20ff PDF
20ff PDF
20ff PDF
Lin Liu1, Taiwen Huang1, Yuhua Xiong2, Aimin Yang3, Zhilong Zhao1, Rong Zhang1, and Jinshan Li1
1
State Key Laboratory of Solidification Processing, Northwestern Polytechnical University, Xiƍan 710072, P.R. of China
2
Beijing General Research Institute for Non-Ferrous Metals, Beijing 100088, P.R. of China
3
Department of Mechanical Engineering, Xiƍan Petroleum Institute, Xiƍan 710065, P. R. of China
Keywords: superalloy, grain refinement, microstructure, tensile properties, LCF, grain size
493
like microstructure. At the same time, Air Research Casting on fine-grained superalloys, the results were obtained mainly from
Company [9] developed a kind of fine grain casting process (FGP). the specimens prepared by thermal or mechanical methods. No
The process was established to use low pouring temperature research was reported by chemical methods.
(melting point +22ć), low melt superheat temperature, low
mould preheating temperature (1100ć) and local chills. For the In the present study, two ternary intermetallic compounds Co-Fe-
IN 713 LC and MAR-M 247 alloy, grain size in the range of Nb and Cr-Mo-Nb have been chosen as inoculants to investigate
ASTM 1-2 (0.25-0.18 mm) was obtained. The similar results were their effects on the grain structures of cast superalloy K4169
obtained for In-713LC alloy recently [10]. Like FGP, Microcast-X under various melting and casting conditions. Tensile properties
process also reduces molten-metal fluidity, and has less ability to and low cycle fatigue properties were compared for both
fill thin sections. Furthermore, the dispread of microporosities conventional and fine grained test bars.
within castings due to low poring temperature must be eliminated
by IHPing process, which increases production costs. Materials and Experimental Procedures
494
Macro- and Micro-Structure Analysis mechanical fatigue equipment at room temperature in static air
under whole strain control pull-press loading in all re-direction
The alloy ingots were sectioned along the transverse cross-section axis at namely strain amplitude r0.3-0.6%, with triangle wave at
and samples were ground, polished and subsequently chemically strain ratio R= -1 and loading rate 1×10-4 mm/sec.
etched with a solution consisting of 15g CuSO4, 3.5ml H2SO4 and
50ml HC1 to expose grain structures. The average equiaxed grain Experimental Results
size and equiaxed fraction at transverse cross-section were
determined by a standard quantitative metallographic technique Grain Structure
by the line intercept method [22]. The grain size was also
estimated with reference to the ASTM standard [23]. The Table II illustrates the experimental scheme and processing
distribution of the alloying elements was determined with a 733 parameters, together with characteristic parameters of grain
type electron microprobe. For observing microporosity, the structures under various refining processing. The corresponding
samples need to be polished but need not to be etched. grain structures are shown in Figure. 1. It can be seen that
lowering the pouring temperature and adding refiner to the melt
Mechanical Property Tests can reduce grain size (D) and improve equiaxed fraction at
traverse cross-section (P). In addition, the grain size increases and
The test bars for mechanical property test were processed into proportion of equiaxed grains reduces with increasing the
property samples with reference to Chinese National Standard GB superheating temperature TH. It must be noticed that finer grains
6397-860. Tensile properties at room temperature and 700qC were (less than 0.2mm) can be achieved both by addition of refiners
tested with a WJ-10B mechanical universal strength tester and a and by lowering pouring temperature without melt superheat, as
ZDM-5 tensile tester, respectively, seen in sample 1, 3 and 6 in Table I. Particularly, fine grains were
obtained at conventional melting and casting conditions, see
The fatigue test samples were prepared with 6mm in diameter and sample 8 in Table I.
the fatigue tests were carried out with a 100kN MAYES
Figure 1 Grain structures of cylindrical ingots obtained under various refining processing
(a)a(h) correspond to specimen No. 1, 2, 4, 7, 3, 5, 6 and 8, respectively
495
Dendrite Structure granular so that it is impossible to test their Laxe and SDAS.
Lowering pouring temperature and addition of refiner not only It can be seen from Table II that the Laxe decreases with the
reduce the grain size, but also change the morphologies of refinement of the grains. Whereas the value of SDAS keeps
dendrites. As shown in Figure2, developed side branches were almost the same at the same pouring temperature despite that
observed at higher pouring temperatures and without additions. In grain has been refined. In such a case, a great number of grains
contrast, dendrites tend to granulate at lower pouring temperature nucleate simultaneously and grow competitively at a velocity
and with refiner additions. In order to describe dendrite determined by their local temperature, which makes the region of
morphologies quantitatively, we use the length of the average constitutional supercooling decrease. These grains soon impinge
primary dendrite axes (Laxes) and the secondary dendrite arm on one another and prevent further growth. As a result, the laxe
spacing (SDAS) to characterize the morphological features of decreases after grains in the castings were refined.
dendrites. The grains of samples No. 3 and No. 6 are almost
(a) (b)
(c) (d)
200m
Figure2 Microscopic grain structure of cylinder ingots under various refining processing
(a) Sample No.7, pouring at 1420ć without addition (b) Sample No.8, pouring at 1420ć with addition
(c) Sample No.4, pouring at 1400ć without addition (d) Sample No.6, pouring at 1400ć with addition
Interdendritic Segregation
4.0
3.8
Fe
The segregation ratio is defined as the maximum concentration 3.6 Cr
in interdendrite areas over the minimum concentration in the 3.4 Mo
3.2 Nb
center of dendrite trunk. Figure3 illustrates the segregation ratio of 3.0 Ti
elements Fe, Cr, Mo, Nb and Ti of samples No. 4, 5, 6, 7 and 8
Segregation ratio
2.8
2.6
determined by microprobe. It can be concluded that Mo, Nb and Ti 2.4
are positive segregation elements, Fe and Cr are negative ones. 2.2
2.0
The results also indicate that segregation alleviates after grain 1.8
refinement. It is suggested that spatial scope for the growth of 1.6
1.4
dendrites diminishes along with grain refinement, resulting in the 1.2
reducing of segregation in distance and amplitude, which is the 1.0
0.8
most important reason for the lessening of segregation. Meanwhile,
0.0 0.5 1.0 1.5 2.0 2.5
it is known from the above observation on the morphologies of Grain size, mm
grains that the number of granular grains increases after grain
refinement. Thus, the elimination of dendrite structures makes it Figure3 Variation of segregation ratio of alloying elements with
inconvenient for elements to be rejected in interdendritic areas, grain size
which helps to reduce segregation.
496
Microporosity
0.70
0.65
Superalloy K4169 is subject to forming microporosity due to wide fine grains
coarse grains
0.60
freezing range. Figure 4 illustrates the effect of pouring
0.55
temperature and refiner additions on microporosity. It can be
(a) and (c), we can see that microporosities reduce in great extent 0.40
although grain size is almost the same, implying that both fine 0.35
also reveal two important phenomena: Firstly, the liquid within 0.20
497
n, the number of equiaxed nuclei per cubic meter, leads to the
increase of the right-hand side of equation (1). As a result, it is
a beneficial for the forming of equiaxed grain zone. Additionally,
refiner particles dispersed uniformly in the melt produce two
effects: (1) they cause a large quantity of equiaxed grains to form
and release a great amount of latent heat which prohibits the
crystal growth. (2) The formation and growth of many equiaxed
grains impede the growth of columnar grains.
Figure 6 Schematic view of lattice matching between K IC 2.9V 0.2 [exp(V b / V 0.2 1) 1]1 / 2 U 1 / 2 (3)
inoculants and J matrix Where U is the radius of end of crack. Obviously, the larger the
value of ˄ıb /ı0.2˅, the larger of fracture toughness is and the
smaller the brittle of the grain structure is. From the experimental
Mechanism of Increasing in the Proportion of Equiaxed Grains results in Table III, the values of ˄ıb /ı0.2˅of refined structure
vary from 1.45 to 1.20 as temperature changes from room
It can be seen from Table II and Figure1 that lowering the pouring temperature to 700qC. That is refined structure becomes more
temperature or adding refiner to the melt make the equiaxed brittle at elevated temperature. While the values of ˄ıb /ı0.2˅of
fraction increase along with the grain refinement. It is known that coarse grain structure has no change (from 1.27 to 1.27). Its
the condition that equiaxed grains appear ahead of the columnar fracture toughness KIc changes little and keeps larger value, so
front can be expressed in the following equation [26-27], that refined structure produces crack propagation quickly and its
1 1
'T CV (1) deformation amount is less. That is why elongation and reduction
G Kn 3 [1 ( N ) 3 ][8*m(1 k 0 ) 0 ] 2 of area of refined structure were less than coarse grain structure.
'TC D
where G is the temperature gradient (K/m), Ș a columnar to
Another explanation of the facts above is that the precipitates at
equiaxed transition coefficient relative to the alloy, n the number
grain boundaries have lower capability to resist oxidation than J
of sites per cubic meter (m-3), ǻTN the undercooling at the
matrix at 700ć, resulting in crack propagation rate to be faster at
heterogeneous nucleation temperature (K), ǻTC an undercooling
grain boundaries and crack to occur at surfaces of grain
equal to that of the columnar growth front (K), Ƚ Gibbs-Thomson
boundaries. As a result, the ductility level of fine-grained sample
coefficient (m.K), m the liquidus slope(K/wt.%), ko the
with much more grain boundaries is reduced.
distribution coefficient, and D the liquid diffusion parameter
1
.
(m2/s). There exists the relation ' T C v (VC 0 ) 2 , with V the Conclusions
growth velocity (m/s), and C0 alloy composition (wt. %). (1) Intermetallic compound grain refiners for superalloys were
Equation (1) reveals that lowering G and increasing in n, V, and developed, which have high melting point, good dispersion in the
mC0(1-k0) have an advantageous for the formation of equiaxed alloy melt, reasonable density and less mismatch with the matrix
grains. Low pouring temperature Tp reduces temperature gradient and do not bring inclusions into to the alloy. Therefore, the refiner
G, which helps to increase the constitutional undercooling. In is ideal for the grain refinement of superalloys.
addition, low Tp could increase cooling rate, resulting in the
increase of melt supercooling and V. That is to say, low Tp can (2) While adding refiner to the melt, a uniform and fine grains of
reduce the left-hand side of equation (1) and increase the right- superalloys can be realized under conventional melting and
hand side, contributing to the formation of equiaxed grain zone. casting conditions. The average size of equiaxed grains could be
refined to the order of ASTM 3.2. Equiaxed fraction at transverse
From the well-known criterion of constitutional undercooling cross-section could be improved from 56% to 99%.
G mC 0 (1 k 0 ) (2)
t
V Dk 0 (3) The segregation ratios of main elements such as Fe, Cr, Nb,
we can see that lowering in G and increasing in V and -mC0(1-k0) Mo and Ti are close to 1 progressively with the decrease in grain
are helpful for forming equiaxed grains. However, the effect of n size, which is beneficial for the improvement of mechanical
is not considered. The addition of refiner results in an increase in properties.
498
(4) The morphology of grains is transformed from dendrite in 8. J. M Lane, “Investment Cast Superalloys Challenge Wrought
conventional castings to granulation in fine-grained castings. Materials,” Advanced Materials & Processes. 138 (4) 1990, 107-
108.
(5) Low pouring temperature reduces temperature gradient,
increases cooling rate, melt supercooling and nucleation 9. M. Woulds, and H. Benson, “Development of a Conventional
undercooling, resulting in grain refinement. Refiner particles of Fine Grain Casting Process,” Superalloy 1984, eds. M. Gell, C. S.
good lattice compatibility with Ȗ matrix act as substrata of Ȗ Kortovich, R. H. Bricknell, W. B. Kent, and J. F. Radavich, (The
matrix, thereby causing grain refinement. Metallurgical Society of AIME,1984), 3-12.
(6) All tensile properties for fine grained samples are more 10. C. N. Wei, H. Y. Bor, C. Y. Ma, and T. S. Lee, A Study of In
superior to those of the coarse grained ones at room temperature. 713LC Superalloy grain Refinement Effects on Microstructure
At 700ć, the situations are the same for both strength properties, and Tensile Properties,” Materials Chemistry and Physics, 80
but the values of elongation and reduction of area drops for (2003), 89-93.
refined grains. This is caused by the different value of ıb /ı0.2 at
various temperatures. 11. B. S. Murty, S. A. Kori, and M. Chakraborty, “Grain
Refinement of Aluminium and its Alloys by Heterogenoous
(7) Finer grain improves the LCF lives significantly at room Nucleation and Alloying,” International Materials Review, 47 (1)
temperature, ranging from 20% to 100%, according to the strain (2002), 3-47.
amplitude. In addition, the dispersion degree of LCF data of fined
grain samples is diminished. 12. J. Fang, and B. Yu, “Investment of Surface Grain Refinement
for Superalloy Castings,” High Temperature Alloys for Gas
Acknowledgments Turbines 1982, eds. R. Brunetaud et al., (D. Reidel Pub., 1982),
987-997.
This work was supported by State Key Fundamental Research
Plan of China under grant No.G2000067202. The authors would 13. E. Chang, B. Tseng. J. C. Chou, and V. Yin, “Processing,
like to thank Dr.Yipng Guo, Dr. Wei Liu, Dr. Jinagbang Zheng Structure and Mechanical Property of Investment Cast In 713C
for participating in partial experimental and theoretical work. Superalloy,” AFS Transaction, 88-124, 47-54.
499
Materials Science and Technology, 3 (1987), 268-274
28. D.E. Macha, G.R. Cole, and J.A. Butzer, “Fine Grain,
Investment-Cast Integral Turbine Wheels,” Grain Refinement in
Casting and Welds, eds. G. J. Abbaschian, and S. A. David, (The
Metallurgical Society of AIME, New York, 1983), 197-219
500