This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Nickel-based superalloys are usually employed for turbine blades because they have excellent high temperature resistance, whilst retaining high strength at operating temperature, which are best suited for aero engine applications.
In this literature review, some important issues are considered. Firstly, a historical review of the development of the superalloys and the driving force are generally discussed. Then the major phases of nickel-based superalloys are introduced, with the effects of chemical composition. Moreover, as a major part of this literature review the changes of microstructure and mechanical properties during the service are discussed in detail. Finally, a general introduction of the material CMSX-4, which is used in the experimental at work and the rejuvenation heat treatment.
Historical development of the superalloys
Superalloys are a group of materials designed to have high performance at elevated temperature. They provide high strength and corrosion resistance combined with good low temperature ductility and excellent surface stability
[1, 2, 3, 4].
Most of superalloys are based on Group VIIIB elements and they consist of various combinations such as Fe, Ni, Co, and Cr, as well as small amounts of W, Mo, Ta, Nb, Ti, and Al . The major classes of superalloys are based on iron, cobalt and nickel, and each is designed for specific applications.
Nickel-based superalloys are the most complex and widely used high-temperature materials. The most famous application is aircraft and land-based gas turbine blades. These superalloys are employed in the hottest sections of the turbines, under the large and frequent loads. This is because they can retain most of their strength even after long exposure times at elevated temperature .
It is important to know the history of these materials, since much can be learned from it. The first nickel-based superalloys, Ni-20wt% Cr alloys, were developed in Great Britain in 1941. Since then, the performance of materials
has improved dramatically, especially in operating temperature. This is because a higher temperature results in improvements to the efficiency of the engine and therefore lower fuel burn . The main superalloys developed in the last 60 years and their operating temperatures are shown in Fig.1.
Fig.1. Evolution of the high-temperature capability of the superalloys over a 60 year period since their emergence in the 1940s .
According to the references [2, 5, 6], there are four types of nickel-based superalloys: wrought, conventionally cast, directionally solidified and single crystal followed in the time sequence of development.
The wrought alloys were very popular in 1940s and 1950s, and were developed for use in aircraft turbine engines. The wrought alloys have excellent high-temperature tensile strength and creep resistance at 650ï¿½ï¿½C, but this decreases sharply at 900 ï¿½ï¿½C. The poor strength and oxidation resistance at high temperature limits the applications of these materials. Although, significant amounts of chromium and cobalt were added for solid solution strengthening, the strength of materials at high temperature is still weak due to
the low volume fraction of gamma prime precipitates (ï¿½ï¿½'), the volume fraction of which determines the high temperature capability .
In order to further increase the operating temperature by increasing ï¿½ï¿½' volume fractions, a casting process was employed in 1950s . The creep performance of superalloys improved due to the use of vacuum casting technologies with the help of high quality and clean alloys. However, the components produced by conventionally casting have equiaxed structures (Fig.2.a), which results in poor creep performance of materials.
The significant improvement in the performance is contributed to the using of the new casting process- directional solidified casting. It is found that the creep strength and ductility of superalloys are improved dramatically by using directional solidified casting, which were developed in 1970s . This is because a columnar grained structure (Fig.2.b) is produced, which removes the transverse grain boundaries and aligns the grains parallel to the blade axis . With this structure, the materials can afford more strength compared with equiaxed structure. However, the grain boundaries are still weak point of materials.
Further improvement is achieved by removal of grain boundaries. In single crystal alloys (Fig.2c), the grain boundaries are eliminated together with grain boundary strengthening elements such as carbon, boron and zirconium, since they are no longer necessary . The first industrial single crystalsï¿½ï¿½PWA1480 were developed in 1970s . With excellent creep properties and ductility, the number of single crystal always has increased dramatically in recent years.
However, cost is a big issue for this material. The price of a modern single crystal turbine blade is several ten times of directionally solidified one. This is not only because of the high cost of rare elements, such as Re, but more particularly the high degree of technological sophistication attained .
Fig.2. Turbine blades in the (a) equiaxed, (b) columnar and (c) single-crystal form .
However, cost cannot stop the new classes of single crystal superalloys are continually being sought by gas turbine manufacturers. The driving force is which the increasing demand for higher efficiency of gas turbines engine leads to rising temperatures and stresses . Today, there are four generations of single crystal superalloys (see Table 1).
Table 1. The generations of single crystal superalloys 
First Generation PWA 1480, Rene N4 and SRR99 Contain modest amount of ï¿½ï¿½' hardening elements, like Al, Ti and Ta
Second Generation PWA1484, CMSX-4 and Rene N5 Contain 3 wt% Re
Third Generation CMSX-10 and Rene N6 Contain 6 wt% Re
Lower concentration of Cr Higher concentration of Al
Fourth Generation MC-NG, EPM-102 and TMS-162 Addition of Ruthenium
Compositionï¿½ï¿½Microstructure Relationships In Nickel-based Superalloys
It is well known that the composition and microstructure are close related. The change of chemical composition directly influences the microstructure of nickel-based superalloys and later the mechanical properties. In order to get excellent properties, it is essential to understanding the relationship between composition and microstructure.
Composition ï¿½ï¿½ Role of The Different Alloying Elements
The chemical compositions have a great effect on the microstructure of nickel-based superalloys. The effects of the alloying elements commonly used in superalloys are shown in Table 2.
Table 2. Effects of the major alloying elements in nickel-based superalloys 
Strengthening Increase in ï¿½ï¿½' volume fraction Grain Boundaries Other Effects
Cr moderate Moderate M23C6 ,M7C3 Improves corrosion resistance;
Promotes TCP phases
Mo, W high Moderate M6C and MC Increases density;
Promotes TCP phases ï¿½ï¿½ and ï¿½ï¿½
Nb high Moderate NbC Promotes ï¿½áï¿½ and ï¿½ï¿½ phases
Ti moderate Very large TiC
Al moderate Very large --
Re moderate -- Retards coasening
C, B, Zr moderate -- Carbides Improves grain boundary strength; Improves creep strength and ductility
Chromium provides the corrosion resistance, whilst strengthens the matrix of nickel-based superalloys. However, the amount of the chromium should be well controlled. This is because an excessive amount of chromium promotes the formation of topologically close-packed (TCP) phases, such as ï¿½ï¿½, ï¿½ï¿½ and R, which are embrittling second phases and may cause potential consequences in application .
Heavy elements such as molybdenum and tungsten are the most efficient matrix strengtheners. They increase the strength of the materials by solid solution strengthening. The problem of these elements is that they increase the density of the components, which is undesirable for aerospace applications. Promoting the formation of TCP phases is another disadvantage of these elements.
Aluminium and titanium are ï¿½ï¿½'-forming elements. The largely increase the volume fraction of gamma prime particles, which are major strengthening of the superalloys at high temperature. Typically, the nickel-based superalloys contain 10%~60% of this strengthening phase . Aluminium not only enhances the strength but also provides the oxidation resistance to the materials.
Cobalt is widely used in nickel-based superalloys, which contributes the oxidation/corrosion resistance and enhances strength as well. This is because the cobalt modifies the solubility of ï¿½ï¿½' and increases the solidus temperature, which leads a higher volume fraction of ï¿½ï¿½' .
Carbon, boron and zirconium are the primary grain boundary elements. They enhance the strength of the grain boundary and improve the creep resistance and ductility of materials as well .
Although a variety of alloying elements are used in nickel-based superalloys, It is very common that most of the nickel alloys contain significant amounts of chromium(10-20wt%), cobalt(5-10wt%), aluminium and titanium (up to about 8wt%), and small amounts of carbon, zirconium and boron [4,9,10].
The Microstructure of Nickel-based Superalloys
The microstructures of nickel-based superalloys are very complex and each phase has different effects on the mechanical properties. Generally, there are
four major phases present in nickel-based superalloys, which can be summarised as the gamma phase (ï¿½ï¿½), the gamma prime precipitate (ï¿½ï¿½'), carbides and borides and other phases.
The Gamma Phase (ï¿½ï¿½)
The gamma phase is a continuous matrix, which is a face-centred-cubic (FCC) nickel-based austenitic phase (Fig. 3.a). It usually contains a high percentage of solid-solution elements such as cobalt, chromium, molybdenum and tungsten . The ï¿½ï¿½ phases are favoured by most gas turbine designers due to their excellent stability even at severe temperature for a long time.
The Gamma Prime Precipitate (ï¿½ï¿½')
The primary strengthening phase in nickel-based superalloys is Ni3(Al,Ti), which is called the gamma prime precipitate. It is an ordered L12 crystal structure (Fig.3b), with nickel atoms at the centres of faces and aluminium or titanium atoms at the cube corners .
FCC (b) L12
Fig.3. The crystal structure of ï¿½ï¿½ and ï¿½ï¿½' .
Fig.3 indicates that the L12 structure is similar to FCC structure. The close match in matrix lattice parameter leads to coherent interface between ï¿½ï¿½/ï¿½ï¿½' when the precipitate size is small. The properties of the nickel-based superalloys are very dependent on the coherency of ï¿½ï¿½/ï¿½ï¿½' interface. The
coherency of the ï¿½ï¿½/ï¿½ï¿½' interface allows homogeneous nucleation of gamma prime precipitate with low surface energy and long-time stability [6, 10] .The coarsening of the ï¿½ï¿½' leads to loss of coherency thus a decrease the performance of materials.
The ï¿½ï¿½' is a unique intermetallic phase. It contributes strength by inhibiting the movement of dislocations. The more interesting thing is that the strength of ï¿½ï¿½' increases as temperature increases. However, it has a limit, after which strength decreases again. This is why the ï¿½ï¿½' precipitates provide the most strength of superalloys at elevated temperature. Furthermore, the inherent ductility of ï¿½ï¿½' prevents it forming being a source of fracture .
Carbides and Borides
Various carbides and borides can form in the superalloys; the types of carbide and boride very depend on the alloy composition and the processing conditions employed. The most important carbides and borides are MC, M6C, M23C6 and M3B2, where M stands for a metal atom such as Cr, Mo, Ta and Ti [6, 10].
The carbide is often found interdendritic spaces and no distinct orientation relationship with the matrix is displayed . The presence of carbides and borides tends to a lower the solidus of the alloy, which limits the heat treatment temperature of materials .
It is believed that carbides have beneficial effects on rupture strength and creep properties at high temperature. This is because the carbides and borides prefer to from at grain boundaries and inhibit the grain-boundary sliding . Therefore, carbon and boron are often regarded as grain-boundary strengtheners. It is well known that carbides influence ductility and chemical stability of the matrix through the removal of reacting elements.
MC carbides usually take a coarse random cubic or script morphology and precipitate at high temperature, but during the heat treatment or service, these carbides begin to decompose and form ï¿½ï¿½lowerï¿½ï¿½ carbides such as M23C6 and M6C. M6C carbides form at 980 ï¿½ï¿½C, while M23C6 carbides form at lower temperaturesï¿½ï¿½around 750ï¿½ï¿½C. These carbides tend to form at grain boundaries, which enhances the grain boundaries strength .
The borides form at grain boundaries, where they block the onset of grain boundary tearing under creep rupture loading, which leads to an increase in creep strength.
An excessive amount of chromium, molybdenum and rhenium promotes the formation of Topologically Close-Packed (TCP) phases, such as ï¿½ï¿½, ï¿½ï¿½, P and Laves phases, which can result in a deterioration of stress-rupture properties .
The name TCP comes from the structure of the phase. A TCP phase has close-packed atoms in layers separated by relatively large inter-atomic distances. Though the TCP phases have similar layers of close packed atoms, the atoms are not as close-packed as nickel atoms [1, 2]. Due to this 'topological' structure, these phases are called Topologically Close- Packed Phases.
TCP phases are undesirable brittle phases, which can form either during the heat treatment or more commonly during service. Usually, they appear as thin linear plates in microscopy. TCP phases have negative effects on the mechanical properties and may cause potential damage for the materials. Thus, the amount of TCP phases should be reduced as much as possible .
Changes In Microstructure During High Temperature Exposure
Microstructure changes or microstructural degradation can occur during heat treatment or more commonly during service. According to the microstructural evolution mechanisms, these changes are very dependent on the thermal exposures and applied loads. The attention of this section focuses on the microstructure evolutions under different conditions.
Dramatic improvements in the high temperature capability of nickel-based superalloys have resulted from the development of heat treatments . The effects of heat treatment can be summarised as follows:
Homogenising: to eliminate the chemical segregation and dissolve the ï¿½ï¿½' and ï¿½ï¿½-ï¿½áï¿½ eutectic into the matrix.
Carbide reactions: to decompose MC carbides to form M23C6 and M6C carbides
Aging: to precipitate ï¿½ï¿½' particles, carbides and other phases for precipitation hardening [13, 14]
Generally, nickel-based superalloys are given two types of heat treatments: solution treatment and ageing. The solution treatment, typically 1280 ï¿½ï¿½C, is used to dissolve the ï¿½ï¿½' and ï¿½ï¿½-ï¿½ï¿½' eutectic resulted from the casting process, whilst eliminating the chemical segregations present from casting. It is also called a homogenising treatment, because it homogenises microstructure and chemical composition. Aging heat treatment, typically 870 ï¿½ï¿½C, is applied to precipitate additional ï¿½ï¿½' and other phases such as carbides and borides on grain boundaries, which strengthens the materials [4, 13, 14, 15].
Chemical segregation during casting leads to a non-uniform precipitate distribution and the formation of secondary phases in the interdendritic region. These significantly reduce the mechanical properties of materials . In order to reduce or eliminate the segregation, a solution treatment is applied, which homogenises the chemical composition and microstructure.
It should be noted that the temperature of solution treatment should be well controlled. It must high enough to dissolve the interdendritic particles and precipitates (above the ï¿½ï¿½' solvus), but not too high to melt the materials (below the incipient melting temperature). Commonly, a solution treatment ï¿½ï¿½windowï¿½ï¿½ is used to identify the temperature range of treatment. The typical CMSX-4 solution treatment window is 1280-1327 ï¿½ï¿½C [4, 17]. For single crystal materials, the elimination of the grain boundaries allows the removal of grain boundary strengtheners such as carbon, boron and zirconium, which increases the melting temperature and homogenises the structure prior to aging .
In order to fully dissolve the segregations, the solution treatment usually involves several steps. Table 3 shows a standard heat treatment for CMSX-4, which comprises a three-step solution treatment and a two-step aging treatment. This reason for this is that the ï¿½ï¿½' particles are dissolved at a lower temperature and shorter time, whereas chemical segregation is much more difficult to dissolve, which requires a higher temperature and much longer processing time. Take CMSX-10 for example, the ï¿½áï¿½ and the eutectic ï¿½ï¿½/ï¿½áï¿½ are completely solutioned at approximately 1340ï¿½ï¿½C, while the chemical segregations will not dissolve until the temperature reaches 1360ï¿½ï¿½C . It needs an extremely long time to eliminate all the chemical segregations. The residual segregations of W and Re in the dendrite cores will lead to an unstable microstructure and possibly the formation of TCP phases . The temperature of solution treatment at each steps are gradually increased due to the increasing of incipient melting temperature, which is contributed by homogenisation.
Table 3. The standard heat treatment for CMSX-4
Material Solution Treatment Aging
CMSX-4 1300 ï¿½ï¿½C for 4 h + AC*
1305 ï¿½ï¿½C for 4 h + AC
1310 ï¿½ï¿½C for 4 h + AC 1080 ï¿½ï¿½C for 4 h + AC
870 ï¿½ï¿½C for 24 h + AC
*AC is air-cooling.
The aging process is a lower temperature but longer time heat treatment compared with solution treatment. The typical aging process for single crystal alloys is holding at 980-1100 ï¿½ï¿½C for 4-16 h . The detail of the heat treatment condition is sensitive to the mechanical properties required.
Usually, there are two aging treatments. A higher temperature treatment, sometimes called a precipitation treatment, is used to optimise the mechanical properties [4, 19]. Another treatment is carried out at a lower temperature to precipitate additional ï¿½ï¿½' particles and other phases such as carbides and borides .
Changes In Gamma Prime Phase
As it is the main strengthening phase at high temperature, the microstructural change associated with the ï¿½ï¿½' precipitates is one of most important changes in nickel-based superalloys, which directly affects the mechanical properties of materials. Therefore, it is essential to understand the microstructure evolution of ï¿½ï¿½' precipitates. The microstructure changes are very dependent on the heat-treatment conditions. Therefore, the changes can be separated into two parts: changes during continuous cooling and changes under isothermal exposure.
Changes during continuous cooling
When nickel-based superalloys are continuously cooled after solution treatment, a variety of changes can be observed under microscopy . Grain size, shape and volume fraction changes are the three main changes observed during the cooling in relation to the ï¿½áï¿½ particles.
Fig.4. The microstructure of AM1 alloy during the precipitation process after a continuous cooling at 0.16ï¿½ï¿½C/s. The right image shows the microstructure obtained after isothermal holding for 40 min at the temperature indicated .
Fig.4 indicates the microstructure evolution during continuous cooling. The coarsening behavior of the ï¿½áï¿½ particles is easily can be seen and the shape of the precipitates is significantly changed [4, 20]. The ï¿½áï¿½ shapes are changed by following sequence:
Sphere ï¿½ï¿½ cube ï¿½ï¿½ octocube ï¿½ï¿½octodendrite ï¿½ï¿½ dendrite (see Fig.5)
Fig.5. Schematic showing the evolution of ï¿½ï¿½' morphology during continuous cooling .
The driving force for the shape changes is the minimisation of the interfacial and elastic energy. When the precipitate is very small, the interfacial energy is prominent, which leads to a spherical shape due to minimisation of the interfacial energy. As the precipitate grows, the elastic energy dominates and a cubic shape is preferred. The octocube structure is formed by the splitting of single precipitate. It is believed that the elastic energy associated with the octocubic morphology is lower than the one associated with the cubic morphology [21, 22]. Further growth leads to the formation of a ï¿½áï¿½ octo-dendrite, which largely reduces the coherency of the ï¿½ï¿½/ï¿½áï¿½ . This process is controlled by the diffusion of solute. With coarsening of the dendrites, the diffusion distance of solute increases, which reduces the growth rate of dendrite and finally forms a stable dendrite in alloys.
In nickel based superalloys, the evolution of ï¿½áï¿½ shape and distribution during the cooling is strongly affected by elastic distortions associated with the ï¿½ï¿½/ï¿½ï¿½' misfit . A larger misfit will introduce a larger elastic energy in the material, which promotes the evolution of ï¿½ï¿½' morphology.
In single crystal superalloys, a high level of rupture strength and creep resistance results from the high volume fraction of ï¿½ï¿½' obtained. The relationship between ï¿½ï¿½' volume fraction and temperature is shown in Fig.6. It can be seen that the ï¿½ï¿½' volume fraction is very dependent on temperature. For single crystal alloys, the ï¿½ï¿½' volume fraction can reach 70% at room temperature. However, it decreases with increasing temperature, particularly after 1000 ï¿½ï¿½C. This is
because the ï¿½ï¿½' can dissolve in the matrix and the dissolution ability increases with temperature.
Fig.6. ï¿½ï¿½' volume fraction as a function of temperature for AM1 alloy, measured by image analysis (IA) and neuron diffraction (DN) .
The size and shape of ï¿½áï¿½ precipitates are strongly dependent on the cooling conditions. Fig.7 shows microstructures after various cooling rates. It can be seen that at very fast cooling (cooling rate >150 ï¿½ï¿½C/s), the ï¿½áï¿½ precipitates are round with a diameter less than 150 nm. For a cooling rate of 5 ï¿½C 100 C/s, the ï¿½áï¿½ precipitates are well distributed with a cubic shape. The size of the cubic ranges from 150 nm to 350 nm. When the cooling rate is lower than 2 ï¿½ï¿½C/s, the cubic shape becomes irregular and more complex shapes form .
Fig.7. CCT curve of the CMSX-2 alloy and their microstructure for various cooling rates .
Changes during isothermal exposure
Most nickel-based superalloys are precipitation hardened by a dispersion of fine ï¿½áï¿½ particles. The mechanical properties of an alloy are strongly dependent on the size and distribution of the ï¿½áï¿½ precipitates. The changes in ï¿½áï¿½ particles during isothermal exposures, such as ageing or long-term thermal exposures in service will significantly affect the mechanical properties of alloys.
Microstructural changes of ï¿½ï¿½' phases are very dependent on temperature and time. Fig.8 indicates that the ï¿½ï¿½' precipitates have an obvious growth with increasing ageing temperature and time, particularly with temperature increasing.
Fig.8. Coarsening of ï¿½áï¿½ precipitates in the alloy .
Time is an important factor for the ï¿½ï¿½' precipitate evolution. A uniform distribution of cuboidal ï¿½ï¿½' particles with a size range between 350 and 600 nm is produced after aging. However, the size and shape will change as time increases. Fig.9 illustrates the ï¿½ï¿½' microstructure evolution of CMSX-2 with time. The ï¿½ï¿½' precipitates gradually coarsen and their morphology retains a cubic shape until 1000 h. After 1000h, the adjacent particles join together to form plates or rafts. This is might be caused by chemical gradients resulting from segregation or internal stress due to misfit of ï¿½ï¿½/ï¿½ï¿½' at interface .
Fig.9. Morphology of ï¿½áï¿½ phase of CMSX-2 in different aging time at 900 ï¿½ï¿½C; (a) t=0 h; (b) t=500 h; (c) t=1000 h; (d) t=1500 h .
The reason for these changes can be explained as follows: For the cubic shape, there is high strain energy gradient in  direction. In order to minimise the elastic and interfacial energy, the ï¿½áï¿½ particles are coalesced to form plates or rafts aligned along  direction, which not only largely reduces the elastic energy but also reduces the interfacial energy. The lower interfacial energy is attributed to reduction of interface area .
This rafted structure is more commonly generated under an applied load. This is because when a load applied, there is a large elastic energy introduced along the load direction, which forces the rafts to develop perpendicular or parallel to the direction of applied stress [18, 25].
Temperature is another important factor for the coarsening of ï¿½áï¿½ precipitates coarsen. It is obvious that the coarsening rate is higher at higher temperature. The evidence is shown in Fig.9 and Fig.10. The ï¿½áï¿½ particle coarsens more rapidly at 1000ï¿½ï¿½C compared with aging at 900ï¿½ï¿½C.
Fig.10. Morphology of ï¿½áï¿½ phase of CMSX-2 in different aging time at 1000 ï¿½ï¿½C; (a) t=0 h; (b) t=500 h; (c) t=1000 h; (d) t=1500 h .
The growth rate of rafting is controlled by the diffusion of alloying elements [26, 27]. Therefore, with a high diffusion rate, the rafting is faster at higher temperature.
Changes In Carbides During Cooling
In nickel-based superalloys, about 0.05-0.2 wt % carbon is added to combine with refractory elements such as molybdenum, tungsten and rhenium to form carbides . The carbides have beneficial effects on the rupture strength and creep properties. However, the type of carbides change in different heat conditions [10, 28]. Therefore, it is very important to understand the changes in carbides and carbide reactions during cooling. Chemical composition and temperature are two key factors affecting these changes.
MC, M23C6 and M6C are the three common carbides in nickel-based superalloys. MC is mainly composed of Ta and Hf at higher temperature. When temperature reduces, it begins to decompose to form M6C and M23C6 [10, 29]. M6C has a cubic structure and forms at higher temperatures (815ï¿½C980 ï¿½ï¿½C) than M23C6.The reaction is shown as follows:
M6C carbides tend to form at W and Mo rich places and they usually prefer to precipitate on grain boundaries, which are confirmed by the EDX analysis . Since M6C carbides are more stable at higher temperature, M6C is more beneficial as a grain boundary precipitate to control grain size in process.
M23C6 is similar to M6C but it more likely to be found at Cr rich places and they form during lower temperature (around 760ï¿½ï¿½C) heat treatment or service. The reaction is shown as follows:
Both M23C6 and M6C prefer to precipitate on the grain boundaries, which have a significant effect on nickel alloy properties. Their critical location at grain boundaries increases rupture strength by inhibition of grain boundary sliding .
For modern nickel based single crystal alloys, a high level of mechanical properties is achieved due to the large contents of refractory elements such as Mo, W and Re. However, the excessive amount of these elements promotes the formation of TCP phases, which were shown to have deleterious effects on the mechanical properties . This problem has been solved by reintroduction of carbon in the single crystal alloys. The addition of carbon
reduces the extent of segregation for refractory elements by forming of carbides instead of TCP phases .
Changes In Mechanical Properties
Since the mechanical properties are determined by the microstructure, the microstructural degradation during service, and in particular the coarsening of ï¿½áï¿½ precipitates, significantly affect mechanical properties of materials . Creep behaviour is one of the most important life-limiting factors and is discussed in detail below.
The high level of creep strength of superalloys is mainly due to the presence of fine and well distributed ï¿½áï¿½ precipitates in the matrix . However, the reduction of creep resistance during service results from microstructural degradation such as ï¿½áï¿½ precipitate coarsening and grain boundary cavitations. The ï¿½áï¿½ precipitates provide creep strength by inhibiting the movement dislocations . The size and distribution of ï¿½áï¿½ precipitates have significant influence on the creep properties of materials.
After normal heat treatment, there are two kinds of ï¿½áï¿½ precipitates: the primary ï¿½áï¿½ precipitates, which are cubodial in shape with a diameter around 0.5 ï¿½ï¿½m, and secondary ï¿½áï¿½ precipitates, which are spheroidal in shape with a diameter around 0.05ï¿½ï¿½m. The spheroidal precipitates are located in the channels between the primary precipitates .
When the size and spacing of primary ï¿½áï¿½ precipitates is small, the dislocation moves across the precipitates by shearing or cutting. The small spacing and presence of secondary ï¿½áï¿½ precipitates largely inhibit the movement of dislocations. This structure improves the creep properties of materials by providing effective barriers to prevent dislocation gliding and climbing around ï¿½áï¿½ precipitates [32. 33].
The coarsening of primary ï¿½áï¿½ precipitates increases their size and spacing, which makes the channel wide enough for dislocation to bow out between
them. In addition, the coarsening of primary precipitates is contributed to expense of secondary ï¿½áï¿½ particles. These changes make it much easier for the dislocation to pass through and finally lead a decreasing of creep behaviour [34, 35].
The formation of grain boundary cavitations is another important reason for the loss of creep properties. During the service, cavities can grow due to the stress directed vacancy flow, especially at high temperature because this is diffusion-controlled process. The growth and coalescence of cavities along the grain boundary produces cracks, which finally leads a creep fracture in materials .
The increasing demand for higher efficiency and performance of the gas turbine engine leads the designers to seek higher performance materials. The problem is solved by using single crystal nickel-based superalloys. CMSX-4 is widely used because it improves fuel efficiency and performance with a lower life cycle costs.
CMSX-4 superalloys were developed by Cannon Muskegon Corporation for the use of solar turbines in late 1980s. CMSX-4 is a second-generation single crystal nickel-based superalloy and it can be operated at temperature up to 1163ï¿½ï¿½C . It provides high strength and corrosion resistance combined with good fatigue resistance. The excellent properties are achieved by the balance of chemical composition. Table 4 shows the typical chemical composition for CMSX-4.
Table 4. The typical chemical composition for CMSX-4 
Al C Co Cr Fe Mo Ni Re Si Ta Ti W
Wt% 5.6 0.006 10 6.5 0.15 0.6 60.5 3.0 0.04 6.5 1.0 6.4
CMSX-4 contains about 3 wt% rhenium, low concentration of chromium and higher concentration of aluminium. Excellent physical stability is due to the addition of rare earth elements [4, 37]. The high volume fraction of ï¿½áï¿½results from the high concentration of Al and Ti added in the materials. No carbon or
very little carbon, is another interesting feature. The reason for that is the removal of the grain boundary makes these grain boundary strengthers become unnecessary .
Nickel-based superalloys are widely used for turbine blades. The blades work at temperatures up to 1050ï¿½ï¿½C under high loading in an aggressive environment. With a long-term exposure at such aggressive operating conditions, the service lives of components are limited by creep, mechanical fatigue and corrosion due to microstructural degradation. The high cost of the replacement components has tended to lead to the use of reconditioning components, which can extend their service lives by appropriate rejuvenation procedures .
The rejuvenation heat treatment restores the microstructure to the levels equivalent to the original state, which leads to a recovery of the mechanical properties such as creep resistance and fatigue properties. The experimental work will focus on the restoration of microstructure and recovery of creep and fatigue properties.
Restoration of microstructure
The primary microstructural degradations resulting from long term thermal exposures at elevated temperature are: the formation of cavities and voids, coarsening of ï¿½áï¿½ particles; the changes in grain boundary carbides and the formation of topologically close packed (TCP) phases [37,38]. It is claimed that the rejuvenation can restore the degradation and bring the microstructure to the original states.
Grain boundary cavities
It is well known that the creep fracture usually occurs due to the formation and coalescence of grain boundary cavities. This is more common when there is a stress applied on it [39, 40]. The rejuvenation heat treatment eliminates or reduces grain boundary cavities by two ways: firstly, it inhibits the further
growth of creep cavitations. Secondly, it removes the cavitations completely by means of a sintering heat treatment. However, the recovery of the grain boundary cavities is very dependent on the cavities size. It only works when the cavities are small. Therefore, the rejuvenation heat treatment must be carried before significant failure occurs.
The coarsening of ï¿½áï¿½ precipitates is one of the most important life limiting factors for the single crystal nickel-based superalloys. This is because when the particles are small and well distributed, the dislocations find it difficult to shear between the ï¿½áï¿½ particles. While the coarsening of ï¿½áï¿½ precipitates or rafting increases the channel between the precipitates, which makes the dislocations much easier to move across them . The formation of dislocation networks leads finally failure of the components.
The common rejuvenation heat treatment has a complete solution treatment and reprecipitation procedure, by which it dissolves the coarse ï¿½áï¿½ precipitates and redistributes the precipitates in a similar manner to the original microstructures .
Carbides have either a beneficial or a deleterious effect on the mechanical properties. The beneficial effect is that M23C6 and M6C grain boundary carbides pin the grain boundaries, which prevent the grain boundary sliding and increase creep properties. While the formation of a continuous MC carbide film along the grain boundary has detrimental effects on the creep properties. Rejuvenation heat treatment leads to the dissolution of the continuous carbides, which is good for the creep properties. However, the rejuvenation heat treatment also dissolves the carbides on the grain boundaries, which may results a reduction of the creep properties. Therefore, the effect of rejuvenation for the carbides is unpredictable and is also dependent on the exact type and amount of the carbides present .
Embrittling TCP phases such as ï¿½ï¿½ and ï¿½ï¿½ can form during service exposure,
which is deleterious for the materials. During a rejuvenation heat treatment, the TCP phases can be taken to in solution. However, the elemental segregation resulting from the prolonged exposure at high temperature may lead to these phases reprecipitating very rapidly once parts are returned to service .
Recovery of mechanical properties
The rejuvenation also helps the recovery of the mechanical properties. The recovery of creep and fatigue properties are discussed as follows:
Regeneration of creep properties is contributed to the rejuvenation heat treatment. The restoration of creep behaviour is achieved by the sintering and reduction of grain boundary cavities.
The rejuvenation treatment also has beneficial effect on the fatigue properties, which is contributed to the recovery of dislocations during heat treatment . The high dislocation density developed during long-term thermal exposures is recovered during annealing stages.
However, it is very difficult to get an agreement as to the benefits of rejuvenation because the results are not well repeatable . This is because the fatigue properties may be affected by the surface defects. The introduction of surface damage from processing largely affects fatigue results.
Considerations for rejuvenation procedures
The rejuvenation heat treatment must be tailored because it is very dependent on the type of alloy, the stage of damage and the cost of repairs .
Temperature is one important issue for the rejuvenation heat treatment. Usually it is performed at a temperature above the ï¿½áï¿½ solvus temperature to regenerate a microstructure close to the original state . Another important issue is applying the rejuvenation procedures at the optimum stage in service.
The rejuvenation procedures must be applied early enough to prevent irreparable damages but late enough to give a cost-effective benefit. Usually, it is applied at the end of the secondary or early in the tertiary stage creep [40, 41, 42].