Materials Sciences and Applicatio ns, 2011, 2, 1243-1255
doi:10.4236/msa.2011.29168 Published Online September 2011 (http://www.SciRP.org/journal/msa)
1243
The Role of Magnesium in Superalloys—A Review
Kumkum Banerjee
Research and Development Division, Tata Steel Limited, Jamshedpur, India.
Received February 17th, 2011; revised May 20th, 2011; accepted June 17th, 2011.
ABSTRACT
The role of magnesium (Mg) in improving the high temperature mechanical properties of the superalloys, like creep,
fatigue, tensile ductility, impact toughness etc. have been vividly studied by several authors. On the other hand, very few
authors have contradicted the view of any beneficial effect of Mg on the mechanical properties. This review presents a
summary of the open literature related to the effect of Mg on the microstructu re and mechanical properties of su peral-
loys and from whi ch f urt her metallurgi cal re search on the unexamined topics are prop o sed.
Keywords: Superalloy, Magnesium
1. Introduction
Over more than three decades several studies have been
made on the influence of micro alloying of wrought and
cast superalloys with magnesium. It was well recognized
that a small addition of minor elements such as B, Mg,
Ca and rare earths etc. could improve the stress rupture
and creep properties significantly [1]. The results with
Mg addition indicate that an optimum addition of mag-
nesium causes enhancement in high temperature me-
chanical properties. e.g. creep life, creep rupture duc-
tile= ity, high temperature tensile ductility, cyclic stress
rupture properties creep-fatigue interaction, crack propa-
gation and hot workability [2-4]. Mg has been shown to
improve the creep properties and particularly the high
temperature ductility of the wrought alloys due to re-
finement of the grain boundary carbides and equilibrium
segregation [5]. Magnesium addition can also prolong
secondary creep stage and develop tertiary creep stage
and simultaneously alloys possess longer stress rupture
life than the alloys without Mg [6]. The detrimental ef-
fect of sulfur has been found to be reduced by the addi-
tion of Mg. The results on cast IN-718 superalloy
showed that a small amounts of Mg improved impact
toughness and decreased Nb segregation by decreasing
secondary arm spacing, which resulted in less and
smaller interdendritic Laves and MC eutectics [5].
Moreover, magnesium also decreased the quantity of 
eutectic by segregating to the phase boundaries and thus
refining the  eutectic. Small amounts of Mg produced a
more spherical as well as more dispersive MC phase. In
contrast to this, detrimental effects of Mg on creep life
and ductility has been reported in some studies [7,8]. In
the attempt to achieve favourable properties many invest-
tigations have been made to assess the optimum Mg
concentration in the alloys and its relationship to micro-
structure and properties [1,9-16]. However, common
conclusions have not yet been achieved on this problem.
This might be due to the different procedures used to
determine the Mg percent and lack of comparability of
the influence of magnesium in small-scale laboratory test
and industrial ingot.
Thus, the unique effect of Mg has been established for
many superalloys and the beneficial influence of Mg has
been attributed to the refinement of carbides, -phase and
Laves phase on grain boundary and reducing the detri-
mental effect of sulfur. In this paper, the existing findings
in this area are critically reviewed and the unexamined
areas are pointed out for carrying out further research in
this direction.
2. Influence of Mg on Mechanical
Properties and Weldability of
Superalloys
2.1. Mechanical Properties
The high temperature symmetrical low cycle fatigue
(LCF) (R = –1) and unsymmetrical LCF (R = 0.42) were
studied by Xie et al. [15] on IN- 718 iron and nickel-base
superalloys and it was reported that the micro alloying of
Mg (30-50 ppm) showed beneficial effect on unsymmet-
rical LCF, where crack growth rate decreased by a factor
of 3 - 7. However, symmetrical tension-compression stress

The Role of Magnesium in Superalloys—A Review
1244
controlled high temperature LCF properties remained
unaffected.
In the research by Ma et al. [4,17] it was observed that
an optimal addition of Mg to the wrought superalloys
(GH33, GH220) prolonged the secondary and especially
the tertiary creep (700˚C/392 MPa) at low strain rate of
steady state creep stage ( < 10–5 mmmm–lh–1). How-
ever, at greater strain rate the effect of Mg vanished. In a
different study, Zhong et al. [2] explained about the
similar observation on the effect of Mg on creep behavior
(700˚C/343 MPa) of a Ni-base superalloy (GH33). In this
case also an optimum Mg addition prolonged the time for
creep crack nucleation and growth rate and which in turn
increased the rupture life. In support to this, G. Chen and
H. Ge [18] have showed the improvement of creep life
and elongation of GH 698 (Figure 1) due to prolongation
of secondary creep stage and the development of tertiary
creep stage. However, no evidence of influence of creep
rate on the alloy was observed. Likewise, several studies
have been made by other researchers [1,19,20] and they
have also mentioned the beneficial effect of Mg on creep
properties, which has been manifested by the increase in
secondary and development of tertiary creep stages and
thus improving the rupture life and the elongation. The
improved influence of Mg on creep is attributed to the
improved grain boundary ductility arising from the in-
fluence of Mg on the formation and growth of creep
cavities. The stress rupture life and ductility observed to
be increasing with the increase in Mg content up to an
optimum level followed by a decrease in the properties
with further addition of Mg [18]. It is noteworthy that the
optimum Mg level is different for different alloys. The
creep properties, rupture life and stress component of
creep of the superalloys containing Mg were examined
by Danien Ke [21] and reported to have significantly
improved ductility and rupture life by the impediment of
grain boundary sliding and the prevention of nucleation
Figure 1. Creep curves of alloy A (containing 0.005% Mg)
and B (no Mg) at 750˚C and 343 MPa [18].
and propagation of creep crack growth. The magnesium
(0.0094%) addition to GH 169 improved stress rupture
life and stress rupture ductility at temperature 650˚C
(Figure 2) [1]. The alloy with 0.0094% Mg gradually
softened during long time exposure at 650˚C, therefore
stress rupture life decreased and ductility increased
mildly. However, the properties were still better than the
alloy containing negligible amount of Mg (0.0001% Mg).
Bor et al. [22-24] extensively worked on the effect of Mg
on creep characteristics, fracture mechanism and carbide
characteristics of MAR-M247, a Ni-base superalloy un-
der 1255K/200 MPa and 1033K/724 MPa conditions.
They observed that under the former testing condition
0.005% Mg refined and spheroidized MC carbide at GB,
which enhanced the creep properties, whereas 0.008%
Mg addition increased the number of MC at GB, which
significantly decreased rupture life and elongation.
However, under the condition of 1033K/724 MPa, it was
observed that the rupture life and elongation both sig-
nificantly improved to 3 - 5 times with 80 ppm Mg. In
this work the transition of fracture initiation site from
carbide matrix to a more ductile interface, /' has been
reported [24]. A decrease in the steady state creep rate in
the Mg containing alloys is reported in the literature.
Conversely, Chen et al. [25] have emphasized about the
insignificant influence of Mg addition on steady state
creep rate, but mentioned that by Mg addition creep life
and elongation (750˚C/1343 MPa) were improved. The
segregation of Mg on the cavity surface was reported to
be instrumental in lowering the creep cavity growth rate.
Chen et al. [26] in their work on three wrought super-
alloys including IN-718, studied the effect of Mg on
creep properties (650˚C - 750˚C at different stresses),
LCF (700˚C), cyclic stress rupture (650˚C) and fatigue-
creep interaction properties (1 HZ, 650˚C, in air). They
Figure 2. Effect of long time exposures at 650˚C on 650˚C
/686 MPa stress rupture [1].

The Role of Magnesium in Superalloys—A Review 1245
reported that Mg did not affect the steady state creep rate.
The LCF properties and crack growth rate of LCF at
700˚C were also unaffected by micro addition of Mg.
Conversely, creep rupture life and fatigue failure life
were increased in the alloy containing magnesium by
reducing the cavity growth rate of the crept samples and
increasing the number of cycles for fatigued specimens
respectively. Xie et al. [27] in their investigation de-
scribed that Mg helped increase the stress rupture life of
both the smooth and notched specimen (Figure 3), cyclic
stress rupture and also LCF (Figure 4), cyclic stress rup-
ture and also LCF at any grain size. It implies therefore,
that the views on the effect of Mg on LCF are at vari-
ance.
There are various literatures explaining the enhance-
ment of stress rupture ductility [28-30] improvement by
small addition of Mg (1 - 350 ppm). In the early 70s
Couts et al. [28] studied the effect of Mg (1 - 350 ppm)
on the mechanical properties of alloy IN-718 and showed
stress rupture ductility improvement in the range of
Figure 3. Grain size and Mg effect on stress rupture life and
elongation at 650˚C and 686 MPa [27].
Figure 4. Grain size and Mg effect on cyclic stress rupture
life with different holding times at maximum stress of 686
Mpa at 650˚C (1, 2-5 sec, 3, 4-180 sec, 5, 6-1800 sec) [27].
30 - 200 ppm but little was presented in the lower range
(up to 100 ppm) of Mg. In 1971 Muzyka et al. [29]
showed beneficial stress rupture ductility improvement at
30 ppm Mg. In 1984, Moyer [30] in his study with extra
low-carbon alloy showed a remarkable stress rupture
ductility and life improvement with a small addition of
Mg (13 - 19 ppm). The beneficial effect of Mg in alloy
IN-718 was observed to be maintained even after long
time exposure at 650˚C [27]. Liu et al. reported [7] Mg
had no beneficial effect on the stress rupture properties as
well. This was accounted for the small Mg (O, S) parti-
cles, which increased void nucleation sites demonstrating
negative influence on the mechanical properties. Addi-
tionally due to low percent of sulfur, Mg formed Ni2Mg-
Laves phase, which was harmful to the properties of the
alloys. They have also suggested that if the sulfur content
can be controlled to a low level, there is no necessity to
add Mg from the industrial point of view.
The stress rupture (smooth and notched), creep-fatigue
interaction, notched cyclic stress rupture-all these prop-
erties were reported to be improved with the addition of
micro alloying of Mg (30 - 70 ppm) by the researchers [3,
31]. Crack propagation rate was found to be decreased as
well. The stress rupture life of notched 70 ppm Mg alloy
increased from 80 h to 350 h whereas rupture life for
smooth specimen increased from 70 h to 100 h only. In
creep-fatigue interaction cycle, life to failure of higher
Mg (70 ppm) alloy was apparently longer than lower Mg
containing alloy. Similarly notched stress rupture life of
higher Mg (70 ppm) alloy was 2158 cycles whereas, it
was 653 cycles for lower Mg containing alloy. These
beneficial effects were attributed to the improvement in
distribution, morphology and quantity of -phase (Ni3Nb)
at the grain boundary due to the addition of Mg.
The effect of grain size and Mg addition on stress rup-
ture and notched cyclic stress rupture were studied [32]
and it was concluded that a minute amount (59 ppm) of
Mg and grain refinement might improve high tempera-
ture stress rupture notch sensitivity and decrease failure
lives of Ni-Fe base superalloy IN-718 even in the pres-
ence of duplex grain structure.
A small percent of Mg also prolonged the secondary
creep and tertiary stages of creep resulting in increased
stress rupture life and ductility by ~ 1.5 - 2.5 times [1].
The optimum content of Mg moved down to lower level
by the high temperature long time exposure. There was
not a significant effect observed by the addition of Mg on
the steady state creep rate. Likewise, Magnesium was
reported to have no beneficial effect on the stress rupture
properties of Inconel 718 at 650˚C, 686 MPa while the
sulfur content is not more than 10 ppm [33].
The effect of Mg on cast superalloys has been investi-
gated in the late 80 s by Chen et al. [5] and favourable

The Role of Magnesium in Superalloys—A Review
1246
reason for this effect was not explained although. Bor et
al. [22-24] observed that in MAR-M247 nickel based
alloy, 0.008% Mg increased the number of MC on grain
boundary, which significantly decreased creep rupture
life and elongation of the alloy.
effect of Mg on impact toughness was reported which is
shown in Figure 5, which is possibly due to the decrease
in interdendritic Laves eutectic (Figure 6) and the re-
finement of interdendritic segregation of Nb and Ti,
which allowed shorter homogenization cycle. In Figure
7 it is indicated that as the Mg content increases the
Laves phase decreases, the amount of  plates around the
Laves islands increases. This occurs because the amount
of Nb necessary for Laves phase formation is too small
but the Nb content as high enough for -phase formation.
The - plates can be eliminated by homogenization much
easier than the Laves islands. A large number of frac-
tured MC and Laves particles were observed in the alloy
without Mg, whereas a negligible amount of the same
was noticed in the alloy containing Mg (Figure 8).
Further, Liu et al. [7] reported about the detrimental or
no effect of Mg (76 ppm - 94 ppm) in their recent studies
(in 2001) on IN-718 containing low amount of sulfur
content (<10 ppm). Tensile properties did not respond to
the addition of Mg content while the tests were con-
ducted at room temperature, 500˚C and 650˚C and the
Mg had no beneficial effect on the stress rupture proper-
ties as well.
2.2. Weldability
ln another study of Ford [8] it was declared that Mg
had a dual influence on the material property. He pointed
out that although Mg reduced micro porosity in superal-
loys and increased castability but on the other hand, it
decreased high temperature creep life and ductility. The
The influence of Mg on IN-718 weld micro fissuring was
investigated by Morrison et al. [34]. The evaluation dis-
closed that magnesium had some positive influence on
the weldability of the alloy. The tests carried out by them
in the year 1966 showed no micro fissuring in their lon-
(a)
Figure 5. Effect of Mg addition on impact toughness in cast
alloy IN-718 [5].
(b)
Figure 7. Effects of mg and  plates around the Laves is-
lands in cast alloy IN-718 (a) alloy without Mg and (b) alloy
with Mg [5].
Figure 6. Effect of Mg contents on the quantity of laves
eutectic in cast alloy IN-718 [5].

The Role of Magnesium in Superalloys-A Review 1247
(a) (b)
Figure 8. Micro cracks in impact samples (cast IN-718) (a) alloy without Mg and (b) alloy with Mg [5].
gitudinal and transverse sections of the alloy by the addi-
tion of Mg. The tests carried out in 1967 on the
prewelded specimens (1150 ˚C/1 hr./AC), welded by TIG
showed no cracking with 37 ppm of Mg. However, when
the tests were conducted at the same preweld heat treated
condition with 52 ppm Mg, micro fissuring was observed.
Whereas the preweld temperature of 1060˚C was com-
patible with both the Mg contents. Their result indicated
that around 20 ppm of Mg weldability improvelnent
started and at around 30 ppm Mg and a plateau value was
observed.
Liu et al. [7] reported about the detrimental or no ef-
fect of Mg (76 ppm-94 ppm) in their recent studies (in
2001) on IN-718 containing low amount of sulfur content
(<10 ppm). Tensile properties did not respond to the ad-
dition of Mg content while the tests were conducted at
room temperature, 500˚C and 650˚C. High temperature
tensile ductility of wrought heat resistant alloy EP 199,
showed a consistency in ductility in the work of Topilin
and Tsvetkeva [35]. A fairly high strength and ductility
characteristics at 20˚C and 900˚C were obtained with Mg
concentration of 150 - 290 ppm. They declared that there
was a critical amount of Mg (0.005%) below which the
ductility of the alloy decreased sharply. An improved
high temperature tensile ductility in a Mg containing
Ni-base alloy was also reported by Xu et al. [31], but a
little effect was realized on tensile strength. The tensile
properties at high temperature of 650˚C were reported to
be improved with 30-70 ppm alloying of Mg [3,31]. A
significant increase in reduction in area (45%) and elon-
gation (27%) was reported. Further, the beneficial effect
of higher yield and ultimate tensile strength observed at
ambient temperature, due to a small amount of Mg addi-
tion, was observed to be disappeared at higher tempera-
ture [1]. Further, Xie et al. reported that Magnesium had
almost no influence on tensile strength and ductility in
Inconel 718 alloys with low content of sulfur at room
temperature, 500˚C and 650˚C [33].
2.3. Hot Ductility Behavior of Superalloys
Liu and his group [36] performed a series of hot tensile
tests to study the effect of S and Mg on hot ductility of
IN-690 alloy and reported about the beneficial effect of
an appropriate addition of Mg. It was observed that with
the increase in S content (0 - 80 ppm) there was sharp
decrease in ductility at high temperatures 900˚C - 1150˚C
in the absence of Mg. However, a very low level of sul-
fur content (<10 ppm) showed excellent ductility in the
temperature range 900˚C - 1200˚C. But when the sulfur
content exceeded 20 ppm, the value of reduction in area
was less than 50% at 900˚C - 1000˚C, which corre-
sponded to poor hot ductility. It was indicated that sulfur
had less influence on ductility at relatively higher tem-
peratures (>1050˚C), which was accounted for softening
of the alloy matrix and the possibility of increase in sul-
fur solubility at higher temperature. When sulfur was
about 40 ppm, the ductility was reported to be better with
160 ppm of Mg. But at the same sulfur level higher Mg
content had detrimental effect on the ductility. In the late
70’s Yamaguchi et al. [37] studied the effect of minor
elements (S, Ca, Mg, Y and Zr) on hot workability of
solid solution strengthened Ni-base superalloys. The hot
workability of Ni-base superalloys is mainly controlled
by dS = %S – 0.8 X %Ca – 0.3 X Mg % – 0.5 X% Y –
0.1 X %Zr. The dS corresponds to the residual amount of
sulfur that is not fixed by strong sulfide forming ele-
ments-Ca, Mg, Y and Zr. An excellent hot ductility can
be achieved with 0.003 > dS > 0.004. The ductility de-
creased gradually with dS < 0.004 and became extremely
poor when dS > 0.003. Both dS > 0 and dS < 0 lowered
hot ductility. Hence in order to achieve superior ductility
dS should be nearly zero.
In order to strengthen the grain boundaries and change
the type of fracture from intergranular to transgranular

The Role of Magnesium in Superalloys—A Review
1248
with a corresponding increase in ductility alloying boron,
zirconia, magnesium and hafnium is recommended [38,
39] Boron atoms, which are not chemically bonded, to-
gether with the atoms of zircon, hafnium and magnesium
segregate to grain boundaries and contribute significantly
to improving the high temperature properties of superal-
loys at elevated temperatures [40].
From the aforementioned reported results it is implied
that a small amount of Mg can significantly improve high
temperature mechanical properties of some Ni and
Fe-base superalloys. However, to maintain the optimum
Mg concentration in various alloys, different techniques,
e.g., VIM, VAR, ESR, VADER (Vacuum Arc Double
Electrode Remelting) are being used and have been dis-
cussed in the literature [41-43]. Yeniscavich and Fox [44]
served a detrimental effect of Mg and Si on zero ductility
while working on Hastelloy X.
3. Influence of Mg on Microstructure and
Mechanisms Involved
Magnesium has a great influence on the size, morphol-
ogy, distribution and quantity of carbide and phases at
grain boundaries. The studies indicated that Mg is a sur-
face-active element and segregation of Mg takes place at
the grain boundaries and the interphases like /carbide
and / [45,46]. The segregation of Mg at the interfaces
and interphases in turn accounted for the surface-active
characteristic of Mg and hence its high propensity for
segregation to interfaces and interphases [22]. Mg
formed an enriched layer of Mg surrounding the carbide
and might influence the transport of carbon and carbide
forming elements. This resulted in an isotropic growth of
carbides during solidification and eventually led to the
refinement and spheroidization of MC carbides both at
the grain boundary and within the grain interior [4,
22-24]. The spheroidization resulted from Mg addition
improves the stress distribution state at grain boundary
thus decreases the possibility of wedge crack at the grain
boundary [17].
Hence the coalescence of grain boundary micro voids
generated by carbides became the prominent mechanism
of creep crack initiation and the crack initiation and
propagation in the tertiary stages of creep were retarded
and thus the rupture life was subsequently prolonged.
The spheroidization of grain boundary carbides resulting
from the micro alloying of Mg resisted the propagation
of cracks and the carbides at the tip of the crack may
become the nucleating centre for further growth of cracks.
However, the stress concentration around the spher-
oidized carbides is smaller than that of the plate-like car-
bides and hence the possibility of crack nucleation is
decreased lowering the crack growth rate. An optimal
Mg addition inhibits the precipitation of coarse MC car-
bide and causes the formation of a number of discrete
M23C6 carbides at grain boundary. The morphology and
distribution of carbides at grain boundaries are the major
factors in determining the creep behaviour. These struc-
ture changes decreased stress concentration at the inter-
faces/grain boundaries during creep and hence prolong
secondary and tertiary creep stages and simultaneously
increases ductility at fracture. Generally a large amount
of fine and discrete and less number of MC on grain
boundary is beneficial for grain boundary strengthening
and for inhibiting grain boundary migration. The discrete
GB carbides are generally considered beneficial since
they inhibit GB sliding and retard the onset of creep
cavitation and rupture. Typically over addition of mag-
nesium resulted in an exceptionally high amount of MC
carbide at the grain boundary resulting in a sharp de-
crease in M23C6 carbides. It has been reported that coarse
GB MC carbide could weaken the boundaries and enhance
intergranular cracking [47,48]. TEM results showed [6]
that Mg addition to alloy GH36 could change grain boun-
dary carbide distribution from continuous plate form to
globular shapes and the M23C6 could be distributed in a
non-continuous chain form (Figure 9(a), (b)). Fracto-
graphic analysis showed that the intergranular cracks
changed from wedge type to cavity type to cavity type
(Figures 9(c), (d),) in a stress rupture test conducted at
650˚C, which implied Mg retarded intergranular crack
propagation. SEM observation showed ductile fracture
characteristics in modified GH 36 alloy in comparison
with unmodified GH 36 (Figures 9(e), (f)). All these
indicate that Mg addition strengthens the grain bounda-
ries mainly due to the retardation in intergranular void
initiation and the decrease in creep crack growth rate.
Another mechanism of Mg micro addition in superal-
loy causing the ductility enhancement, has been put for-
ward by many researchers [1-2,6,49-51], is to purify the
grain boundary through binding the detrimental elements
like S, P, O etc. However, if the amount of the detrimen-
tal species is far below 5 ppm, then this mechanism
might not be applicable.
The energy concept is a well-accepted mechanism by
many researchers [2,42,50-52]. According to this mecha-
nism, Mg segregates to the grain boundary and carbide/
matrix interfaces leading to the lowering of interfacial
energy and consequently increasing the cohesion energy/
rupture energy between the carbide and the matrix or the
boundary and the matrix [53,54]. Under this circum-
stances crack path changes from the carbide/matrix in-
terface to  interface, a region having superior ductil-
ity [55]. The crack initiation and propagation are even-
tually retarded and rupture life and elongation are im-
proved.
The beneficial influence of Mg in changing the GB

The Role of Magnesium in Superalloys-A Review 1249
(a)
(b)
(c)
(d)
(e)
(f)
Figure 9. Effect of magnesium on grain boundary carbide morphology (a, b), mode of grain boundary cracks (c, d) and char-
acter of intergranular fracture (e, f) of GH 36 [6].
-Ni3Nb cellular precipitates is also reported in the lit-
erature [1,4]. The continuous cellular precipitate/plate
like precipitates of -Ni3Nb transforms to small amount
of phases or discrete globular shapes in the presence of
Mg hence retards intergranular crack growth, which si-
multaneously increases stress rupture ductility and pro-
longs failure life. The amount of -Ni3Nb depends on
grain size, amount of Mg and heat treatment. However,
the addition of Mg does not reported to have any signifi-
cant influence on main strengthening phases  and  [1,
27]. Micro mechanical phase analysis results showed that
the amount of strengthening phases of  and  phase is
not affected by Mg addition or grain size in alloy IN-718
and 718M as shown in Figure 10 [27]. Mg-free and
Mg-containing 718 M, both contained approximately
14% + independent of grain size. However, -Ni3Nb
precipitation at grain boundaries increased with grain
refinement and increase in amount of Mg.
It has also been suggested that Mg with large atomic
radius in the grain boundary areas might decrease the
vacancy density and the diffusion coefficient of vacan-
cies. The initiation and propagation rate of creep voids

The Role of Magnesium in Superalloys—A Review
1250
Figure 10. Grain size and Mg effect on the amount of  + 
and -Ni3Nb [27].
are proportional to the grain boundary diffusion coeffi-
cient of vacancies [56,57], therefore, the segregation of
Mg at grain boundary retards the initiation and propaga-
tion of creep voids. Concentration of Mg at grain bound-
ary plays a strengthening role on grain boundaries [58].
The reduction of steady state creep rate at low strain
rates pointed out by some researchers is considered to be
due to the presence of Mg in the matrix [4,17]. Lagne-
berg et al. [59] in their study of the creep behaviour of
precipitation strengthening types alloys under various
stresses at intermediate temperature reported that at low
strain rate condition ( < 10–5 mmmm–1h–1) the disloca-
tion movement could not be realized by Orwan mecha-
nism or shearing particles but dislocation climbing over
' particles. At low strain rate climbing of dislocation on
' depends upon vacancy concentration in the matrix. The
increase in vacancy activation energy in Mg containing
alloys causes the decrease in vacancy concentration.
Therefore, the climbing velocity of dislocation over ' a
particle reduces and a low steady state creep rate pro-
ceeds. At higher strain rate the vacancy concentration
arising from the Mg disappears.
3.1. Formation of Laves Phase
With an appropriate amount of Mg in IN-718, the shape
of the carbides at the interfaces was found to be smoother
than that was without Mg [54]. The incoherent interfaces
like carbide/matrix (or '-phase) promoted the formation
of small Ni2Mg, wrapped around the carbides and thus
smoothened the carbide shape improving high tempera-
ture creep rupture and fracture toughness properties.
They have emphasized on the tact that a very small sized
Laves phase (~10 nm) can improve high temp creep
properties and this was detected by high resolution elec-
tron microscope with ~20 nm probe size. Zhu et al. [60]
also noticed, about the similar wrapping of Ni2Mg phase
around the carbide, while using TEM. However, the
presence of a Laves phase, Ni2Mg, is considered to be
detrimental for mechanical properties.
3.2. Mg and S Interaction
The beneficial effect of Mg on intergranualar fracture has
been established by many methods [1]. The Mg influ-
ences the growth rate of creep cracks in steady state
range. To understand the segregation behaviour of Mg
during creep AES analysis was carried out at the crept
specimens at different stages [4]. The tests were inter-
rupted at different stages of the creep test and the speci-
mens were being taken out and broken keeping them in
liquid nitrogen to obtain intergranular fracture and hence
measured the segregation behaviour of Mg by AES. The
Mg distribution at the grain boundaries was found to be
quite inhomogeneous possibly due to the variation of
grain boundary structure, or the microstructure near the
grain boundary. At the initial stage of creep under the
action of applied stress, Mg and S solutes redistribute
themselves. Hence the Auger peaks of Mg and S at many
GBs were eliminated/reduced. The creep cavities then
nucleated at the segregation free condition of Mg so it
was assumed that there is no effect of Mg on the initia-
tion of creep cavities, which contradicted the results
mentioning about the active role of Mg in crack initiation.
However, as cavities formed, Mg and S segregated to the
cavity surfaces due to the high surface activity of the
elements and stress free condition at cavity surface. It has
been reported that the segregation of sulfur at cavity sur-
face with Mg is less than without Mg containing cavities.
According to the cavity growth theory surface tension
and surface diffusion coefficient both decrease by Mg
segregation and hence process of cavity growth and link
between the cavities are influenced by Mg segregation,
which lowers the cavity growth rate [61].
The fracture surface analysis by SEM while con-
ducting hot ductility testing showed ductile appearance at
very low level of sulfur [35]. When sulfur increased the
fracture surface appeared to be intergranular and sulfur
was found on grain boundaries. However, at relatively
higher sulfur level the fracture surface showed partially
ductile fracture with appropriate Mg addition and no
segregation of Mg and S was observed. Furthermore,
excessive Mg addition produced Mg particles in the alloy
matrix and sulfur segregation at grain boundary reduced
with ductile fracture but ductility was not improved.
There are some negative views regarding Mg-S inter-
action. The detrimental interaction of these elements
have been reported in the work of Liu et al. [7] (2001)

The Role of Magnesium in Superalloys—A Review 1251
containing low sulfur content. The fractographic and
micrographic analysis by SEM reveal that Mg had little
effect on the microstructure of IN-718, i.e. grain size,
-phase etc. The Auger research revealed no segregation
of Mg at GB. The phase diagram of Ni-Mg shows no
solution of Mg in Ni-matrix and Mg tends to segregate
at the GB of the matrix. However, it is a matter of con-
troversy since IN-718 is a multicomponent system and
there is complex interaction among the alloying ele-
ments. The sites at grain boundaries, which can absorb
solute atoms, are limited. Hence the diffusion velocity of
Mg is lower than that of the other elements and there is
no room for Mg segregation. In this case, Mg appears
either as MgO, MgS or Mg (O, S). However, Mg (O, S)
is supposed to increase void nucleation sites, which is
detrimental for, mechanical properties. This is the reason
put forward by the authors [7] for the opposite effects on
mechanical properties of high and low sulfur levels.
Moreover, if the sulfur is very low, Mg forms Ni2Mg
type of Laves phase. Magnesium had no effect on stress
rupture properties of Inconel 718 at 650˚C, 686 MPa
while the S ≤ 10 ppm [33]. Liu et al. [7] also reported
about the detrimental or no effect of Mg (76 ppm - 94
ppm) in their recent studies on IN-718 containing S < 10
ppm).
3.3. Magnesium Segregation
The segregation behaviour of Mg to phase interfaces
has been established by many using electron micro-
probe technique, AES and EDS analysis on the TEM
thin film and the results showed Mg segregation at
MC/, '/ [62,63]. The beneficial effects of Mg addi-
tion on grain boundaries are related to the grain bound-
ary segregation behaviour and its influence on the grain
boundary properties. Due to the segregation of Mg to
the grain boundary, the following changes in the grains
may occur [18]:
1) Grain boundary cohesive bond is intensified in
terms of the increase and more homogeneous distribu-
tion in electron density due to Mg segregation.
2) Grain boundary dislocation mobility decreases due
to Mg atom segregation to dislocation cores and which
may influence the creep rate if the grain boundary creep
predominates over the creep in grains.
3) The vacancy formation energy is increased due to
Mg segregation, which results in the decrease in grain
boundary vacancy concentration and grain boundary
diffusion coefficient.
4) The morphology of grain boundary precipitates
may be changed due to the decrease in grain boundary
energy, and thus the uniform granular grain boundary
precipitates decreased the mobility of grain boundary
migration, which results in the retardation of creep void
initiation and growth.
By EPM analysis, Mg concentration at the phase in-
terface, interior of MC and the matrix of a Ni base su-
peralloy has been measured –2.55 × 10–2, 0.437 × 10–2
and 0.956 × 10–2 % respectively. This indicates that Mg
is enriched at the interface of the phases and not the in-
terior of it. TEM analysis of ' interface, interior of '
and in the matrix were respectively 0.555, 0.252 and
0.256. Again the segregation of Mg at the interface is
pronounced. Furthermore, AES results of phase interface
also showed Mg segregation at the interface. The Mg
segregation thickness at MC carbide interface is thicker
than ' phase interface. This is because the lattice mis-
match between the carbide and matrix is larger. The
structure of MC carbide interface is incoherent and has a
large deformed area. It creates a favourable thermody-
namic condition for the Mg atom to segregate to MC
phase interface. If Mg atom enters MC phase it increases
Gibbs free energy, as a result it segregates at the inter-
faces instead of the interior of the carbide. On the other
hand, ' phase interface is semi-coherent and the struc-
ture of the interface consists of dislocation arrays and
impurities and the degree of distortion is also less-hence
the segregation of Mg is easier in this case.
Basically, the driving force for grain boundary segre-
gation of an element is the elastic distortion energy,
which is proportional to the square of the mismatch be-
tween the atomic radius and the unoccupied hole radius.
Hence, Mg should have a strong tendency to segregate
to the GB due to its larger atomic radius compared with
Nb and Mo. In addition, from the segregation theory, the
solubility of an element in a matrix is an indication of
the ability of an element to segregate, i.e., lower the
solubility the stronger the tendency to segregate. Since
the solubility of Mg is much less than Nb, Mo and W,
Mg segregates to the grain boundary. As the mismatch
of' Mg and Nb both are positive, a repulsive interaction
between Mg and Nb segregation can be expected. Mg
causes an additional lattice distortion and this distortion
assumed to have promoted “B” and “C” segregation at
GB. It is therefore, assumed that Mg segregation causes
Nb segregation to decrease and the segregation of ‘B’
and “C” to increase. Additional potential energy gener-
ated from the lattice distortion caused by Mg atoms may
accelerate the segregation of' “B” at the GB [64]. Hence
Mg helps eliminate the thick lamellae [65] or brittle
strips of NbC and -Ni3Nb along the GB [52,65]. An
example of globular type Ni3Nb along grain boundary of
Mg (0.0094) containing GH 169 alloy is presented as
opposed to cellular precipitates at grain boundary of
alloy with negligible amount of Mg (0.0001) (Figure 11)
[1].

The Role of Magnesium in Superalloys—A Review
1252
(a)
(b)
Figure 11. The influence of Mg on grain boundary -Ni3Nb
precipitation in GH 169 (a) without Mg (0.0001) (b) with
Mg (0.0094) [1].
3.4. Mg in Carbides and Matrix
The literatures [17,52,66] indicate that Mg not only seg-
regates along GB but also into the carbides of GB and '
phase either along grain boundary or in the grains. The
Mg dissolved in the ' phase and carbides lead to the
composition change, i.e., increases the contents of W,
Mo and Ti in the phases and therefore changes their lat-
tice constants and increases the elastic energy, which is
proportional to the increase in M23C6 thickness by caus-
ing a large distortion due to the large atomic radius of
Mg. Thus as the thickness of M23C6 reaches a critical
value, the coherent interface becomes semi or incoherent,
resulting in decrease in long range stress field and a de-
crease in total energy [66]. It is noteworthy that a pre-
cipitate phase always tends to minimize its surface en-
ergy. Hence the M23C6 with incoherent interface gradu-
ally becomes granular-which is beneficial for mechanical
properties. However, over addition of Mg caused M23C6
lamellae to precipitate due to the increase of supersatura-
tion of carbon at the GB [1].
3.5. Behaviour of Mg in Cast Superalloys
Mg improves the solidification structure of cast superal-
loys [5]. It segregates to phase boundaries and refines the
interdendritic MC carbides and ' eutectic-decreasing
quantity of ' eutectic. The addition of Mg influences the
grain size and decrease the secondary arm spacing. As a
result, the interdendritic precipitates like Laves phase and
MC eutectic are reduced. A decrease in Nb segregation
was also reported by using electron microprobe analysis
and this was considered good since MC eutectic and
Laves phase were the results of Nb segregation. The
morphology of MC also changed to spheroidal due to the
addition of Mg. Hence an optimum amount of Mg de-
creases Nb segregation and decreases initial cast segre-
gation. which shortens homogenization cycle? However,
the amount of -Ni3Nb plate around laves island in-
creased. This occurs because the Nb available for Laves
phase formation is too low but the Nb content for  plate
formation is high. Although both the phases are generally
considered to be detrimental, however, the elimination of
-Ni3Nb by homogenization is much easier than the
elimination of laves islands.
4. Level of Mg in Superalloys
The content of Mg from 30 - 70 ppm showed beneficial
effect in improving plasticity, high temperature tensile
ductility, stress rupture life, notched cyclic stress
ru~1ture life and creep fatigue interaction ability [3]. The
effect of Mg from 1 - 350 ppm on mechanical properties
was studied and 30 - 200 ppm was reported to have
beneficial effect on stress rupture ductility improvement
—no data was presented in the lower range (up to 100
ppm) [28]. An enhanced stress rupture ductility effect
was reported by addition of 30 ppm Mg [29]. A remark-
able stress rupture ductility and life improvement was
noticed with a small addition of Mg (13 - 19 ppm) [30].
The mechanical properties like tensile and stress rupture
ductilities, smooth and notch stress-rupture lives, fatigue-
creep interaction properties were reported to have in-
creased by 59 ppm addition of Mg (S-40 ppm) [27].
Stress rupture and creep properties improve by the addi-
tion of 94 ppm Mg in a vacuum melted IN-718 (S-50
ppm) [1]. The effect of Mg on cast alloys was reported to
be beneficial in the optimum range of 40 - 80 ppm Mg
addition [5]. By the addition of 20 ppm Mg impact

The Role of Magnesium in Superalloys—A Review 1253
toughness was reported to have improved and influenced
favourably the distribution of interdendritic Laves and
MC particles. The materials contained Mg in the range of
0 - 110 ppm. In a study with 50 ppm Mg level it was
concluded that Mg had no effect on LCF but creep prop-
erties enhanced (S-40 ppm) [26]. High temperature creep
properties were increased by the addition of 59 ppm Mg
[32]. An improved influence of Mg on stress rupture,
high tem- perature tensile ductility, fatigue creep interac-
tion properties were reported with 59 ppm Mg [5].
No effect or detrimental effect of Mg has been re-
ported in some literature [7]. The Mg did not show any
influence on tensile strengths and ductilities and stress
rupture life reported to have decreased when sulfur is less
than 10 ppm. Mg added was 0, 76 and 94 ppm (S-10
ppm).
5. Thrust Areas for Further Work
From the literature review it is obvious that the studies
have been made on the high temperature tensile, creep,
fatigue and impact toughness properties of Mg contain-
ing superalloys, including IN-718 alloy in question.
However, the studies on weldability of the material are
scarce if non-existent. In late 1960s the weldability
measurement by conducting gleeble hot ductility test was
performed by Morrison et al. [34], and the beneficial role
of Mg on weldability was reported. However, there was
no mention about the production route of the material.
Since there is propensity for the material to pick up Mg
during melting practice—from the furnace lining, cruci-
ble or the slag-the amount of optimum Mg reported by
their study, which was considered to have beneficial in-
fluence on the weldability of the Mg containing IN-718
was debatable. Moreover, Laves phase formation present
in the dendritic structure poses difficulty during welding
of IN-718 and as it has been observed in the literature
that an optimum amount of Mg can modify the Laves
phase and improve mechanical properties it is again
worthwhile to study the role of Mg in influencing the
weldability of IN-718. Furthermore, Mg being a sur-
face-active element has the tendency to segregate at the
interfaces and grain boundaries and hence it might have
great impact in affecting intergranular liquation cracking
in HAZ region of the weldments. In addition it is impera-
tive to study the interactive effect of Mg with the detri-
mental elements like S, P, and especially B in assessing
the weldability behaviour of superalloys produced by
using the vacuum technological routes (VIM, VAR and
VADER).
6. Conclusions
1) Microalloying of Mg shows beneficial effect of de-
creased low cycle fatigue crack growth rate by a factor of
3-7.
2) Mg prolongs secondary and tertiary creep at low
strain rate ( < 10–5 mmmm–1s–1) of steady state and
also prolongs rupture creep strength and elongation.
3) Mg refines and spheroidizes MC carbide at GB,
thus improving the stress distribution state at GB, which
in turn has decreases the possibility of wedge crack at the
GB. This enhances the creep properties.
4) Beneficial and detrimental effects of Mg addition
on rupture life and elongation, depend upon the operating
stress and temperature.
5) Mg can improve tensile strength and ductility at
high as well as low temperature.
6) Mg has been found to improve impact toughness,
which is possibly due to the decrease in interdendritic
Laves eutectic and the refinement of interdendritic seg-
regation of Nb and Ti, which allowed shorter homogeni-
sation cycle.
7) The effect of Mg on hot ductility behaviour is re-
ported to be controversial.
8) Ni2Mg (Laves phase) wrapping around MC has
been found to smoothen the carbide shape, thus improv-
ing high temperature creep and fracture properties.
9) Mg scavenges sulfur giving ductile fracture with no
magnesium and sulfur segregation on GB.
10) Mg helps eliminate thick lamellae or brittle strips
of NbC and -Ni3Nb along the GB.
11) Segregation of Mg has been shown to occur into
M23C6 and  apart from along grain boundary and other
surfaces. Thus the Mg containing M23C6 phase with in-
coherent interface becomes granular, which is beneficial
for mechanical properties.
12) In cast superalloys the segregation of Mg to phase
boundaries improves the solidification structure and thus
refining MC and  eutectic. Mg also decreases secondary
arm spacing and thus reducing interdendritic precipitates
like MC and Laves eutectic.
13) The study of the effect of Mg in influencing HAZ
liquation cracking and weldability of superalloys has
been proposed.
REFERENCES
[1] G. Chen, D. Wang and Z. Xu, “Proceedings Superalloys
1984,” In: M. Gell, C. S. Kortovich, R. H. Bricknell, W.
B. Kent and J. F. Radavich, Eds., The Metallurgical Soci-
ety, AIME, Warrendale, PA 15086, 1984, pp. 611-620.
[2] Z. Zhong, P. Ma, J. Zhuang, Y. Yuan and N. Jim, Pro-
ceedings Superalloy 1992, TMS, Warrendale, PA, 1992,
pp. 629-635.
[3] J. Dong, W. Xie, Z. Xu, X. Xie and S. Zhang, Acta
Metallurgica Sinica, Vol, 6. No. 6, 1993, pp. 405-409.
[4] P. Ma, Y. Yuan and Z. Zhong, Acta Metallurgica
Sinica,Vol. 3, No. 3, 1990, pp. 208-212.

The Role of Magnesium in Superalloys—A Review
1254
[5] G. Chen, Q. Zhu, D. Wang, X. Xi and J. F. Radavich,
Proceedings Superalloy 718, E. A. Loria, Ed., TMS,
Warrendale, PA, 1989, pp. 545-551.
[6] X . Xie, J. Liang and H. Jiang, Elsevier Applied Science,
1987, pp. 719-723.
[7] X. Liu, J. Dong, X. Xie and K. Chang, “The Appearance
of Magnesium and Its Effect on the Mechanical Proper-
ties of Inconel 718 with Low Sulfur Content,” Materials
Science and Engineering A, Vol. 303, 2001, pp. 262-266.
doi:10.1016/S0921-5093(00)01789-5
[8] D. A. Ford, Metals Technology, Vol. 11, No. 10, 1984, pp.
438-445.
[9] Z. M. Kalinina, et al., Production of Electric Steel’ (Pro-
izvodstvo Clektrostali), 1972, pp. 94-100.
[10] Z. M. Kalinina, et al., Production of Electric Steel' (Pro-
izvodstvo Clektrostali), 1972, pp. 89-94.
[11] Z. M. Kalinina, et al., “Effect of Microalloying with
Magnesium on Structure and Properties of Nickel Base
Alloys,” Proceedings Production of Steels and Alloys in
Vacuum, Electroslag and Electron Beam Furnaces (in
Russian), Part 2, Chermetinformatsiya, 1971, pp. 30-32.
[12] O. V. Rutes, G. S. Chernyak and S. B. Maslenkov, “Ef-
fect of Mg, Ca and Ba on structure and properties of
heat-resistant alloys hardened with Ni3Al,” Special Steels
and Alloys (in Russian), Metallurgiya, No. 2, 1973, pp.
41-47.
[13] G. S. Chyernyak, O. V. Rutes and S. B. Maslenkov, “Ef-
fect of Magnesium and Calcium on Structure and Proper-
ties of Heat-Resistant Alloys,” Special Steels and alloys
(in Russian), Metallurgiya, No. 1, 1972, pp. 89-97.
[14] Z. M. Kalinina, et al., Metall (Metalli) SSSR, No. 4, 1973,
pp. 193-196.
[15] V. V. Topilin and V. K. Versina, Quality Steels and Al-
loys, Moscow, 1976, pp. 108-113.
[16] V. V. Topilin and V. K. Versina, Metalloved. Term
Obrab Met., Vol. 11, 1977, pp. 10-13.
[17] P. Ma, Y. Yuan and Z. Zhong, “Creep Behavior of Mag-
nesium Microalloyed Wrought Superalloys,” Proceedings
Superalloys 1988, The Metallurgical Soci ety/A IME, 1988,
pp. 625-633.
[18] G. Chen and H. Ge, Journal of University of Science &
Technology, Beijing, Vol. 2, No. 2, 1995, pp. 84-91.
[19] Y. Li, C. Sun, J. Liu and G. Chen, High Temperature
Technology, Vol. 5, No. 4, November 1997, pp. 201-204.
[20] Guo E. and F. Xu, Acta Metallurgica Sinica, October
1988, pp. 957-962.
[21] D. Ke, Acta Metallurgica Sinica, Vol. 19, No. 5, October
1983, pp. a377-a384.
[22] H. Y. Bor, C. Y. Ma and C. G. Chao, “The Influence of
Mg on Creep Properties and Fracture Behaviors of
Mar-M247 Superalloy under 1255 K/200 MPa”, Metal-
lurgical and Materials Transactions A, Vol. 31A, May
2000, pp. 1365-1373.
[23] H. Y. Bor, C. G. Chao and C. Y. Ma, “The Influence of
Magnesium on Carbide Characteristics and Creep Be-
havior of the Mar-M247 Superalloy,” Scripta Materialia,
Vol. 38, No. 2, 1998, pp. 329-335.
[24] H. Y. Bor, C. G. Chao and C. Y. Ma, “The Effects of Mg
Micro Addition on the Mechanical Behavior and Fracture
Mechanism of MAR-M247 Superalloy at Elevated Tem-
peratures,” Metallurgical and Materials Transactions A,
Vol. 30A, No. 3, 1999, pp.55l-561.
[25] G. L. Chen, T. H. Zhang and W. Y. Yang, High Tem-
perature Technology, Vol. 6, No. 3, 1988, pp. 149-152.
[26] G. Chen, X. Xie, Z. Xu, J. Zhang, Chinese Journal of
Metal Science and Technology, Vol. 7, No. 6, 1991, pp.
435-441.
[27] X. Xie, Z. Xu, B. Qu, G. Chen and J. F. Radavich, Super-
alloys 1988, S. Reichman, D. N. Duhl, G. Maurer, S. An-
tolovich and C. Lund, Eds., AIME, Warrendale, PA, ,The
Metallurgical Society/AIME, 420 C'wealth Dr., Warren-
dale, PA, 1988, pp. 635-642.
[28] W. H. Couts Jr., et al, “Effect of Mg as an Alloying Ele-
ment in IN-718,” Report AEML-TR-7s-76, 1971.
[29] D. R. Muzyka and C. R. Whitney, “Process for Making
Ni-Base Precipitation Hardenable Alloys,” U.S. Patent
3,575,734, 20 April 1971.
[30] J. M. Moyer, “Extra Low Carbon Alloy 718,” Proceed-
ings Superalloys 1984, The metallurgical Society/AIME,
420 C'wealth Dr. Warrendale, PA 15086, pp. 223-254.
[31] Z. Xu, X. Xie, B. Qu, G. Chen and J. F. Radavich, Sci-
ence and Technology Beijing, Vol. 2, No. 6, 1989, pp.
560-567.
[32] Z. Xu, B. Xu, X. Xie, S. Zhou and X. F. Cheng, Journal
of Beijing University of Iron and Steel Technology, Vol. 9,
No. 4, 1987, pp. 38-44.
[33] X. S. Xie, J. X. Dong and M. C. Zhang, “Research and
Development of Inconel 718 Type Superalloy,” Materials
Science Forum, Vol. 539-543, March 2007, pp. 262-269.
[34] T. J. Morrison, et al, Proceedings Symposium of Welding
Research Council, Department of Mechanical Engineer-
ing, University of Manitoba, Winnipeg, Canada R3T 2N2,
1969, pp. 47-67.
[35] V. V. Topilin and Tsvetkova, Metalloved. Ternl. Obrab.
Met., Vol. 2, February 1980, pp. 62-64.
[36] K. Liu, et al., Proceedings 7th Symposium on Environ-
mental Degradation, Breckenridge, Co, USA, 1995, pp.
509-517.
[37] S. Yamaguchi, H. Kobayashi, T. Matsumiya and S. Ha-
yami, Metals Technology, May 1979, pp. 170-175.
[38] E. O. Ezugwu. J. Bonney and Y. Yamane, “An Overview
of the Machinability of Aeroengine Alloys,” Journal of
Materials Processing Technology, Vol. 134, 2003, pp.
233-253. doi:10.1016/S0924-0136(02)01042-7
[39] T. Baumgartner, K. Bothe, S. Hurta, W. M. Laanemae
and V. Gerold, “Thermomechanical Fatigue of Nimonic
80A,” Advanced Materials and Processes—Proceedings
of the First European Conference. EUROMAT ’89. Vol. 1,
Aachen, FRG, 22-24 November 1989. pp. 535-540.
[40] R. Sunulahpasic and M. Oruc, Metallurgica, Vol. 50, No.

The Role of Magnesium in Superalloys—A Review
1255
3, 2011, pp. 155-158.
[41] J. Fu, H. Wang and E. P. Chen, Proceedings 7th ICVM,
Tokyo, Japan, 1982, p.1226.
[42] J. Fu and L. Gao, J. Vacuum Science and Technology,
Vol. 5A, No. 4, Part 4, 1987, p. 2687.
[43] J. Fu and L. Song, Proceedings 10th International Con-
ference on Vacuum, Metallurgical Industry Press-Peoples
Republic of China, Beijing, China, 1991, pp. 160-168.
[44] W. Yeniscavich and C. W. Fox, Proceedings Symposium
of Welding Research Council, Department of Mechanical
Engineering, University of Manitoba, Winnipeg, Canada
R3T 2N2, 1969, pp. 24-35.
[45] F. Wang, Metal Physics, Mechanical Industry Press, Bei-
jing, 1981, p. 47.
[46] D. Ke and Z. Zhong, Acta Metallurgica Sinica, 1983, p.
A377.
[47] P. S. Kotval, J. D. Venables and R. W. Calder, Metallur-
gical Transactions A, Vol. 3, 1972, pp. 453-458.
[48] G. R. Leverant and M. Gell, Transactions TMS-AIME,
Vol. 245, 1969, pp. 1167-1173.
[49] G. S. Chyernyak, A. V. Smirnova and S. B. Maslenkov,
Metall.(Metalli.) SSSR, No. 1, 1973, pp. 98-104.
[50] D. Ke, Masters Thesis, CISRI, Beijing, China, 1981.
[51] V. V. Topilin, Steel USSR, 1978, pp. 643-646.
[52] D. Ke and Z. Zhong, Acta Metallurgica Sinica, Vol. 19,
1983, p. A377.
[53] P. Ma J. Zhuang, J. Yang and L. Gao, Acta Metallurgica
Sinica, Vol. 23, 1987, p. A195.
[54] D. Ke and Z. Zhong, CISRI Technical Bulletin, Vol. 73,
1983, p. 3.
[55] J. Zhu, Z. Y. Cheng and H. Q. Ye, “The Distribution and
Morphology of Trace Mg at a Grain Boundary in a
Ni-Base Superalloy,” Scripta Metallurgica, Vol. 23, 1989,
pp. 1537-1542. doi:10.1016/0036-9748(89)90124-5
[56] K. L. Gasko, G. M. Janowski and B. J. Pletka, “The In-
fluence of γ-γ′ Eutectic on the Mechanical Properties of
Conventionally Cast MAR-M247,” Materials Science
and Engineering, Vol. 104A, October 1988, pp. 1-8.
[57] A. S. Argon, I-W. Chen and C.-W. Lan, Proceedings
Creep-Fatigue Environmental Interaction, AIME, New
York, 1980, pp. 46.
[58] R. J. Raj, Engineering Materials Technology, Vol. 98,
1976, p. 132. doi:10.1115/1.3443355
[59] R. Lagneberg and B. Bergman, “The Stress/Creep Rate
Behaviour of Precipitation-Hardened Alloys,” Metals
Science, Vol. 10, No. 1, 1976, pp. 20-28.
doi:10.1179/030634576790431462
[60] J. Zhu, P. Ma, S. Zhang and L. Gao, Proceedings XIth
International Conference on Electron Microscopy, Kyoto,
Japan, 1986, pp. 851-852.
[61] T-J. Chuang and J. R. Rice, “The Shape of Intergranular
Creep Cracks Growing by Surface Diffusion,” Acta Met-
allurgica, Vol. 21, 1973, pp. 1625-1628. .
doi:10.1016/0001-6160(73)90105-3
[62] Q. Zhu, D. Wang, H. Ge and G. Chen, Acta Metallurgica
Sinica, June 1989, pp. A185-A189.
[63] Q. Zhu, D. Wang, H. Ge and G. Chen, Acta Metallurgica
Sinica, Vol. 2A, No. 6, 1989, pp. 408-411.
[64] V. V. Topilin, Stahl, Vol. 11, 1978, p. 1074.
[65] Y. Li, B. Xu, High Temperature Technology, Vol. 6, No.
4, November 1988, pp. 203-209.
[66] R. Ming, “Physical Basis of Crystal Growth”, Shanghai
Science & Technology Press, Shanghai, 1982, p. 364.