Materials Science and Engineering A319– 321 (2001) 792– 795
www.elsevier.com/locate/msea
Influence of tungsten and rhenium concentration on creep
properties of a second generation superalloy
S. Wöllmer a, T. Mack b, U. Glatzel a,*
a
b
Metallische Werkstoffe, Uni6ersität Jena, Löbdergraben 32, D-07743 Jena, Germany
Daimler-Chrysler Aerospace, MTU München, PF 500640, D-80976 München, Germany
Abstract
Creep properties at three different temperatures (1123, 1253 and 1373 K) for five single crystal alloys with varying Re and W
content are presented. Tungsten content has been varied between 3.0 and 3.6 wt.%, rhenium between 2.3 and 2.9 wt.%. Creep
properties, such as minimum creep rate, time to reach 1, or 2% strain and time to failure have been evaluated. For the low
temperature regime, two parts of rhenium can be replaced by one part of tungsten. In the intermediate range, the ratio
cW/cRe :1.3 has to be fulfilled with the constraint that cRe ] 2.3 wt.%. At high temperature the sum of W plus Re concentration
determines the time to failure. © 2001 Elsevier Science B.V. All rights reserved.
Keywords: Rhenium; Superalloy; Creep strength; Alloy design
1. Introduction
Second and third generation nickel based superalloys
(3 and 6 wt.% rhenium content) have superior creep
strength than first generation alloys without Re. This
increase in creep strength is paid by an increasing
density (8.7 and 9.1 g cm − 3 as compared to 8.0 without
Re).
The element Re has several beneficial effects on the
superalloy. The large and slow diffusing element Re
segregates mainly in the matrix phase [1]. Additionally
to the solid solution strengthening effect of Re in the
matrix, rhenium atoms tend to cluster [2,3], thereby
hindering dislocation movement. The high melting
points and low diffusion coefficients of both Re and W
lead to an increase in melting temperature of the superalloy, as well as a reduced velocity of k% morphology
changes. Finally, since Re segregates mainly in the
matrix phase and has a large atomic radius, it influences the misfit between matrix and k% lattice parameter
[4 –9].
Therefore an alloy investigation has taken place in
order to minimize the W and Re concentration by
* Corresponding author. Tel.: + 49-3641-947790; fax: + 49-3641947792.
E-mail address: [email protected] (U. Glatzel).
keeping the creep strength [10]. During service, turbine
blades are usually strained below 1% [11], therefore the
criterion ‘time to reach 1 or 2% strain’, t1%, respectively
t2% was chosen as a main indication for creep strength.
2. Results
Fig. 1 shows the variation in the concentration of Re
and W for the alloys used. Alloys 1A –C have constant
W content and are tested only at 1253 K. For the alloys
2–6 both concentrations Re and W have been varied.
All other elemental concentrations, such as Ni, Co, Al,
Cr, Ta, Mo and Ti are in-between the values used with
the commercial superalloys CMSX-6® (first generation,
cRe = 0) and CMSX-4® (cRe = 3 wt.%). A three stage
heat treatment was carried out in order to ensure
homogeneity and the presence of 500 nm cubical k%
particles. Out of respect for our industrial partner’s
wishes, detailed information on heat treatment and
composition cannot be provided.
Creep specimens were cast as single crystals. Orientation of load axis was determined by Laue X-ray
backscattering technique. The small deviations of the
tensile axis from the ideal [001] crystallographic axis of
B 5°, except for alloy 1B with 13° deviation, allow the
0921-5093/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 9 2 1 - 5 0 9 3 ( 0 1 ) 0 1 0 6 3 - 2
S. Wöllmer et al. / Materials Science and Engineering A319–321 (2001) 792–795
793
Fig. 2. Creep behavior at 1123 K and 500 MPa for alloys 2 –6.
Fig. 1. Compositional variations of the alloys investigated. Alloys
1A– C tested only at 1253 K. Densities vary from 8.1 g (alloy 3) to 8.5
g cm − 3 (alloy 2).
influence of anisotropic creep behavior [12,13] to be
neglected.
2.1. Temperature of 1123 K and stress of 500 MPa
Fig. 2 shows the creep behavior plotted with the log
of strain rate versus strain as an example for the
temperature of 1123 K and an applied stress of 500
MPa. Typical for these conditions is a prolonged primary creep which can extend to up to 0.5% strain
[12,13]. Due to the large scatter in the strain to failure
data, the ‘time to reach 1% strain’ is used instead.
Fig. 3 shows the time to reach 1 and 2% strain, for
each alloy relative to alloy 2. Note that in most cases
alloy 2 has the superior creep strength. Despite the
small number of data points a contour line has been
estimated, indicating alloy compositions leading to similar 1 or 2% creep strength. The replacement of two
parts of Re by one part of W leads to an alloy with
similar t1%. By this a reduction in density of 1–2% is
achieved. Considering a stress exponent of 5–6 for
these temperatures, the stress state arising from centrifugal forces by the weight of turbine blade can be
lowered by about 5– 12%.
2.2. Temperature of 1253 K and stress of 230 MPa
Fig. 4 shows the dependence of minimum creep rate
on Re concentration at 1253 K and 230 MPa. A
step-like behavior can be determined with sharply increasing minimum creep rates with a Re concentration
lower than 2.3 wt.%. As can be seen in Fig. 5, a
maximum of creep strength is detected for a ratio of
cW/cRe :1.3. The underlined numbers in Fig. 5 indicate
Fig. 3. Time to reach 1 and 2% strain at 1123 K and 500 MPa.
the ratio cW/cRe. Considering these two observations,
an optimum alloy composition at this temperature is
given by: cRe = 2.3 wt.% and cW = 3.0 wt.%.
2.3. Temperature of 1373 K and Stress of 120 MPa
At the high temperature of 1373 K the creep response
is fully determined by rafting of the k% phase. The
matrix ‘channels’ in-between the k% rafts widen [14],
thereby reducing the Orowan stress for dislocations
entering these channels. After reaching a minimum
value, the creep rate steadily increases until rupture
occurs. In this case, time to failure and time to reach
1% strain are mainly determined by the slow diffusing
elements W and Re. Therefore, with an increasing sum
of (cW + cRe), time to failure increases, as indicated in
Fig. 6.
A more detailed analysis can be made from Fig. 7.
Similar to the low temperature of 1123 K (Fig. 3) a
794
S. Wöllmer et al. / Materials Science and Engineering A319–321 (2001) 792–795
Fig. 4. Strong dependence of minimum and 1% creep rate on cRe at
1253 K and 230 MPa.
Fig. 6. 1373 K and 120 MPa: Regarding the total life time the sum of
W and Re is of importance since k rafting is the dominating process.
Fig. 7. Time to 1% and to failure.
Fig. 5. 1253 K and 230 MPa: Time to reach 1 and 2% strain is
maximal if cw/cRe :1.3.
rough estimate can be made for contour lines with similar
creep behavior, resulting again in a possible replacement
of two parts of Re by only one part of W, leading to an
alloy with comparable creep strength but lower density.
If the tested alloys are compared to commercial alloys
by a Larsen-Miller plot [15]. They fall in-between the
boundaries of the alloys CMSX-4 and CMSX-6 with a
temperature gain of 10 – 30 K as compared to
CMSX-6.
part of tungsten, thereby lowering the density. A stress
reduction of 6–12% can be achieved. (2) Intermediate Temperature (1253 K): The ratio of tungsten to
rhenium concentration has to be kept constant in order
to achieve similar creep strength. A sharp drop in creep
strength is observed, if the Re concentration is lowered
below 2.3 wt.%. This leads to the optimal concentrations of cRe = 2.3 wt.% and cW = 3.0 wt.%. (3) High
temperature (1373 K): Time to failure steadily increases
with (cW + cRe). Similar to 1123 K, two parts of rhenium can be replaced by one part of tungsten.
References
3. Discussion and conclusions
Despite the small number of data points, some interesting conclusions can be drawn: (1) Low temperature
(1123 K): Two parts of rhenium can be replaced by one
[1] R. Darolia, D.F. Lahrmann, R.D. Fields, Superalloys 1988, The
Metallurgical Society, 1988, p. 255.
[2] D. Blavette, A. Bostel, Acta Metall. 32 (1984) 811.
[3] N. Wanderka, U. Glatzel, Mater. Sci. Eng. A A203 (1995) 69.
[4] A.F. Giamei, D.L. Anton, Met. Trans. A 16A (1985) 1997.
S. Wöllmer et al. / Materials Science and Engineering A319–321 (2001) 792–795
[5] L. Müller, T. Link, M. Feller-Kniepmeier, Scripta Met. Mater.
26 (1992) 1297.
[6] U. Glatzel, A. Müller, Scripta Met. Mater. 31 (1994) 285.
[7] U. Glatzel, Scripta Met. Mater. 31 (1994) 291.
[8] R. Völkl, U. Glatzel, M. Feller-Kniepmeier, Acta Mater. 46
(1998) 4395.
[9] C. Schulze, M. Feller-Kniepmeier, Mater. Sci. Eng. A A281
(2000) 204.
[10] T. Mack, J. Wortmann, U. Glatzel, S. Wöllmer, Patent pending
submitted to ‘‘Deutsches Patentamt’’ January 2001.
795
[11] F. Lubahn, Plasticity and Creep of Metals, Wiley, New York,
1968.
[12] U. Glatzel, Microstructure and Internal Strains of Undeformed
an Creep Deformed Samples of a Nickel-Base Superalloy, Verlag
Köster, Berlin, 1994.
[13] V. Sass, U. Glatzel, M. Feller-Kniepmeier, Acta Mater. 44
(1996) 1967.
[14] Z. Peng, U. Glatzel, T. Link, M. Feller-Kniepmeier, Scripta
Mater. 34 (1996) 221.
[15] S. Wöllmer, PhD thesis, Friedrich-Schiller-Universität Jena, Germany, 2000.