- Email: [email protected]
Nonequilibrium Grain Boundary Segregation of Phosphorus in Ni-Cr-Fe Superalloy
Available online at www.sciencedirect.com
.-"
-.;- ScienceDirect JOURNAL OF IRON AND STEEL RESEARCH, INTERNXfIONAL. 2011, 18(1): 61-67
Nonequilibrium Grain Boundary Segregation of Phosphorus in Ni-Cr-Fe Superalloy WANG Kai' ,
SI Hong" ,
YANG Chun' ,
xu Ting-dong'
O. Superalloy Department, Central Iron and Steel Research Institute, Beijing 100081, China; Center of Iron and Steel, Central Iron and Steel Research Institute, Beijing 100081, China)
2. National Analysis
Abstract: In virtue of Auger electron spectroscopy, the grain boundary concentrations of phosphorus in Ni-Cr-Fe superalloy are measured after solution treatment at 1180 'C for 45 min. The results show that a peak of phosphorus concentration occurs at about 180 min during isothermal ageing at 500 'C, and a maximum concentration of phosphorus appears also at about 500 'C for all specimens aged for 20 min at temperatures of 200, 400, 500. 700 and 800 ·C. The results are analyzed with the laws of nonequilibrium grain boundary segregation. It is found from the analysis that peaks are related to critical time for nonequilibrium grain boundary segregation of phosphorus. Key words: nonequilibrium segregation; grain boundary; phosphorus; nickel based superalloy
At the present time there are two distinct viewpoints of the effect of phosphorus on the mechanical properties of Ni-base superalloy. Some investigators regard phosphorus as a detrimental element which can degrade mechanical properties[I-3]. On cast samples of Hastelloy X, W Yeniscavich et al[2] have already shown that phosphorus promotes high-temperature (760 - 980 'C) cracking and hot tearing during Gleeble tests. In 1975, D A Vermilyea et al[3] have also shown that additions of phosphorus to Inconel 600 type alloys cause susceptibility to intergranular attack in strongly oxidizing media. This attack occurred in solution-annealed materials and was believed to be associated with segregation of the phosphorus to grain boundaries. While some investigators consider that phosphorus is a beneficial element which can promote mechanical properties greatly':"?". For example, G S Was and his co-workers[4-6] confirmed experimentally that the addition of phosphorus to Ni-Cr-Fe alloy 600 improves the creep and intergranular crack propagation resistance. Recently experimental and simulative results[7-9] not only showed that phosphorus can enhance the stress rupture life, but also demonstrated that segregated phosphorus at grain boundaries affects these properties directly[7-S].
The theory of micro-segregation to grain boundary segregation of solute atoms can be divided into equilibrium and nonequilibrium[IO-l1]. Equilibrium grain boundary segregation (EGS) occurs as a result of inhomogeneities in the solid giving rise to sites of grain boundaries for which solute atoms have a lower free energy[l1]. Nonequilibrium grain boundary segregation (NGS) arising from thermal effects was first reported by K T Aust[12] and T R Anthony'Pl, The model and equations for critical time and kinetics of NGS were given by T Xu et al[IO.14] and R G Faulkner-P' , which have been approved by many researchers[16-IS]. It is well known that apprehending the effect of grain boundary segregation on properties of alloys correctly requires determining the EGS or NGS of solute atoms experimentally. To our knowledge, there is very little study on the NGS of phosphorus for Ni-base superalloy. Recently, the author[19] analyzed Briant's experimental results[20-21] with the laws of NGS. A grain boundary segregation peak of phosphorus in austenitic steel appeared at about 24 h during the segregation isotherm at temperature 700 'C, and a segregation peak occurred also at 600 'C for all specimens aged for 100 h at temperatures of 500, 550, 600, 650 and 700 'C after annealing at 1100 'C for 1 h,
Foundation Item: Item Sponsored by National Natural Science Foundation of China (50771036. 51001030) Biography: WANG Kai0978-). Male. Doctor; E-mail: [email protected]. en; Received Date: December 15. 2009
62 •
were induced by the critical time of phosphorus NGS[19]. In the present work, the critical time of phosphorus NGS in Ni-Cr-Fe superalloy is studied with Auger electron spectroscopy (AES), and the critical time constant are first proposed experimentally in this superalloy.
1
Vol. 18
Journal of Iron and Steel Research, International
Experimental Procedure
The investigations were conducted on wrought rods with diameter of 14 mm of a commercial Ni-CrTable 1
Fe superalloy. The chemical composition of this alloy is listed in Table 1. The samples were homogenized at 1180 'C for 45 min and then water quenched. The ASTM grain size was about 66. 4 fim. Some samples were then-aged at 500 'C for 20, 180, 6000 and 60000 min respectively, other samples were aged at 200, 400, 500, 700 and 800 'C for 20 min respectively. After that, the sarqples were immediately water quenched to restrain elemental segregation during cooling.
Chemical composition of the
Ni-C~Fe
superalloy
C
Cr
Fe
Mo
P
S
Ti
Cu
Ni
Mass percent! %
0.091
21. 42
17.78
8. 74
0.0062
0.001
0.082
<0.10
Balance
Atomic percent! %
0.094
24.09
18.57
5.32
0.012
0.0018
0.099
<0.094
Balance
To evaluate the concentration of phosphorus at the grain boundaries, Auger electron spectroscopy (AES) was employed. The standard AES cylindrical specimens of ,p3. 6 mmX 31. 7 mm with sharp notches were prepared after above ageing. Subsequently, the notched Auger samples were cathodically hydrogen charged for 120 -168 h at room temperature in O. 5 mol/L H 2S04 and 250 mg/L As 203 solution with a current density of 50 mAl em" to obtain intergranular fractures. After hydrogen charging, the samples were immediately loaded into the AES chamber with a vacuum of 1 X 10- 9 Torr and then fractured at room temperature by impact for subsequent AES analysis. Auger measurements on the fracture surfaces were carried out using a PHI595 Auger spectrometer. Typical parameters for the Auger measurements were: primary energy = 5 ke V, emission voltage = 170 V, filament current = 1. 45 A. The primary beam size was about O. 8 fim in diameter. All spectra were recorded in the differential mode dN (E) I dE. Assuming that the Auger peak height associated with an element is proportional to its concentration, this concentration (atomic fraction), Co, can be approximately evaluated byC22]
c= .
Hj5 i
~kH kl5k
account in the analysis because they could have an apparent contribution from contamination from the residual atmosphere in the AES system.
2
Results
2. 1 The AES measurement of phosphorus at grain boundaries Fig. 1 shows the freshly intergranular fracture produced by cathodic charging in AES chamber for experimental superalloy. It can be seen from Fig. 1 that grain boundaries are smooth and there is no evident precipitation on them. The typical Auger spectrum for the specimens aged at 500 'C for 20 min after homogenizing at 1180 'C indicates small amount of phosphorus at grain boundaries, as shown in Fig. 2. The concentration of phosphorus at grain boundaries calculated with Eqn, (1) in Fig. 2 is 3.43% (in atomic percent) or approximately 300 times the bulk concentration, the degree of segregation are generally considered as the results of segregation for solute.
(1)
where Hi and S, denote the Auger peak height and sensitivity factor of element i , respectively. H, and S, have the same meaning as Hi and S, for any element identified by AES. In the calculations, the Ni 848 eV, Fe 703 eV, Cr 529 eV and S 152 eV peaks were considered using their Auger sensitivity factors of 0.281, O. 246, 0.359 and 1. 041 [22J. The carbon and oxygen peaks in the spectra were not taken into
Fig. 1
The freshly intergranular fracture obtained by impact at room temperature in AES
300
500 700 Kineticenergy/eV
lative average levels of phosphorus calculated for more than 20 individual grain boundaries in steel. Here the error in measuring the mean concentration of phosphorus at a grain boundary was approximately ±2.5%. In this study, based on the Eqn. (2), the cumulative average levels would be calculated.
900
Fig. 2 Typical Auger electron spectra from an Auger sample aged at 500 'C for 20 min after annealing at 1180 'C
Fig. 3 gives the depth profile of phosphorus at the grain boundary facet. It is exhibited that phosphorus concentration decreases smoothly with increasing argon ion sputtering time, which suggests that phosphorus is present through a segregation mechanism and is not as a precipitate film[23-24]. Just as E D Hondros and M P Seah[24] mentioned that a depth composition profile, showing the' element distribution with distance from the fracture surface is normally carried out by combining AES with controlled sputtering of the surface by inert gas ion bombardment. The procedure has been successful in defining the extent over which grain boundary segregation exists[24]. 2.0 , - - - - - - - - - - - - - - - - - - ,
K
X=VC/K
o
20
40
60
Time/s
80
500
Concentration of phosphorus in atomic pencent lis a function of sputtering time at a grain boundary
2. 2 The scatter of the cumulative average concentration at grain boundaries The amount of segregation to individual grain boundary, as determined by Auger spectroscopy in samples of a given condition, could be quite variable[25-27]. The deviation in grain boundary structure, fracture path and bulk chemical inhomogeneity can attribute to this variation'P'" , In order to acquire the segregation levels for samples of a given condition, the cumulative average levels of many individual grain boundaries often are adopted. For examples, T Shinoda and T Nakamura[29] had adopted the cumu-
(2)
where, Xi is the individual value of concentration at i-th grain boundary facet; and K is the number of Auger measurements. The Auger measurements are made on five separate Auger samples machined from the same rod that had aged for 20 min at 500 ·C after annealing at 1180 ·C for 45 min. Fig. 4 shows the scatter of a cumulative average levels, X, with number of Auger measurements K. It is found that the maximum and minimum levels derived from the cumulative average levels ofX=21, 22, ,·,,30 and 31 are 1. 753 % and 1. 705 % (in atomic percent) respectively. The average of these two values, 1. 729 %, is used as a most probable value of the segregation concentration of grain boundary for this condition. As can be seen from this figure, the scatter of the cumulative average of X from the levels of 1. 729 % is comparatively large when the number of K is smaller, but it always fell within the ± 1 % error band around 1. 729 %, when the number of K is increased to more than 20. Therefore, over 20 individual grain boundaries are measured for each condition to obtain the average in this study. 1.9 , - - - - - - - - - - - - - - - - , ± 1% error band for 1.729 1.8
?------I'!-
Fig. 3
• 63 •
Nonequilibrium Grain Boundary Segregation of Phosphorus in Ni-Cr-Fe Superalloy
Issue 1
i~ ::s .... 1.7
1.729
o 0 .0 0::
,~
:8
1.6 6i>E
~
~
§
'"
1.5 1.4
0
5
10 15 20 25 Number of measurementsIK
30
35
Fig. 4 Scatter of the cumulative average of phosphorus concentration in atomic percent, X, as a function of number of AES measurements K for samples aged at 500 'C for 20 min after annealing at 1180 °C
2. 3 Phosphorus concentration at grain boundary for different heat treatment conditions The variations of grain boundary concentrations (in atomic percent) of phosphorus with ageing time at
Journal of Iron and Steel Research, International
• 64 •
500 'C for this alloy after quenching from 1180 'C are shown in Table 2. Obviously, a peak of grain boundary concentration of phosphorus appears at about 180 min during isothermal ageing at 500 'C. When the ageing time is shorter than 180 min, the grain boundary concentrations of phosphorus increase with ageing time; when the ageing time is longer than 180 min, the grain boundary concentrations of phosphorus decrease with ageing time. Here the grain boundary concentration of phosphorus at 60000 min is the cumulative average levels of 6 individual grain boundary facets. It is shown from Fig. 4 that the cumulative average levels of 6 individual grain boundary facets have larger error than that of 20 individual grain boundary facets. Even so, the cumulative average levels at 60000 min also are lower than at 180 min. Table 2 The variation of phosphorus concentration at grain boundary with ageing time at 500 'c for Nl-Cr-Fe superalloy Ageing time/min
0
Concentration of P / %
1. 80
Note:
I)
20
180
6000
60000
1. 73
2.06
1. 17
1. 42
The initial segregation concentration is for the homogenized samples. 1)
Fig. 5 shows the grain boundary concentration Cin atomic percent) of phosphorus for the alloy aged
for 20 min at 200, 400, 500, 700 and 800 'C after quenching from 1180 'C, and indicates a maximum of concentration occurs at 500 ·C. When the ageing temperatures are lower than 500 'C, the grain boundary concentration of phosphorus decrease with decreasing ageing temperature. The grain boundary concentration of phosphorus at 700 'C and 800 'C is lower than that of phosphorus at 500 'C, and they have no obvious differences, which attribute to the limit of sensitivity of the Auger technique.
1.0 L....... 200
---'
-'---
--o-J
600 Temperaturel'C
800
400
Fig. 5 Phosphorus concentration at grain boundary as a function of ageing temperature for Ni-Cr-Fe superalloy aged for 20 min after annealing at 1180 'c
3
Vol. IS
Discussion
3.1 Critical time of phosphorus NGS in Ni-Cr-Fe superalloy According to the formation of solute atom-vacancy complexes within the matrix[12-13], R G Faulkner-P'' suggested the concept of a critical time. Subsequently, T XU[IO.30] experimentally indicated that the critical time should correspond to the peak in the solute grain boundary concentration during isothermal ageing after quenching. When the ageing time is shorter than the critical time, the grain boundary concentration of solute will increase with ageing time, which is referred to as the segregation process. When the ageing time is longer than the critical time, the grain boundary concentration of solute will decrease with ageing time, which is referred to as the de-segregation process. It is clear that the NGS concentration depends on how the ageing time approaches the critical time. When the ageing time is closer to the critical time, the NGS concentration is high even if the material is over-or under-aged compared to the critical time, as shown in Fig. 1 in Ref. [10]. This is the characteristic aspect of NGS, which has been verified for variant solutes in various alloy systems[31-33]. The formula for critical time was worked out by R G Faulkner[l5] and T XU[IO.14,30] :
r 21nCDc / D )
t c = 8CDc - D )
(3)
where, D i and D, are the diffusion coefficients for solute atoms and complexes respectively; r is the grain radius; and 8 is the critical time constant. The relationship between critical times and ageing temperature was calculated out of Eqn. (3) and shown in Fig. 3 and Fig. 4 of Ref. [10]. The critical time increases with the decrease of ageing temperature, which has been confirmed a lot of times such as In Ref. [10J, Ref. [19J, and Ref. [34]. The variations in Table 2 show that the grain boundary concentration of phosphorus reaches a peak after about 180 min during ageing at 500 'C following quenching from 1180 'C. This means the phosphorus has characteristic of NGS, and the critical time of phosphorus at 500 'C is at about 180 min for Ni-Cr-Fe superalloy. When the ageing time is shorter than 180 min, the process of segregation for NGS will be reached. So the grain boundary concentrations of phosphorus increase with increasing ageing time. When the ageing time is longer than 180 min, the process of de-segregation for NGS will be
Nonequilibrium Grain Boundary Segregation of Phosphorus in Ni-Cr-Fe Superalloy
Issue I
reached. So the grain boundary concentrations of phosphorus decrease with increasing ageing time.
3. 2 The critical time constant (; of phosphorus in NiCr-Fe superalloy Data on the diffusion coefficient of phosphorusvacancy complexes and the critical time constant ~ of phosphorus for Ni-base superalloy are scarce, which will limit the study on the segregation of phosphorus in Ni-base superalloy. Based on Ref. [35J, the self-diffusion coefficient of Ni , DNi-self, m 2 / s , is: DNi-self=1. 27 X 10- 4exp( -2. geV /kT) (4) which is characteristic of lattice diffusion in high-purity nickel. The most likely mechanism of self-diffusion in high-purity metals is that of continual migration of vacancies (vacancy diffusion) [34J. With vacancy diffusion, the probability that ali atom may jump to the next site will depend on: 1) the probability that the site is vacant (which in turn is proportional to the fraction of vacancies in the crystal) and 2) the probability that it has the required activation energy to make the transition. For self-diffusion, where no complications exist, the diffusion coefficient is, therefore, given by[36J : (5) DNi-self=Doexp[ -(Ef+Em)/kTJ where Do is a constant involving the frequency of atomic vibration or vacancy vibration. The frequency of vacancy diffusion is Do = 1. 27 X 10- 4. E, is the energy of formation of a vacancy, Em is the energy of migration of a vacancy, and the sum of the two energies, Q= E, Em' is the activation energy for self-diffusion in high-purity metals[36 J. The energy of formation of a vacancy is about 1. 5 eV in Ni-based alloys[37J. Therefore, the activation energy for vacancy diffusion, i. e. energy of migration of a vacancy, is Em = Q - E, = 2. 9 - 1. 5 = 1. 4 eV)n pure Ni. Therefore, the diffusion coefficient of vacancies in Ni-base alloys should be: D, =1. 27 X 10- 4exp( -1. 4 eV /kT) m 2 / s (6) The diffusion coefficient of phosphorus atom, D; in a Ni-base alloy is[3BJ : (7) D p=lX10- Bexp(-1. 78 eV/kT) m 2/s Solute phosphorus atoms in Ni-base alloys are in substitutional sites. For the same reason, the activation energy of phosphorus diffusion in Ni-base, alloys, Qp, is also suggested to be Qp = E, E p, where E, is still the energy of formation of a vacancy and E; is the energy of migration of phosphorus atoms in aNi-base alloy. Therefore, the energy of migration of phosphorus atoms is E p=Qp-Ef=1. 78,1, 5=0. 28 eV.
+
+
• 65 •
According to the mechanism suggested by S H Song and L Q Weng[39J, the activation energy for a vacancy-phosphorus complex in Ni-base alloys is the energy of migration of a vacancy or the energy of migration of phosphorus atoms. Selection of the former or latter depends on which is higher. Therefore, the activation energy for a vacancy-phosphorus complex in a Ni-base alloy should be 1. 4 eV. With reference to the concept of non-dissociation mechanism suggested by S H Song and L Q Weng[39J , for each step in the migration process of a vacancy-phosphorus complex, the vacancy requires three jumps, of which two are towards a nearest lattice site occupied by a matrix atom and the other jump is towards that occupied by phosphorus. As a consequence, it is reasonable to assume that the pre-exponential constant for diffusion of vacancy-phosphorus complexes is one-third of the pre-exponential constant for diffusion of vacancies. The pre-exponential constant for complex diffusion is, therefore, (1. 27 X 10- 4 )/3 = 4. 23 X 10- 5 • The diffusion coefficient of vacancyphosphorus complexes in a Ni-base alloy can be shown to be: Dv-p=4. 23X10- 5 e x p ( - 1 . 4 eV/kT) m 2/s (8) The critical time constant (J in Eqn. (3) can be calculated to be about (J = 47 from the critical time, 180 min, and the relative parameters above for this superalloy.
3. 3 The maximum segregation of phosphorus at 500°C for ageing for 20 min Fig. 5 shows that, for the experimented alloy aged for 20 min at 200, 400, 500, 700 and 800 'C after quenching from 1180 'C, the concentration of phosphorus at grain boundaries reaches a maximum at 500 ·C. From the critical time constant (J = 47 and the relative diffusion coefficients for phosphorus and complexes in Ni-base alloys, the critical times of different ageing temperatures t; can be calculated out of Eqn, (3). The calculated results are shown in Table 3, where it can be seen that the critical time decreases rapidly with increasing ageing temperature, which accords with the relationship between critical times and ageing temperatures mentioned above. It can be seen from Table 3 that when the ageing temperatures are lower than 500 'C, the critical times of phosphorus NGS increase rapidly. They are 4. 36 X 103 min at 400 'C and 1. 43 X 108 min at 200 'C, which are considerably longer than 20 min. Therefore, the grain boundary concentration of phosphorus will get lower and lower when the ageing time adopted is
Journal of Iron and Steel Research, International
• 66 • 'Thble 3
Critical time calculated at different ageing temperatures
Ageing temperature/"C Critical time/min
200
400
1. 43 X 10· 4. 36 X 10 3
500
700
800
180
2
0.4
still 20 mm , as shown in Fig. 5. It can be seen also from Table 3 that when the ageing temperatures are higher than 500 "C, the critical times of phosphorus NGS decrease rapidly. They are 2 min at 700 "C and o. 4 min at 800"C, which are considerably shorter than 20 min. Therefore, the grain boundary concentration of phosphorus will get lower and lower when the ageing time adopted is still 20 min, as shown in Fig. 5. As mentioned above, the NGS concentration of solute depends on how the critical time approaches the ageing time. Table 3 shows that the critical time of phosphorus at 500 "C is 180 min , which is closest to the ageing time of 20 min than at other ageing temperatures. Thus, the grain boundary concentration of phosphorus can attain maxima for an ageing of 20 min at 500 "C. Therefore, it may be stated that for a alloy system in which the solute has NGS characteristic, an ageing temperature must exist at which the NGS concentration of solute has a maximum for the alloy quenched from a high solution temperature and then aged at various temperatures for a certain time. The critical time at this ageing temperature is equal or close to the ageing time. This is another important characteristic of NGS, and has been confirmed for P in 304L stainless steel[19], Mg in Ni-Cr-Co alloyE34] , and S in Ni-Cr-Fe superalloy'<'".
4
Conclusions
1) The phosphorus of Ni-Cr-Fe superalloy has characteristic of NGS. 2) The critical time constant for phosphorus in this superalloy is about 47. 3) For the thermal cycle of Ni-Cr-Fe superalloy in this study, an ageing temperature 500 "C exists at which the NGS concentration of phosphorus has a maximum. The critical time of phosphorus at the ageing temperature 500 "C is very close to the ageing time 20 min, which was adopted at all the ageing temperatures. References: [lJ [2J
[3J
Holt R T. Wallace W. Impurities and Trace Elements in Nickel-Base Superalloys [JJ. Inter Metals Rev. 1976,21: 1. Yeniscavich W. Fox C W. Effect of Minor Elements on the Weldability of High Nickel Alloys [M]. New York: Welding Research Council. 1969. Vermilyea D A. Tedmon C S. Broecker D E. Effects of Phos-
Vol. 18
phorus and Silicon on the Intergranular Corrosion of a NickelBase Alloy [JJ. Corrosion. 1975. 31: 222. [4J Was G S. Kruger RM. A Thermodynamic and Kinetic Basis for Understanding Chromium Depletion in Ni-Cr-Fe Alloys [J]. Acta Metall. 1985. 33(5): 841. [5J Sung J K. Was G S. Intergranular Cracking of Ni-16Cr-9Fe Alloys in High-Temperature Water [J]. Corrosion. 1991. 47 (11): 824. [6J Was G S. Sung J K. Angeliu T M. Effects of Grain Boundary Chemistry on the Intergranular Cracking Behavior of Ni-16Cr9Fe in High-Temperature Water [J]. Met Trans. 1992. 23A (12): 3343. [7J Cao W D. Kennedy R L. Superalloys 718. 625. 706 and Various Derivatives [MJ. Warrendale: TMS. 1994. [8J Liu X. Liu H. Dong J X. et al. Molecular Dynamics Simulation on Phosphorus Behavior at Ni Grain Boundary [JJ. Scripta Mater. 2000. 42(2): 189. [9J Guo J T. Zhou L Z. Superalloys 1996 [M]. Warrendale: TMS. 1996. [10J XU T. Cheng B. Kinetics of Non-Equilibrium Grain Boundary Segregation [JJ. Prog Mater Sci. 2004. 49(2): 109. [l1J McLean D. Grain Boundaries in Metals [MJ. Oxford: Oxford University Press. 1957. [12J Aust K T. Hanneman R E. Niessen P. et al. Solute Induced Harding Near Grain Boundaries in Zone Refined Metals [JJ. Acta Metall. 1968. 16: 291. [13J Anthony T R. Solute Segregation in Vacancy Gradients Generated by Sintering and Temperature Changes [JJ. Acta Metall. 1969. 17(5): 603. [14J XU T. Song S. A Kinetic Model of Non-Equilibrium Grain Boundary Segregation [JJ. Acta Metall. 1989. 37: 2499. [15J Faulkner R G. Non-Equilibrium Grain-Boundary Segregation in Austenitic Alloys [JJ. J Mater Sci. 1981. 16: 373. [16J Li Q. Yang S. Li L. Experimental Study on Non-Equilibrium Grain-Boundary Segregation Kinetics of Phosphorus in an Industrial Steel [JJ. Scripta Mater. 2002. 47: 389. [17J Seve P. Janovec J. Lucas M. Kinetics of Phosphorus Segregation in 2. 7Cr-0. 7Mo-0. 3V Steels With Different Phosphorus Contents [JJ. Steel Res. 1995. 66(12): 537. [18J Chen W, Chaturvedi M C. Richards N L. Effect of Boron Segregation at Grain Boundaries on Heat-Affected Zone Cracking in Wrought Inconel 718 [JJ. Metal Mater Trans. 2001. 32A (4):931. [19J Wang K. Wang M Q. Si H. et al. Critical Time for Non-Equilibrium Grain Boundary Segregation of Phosphorus in 304L Stainless Steel [JJ. Mater Sci Eng. 2008. 485A: 347. [20J Briant C L. Grain Boundary Segregation of Phosphorus in 304L Stainless Steel [JJ. Metall Trans. 1985. 16A: 2061. [21J Briant C L. Grain Boundary Segregation of Phosphorus and Sulfur in Type 304L and 316L Stainless Steel and Its Effect on Intergranular Corrosion in the Huey Test [JJ. Metall Trans. 1987. 18A: 691. [22J Davis L E. McDonald M C. Palmberg P W. et al. Handbook of Auger Electron Spectroscopy [MJ. 2nd ed. Minnesota: Physical Electronics Division of Perkin-Elmer Corporation. [23J
[24J [25J
1976. Dong J. Zhang M. Xie X, et al. Interfacial Segregation and Cosegregation Behaviour in a Nickel-Base Alloy 718 [JJ. Mater Sci Eng. 2002. 328A: 8. Hondros E D. Seah M P. Segregation to Interfaces [JJ. Int Met Rev. 1977. 222: 262. Watanable T. Kitamura S. Karashima S. Grain Boundary Hardening and Segregation in Alpha Iron-Tin Alloy [JJ. Acta
Issue 1
[26J
[27J
[28J [29J
[30J
[31J [32J
Nonequilibrium Grain Boundary Segregation of Phosphorus in Ni-Cr-Fe Superalloy Metall. 1980. 28: 455. Hall E L. Imeson D. Vander sande J B V. On Producing High-Spatial-Resolution Composition Profiles via Scanning Transmission Electron Microscopy [n. Phi Mag. 1981. 43A: 1569. Kimeda J. Mcmanhon Jr C J. The Effects of Sb, Sn , and P on the Strength of Grain Boundaries in a Ni-Cr Steel Metall Trans. 1981. 12A: 31. Briant C L. Source of Variability in Grain Boundary Segregation [n. Acta Metall. 1983. 31(2): 257. Shinoda T. Nakamura T. The Effects of Applied Stress on the Intergranular Phosphorus Segregation in a Chromium Steel [n Acta Metall. 1981. 29: 1631. Xu T. The Critical Time and Critical Cooling Rate of Non-Equilibrium Grain-Boundary Segregation [n. J Mater Sci Let. 1988. 7: 241. Ding R G. Rong T S. Knott J F. Phosphorus Segregation in 2. 25Cr-1Mo Steel [n. Mater Sci Tech. 2005. 21(1): 85. Li Q. Liu E. Liu D. et al. Temper Embrittlement Dynamics Induced by Non-Equilibrium Segregation of Phosphorus in
[n.
[33J [34J [35J [36J [37J
[38J [39J [40J
• 67 •
Steel 12CrlMoV [n. Scripta Mater. 2005. 53(3): 309. Zhang Z L. Lin Q Y. Yu Z S. Grain Boundary Segregation in Ultra-Low Carbon Steel DJ. Mater Sci Eng. 2000. 291A: 22. Xu T. Critical Time for Mg Grain-Boundary Segregation in Nic-ce alloy [n. Phi Mag Lett. 2006. 86(8): 501. Hoffman R E. Pikus F W. Ward R A. Self-Diffusion in Solid Nickel [n Trans AIME. 1956. 206: 483. SmallmanR E. Modern Physical Metallurgy [M]. London: Butterworth. 1963. Kim S M. Buyers W J L. Vacancy Formation Energy in Iron by Positron Annihilation [1]. J Phys F. Metal Phys , 1978. 8L: 103. Brand E A. Brook G B. Smithells Metals Reference Book [MJ. 7th ed. London: Butterworth-Heinemann Ltd. 1992. Song S a. Weng L Q. Diffusion of Vacancy-Solute Complexes in Alloys [n. Mater Sci Tech. 2005. 21: 305. Wang K. Xu T. Wang Y Q. et al. Intermediate-Temperature Embrittlement Induced by Non-Equlibrium Grain-Boundary Segregation of Sulfur in Ni-Cr-Fe Alloy [1]. Phi Mag Lett. 2009. 89(11): 725.
.-"
-.;- ScienceDirect JOURNAL OF IRON AND STEEL RESEARCH, INTERNXfIONAL. 2011, 18(1): 61-67
Nonequilibrium Grain Boundary Segregation of Phosphorus in Ni-Cr-Fe Superalloy WANG Kai' ,
SI Hong" ,
YANG Chun' ,
xu Ting-dong'
O. Superalloy Department, Central Iron and Steel Research Institute, Beijing 100081, China; Center of Iron and Steel, Central Iron and Steel Research Institute, Beijing 100081, China)
2. National Analysis
Abstract: In virtue of Auger electron spectroscopy, the grain boundary concentrations of phosphorus in Ni-Cr-Fe superalloy are measured after solution treatment at 1180 'C for 45 min. The results show that a peak of phosphorus concentration occurs at about 180 min during isothermal ageing at 500 'C, and a maximum concentration of phosphorus appears also at about 500 'C for all specimens aged for 20 min at temperatures of 200, 400, 500. 700 and 800 ·C. The results are analyzed with the laws of nonequilibrium grain boundary segregation. It is found from the analysis that peaks are related to critical time for nonequilibrium grain boundary segregation of phosphorus. Key words: nonequilibrium segregation; grain boundary; phosphorus; nickel based superalloy
At the present time there are two distinct viewpoints of the effect of phosphorus on the mechanical properties of Ni-base superalloy. Some investigators regard phosphorus as a detrimental element which can degrade mechanical properties[I-3]. On cast samples of Hastelloy X, W Yeniscavich et al[2] have already shown that phosphorus promotes high-temperature (760 - 980 'C) cracking and hot tearing during Gleeble tests. In 1975, D A Vermilyea et al[3] have also shown that additions of phosphorus to Inconel 600 type alloys cause susceptibility to intergranular attack in strongly oxidizing media. This attack occurred in solution-annealed materials and was believed to be associated with segregation of the phosphorus to grain boundaries. While some investigators consider that phosphorus is a beneficial element which can promote mechanical properties greatly':"?". For example, G S Was and his co-workers[4-6] confirmed experimentally that the addition of phosphorus to Ni-Cr-Fe alloy 600 improves the creep and intergranular crack propagation resistance. Recently experimental and simulative results[7-9] not only showed that phosphorus can enhance the stress rupture life, but also demonstrated that segregated phosphorus at grain boundaries affects these properties directly[7-S].
The theory of micro-segregation to grain boundary segregation of solute atoms can be divided into equilibrium and nonequilibrium[IO-l1]. Equilibrium grain boundary segregation (EGS) occurs as a result of inhomogeneities in the solid giving rise to sites of grain boundaries for which solute atoms have a lower free energy[l1]. Nonequilibrium grain boundary segregation (NGS) arising from thermal effects was first reported by K T Aust[12] and T R Anthony'Pl, The model and equations for critical time and kinetics of NGS were given by T Xu et al[IO.14] and R G Faulkner-P' , which have been approved by many researchers[16-IS]. It is well known that apprehending the effect of grain boundary segregation on properties of alloys correctly requires determining the EGS or NGS of solute atoms experimentally. To our knowledge, there is very little study on the NGS of phosphorus for Ni-base superalloy. Recently, the author[19] analyzed Briant's experimental results[20-21] with the laws of NGS. A grain boundary segregation peak of phosphorus in austenitic steel appeared at about 24 h during the segregation isotherm at temperature 700 'C, and a segregation peak occurred also at 600 'C for all specimens aged for 100 h at temperatures of 500, 550, 600, 650 and 700 'C after annealing at 1100 'C for 1 h,
Foundation Item: Item Sponsored by National Natural Science Foundation of China (50771036. 51001030) Biography: WANG Kai0978-). Male. Doctor; E-mail: [email protected]. en; Received Date: December 15. 2009
62 •
were induced by the critical time of phosphorus NGS[19]. In the present work, the critical time of phosphorus NGS in Ni-Cr-Fe superalloy is studied with Auger electron spectroscopy (AES), and the critical time constant are first proposed experimentally in this superalloy.
1
Vol. 18
Journal of Iron and Steel Research, International
Experimental Procedure
The investigations were conducted on wrought rods with diameter of 14 mm of a commercial Ni-CrTable 1
Fe superalloy. The chemical composition of this alloy is listed in Table 1. The samples were homogenized at 1180 'C for 45 min and then water quenched. The ASTM grain size was about 66. 4 fim. Some samples were then-aged at 500 'C for 20, 180, 6000 and 60000 min respectively, other samples were aged at 200, 400, 500, 700 and 800 'C for 20 min respectively. After that, the sarqples were immediately water quenched to restrain elemental segregation during cooling.
Chemical composition of the
Ni-C~Fe
superalloy
C
Cr
Fe
Mo
P
S
Ti
Cu
Ni
Mass percent! %
0.091
21. 42
17.78
8. 74
0.0062
0.001
0.082
<0.10
Balance
Atomic percent! %
0.094
24.09
18.57
5.32
0.012
0.0018
0.099
<0.094
Balance
To evaluate the concentration of phosphorus at the grain boundaries, Auger electron spectroscopy (AES) was employed. The standard AES cylindrical specimens of ,p3. 6 mmX 31. 7 mm with sharp notches were prepared after above ageing. Subsequently, the notched Auger samples were cathodically hydrogen charged for 120 -168 h at room temperature in O. 5 mol/L H 2S04 and 250 mg/L As 203 solution with a current density of 50 mAl em" to obtain intergranular fractures. After hydrogen charging, the samples were immediately loaded into the AES chamber with a vacuum of 1 X 10- 9 Torr and then fractured at room temperature by impact for subsequent AES analysis. Auger measurements on the fracture surfaces were carried out using a PHI595 Auger spectrometer. Typical parameters for the Auger measurements were: primary energy = 5 ke V, emission voltage = 170 V, filament current = 1. 45 A. The primary beam size was about O. 8 fim in diameter. All spectra were recorded in the differential mode dN (E) I dE. Assuming that the Auger peak height associated with an element is proportional to its concentration, this concentration (atomic fraction), Co, can be approximately evaluated byC22]
c= .
Hj5 i
~kH kl5k
account in the analysis because they could have an apparent contribution from contamination from the residual atmosphere in the AES system.
2
Results
2. 1 The AES measurement of phosphorus at grain boundaries Fig. 1 shows the freshly intergranular fracture produced by cathodic charging in AES chamber for experimental superalloy. It can be seen from Fig. 1 that grain boundaries are smooth and there is no evident precipitation on them. The typical Auger spectrum for the specimens aged at 500 'C for 20 min after homogenizing at 1180 'C indicates small amount of phosphorus at grain boundaries, as shown in Fig. 2. The concentration of phosphorus at grain boundaries calculated with Eqn, (1) in Fig. 2 is 3.43% (in atomic percent) or approximately 300 times the bulk concentration, the degree of segregation are generally considered as the results of segregation for solute.
(1)
where Hi and S, denote the Auger peak height and sensitivity factor of element i , respectively. H, and S, have the same meaning as Hi and S, for any element identified by AES. In the calculations, the Ni 848 eV, Fe 703 eV, Cr 529 eV and S 152 eV peaks were considered using their Auger sensitivity factors of 0.281, O. 246, 0.359 and 1. 041 [22J. The carbon and oxygen peaks in the spectra were not taken into
Fig. 1
The freshly intergranular fracture obtained by impact at room temperature in AES
300
500 700 Kineticenergy/eV
lative average levels of phosphorus calculated for more than 20 individual grain boundaries in steel. Here the error in measuring the mean concentration of phosphorus at a grain boundary was approximately ±2.5%. In this study, based on the Eqn. (2), the cumulative average levels would be calculated.
900
Fig. 2 Typical Auger electron spectra from an Auger sample aged at 500 'C for 20 min after annealing at 1180 'C
Fig. 3 gives the depth profile of phosphorus at the grain boundary facet. It is exhibited that phosphorus concentration decreases smoothly with increasing argon ion sputtering time, which suggests that phosphorus is present through a segregation mechanism and is not as a precipitate film[23-24]. Just as E D Hondros and M P Seah[24] mentioned that a depth composition profile, showing the' element distribution with distance from the fracture surface is normally carried out by combining AES with controlled sputtering of the surface by inert gas ion bombardment. The procedure has been successful in defining the extent over which grain boundary segregation exists[24]. 2.0 , - - - - - - - - - - - - - - - - - - ,
K
X=VC/K
o
20
40
60
Time/s
80
500
Concentration of phosphorus in atomic pencent lis a function of sputtering time at a grain boundary
2. 2 The scatter of the cumulative average concentration at grain boundaries The amount of segregation to individual grain boundary, as determined by Auger spectroscopy in samples of a given condition, could be quite variable[25-27]. The deviation in grain boundary structure, fracture path and bulk chemical inhomogeneity can attribute to this variation'P'" , In order to acquire the segregation levels for samples of a given condition, the cumulative average levels of many individual grain boundaries often are adopted. For examples, T Shinoda and T Nakamura[29] had adopted the cumu-
(2)
where, Xi is the individual value of concentration at i-th grain boundary facet; and K is the number of Auger measurements. The Auger measurements are made on five separate Auger samples machined from the same rod that had aged for 20 min at 500 ·C after annealing at 1180 ·C for 45 min. Fig. 4 shows the scatter of a cumulative average levels, X, with number of Auger measurements K. It is found that the maximum and minimum levels derived from the cumulative average levels ofX=21, 22, ,·,,30 and 31 are 1. 753 % and 1. 705 % (in atomic percent) respectively. The average of these two values, 1. 729 %, is used as a most probable value of the segregation concentration of grain boundary for this condition. As can be seen from this figure, the scatter of the cumulative average of X from the levels of 1. 729 % is comparatively large when the number of K is smaller, but it always fell within the ± 1 % error band around 1. 729 %, when the number of K is increased to more than 20. Therefore, over 20 individual grain boundaries are measured for each condition to obtain the average in this study. 1.9 , - - - - - - - - - - - - - - - - , ± 1% error band for 1.729 1.8
?------I'!-
Fig. 3
• 63 •
Nonequilibrium Grain Boundary Segregation of Phosphorus in Ni-Cr-Fe Superalloy
Issue 1
i~ ::s .... 1.7
1.729
o 0 .0 0::
,~
:8
1.6 6i>E
~
~
§
'"
1.5 1.4
0
5
10 15 20 25 Number of measurementsIK
30
35
Fig. 4 Scatter of the cumulative average of phosphorus concentration in atomic percent, X, as a function of number of AES measurements K for samples aged at 500 'C for 20 min after annealing at 1180 °C
2. 3 Phosphorus concentration at grain boundary for different heat treatment conditions The variations of grain boundary concentrations (in atomic percent) of phosphorus with ageing time at
Journal of Iron and Steel Research, International
• 64 •
500 'C for this alloy after quenching from 1180 'C are shown in Table 2. Obviously, a peak of grain boundary concentration of phosphorus appears at about 180 min during isothermal ageing at 500 'C. When the ageing time is shorter than 180 min, the grain boundary concentrations of phosphorus increase with ageing time; when the ageing time is longer than 180 min, the grain boundary concentrations of phosphorus decrease with ageing time. Here the grain boundary concentration of phosphorus at 60000 min is the cumulative average levels of 6 individual grain boundary facets. It is shown from Fig. 4 that the cumulative average levels of 6 individual grain boundary facets have larger error than that of 20 individual grain boundary facets. Even so, the cumulative average levels at 60000 min also are lower than at 180 min. Table 2 The variation of phosphorus concentration at grain boundary with ageing time at 500 'c for Nl-Cr-Fe superalloy Ageing time/min
0
Concentration of P / %
1. 80
Note:
I)
20
180
6000
60000
1. 73
2.06
1. 17
1. 42
The initial segregation concentration is for the homogenized samples. 1)
Fig. 5 shows the grain boundary concentration Cin atomic percent) of phosphorus for the alloy aged
for 20 min at 200, 400, 500, 700 and 800 'C after quenching from 1180 'C, and indicates a maximum of concentration occurs at 500 ·C. When the ageing temperatures are lower than 500 'C, the grain boundary concentration of phosphorus decrease with decreasing ageing temperature. The grain boundary concentration of phosphorus at 700 'C and 800 'C is lower than that of phosphorus at 500 'C, and they have no obvious differences, which attribute to the limit of sensitivity of the Auger technique.
1.0 L....... 200
---'
-'---
--o-J
600 Temperaturel'C
800
400
Fig. 5 Phosphorus concentration at grain boundary as a function of ageing temperature for Ni-Cr-Fe superalloy aged for 20 min after annealing at 1180 'c
3
Vol. IS
Discussion
3.1 Critical time of phosphorus NGS in Ni-Cr-Fe superalloy According to the formation of solute atom-vacancy complexes within the matrix[12-13], R G Faulkner-P'' suggested the concept of a critical time. Subsequently, T XU[IO.30] experimentally indicated that the critical time should correspond to the peak in the solute grain boundary concentration during isothermal ageing after quenching. When the ageing time is shorter than the critical time, the grain boundary concentration of solute will increase with ageing time, which is referred to as the segregation process. When the ageing time is longer than the critical time, the grain boundary concentration of solute will decrease with ageing time, which is referred to as the de-segregation process. It is clear that the NGS concentration depends on how the ageing time approaches the critical time. When the ageing time is closer to the critical time, the NGS concentration is high even if the material is over-or under-aged compared to the critical time, as shown in Fig. 1 in Ref. [10]. This is the characteristic aspect of NGS, which has been verified for variant solutes in various alloy systems[31-33]. The formula for critical time was worked out by R G Faulkner[l5] and T XU[IO.14,30] :
r 21nCDc / D )
t c = 8CDc - D )
(3)
where, D i and D, are the diffusion coefficients for solute atoms and complexes respectively; r is the grain radius; and 8 is the critical time constant. The relationship between critical times and ageing temperature was calculated out of Eqn. (3) and shown in Fig. 3 and Fig. 4 of Ref. [10]. The critical time increases with the decrease of ageing temperature, which has been confirmed a lot of times such as In Ref. [10J, Ref. [19J, and Ref. [34]. The variations in Table 2 show that the grain boundary concentration of phosphorus reaches a peak after about 180 min during ageing at 500 'C following quenching from 1180 'C. This means the phosphorus has characteristic of NGS, and the critical time of phosphorus at 500 'C is at about 180 min for Ni-Cr-Fe superalloy. When the ageing time is shorter than 180 min, the process of segregation for NGS will be reached. So the grain boundary concentrations of phosphorus increase with increasing ageing time. When the ageing time is longer than 180 min, the process of de-segregation for NGS will be
Nonequilibrium Grain Boundary Segregation of Phosphorus in Ni-Cr-Fe Superalloy
Issue I
reached. So the grain boundary concentrations of phosphorus decrease with increasing ageing time.
3. 2 The critical time constant (; of phosphorus in NiCr-Fe superalloy Data on the diffusion coefficient of phosphorusvacancy complexes and the critical time constant ~ of phosphorus for Ni-base superalloy are scarce, which will limit the study on the segregation of phosphorus in Ni-base superalloy. Based on Ref. [35J, the self-diffusion coefficient of Ni , DNi-self, m 2 / s , is: DNi-self=1. 27 X 10- 4exp( -2. geV /kT) (4) which is characteristic of lattice diffusion in high-purity nickel. The most likely mechanism of self-diffusion in high-purity metals is that of continual migration of vacancies (vacancy diffusion) [34J. With vacancy diffusion, the probability that ali atom may jump to the next site will depend on: 1) the probability that the site is vacant (which in turn is proportional to the fraction of vacancies in the crystal) and 2) the probability that it has the required activation energy to make the transition. For self-diffusion, where no complications exist, the diffusion coefficient is, therefore, given by[36J : (5) DNi-self=Doexp[ -(Ef+Em)/kTJ where Do is a constant involving the frequency of atomic vibration or vacancy vibration. The frequency of vacancy diffusion is Do = 1. 27 X 10- 4. E, is the energy of formation of a vacancy, Em is the energy of migration of a vacancy, and the sum of the two energies, Q= E, Em' is the activation energy for self-diffusion in high-purity metals[36 J. The energy of formation of a vacancy is about 1. 5 eV in Ni-based alloys[37J. Therefore, the activation energy for vacancy diffusion, i. e. energy of migration of a vacancy, is Em = Q - E, = 2. 9 - 1. 5 = 1. 4 eV)n pure Ni. Therefore, the diffusion coefficient of vacancies in Ni-base alloys should be: D, =1. 27 X 10- 4exp( -1. 4 eV /kT) m 2 / s (6) The diffusion coefficient of phosphorus atom, D; in a Ni-base alloy is[3BJ : (7) D p=lX10- Bexp(-1. 78 eV/kT) m 2/s Solute phosphorus atoms in Ni-base alloys are in substitutional sites. For the same reason, the activation energy of phosphorus diffusion in Ni-base, alloys, Qp, is also suggested to be Qp = E, E p, where E, is still the energy of formation of a vacancy and E; is the energy of migration of phosphorus atoms in aNi-base alloy. Therefore, the energy of migration of phosphorus atoms is E p=Qp-Ef=1. 78,1, 5=0. 28 eV.
+
+
• 65 •
According to the mechanism suggested by S H Song and L Q Weng[39J, the activation energy for a vacancy-phosphorus complex in Ni-base alloys is the energy of migration of a vacancy or the energy of migration of phosphorus atoms. Selection of the former or latter depends on which is higher. Therefore, the activation energy for a vacancy-phosphorus complex in a Ni-base alloy should be 1. 4 eV. With reference to the concept of non-dissociation mechanism suggested by S H Song and L Q Weng[39J , for each step in the migration process of a vacancy-phosphorus complex, the vacancy requires three jumps, of which two are towards a nearest lattice site occupied by a matrix atom and the other jump is towards that occupied by phosphorus. As a consequence, it is reasonable to assume that the pre-exponential constant for diffusion of vacancy-phosphorus complexes is one-third of the pre-exponential constant for diffusion of vacancies. The pre-exponential constant for complex diffusion is, therefore, (1. 27 X 10- 4 )/3 = 4. 23 X 10- 5 • The diffusion coefficient of vacancyphosphorus complexes in a Ni-base alloy can be shown to be: Dv-p=4. 23X10- 5 e x p ( - 1 . 4 eV/kT) m 2/s (8) The critical time constant (J in Eqn. (3) can be calculated to be about (J = 47 from the critical time, 180 min, and the relative parameters above for this superalloy.
3. 3 The maximum segregation of phosphorus at 500°C for ageing for 20 min Fig. 5 shows that, for the experimented alloy aged for 20 min at 200, 400, 500, 700 and 800 'C after quenching from 1180 'C, the concentration of phosphorus at grain boundaries reaches a maximum at 500 ·C. From the critical time constant (J = 47 and the relative diffusion coefficients for phosphorus and complexes in Ni-base alloys, the critical times of different ageing temperatures t; can be calculated out of Eqn, (3). The calculated results are shown in Table 3, where it can be seen that the critical time decreases rapidly with increasing ageing temperature, which accords with the relationship between critical times and ageing temperatures mentioned above. It can be seen from Table 3 that when the ageing temperatures are lower than 500 'C, the critical times of phosphorus NGS increase rapidly. They are 4. 36 X 103 min at 400 'C and 1. 43 X 108 min at 200 'C, which are considerably longer than 20 min. Therefore, the grain boundary concentration of phosphorus will get lower and lower when the ageing time adopted is
Journal of Iron and Steel Research, International
• 66 • 'Thble 3
Critical time calculated at different ageing temperatures
Ageing temperature/"C Critical time/min
200
400
1. 43 X 10· 4. 36 X 10 3
500
700
800
180
2
0.4
still 20 mm , as shown in Fig. 5. It can be seen also from Table 3 that when the ageing temperatures are higher than 500 "C, the critical times of phosphorus NGS decrease rapidly. They are 2 min at 700 "C and o. 4 min at 800"C, which are considerably shorter than 20 min. Therefore, the grain boundary concentration of phosphorus will get lower and lower when the ageing time adopted is still 20 min, as shown in Fig. 5. As mentioned above, the NGS concentration of solute depends on how the critical time approaches the ageing time. Table 3 shows that the critical time of phosphorus at 500 "C is 180 min , which is closest to the ageing time of 20 min than at other ageing temperatures. Thus, the grain boundary concentration of phosphorus can attain maxima for an ageing of 20 min at 500 "C. Therefore, it may be stated that for a alloy system in which the solute has NGS characteristic, an ageing temperature must exist at which the NGS concentration of solute has a maximum for the alloy quenched from a high solution temperature and then aged at various temperatures for a certain time. The critical time at this ageing temperature is equal or close to the ageing time. This is another important characteristic of NGS, and has been confirmed for P in 304L stainless steel[19], Mg in Ni-Cr-Co alloyE34] , and S in Ni-Cr-Fe superalloy'<'".
4
Conclusions
1) The phosphorus of Ni-Cr-Fe superalloy has characteristic of NGS. 2) The critical time constant for phosphorus in this superalloy is about 47. 3) For the thermal cycle of Ni-Cr-Fe superalloy in this study, an ageing temperature 500 "C exists at which the NGS concentration of phosphorus has a maximum. The critical time of phosphorus at the ageing temperature 500 "C is very close to the ageing time 20 min, which was adopted at all the ageing temperatures. References: [lJ [2J
[3J
Holt R T. Wallace W. Impurities and Trace Elements in Nickel-Base Superalloys [JJ. Inter Metals Rev. 1976,21: 1. Yeniscavich W. Fox C W. Effect of Minor Elements on the Weldability of High Nickel Alloys [M]. New York: Welding Research Council. 1969. Vermilyea D A. Tedmon C S. Broecker D E. Effects of Phos-
Vol. 18
phorus and Silicon on the Intergranular Corrosion of a NickelBase Alloy [JJ. Corrosion. 1975. 31: 222. [4J Was G S. Kruger RM. A Thermodynamic and Kinetic Basis for Understanding Chromium Depletion in Ni-Cr-Fe Alloys [J]. Acta Metall. 1985. 33(5): 841. [5J Sung J K. Was G S. Intergranular Cracking of Ni-16Cr-9Fe Alloys in High-Temperature Water [J]. Corrosion. 1991. 47 (11): 824. [6J Was G S. Sung J K. Angeliu T M. Effects of Grain Boundary Chemistry on the Intergranular Cracking Behavior of Ni-16Cr9Fe in High-Temperature Water [J]. Met Trans. 1992. 23A (12): 3343. [7J Cao W D. Kennedy R L. Superalloys 718. 625. 706 and Various Derivatives [MJ. Warrendale: TMS. 1994. [8J Liu X. Liu H. Dong J X. et al. Molecular Dynamics Simulation on Phosphorus Behavior at Ni Grain Boundary [JJ. Scripta Mater. 2000. 42(2): 189. [9J Guo J T. Zhou L Z. Superalloys 1996 [M]. Warrendale: TMS. 1996. [10J XU T. Cheng B. Kinetics of Non-Equilibrium Grain Boundary Segregation [JJ. Prog Mater Sci. 2004. 49(2): 109. [l1J McLean D. Grain Boundaries in Metals [MJ. Oxford: Oxford University Press. 1957. [12J Aust K T. Hanneman R E. Niessen P. et al. Solute Induced Harding Near Grain Boundaries in Zone Refined Metals [JJ. Acta Metall. 1968. 16: 291. [13J Anthony T R. Solute Segregation in Vacancy Gradients Generated by Sintering and Temperature Changes [JJ. Acta Metall. 1969. 17(5): 603. [14J XU T. Song S. A Kinetic Model of Non-Equilibrium Grain Boundary Segregation [JJ. Acta Metall. 1989. 37: 2499. [15J Faulkner R G. Non-Equilibrium Grain-Boundary Segregation in Austenitic Alloys [JJ. J Mater Sci. 1981. 16: 373. [16J Li Q. Yang S. Li L. Experimental Study on Non-Equilibrium Grain-Boundary Segregation Kinetics of Phosphorus in an Industrial Steel [JJ. Scripta Mater. 2002. 47: 389. [17J Seve P. Janovec J. Lucas M. Kinetics of Phosphorus Segregation in 2. 7Cr-0. 7Mo-0. 3V Steels With Different Phosphorus Contents [JJ. Steel Res. 1995. 66(12): 537. [18J Chen W, Chaturvedi M C. Richards N L. Effect of Boron Segregation at Grain Boundaries on Heat-Affected Zone Cracking in Wrought Inconel 718 [JJ. Metal Mater Trans. 2001. 32A (4):931. [19J Wang K. Wang M Q. Si H. et al. Critical Time for Non-Equilibrium Grain Boundary Segregation of Phosphorus in 304L Stainless Steel [JJ. Mater Sci Eng. 2008. 485A: 347. [20J Briant C L. Grain Boundary Segregation of Phosphorus in 304L Stainless Steel [JJ. Metall Trans. 1985. 16A: 2061. [21J Briant C L. Grain Boundary Segregation of Phosphorus and Sulfur in Type 304L and 316L Stainless Steel and Its Effect on Intergranular Corrosion in the Huey Test [JJ. Metall Trans. 1987. 18A: 691. [22J Davis L E. McDonald M C. Palmberg P W. et al. Handbook of Auger Electron Spectroscopy [MJ. 2nd ed. Minnesota: Physical Electronics Division of Perkin-Elmer Corporation. [23J
[24J [25J
1976. Dong J. Zhang M. Xie X, et al. Interfacial Segregation and Cosegregation Behaviour in a Nickel-Base Alloy 718 [JJ. Mater Sci Eng. 2002. 328A: 8. Hondros E D. Seah M P. Segregation to Interfaces [JJ. Int Met Rev. 1977. 222: 262. Watanable T. Kitamura S. Karashima S. Grain Boundary Hardening and Segregation in Alpha Iron-Tin Alloy [JJ. Acta
Issue 1
[26J
[27J
[28J [29J
[30J
[31J [32J
Nonequilibrium Grain Boundary Segregation of Phosphorus in Ni-Cr-Fe Superalloy Metall. 1980. 28: 455. Hall E L. Imeson D. Vander sande J B V. On Producing High-Spatial-Resolution Composition Profiles via Scanning Transmission Electron Microscopy [n. Phi Mag. 1981. 43A: 1569. Kimeda J. Mcmanhon Jr C J. The Effects of Sb, Sn , and P on the Strength of Grain Boundaries in a Ni-Cr Steel Metall Trans. 1981. 12A: 31. Briant C L. Source of Variability in Grain Boundary Segregation [n. Acta Metall. 1983. 31(2): 257. Shinoda T. Nakamura T. The Effects of Applied Stress on the Intergranular Phosphorus Segregation in a Chromium Steel [n Acta Metall. 1981. 29: 1631. Xu T. The Critical Time and Critical Cooling Rate of Non-Equilibrium Grain-Boundary Segregation [n. J Mater Sci Let. 1988. 7: 241. Ding R G. Rong T S. Knott J F. Phosphorus Segregation in 2. 25Cr-1Mo Steel [n. Mater Sci Tech. 2005. 21(1): 85. Li Q. Liu E. Liu D. et al. Temper Embrittlement Dynamics Induced by Non-Equilibrium Segregation of Phosphorus in
[n.
[33J [34J [35J [36J [37J
[38J [39J [40J
• 67 •
Steel 12CrlMoV [n. Scripta Mater. 2005. 53(3): 309. Zhang Z L. Lin Q Y. Yu Z S. Grain Boundary Segregation in Ultra-Low Carbon Steel DJ. Mater Sci Eng. 2000. 291A: 22. Xu T. Critical Time for Mg Grain-Boundary Segregation in Nic-ce alloy [n. Phi Mag Lett. 2006. 86(8): 501. Hoffman R E. Pikus F W. Ward R A. Self-Diffusion in Solid Nickel [n Trans AIME. 1956. 206: 483. SmallmanR E. Modern Physical Metallurgy [M]. London: Butterworth. 1963. Kim S M. Buyers W J L. Vacancy Formation Energy in Iron by Positron Annihilation [1]. J Phys F. Metal Phys , 1978. 8L: 103. Brand E A. Brook G B. Smithells Metals Reference Book [MJ. 7th ed. London: Butterworth-Heinemann Ltd. 1992. Song S a. Weng L Q. Diffusion of Vacancy-Solute Complexes in Alloys [n. Mater Sci Tech. 2005. 21: 305. Wang K. Xu T. Wang Y Q. et al. Intermediate-Temperature Embrittlement Induced by Non-Equlibrium Grain-Boundary Segregation of Sulfur in Ni-Cr-Fe Alloy [1]. Phi Mag Lett. 2009. 89(11): 725.