zn, Cu, Ag, alloys containing Cd.

LME can be caused by very small amounts of metal, such as solder splashes, metal pigments in coatings, smeared metal films (tools, brushes), and metal dust.
The extreme LME-sensitivity of forged superalloys such as Waspaloy is largely unknown. Under the influence of solders made from Cu, gold, or Ag, catastrophic cracking can occur even with only minimal macro-stresses (Fig. "Liquid metal embrittlement (LME)"). It is possible that this occurrence must be seen in connection with micro residual stresses in the grains (type 2 residual stresses) resulting from thermo-mechanical forming (Fig. "Flaws in forged rotor disks").
A remedy for LME is buffer materials in the shape of coatings. One possibility is galvanically applied nickel coatings. Reliably safe prevention of tension residual stresses is usually not possible. One effect that is probably not related to LME, but has similarities at first glance, is cracking in high-strength Ti alloys (e.g. Ti6A14V) in the contact zone where they are pressed against a cadmium-coated surface at temperatures above 150°C. In this case, as well, a relatively high minimum tensile stress of about 1000 N/mm²seems to be necessary. It is possible that hydrogen embrittlement will also have an influence.

Another type of material damage related to soldering is intercrystalline embrittlement (Ref. 16.2.1.4-21) in nickel alloys when using flux. A brittle grain boundary coating forms, becoming deeper with increasing time and temperature. In five minutes at 700 °C, silver solders will not penetrate farther than 0.05 mm. However, even this thin coating can result in micro-cracking during cooling, due to its extreme brittleness. The result is a considerable loss of dynamic strength. Possible solutions are coatings (e.g. galvanic copper-plating) or special fluxes. These fluxes contain traces of a suitable metal that creates a bonding protective film on the soldering surface at the work temperature.
The bottom diagrams show examples of the microscopic appearance (recreated SEM images) of splashed drops or deposited particles. One can see that solder erosion can lead to sharp notches at the edge of the melting zone (bottom left diagram). Microscopic examination (SEM, Fig. "Scanning electron microscopy (SEM)") of the appearance and composition of the melted zone can provide important clues for identifying
the origin of the particles and/or the method by which they were deposited. In order to ensure that this opportunity can be taken advantage of as a prerequisite for targeted solutions and risk assessments (e.g. number of affected parts, time frame and location of origination, Ill. 17-11), one must absolutely refrain from hasty reworking (Fig. "Minimizing scrap rates throuch reworking").

Even without signs of aggressive melting, it is possible that significant diffusion of the fouling into the base material has occurred (Fig. "Embrittling by solder diffusion").

Estimating the potential affected depth:

The affected depth roughly corresponds to the diameter of the melting on the recognizable (unaltered) contact surface. If particles are not melted, i.e. not laminar, then the affected zone can be considerably larger (Fig. "Surface fouling by soldering paste"). Ultimately, verification of complete removal of the affected zone at least requires a metallographic analysis (directly on the part or with the aid of an impression, Fig. "Non destructive microscopic inspection").

Figure "Liquid metal embrittlement (LME)" (Ref. 16.2.1.4-18): This is the connection of a fuel line that broke during operation as a result of a dynamic fatigue crack (bottom detail). The dynamic fatigue crack originated in a gold-colored crack created during the production process. A metallographic cross-section (right diagram) revealed that the discoloration was caused by gold solder that was evidently trapped in the grain boundaries of the pipe material (Waspaloy) during soldering. The damage mechanism is obviously liquid metal embrittlement (LME, Fig. "Cracking due to metal melt drops").
Experience has shown that silver solder can also cause this type of damage on Waspaloy (Ills. 16.2.2.3-5 and 16.2.2.3-10.1). In order for such deep penetration of the solder onto the grain boundaries, sufficiently high tensile stress is necessary, in addition to the LME-sensitive solder/base material combination. The tensile stress can be caused by various different factors:

• Warping of the pipe in the joint area, caused by dimensional problems from production.
• Tension residual stresses from cold bending of the pipe.
• A cold-straightened pipe zone may have developed tension residual stresses when it subsequently sprung back.

Figure "Braze damaged by diffusion coaating": The depicted part is a turbine stator segment made from a cast Ni alloy with an Al diffusion coating on the blades. High-temperature soldering was followed by diffusion coating. This requires temperatures near the soldering temperature over a long period. In some solders (high Si content?), pronounced cauliflower-like bulges occurred. In addition to the strange appearance, it had to be assumed that the soldered joint was unallowably affected. Reworking or resoldering were not possible due to the extent of the damaged area.
Diffusion coating following soldering is also connected with damages such as melting and sticking of part zones. The causes of these problems are probably a lowering of the melting point of the solder in the coating atmosphere (Al absorption by the solder).

Figure "Corrosion of brazed joints": If fluxes are used, it must be strictly ensured that no flux residue remains on the part after soldering. Otherwise, damage can be expected in subsequent finishing stages, storage, and operation.

Corrosion on solders and soldered parts:

Similar appearances as occur during the dezincification of brass, i.e. selective corrosion, can occur on silver solders containing Zn or Cd (e.g. AgCuZn, AgCuZnCd) in acidic water. The solder becomes a sponge-like mass with no Zn/Cd components. This means that it has no notable mechanical strength. Solutions include Zn- and Cd-free solders containing Cu, Ag, or phosphorus. Selective corrosion also occurs in phosphor bronzes (Cu/Sn alloys with P additives) in sulfureous corrosive media (e.g. condensation in industrial atmospheres). The phosphoreous components are subject to more pronounced attack, and the strength of the soldered joint is decisively weakened. Remedies for this damage mechanism are protective coatings (e.g. lacquers).

Dried flux residue can form a very aggressive electrolyte when combined with condensation water. Even “stainless” Cr-Ni steels cannot resist this, especially when they have been sensitized by a soldering process. This type of intercrystalline corrosion at the solder seam transition of an oil nozzle and a distributer is shown in the top diagram, and was probably caused by flux residue. In this case, intergranular attack occurred, during which the entire grain boundaries around individual grains were dissolved, causing grains to fall out (top detail). The result was a dynamic fatigue fracture that was probably promoted by the stiffness notch at the soldered joint. If the loose oil nozzle falls between the gears of the gearbox, or if a bearing overheats due to a lack of oil, extensive safety-relevant consequential damages can be expected.

In brazed steels with no nickel content (e.g. Cr steels), element formation with the solder results in increased pitting corrosion on the base material near the solder seam. Another typical type of corrosion is intensive corrosion in the contact zone of the solder and base material, leading to the separation of the solder. This corrosion-related separation or detachment of solder is observed during operation, for example, in compressor blades made from Cr steels (bottom diagrams). If soldered joints are destroyed in this way, in addition to the general loss of strength, the vibration behavior of the part may also change. As a result of the reduced stiffness, resonances may occur due to the lowering of the vibration frequency. Shifting of the loads in the part may also lead to overloading. A typical example is compressor stator vanes in older engine types, which were assembled using solder. If this type of blade fractures, extensive consequential damages can be expected. These damages are extremely similar to bonding flaws (!) and can occur in a very short time (days) in condensation water or humid air. In this case, as well, lacquer-like protective coatings can offer a solution. In some cases, subsequent galvanizing or lead-coating can be used.

Solders containing nickel are largely immune to this type of corrosion. However, there is a special danger when the solder alloys with the base material. If this process changes the composition of the solder in the soldering gap, the corrosion sensitivity will increase in the direction of flow. One remedy can be nickel-plating of the parts before soldering. However, in this case the bond strength between the nickel coating and the base material determines the strength of the joint. Therefore, this may have to be improved or guaranteed with the aid of special measures (such as diffusion annealing). Corrosion damages to solders and base materials are also often observed in unsuitable cleaning baths during overhauls of the parts.
During operation, pitting corrosion occurs around the soldered joints of built compressor blades made from Cr steels due to element formation with the solder, even without flux residue (bottom diagrams). This corrosion can also occur during improper storage during the finishing or repair process, if there is no appropriate protective coating. The corrosion pittings act as dangerous notches and increase the risk of dynamic fatigue cracks. A proven remedy for this type of operating damage is the use of inorganic Al-powder-filled lacquers with cathodic action.

The middle diagram shows that flux residue can also cause stress corrosion cracking (Ref. 16.2.1.4-18). In this case a pipe made from CrNi steel was affected. Transcrystalline SCC penetrates through the pipe wall (middle diagram). The damage was discovered as a leakage during a pressure test. Micro-analysis revealed fluorine concentrations. These resulted from residue of the flux that had been used.

Figure "Peeling of solders and brazings"(Refs. 16.2.1.4-9, 16.2.1.4-10 and 16.2.1.4-12) The specific strength behavior of a soldered joint must be taken into account by the design engineer.
Fundamentally, soldered joints should not be placed in part zones under high static and dynamic loads. It must be assumed that soldered joints will generally not attain the same creep resistance (thermal strength) and dynamic strength as the base material. This also applies especially to the LCF strength/thermal fatigue behavior. A fracture in the base material during a tensile test (usually at room temperature!) is no verification of the safety of a soldered joint.
Favorable soldering-specific design is a prerequisite for satisfactorily safe operating behavior:

Soldering-friendly designs are especially important in parts under dynamic loads. Examples include

• Dynamic fractures in pipes and oil nozzles (middle right frame, Fig. "Corrosion of brazed joints") resulting from design-related stress concentrations in the solder seam area.
• Dynamic fatigue of soldered joints in V-stator vanes and stators.
• Cracking in stiffness jumps around soldered fasteners for sheet metal structures (e.g. gas ducts, Volume 3, Ill. 12.6.3.3.-18).

Influences that should be avoided in parts under dynamic loads:

• Peeling loads (herring tin effect); a combination of bending and shear (top diagram).
• Butt joints perpendicular to the direction of loading. * Cross section jumps (stiffness notches) in the solder seam.
  

Naturally, the solder seam must have a melting point that is sufficiently higher than the maximum expected operating temperatures. In order to attain this, one can make use of the raising of the melting point by the diffusion melting point-lowering additives such as B, Si, or P during the soldering process. Failure of a soldering joint due to overheating can be observed in turbine stators. Sensitive areas include soldered joints that are used to combine several stator vanes into segments (Fig. "Examples of high temperature brazings"). Other parts that can be mentioned in this context are hot parts with soldered sheet metal constructions (Fig. "Controlling gap filling of a brazing", labyrinth carriers, gas-carrying parts). In these cases, fail safe designs may be required. This may be an additional mechanical safety. It may also be possible to create a positive fit by flanging the sheet metal around an edge or, in rotation-symmetrical parts, a conical soldering surface (Fig. "Controlling gap filling of a brazing"). It is important that this safety acts in a way (direction!) that prevents unallowable shifting. At the same time, failure should be easily recognizable from the outside (e.g. during boroscope inspections).

Examples for the configuration of soldered joints:

The most desirable are large soldering surfaces under compressive loads (frame, top row). Pressure is preferable to tension loads (fourth row). If the pressure surface is subjected to dangerously high loads in spite of an optimized stiffness transition, an additional positive fit can be included (bottom row, Fig. "Controlling gap filling of a brazing").
If soldered joints must be subjected to tension loads, the soldering area must be increased and it must be ensured that there are no acting stress increases caused by stiffness jumps (second row). Even elastic deformations under tension loads must not cause a peeling effect.
If they are unavoidable, peeling effects should be minimized through an elastic design (third row).

Figure "Brazing design of thermal stressed parts": The example shows the optimization of the configuration of a soldered joint. The part is an air-carrying part near the injection nozzle of a combustion chamber. It is made of a very thin, heat-resistant metal sheet and is subjected to high thermal stress, high-frequency vibrations, and gas forces.
After a few hours of test operation, the original version (bottom left) had cracking in the three ligaments between the pipe and the triangular fastening flange (top left diagram). These were evidently caused primarily by thermal fatigue. The cracking was promoted by brittle phases in the diffusion zone of the soldered joint (top right detail). In order to break down the loads, the solder seam area had to be designed more elastically. Suggested variants are shown in the three details (in the frame).

• The first step was a thin triangular flange that was bent around the solder joint.
• In addition, two spot-welded variants were designed.In one of these, the solder was merely replaced by spot welding.
  

In the second, the flanged area was significantly lengthened and the spot weld was positioned so that the considerably longer flange could absorb and balance the strain.

Figure "Controlling gap filling of a brazing": This case concerns an intermediate stage labyrinth of the gas generator turbine of a small gas engine (middle detail, top left). The outer contour of the sheet metal hub was designed as a cylinder, and the solder was deposited on the outside at the gaps. Because of this, it was not possible to tell from the solder seam whether the entire surface required for the joint had been wetted. After the soldered joint failed during operation between the sheet metal membrane and the inner band of the the stator, the membrane was able to press against the rear turbine disk and grind into the hub area. The result was a disk fracture (Volume 2, Ill. 8.2-22). The rotating surface on the shroud side was clearly not wetted, and it was obvious that no solder fusion had occurred here despite the presence of an outer solder fillet.
An improved design with fail safe potential is shown in the top right detail. In this case, the conical external contour of the membrane and shroud with the flange prevents axial shifting of the membrane, even if the soldered joint fails completely. If the solder is deposited in a ring fillet in the radial ligaments of the shroud, the solder flowing out will show wetting of the required joining surfaces in a visual inspection.
The bottom diagrams depict turbine stator vanes with soldered fastening tabs that are intended to prevent rotation relative to the housing. If these fasteners fail, it can result in catastrophic turbine damage when the rotating stator grinds through the housing wall (Volume 2, Ill. 6.2-9).
In the variant depicted at left, the fastening tab is only soldered onto the outer shroud. External inspection of the bonding surfaces is not satisfactory. In case of gradual failure of the soldered joint, the tab cannot brace itself and relieve the stress on the rest of the soldered joint. The blade will be immediately released after the accelerated failure of the soldered joint.
In the next variant at right, the tab is soldered into a throughgoing axial groove. Therefore, the positive fit can relieve the soldered joint and prevents spontaneous relief of the blade even if the soldered joint fails. However, this does not ensure that the tab will not shift axially rearwards and lose its hold.
If the axial groove is closed at the back, it prevents longitudinal shifting of the tab if the solder seam fails. Of course, this is a relatively elaborate variant. If there is still a danger of a released tab moving radially outward and out of the holding area, the outside left variant would be a possibility. This variant has a slit into which the tab is pushed and soldered. In this case, the solder is deposited at the front end of the tab. The solder flowing out of the front of the gap would controllably indicate that the gap was filled.

Figure "Design for braze quality": The position and shape of the solder deposit influences the seam strength. A prerequisite is that the entire soldering gap is filled. In addition, a visual inspection of the soldering seam fillet should be a simple non-destructive test. This requires the solder to flow outward from an inner deposit. Therefore, the solder should fundamentally be forced to flow from one end of the joining surface to the other. In circular surfaces, the flow should occur from the center towards the outside. In soldered joints with soldering foils that cover the whole surface, there is an increased danger of bonding flaws resulting from unfilled hollow spaces. These bonding flaws are promoted by roughness spikes, gas bubbles, or flux inclusions. In this case, the solution is positioning the solder in the center of the cross-section (top diagrams). This allows an even flow front to form. In mortise joints/blind holes, the solder should be deposited at the base of the bore. The solder can be applied as a ring at the bore opening around the bolt. In this case, a ventilation hole must be created at the base of the bore, and must be open during the soldering process and not sealed off by the standing base of the part.
Figure "Brazing problems by trapped air" shows a case in which this fundamental rule was not observed, which ultimately led to a forced landing.
If possible, the solder should flow in the direction of gravity.
This can be achieved through a solder deposit at the base of the bore, if the bore is directed downward. Falling-out due to atmospheric pressure differences or gravity should be appropriately restricted.
A special problem is presented by the non-destructive testing of soldered joints. Thin-walled parts can often be tested with the aid of X-rays (Fig. "Bond face of braze joints"). The quality criteria is the percentage of the soldering surface in which fusion is achieved.
Testing with ultrasonic methods is possible with favorable soldering joint positioning and part geometries.
However, in many cases the design of the soldered joint and parts is such that none of these procedures can be sufficiently reliably implemented. The possibility of inspection of solder seam quality after the soldering process must already be ensured during constructive design of the soldered joint through suitable positioning and shaping of the solder deposit. For example, in internal solder deposits, an even outflow of the solder along the entire seam length may be a usable external indicator of the acceptable quality of the soldered joint (Fig. "Controlling gap filling of a brazing").
The bottom diagram shows an oil filter housing made from low-alloyed steel. In rare cases, the soldered joint of the inset failed. Evidently, an externally located solder deposit that created a solder fillet was selected. However, this did not permit assessment of the filling of the gap below the solder. This situation promotes soldering flaws of the type shown exemplarily in the adjacent detail.

Figure "Separation of brazed coatings": Abradable coatings in labyrinths can be applied to a carrier ring with the aid of soldering (Fig. "Brazing porous coatings"). However, the solder itself can also act as an abradable coating (right diagram).
The separation of this type of solder coating can lead to catastrophic damages (Volume 2, Ill. 7.2.2-4). Separation of soldered abradable coatings is primarily promoted by poor bonding with the base material. There are various causes for this:
A problem with soldering porous coatings is that the solder is absorbed by the coating before it bonds with the base material (Fig. "Brazing porous coatings"). The same is true for connections made with adhesives (Ills. 16.2.1.5-5 and 16.2.1.5-8).
Bonding flaws resulti from oxidation of the base material surface, flux inclusions, or unfavorable process parameters that deviate from the tested soldering processes. Another possible cause of separation of a soldered coating is shown in the diagram. In this case, the cause of the separation during operation was the selection of an unsuitable coating and finishing parameters that were not optimized. Due to very different thermal strain behavior in the solder coating (silver solder) and the base material (heat-treated steel), extreme compressive stresses were created in the solder coating during operation. After longer operating times and many operating cycles, these caused the coating to separate. They may have acted in combination with intensive rubbing (heating) and poor heat removal to the base material. The damage shown in the right diagram occurred around the separated coating. The result was extensive disassembly and repair with high costs.

Figure "Brazing problems by trapped air" (Ref. 16.2.1.4-22): The fuel controller of the engine in the depicted case used an evacuated metallic acceleration bellow that reacts to the surrounding pressure through longitudinal strain. This axial movement acts as the control value. When the vacuum in the bellow collapsed, the fuel feed was restricted so much that the remaining engine power was insufficient, and the aircraft was forced to land.

The bellow consists of sheets of a copper-beryllium alloy. The double-walled design (bottom frame) ensures fail safe behavior if the outer wall leaks. This type of damage occurred during operation in the form of intergranular attack and pitting corrosion. These were caused by aggressive sulphur and chlorine compounds typical of marine and industrial atmospheres. The inner bellow could not fulfill its securing function because the sealing soldered joints at the ends of the bellow had flaws and allowed leakage air into the vacuum.

The development and shape of the flaws revealed that air was trapped in the gap between the bellows when the bellow ends were soldered, and was not able to escape (Fig. "Design for braze quality"). The soldering temperature made the air pressure rise and prevented flawless filling of the soldering gap. During operation, these flaws led to leakages into the vacuum.