- Email: [email protected]
High-temperature fatigue properties of austenitic superalloys 718, A286 and 304L
Available online at www.sciencedirect.com
International Journalof Fatigue
International Journal of Fatigue 30 (2008) 1978–1984
www.elsevier.com/locate/ijfatigue
High-temperature fatigue properties of austenitic superalloys 718, A286 and 304L Kazuo Kobayashi *, Koji Yamaguchi, Masao Hayakawa, Megumi Kimura Fatigue Group, Materials Reliability Center (MRC), National Institute for Materials Science (NIMS), 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Japan Received 1 June 2007; received in revised form 21 December 2007; accepted 10 January 2008 Available online 20 January 2008
Abstract High-temperature fatigue properties of austenitic superalloys 718, A286, and 304L were investigated in region between 102 and 107 cycles. The alloys 718 and A286 are precipitation-hardening type austenitic superalloys, and 304L is a solid solution-hardening type austenitic steel. The fatigue strengths of alloys 718 and A286 in high-cycle region over 104 cycles showed a grain size effect. The initiation sites of the fracture were the crystallographic facets corresponding to the austenitic grain size. Therefore the coarse-grain alloys showed the lower fatigue strength than the fine-grain alloys, because the fatigue strength is generally inversely proportionate to the initiation defect areas for the fatigue cracks. In low-cycle region under 104 cycles, on the other, the initiation sites of the fracture were the surface of the specimens. This type fracture mode is common for fatigue fracture of smoothed specimens. The ratio of the fatigue limit at 107 cycles to the tensile strength is extremely low for the coarse-grain alloys compared with the common case of 0.5. From these results fatigue fracture map is proposed. Ó 2008 Elsevier Ltd. All rights reserved. Keywords: High-cycle fatigue properties; High temperature; Austenitic superalloys; Fatigue strength; Grain size; Fatigue fracture map
1. Introduction Alloys 718 [1,2] and A286 [3] are typical austenitic superalloys of the precipitation-hardening type, which are nickel-based and iron-based, respectively. They are used in high-temperature applications such as blades, disks, and the shafts of jet engines and gas turbines. Austenitic 304L stainless steel is also commonly used in high-temperature applications, but it is a solid solution-hardening type. The above alloys and the steel are also used in the main engine components of the Japanese space rocket H-IIA. The National Institute for Materials Science (NIMS) and the Japan Aerospace Exploration Agency (JAXA) have been conducting the Data Sheet Project to obtain
*
Corresponding author. Tel.: +81 29 859 2248; fax: +81 29 859 2201. E-mail address: [email protected] (K. Kobayashi).
0142-1123/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijfatigue.2008.01.004
the fatigue strength data on space-use materials [4]. These data have been published as NIMS Space Use Materials Strength Data Sheet Nos. 4, 6 and 7, for alloys 718, A286 and 304L steel, respectively [5–7]. This paper is a trial to understand the high-cycle fatigue strength of austenitic superalloy and steel from a viewpoint of austenitic grain size and tensile strength. The fatigue test data of coarse-grain alloys were referred to the Data Sheets [5–7], and those of fine-grain alloys were used in the references [8–10]. 2. Experimental procedures 2.1. Materials The list of materials used in this study and their chemical composition, processing, thermal history and grain size is shown in Table 1. In general, fine-grain materials are
Materials
C
Si
Mn
P
S
Ni
Cr
Mo
V
Co
Cu
Ti
Al
B
Fe
Nb
Ta
Coarse-grain alloy 718 [5] Fine-grain alloy 718 [8] Coarse-grain alloy A286 [6] Fine-grain alloy A286 [9] Coarse-grain 304L steel [7]
0.032
0.09
0.12
0.006
0.0001
53.6
18.64
2.95
–
0.15
0.04
1.02
0.479
–
18.05
5.11
0.017
0.023
0.03
0.03
0.002
0.0017
52.3
18.00
3.18
–
0.02
0.005
0.98
0.49
–
19.5
5.35
–
0.04
0.14
0.16
0.011
0.004
25.01
14.44
1.32
0.33
0.34
0.04
2.06
0.17
0.0033
Balance
–
–
0.032
0.55
1.14
0.0320
0.004
24.65
13.91
1.27
0.15
–
0.14
2.50
0.26
0.0078
Balance
–
–
0.021
0.64
1.79
0.011
0.002
9.55
18.62
0.15
–
–
0.25
–
–
–
–
–
–
Materials
Processing and product form
Thermal history
Grain size(diameter, lm)
Coarse-grain alloy 718 [5]
Forging, rolling, plate
100–200
Fine-grain alloy 718 [8]
Rolling, bar
Coarse-grain alloy A286 [6]
Forging, billet
Fine-grain alloy A286 [9]
Rolling, bar
Coarse-grain 304L steel [7]
Forging, billet
Solution treatment: 1045 °C, l h ? air cooling Aging: (720 °C, 8 h ? furnace cooling 620 °C, 10 h ? air cooling)* 2 Solution treatment: 960 °C, l h ? air cooling Aging: 720 °C, 8 h ? furnace cooling 620 °C, 8 h ? air cooling Solution treatment: 980 °C, 2 h ? oil quenching Aging: 718 °C, 16 h ? air cooling Solution treatment: 982°c, l h ? water quenching Aging: 715 °C, 16 h ? air cooling Solution treatment: 1030 °C, 70 min Water cooling or oil cooling
10–20 60–170 10–70 125
K. Kobayashi et al. / International Journal of Fatigue 30 (2008) 1978–1984
Table 1 Chemical compositions (mass %) and processing details of materials used
1979
1980
K. Kobayashi et al. / International Journal of Fatigue 30 (2008) 1978–1984
produced by the rolling process, and coarse-grain materials are produced by forging process. The 304L steel is one heat, which is a coarse-grain material. The alloy 718 was subjected to a double aging to enhance the precipitation of Ni3Nb. The alloy A286 was subjected to aging to precipitate Ni3Al. The 304L steel was solid solution-treated only. Fig. 1 shows the microstructures of the materials. The microstructure of the 304L steel is shown for coarse-grain material only. The grains are mixed, so the range in the diameter is shown in Table 1. The tensile properties of the materials at high-temperatures are shown in Table 2. 2.2. Fatigue tests Axial load-controlled fatigue tests were conducted using servo-hydraulic fatigue-testing machines. The wave shape
was sinusoidal, and the frequency was 10 Hz. The fatigue tests were continued up to 107 cycles. Fatigue specimens were an hourglass type with a diameter of 6 mm and a cylindrical parallel type with a diameter of 3 mm, depending on the sizes of the materials and the capacities of the testing machines. The test conditions of stress ratio (R = rmin/rmax) and the temperature were determined by considering the parts of the H-IIA rocket engines in which the materials were used, where rmin and rmax are the minimum and maximum stresses of the sinusoidal wave, respectively. The details of the test conditions are listed in Table 3. Testing temperature for fine-grain alloy 718 was 550 °C because the fatigue tests were conducted under Vamas round-robin test program [8]. So the testing temperatures were different for fine- and coarse-grain alloy 718. The specimens were heated in electric resistance furnaces. Temperature of spec-
Fig. 1. Microstructures of materials used.
K. Kobayashi et al. / International Journal of Fatigue 30 (2008) 1978–1984 Table 2 High-temperature tensile properties of materials used Temperature (°C)
Coarse-grain alloy 718 [5] Fine-grain alloy 718 [8] Coarse-grain alloy A286 [6] Fine-grain alloy A286 [9] Coarse-grain 304L steel [7]
600
Tensile strength (MPa)
Elongation (%)
956
1089
21
550
1042
1212
21
600
593
837
19
713
912
22
109
364
44
500
0.2% proof stress (MPa)
1000
Table 3 Fatigue test conditions of stress ratio and temperature Materials
Stress ratio
Temperature (°C)
Coarse-grain alloy 718 [5] Fine-grain alloy 718 [8] Coarse-grain alloy A286 [6] Fine-grain alloy A286 [9] Coarse-grain 304L steel [7]
1 1 and 0.01
600 550 600
0.01
500
Alloy 718 800 Fine-grain
Stress amplitude ; σ a / MPa
Materials
1981
600
Coarse-grain
400
200 : Fine-grain 550˚C
R=-1 0 1 10
10
2
: Coarse-grain 600˚C
10
3
10
4
10
5
10
6
10
7
10
8
Number of cycles to failure ; Nf Fig. 2. Fatigue strengths for fine- and coarse-grain alloy 718.
imen was measured by R-type thermocouples contacted on the specimen surface of the gage portion, although feed-backed thermocouples to control were spot-welded in the shoulder part of the specimen. Temperature gradient was ±5 °C through the gage length for the cylindrical specimen.
1000 A286 600˚C
800
Test results of alloys 718, A286 and 304L steel are shown in Figs. 2–4, respectively. A distinct grain size effect on the fatigue strength was observed, especially in the high-cycle region for alloy 718, shown in Fig. 2, even though the testing temperatures were different by 50 °C. The difference in temperature is not a reason for the difference in the fatigue strength for fineand coarse-grain materials, because the difference in the fatigue strength is not vanished by a normalization of the fatigue strength by the tensile strength at each temperature. For alloy A286 in Fig. 3, testing temperature is the same 600 °C for fine- and coarse-grain materials. There are two test conditions for stress ratio, R. For both the R conditions, grain size effects are observed. For the 304L steel in Fig. 4, the fatigue strength is very low because the material is solid, solution-treated and the test condition of R is 0.01. In this case the data are only four points for the coarse-grain material. The fatigue tests are possible only for limited stress region, in other words, it is impossible to run the tests at higher stress level due to plastic deformation. Fracture surfaces were observed by SEM for the specimens fractured in low- and high-cycle regions.
Stress amplitude ; σ a / MPa
3. Results
600
400
200 : Fine-grain R=-1 : Coarse-grain R=-1 : Fine-grain R=0.01 : Coarse-grain R=0.01
0 1 10
10
2
10
3
10
4
10
5
10
6
10
7
10
8
Number of cycles to failure ; Nf Fig. 3. Fatigue strengths for fine- and coarse-grain alloy A286.
In low-cycle region, the initiation site of fatigue fracture was a surface of the specimens as shown in Fig. 5. This behavior is observed for all the three kinds of materials,
Stress amplitude ; σ a / MPa
1982
K. Kobayashi et al. / International Journal of Fatigue 30 (2008) 1978–1984
400 304L 500˚C 300 200 100 R=0.01 0 1 10
10
2
3
4
5
6
10 10 10 10 10 Number of cycles to failure ; Nf
7
10
8
Fig. 4. Fatigue strengths for coarse-grain 304L steel.
Fig. 6. Fracture surfaces fatigued in high-cycle region for coarse-grain alloy A286 (Stress amplitude: 200 MPa, Nf: 5.48 106 cycles, stress ratio: 0.01).
Fig. 5. Fracture surfaces fatigued in low-cycle region for coarse-grain alloy A286 (Stress amplitude: 500 MPa, Nf: 1.11 104 cycles, stress ratio: 1).
and it is irrespective of the grain size. Fig. 5 is a case of coarse-grain alloy A286. The arrow in the figure indicates the initiation site of the fatigue fracture. It is clear that the initiation site is a surface of the specimen.
In high-cycle fatigue regions, on the other hand, the initiation site was an interior of the specimen as shown in Fig. 6. In this case the material is coarse-grain alloy A286. Dotted circle in the figure indicates a fish-eye pattern [11]. The fish-eye pattern is discriminated from an oxidation color and a roughness of the surface. High magnification reveals that there are many crystallographic facets inside the dotted circle as shown in Fig. 6b. The size of the facets is almost the same as the grain size. As for fine-grain alloys A286, the initiation site was also interior. Fig. 7 is a case of fine-grain alloy A286. Dotted circle in Fig. 7a indicates also the fish-eye pattern [11]. Many crystallographic facets are also observed inside the dotted circle as shown in Fig. 7b. In a case of alloy 718 almost the same fracture surfaces as alloy A286 were observed [10]. In a case of 304L steel, the initiation site of the fracture was a surface of the specimen. It is considered that the 304L steel is not strengthened like the superalloys, and the fracture mode is not interior even in high-cycle region.
K. Kobayashi et al. / International Journal of Fatigue 30 (2008) 1978–1984
1983
The estimated ratio of the fatigue limit at 107 cycles, rcoarse-grain/rfine-grain is 0.86 for alloy A286, and 0.68 for alloy 718. On the other hand, actual ratio of the fatigue strength, rcoarse-grain/rfine-grain is 0.66 for alloy A286 and 0.63 for alloy 718. The value for alloy 718 is normalized by the tensile strength at each temperature. The estimated ratio values are not equal completely to the actual ratio values. The reason is not clear at present, but it seems that this is one of the physical meanings to the effect of austenitic grain size on the fatigue strength. 4.2. Fatigue fracture map for austenitic materials Fatigue limit or fatigue strength at high-cycle region is known to correlate with tensile strength. The results of this study will be compared with reference data for various steels at high and room temperatures [10,12,13]. In this study there are two test conditions for stress ratio R. So the converted fatigue strength derived by the following equation [12] was used. rwðR¼1Þ ¼ 1:317 rwðR¼0:01Þ 17:0;
Fig. 7. Fracture surfaces fatigued in high-cycle region for fine-grain alloy A286 (Stress amplitude: 280 MPa, Nf: 1.57 106 cycles, stress ratio: 0.01).
4. Discussion
ð2Þ
where rw(R=1) and rw(R=0.01) are the fatigue strength for each stress ratio. In this study the fatigue strength is a value at 107 cycles. Eq. (2) is proved to be valid for various carbon steels and austenitic steels [12]. Using this equation makes it possible to evaluate the fatigue strength based on the common test condition of R = 1. The relationship between the fatigue strength and the tensile strength is shown in Fig. 8. The plotted data are the values under R = 1 or converted values from Eq. (2). In Fig. 8 the box shows the results of rotating bending fatigue tests at high-temperatures for various high-temperature materials [9,13]. In this case, the stress ratio of R is 1 and the fatigue strength is a value at 108 cycles. The dotted box shows the results of axial load-controlled fatigue tests for many kinds of steels with a stress ratio of R = 1 at room temperature [12]. In this case the fatigue
In high-strength steels it is well known that fish-eye fatigue fracture occurs in high-cycle region and a non-metallic inclusion exists at a center of fish-eye pattern [11]. Murakami proposed a method for estimating the fatigue limit from a size of the inclusion. The equation is expressed as follows, p 1=6 rw ¼ 1:56ðH v þ 120Þ=ð areaÞ ð1Þ where rwpis an estimated fatigue limit, Hv is Vickers hardness and area is a square root of the defect projected vertically to the load axis. In this study this equation seems applicable to the crystallographic facets fracture. However, it is difficult to determine a location of a center of the fracture inside the fish-eye pattern, because there are many facets everywhere inside the fish-eye pattern. So the maximal grain diameter in Table 1 is used as substitute for the defect area in Eq. (1).
7 Fatigue strength at 10 cycles σ w / MPa
4.1. Effect of grain size 800 700 600
All the data are under R=-1. * means the data converted from Equation (2).
500
a at ld ia x A
300
ta Ro
200
tin
d en g-b
304L
718
ta
400
600
ain se-gr Coar
A286
: Fine-grain A286 (R=-1) : Fine-grain A286 (R=0.01)* : Coarse-grain A286 (R=-1) : Coarse-grain A286 (R=0.01)*
100 200
in
a gd
ain
Equation (3)[12] ( σ w = 0.5 × σ B )
] 13 [9,
400
0 0
r e-g Fin
2] [1
800
: Fine-grain 718 (R=-1) : Coarse-grain 718 (R=-1) : Coarse-grain 304L (R=0.01)*
1000
1200
1400
Tensile strength σ B / MPa Fig. 8. Relationships between fatigue strength and tensile strength.
1984
K. Kobayashi et al. / International Journal of Fatigue 30 (2008) 1978–1984
strength is a value at 107 cycles. These boxes indicate the 95% confidence region. In general, the relationship between the fatigue strength and the tensile strength is expressed as follows [12]. rwðR¼1Þ ¼ 0:5 rB
ð3Þ
where rw(R=1) is the fatigue strength in high-cycle region and rB is the tensile strength at each temperature. The data points such as circles, triangles and squares show the results of this study. The fatigue strength of the 304L steel tested at 500 °C is located in the box. The data of fine-grain alloys 718 and A286 are located near the boxes or near the line of Eq. (3). However, the data of coarse-grain alloys 718 and A286 deviate far from the boxes or the line of Eq. (3). These large deviations are discussed as follows. The reason is thought to be that the fracture site is interior, and the fracture is initiated by the large facets of the crystallographic features. For very fine-grain materials, the fatigue strength is controlled by Eq. (3). Between the fine and the coarse highstrength materials,the grain size effect may appear in the fatigue strength. On the other hand, a low-strength material such as 304L steel shows no grain size effect, because the crystallographic facet fracture does not occur. Therefore, Fig. 8 also shows the fatigue fracture map for austenitic materials as well as a relationship between the fatigue strength and the tensile strength for all kinds of materials. 5. Conclusions Fatigue strength for austenitic superalloys was discussed from the viewpoint of austenitic grain size and the tensile strength. Our conclusions are as follows. High-strength austenitic materials with a coarse-grain showed interior fractures in high-cycle fatigue region. The initiation site was composed of various large facets of the crystallographic features and the facets were as large as the grain size. So the fatigue strength deviated from the equation line of rw(R=1) = 0.5 rB in the relationship between rw(R=1) and rB, where rw (R=1) was the fatigue strength under the stress ratio R of 1, and rB was the tensile strength. On the other hand, high-strength austenitic materials with a fine-grain also showed interior fractures in high-
cycle fatigue region. However, the facets at the initiation sites were as small as the fine-grain size. The fatigue strength was obeyed the equation of rw(R=1) = 0.5 rB. From the above results, Fatigue fracture map for austenitic materials are proposed. This map shows that high-cycle fatigue strength for high-strength austenitic steel or superalloys is determined by the austenitic gain size and the tensile strength, since the high-cycle fatigue strength is controlled by the facet size of the crystallographic features at the initiation area on the fracture surface. References [1] Cozar R, Pineau A. Precipitates and thermal-stability of inconel 718 type alloys. Metall Trans 1973;4(1):47–59. [2] Aerospace Material Specification AMS 5663J. The engineering society for advancing mobility land sea air and space. Warrendale; 1997. p. 1. [3] Bornstein L. Numerical Data and Functional Relationships in Science and Technology. Adv Mater Technol 2006;2:319–21. [4] NIMS. Space Use Materials Strength Data Sheet No.0. Tsukuba: National Institute for Materials Science; 2003. [5] NIMS. Space Use Materials Strength Data Sheet No.4. Tsukuba: National Institute for Materials Science; 2004. [6] NIMS. Space Use Materials Strength Data Sheet No.6. Tsukuba: National Institute for Materials Science; 2005. [7] NIMS. Space Use Materials Strength Data Sheet No.7. Tsukuba: National Institute for Materials Science; 2006. [8] Kitagawa M, Yamaguchi K, Hukuda Y, Komine Y, Hirata H. Analysis on VAMAS low cycle fatigue round robin test in Japan. ISIJ Int 1993;33:817–24. [9] Yoshida S, Kanazawa K, Yamaguchi K, Sasaki M, Kobayashi K, Sato M. Elevated-temperature fatigue properties of 15Cr-26Ni-MoTi-Al-V heat-resisting steel bars for general application. Trans Natl Res Inst Metal 1978;20(6):392–400. [10] Kobayashi K, Yamaguchi K, Hayakawa M, Kimura M. Grain size effect on high-temperature fatigue properties of alloy718. Mater Lett 2005;59:383–6. [11] Murakami Y. Metal fatigue: effects of small defects and nonmetallic inclusions. Oxford: Elsevier; 2002. p. 75. [12] Nishijima S, Ishii A, Kanazawa K, Matsuoka S, Masuda C. Fundamental fatigue properties of JIS steel for machine structural use. National Research Institute for Metals, NRIM Material Strength Data Sheet Technical Document, No.5, 1989. [13] Kanazawa K, Yamaguchi K, Sato S. Elevated temperature, highcycle fatigue properties of engineering materials for high-temperature components. National Research Institute for Metals, NRIM Material Strength Data Sheet Technical Document, No. 6, 1990.
International Journalof Fatigue
International Journal of Fatigue 30 (2008) 1978–1984
www.elsevier.com/locate/ijfatigue
High-temperature fatigue properties of austenitic superalloys 718, A286 and 304L Kazuo Kobayashi *, Koji Yamaguchi, Masao Hayakawa, Megumi Kimura Fatigue Group, Materials Reliability Center (MRC), National Institute for Materials Science (NIMS), 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Japan Received 1 June 2007; received in revised form 21 December 2007; accepted 10 January 2008 Available online 20 January 2008
Abstract High-temperature fatigue properties of austenitic superalloys 718, A286, and 304L were investigated in region between 102 and 107 cycles. The alloys 718 and A286 are precipitation-hardening type austenitic superalloys, and 304L is a solid solution-hardening type austenitic steel. The fatigue strengths of alloys 718 and A286 in high-cycle region over 104 cycles showed a grain size effect. The initiation sites of the fracture were the crystallographic facets corresponding to the austenitic grain size. Therefore the coarse-grain alloys showed the lower fatigue strength than the fine-grain alloys, because the fatigue strength is generally inversely proportionate to the initiation defect areas for the fatigue cracks. In low-cycle region under 104 cycles, on the other, the initiation sites of the fracture were the surface of the specimens. This type fracture mode is common for fatigue fracture of smoothed specimens. The ratio of the fatigue limit at 107 cycles to the tensile strength is extremely low for the coarse-grain alloys compared with the common case of 0.5. From these results fatigue fracture map is proposed. Ó 2008 Elsevier Ltd. All rights reserved. Keywords: High-cycle fatigue properties; High temperature; Austenitic superalloys; Fatigue strength; Grain size; Fatigue fracture map
1. Introduction Alloys 718 [1,2] and A286 [3] are typical austenitic superalloys of the precipitation-hardening type, which are nickel-based and iron-based, respectively. They are used in high-temperature applications such as blades, disks, and the shafts of jet engines and gas turbines. Austenitic 304L stainless steel is also commonly used in high-temperature applications, but it is a solid solution-hardening type. The above alloys and the steel are also used in the main engine components of the Japanese space rocket H-IIA. The National Institute for Materials Science (NIMS) and the Japan Aerospace Exploration Agency (JAXA) have been conducting the Data Sheet Project to obtain
*
Corresponding author. Tel.: +81 29 859 2248; fax: +81 29 859 2201. E-mail address: [email protected] (K. Kobayashi).
0142-1123/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijfatigue.2008.01.004
the fatigue strength data on space-use materials [4]. These data have been published as NIMS Space Use Materials Strength Data Sheet Nos. 4, 6 and 7, for alloys 718, A286 and 304L steel, respectively [5–7]. This paper is a trial to understand the high-cycle fatigue strength of austenitic superalloy and steel from a viewpoint of austenitic grain size and tensile strength. The fatigue test data of coarse-grain alloys were referred to the Data Sheets [5–7], and those of fine-grain alloys were used in the references [8–10]. 2. Experimental procedures 2.1. Materials The list of materials used in this study and their chemical composition, processing, thermal history and grain size is shown in Table 1. In general, fine-grain materials are
Materials
C
Si
Mn
P
S
Ni
Cr
Mo
V
Co
Cu
Ti
Al
B
Fe
Nb
Ta
Coarse-grain alloy 718 [5] Fine-grain alloy 718 [8] Coarse-grain alloy A286 [6] Fine-grain alloy A286 [9] Coarse-grain 304L steel [7]
0.032
0.09
0.12
0.006
0.0001
53.6
18.64
2.95
–
0.15
0.04
1.02
0.479
–
18.05
5.11
0.017
0.023
0.03
0.03
0.002
0.0017
52.3
18.00
3.18
–
0.02
0.005
0.98
0.49
–
19.5
5.35
–
0.04
0.14
0.16
0.011
0.004
25.01
14.44
1.32
0.33
0.34
0.04
2.06
0.17
0.0033
Balance
–
–
0.032
0.55
1.14
0.0320
0.004
24.65
13.91
1.27
0.15
–
0.14
2.50
0.26
0.0078
Balance
–
–
0.021
0.64
1.79
0.011
0.002
9.55
18.62
0.15
–
–
0.25
–
–
–
–
–
–
Materials
Processing and product form
Thermal history
Grain size(diameter, lm)
Coarse-grain alloy 718 [5]
Forging, rolling, plate
100–200
Fine-grain alloy 718 [8]
Rolling, bar
Coarse-grain alloy A286 [6]
Forging, billet
Fine-grain alloy A286 [9]
Rolling, bar
Coarse-grain 304L steel [7]
Forging, billet
Solution treatment: 1045 °C, l h ? air cooling Aging: (720 °C, 8 h ? furnace cooling 620 °C, 10 h ? air cooling)* 2 Solution treatment: 960 °C, l h ? air cooling Aging: 720 °C, 8 h ? furnace cooling 620 °C, 8 h ? air cooling Solution treatment: 980 °C, 2 h ? oil quenching Aging: 718 °C, 16 h ? air cooling Solution treatment: 982°c, l h ? water quenching Aging: 715 °C, 16 h ? air cooling Solution treatment: 1030 °C, 70 min Water cooling or oil cooling
10–20 60–170 10–70 125
K. Kobayashi et al. / International Journal of Fatigue 30 (2008) 1978–1984
Table 1 Chemical compositions (mass %) and processing details of materials used
1979
1980
K. Kobayashi et al. / International Journal of Fatigue 30 (2008) 1978–1984
produced by the rolling process, and coarse-grain materials are produced by forging process. The 304L steel is one heat, which is a coarse-grain material. The alloy 718 was subjected to a double aging to enhance the precipitation of Ni3Nb. The alloy A286 was subjected to aging to precipitate Ni3Al. The 304L steel was solid solution-treated only. Fig. 1 shows the microstructures of the materials. The microstructure of the 304L steel is shown for coarse-grain material only. The grains are mixed, so the range in the diameter is shown in Table 1. The tensile properties of the materials at high-temperatures are shown in Table 2. 2.2. Fatigue tests Axial load-controlled fatigue tests were conducted using servo-hydraulic fatigue-testing machines. The wave shape
was sinusoidal, and the frequency was 10 Hz. The fatigue tests were continued up to 107 cycles. Fatigue specimens were an hourglass type with a diameter of 6 mm and a cylindrical parallel type with a diameter of 3 mm, depending on the sizes of the materials and the capacities of the testing machines. The test conditions of stress ratio (R = rmin/rmax) and the temperature were determined by considering the parts of the H-IIA rocket engines in which the materials were used, where rmin and rmax are the minimum and maximum stresses of the sinusoidal wave, respectively. The details of the test conditions are listed in Table 3. Testing temperature for fine-grain alloy 718 was 550 °C because the fatigue tests were conducted under Vamas round-robin test program [8]. So the testing temperatures were different for fine- and coarse-grain alloy 718. The specimens were heated in electric resistance furnaces. Temperature of spec-
Fig. 1. Microstructures of materials used.
K. Kobayashi et al. / International Journal of Fatigue 30 (2008) 1978–1984 Table 2 High-temperature tensile properties of materials used Temperature (°C)
Coarse-grain alloy 718 [5] Fine-grain alloy 718 [8] Coarse-grain alloy A286 [6] Fine-grain alloy A286 [9] Coarse-grain 304L steel [7]
600
Tensile strength (MPa)
Elongation (%)
956
1089
21
550
1042
1212
21
600
593
837
19
713
912
22
109
364
44
500
0.2% proof stress (MPa)
1000
Table 3 Fatigue test conditions of stress ratio and temperature Materials
Stress ratio
Temperature (°C)
Coarse-grain alloy 718 [5] Fine-grain alloy 718 [8] Coarse-grain alloy A286 [6] Fine-grain alloy A286 [9] Coarse-grain 304L steel [7]
1 1 and 0.01
600 550 600
0.01
500
Alloy 718 800 Fine-grain
Stress amplitude ; σ a / MPa
Materials
1981
600
Coarse-grain
400
200 : Fine-grain 550˚C
R=-1 0 1 10
10
2
: Coarse-grain 600˚C
10
3
10
4
10
5
10
6
10
7
10
8
Number of cycles to failure ; Nf Fig. 2. Fatigue strengths for fine- and coarse-grain alloy 718.
imen was measured by R-type thermocouples contacted on the specimen surface of the gage portion, although feed-backed thermocouples to control were spot-welded in the shoulder part of the specimen. Temperature gradient was ±5 °C through the gage length for the cylindrical specimen.
1000 A286 600˚C
800
Test results of alloys 718, A286 and 304L steel are shown in Figs. 2–4, respectively. A distinct grain size effect on the fatigue strength was observed, especially in the high-cycle region for alloy 718, shown in Fig. 2, even though the testing temperatures were different by 50 °C. The difference in temperature is not a reason for the difference in the fatigue strength for fineand coarse-grain materials, because the difference in the fatigue strength is not vanished by a normalization of the fatigue strength by the tensile strength at each temperature. For alloy A286 in Fig. 3, testing temperature is the same 600 °C for fine- and coarse-grain materials. There are two test conditions for stress ratio, R. For both the R conditions, grain size effects are observed. For the 304L steel in Fig. 4, the fatigue strength is very low because the material is solid, solution-treated and the test condition of R is 0.01. In this case the data are only four points for the coarse-grain material. The fatigue tests are possible only for limited stress region, in other words, it is impossible to run the tests at higher stress level due to plastic deformation. Fracture surfaces were observed by SEM for the specimens fractured in low- and high-cycle regions.
Stress amplitude ; σ a / MPa
3. Results
600
400
200 : Fine-grain R=-1 : Coarse-grain R=-1 : Fine-grain R=0.01 : Coarse-grain R=0.01
0 1 10
10
2
10
3
10
4
10
5
10
6
10
7
10
8
Number of cycles to failure ; Nf Fig. 3. Fatigue strengths for fine- and coarse-grain alloy A286.
In low-cycle region, the initiation site of fatigue fracture was a surface of the specimens as shown in Fig. 5. This behavior is observed for all the three kinds of materials,
Stress amplitude ; σ a / MPa
1982
K. Kobayashi et al. / International Journal of Fatigue 30 (2008) 1978–1984
400 304L 500˚C 300 200 100 R=0.01 0 1 10
10
2
3
4
5
6
10 10 10 10 10 Number of cycles to failure ; Nf
7
10
8
Fig. 4. Fatigue strengths for coarse-grain 304L steel.
Fig. 6. Fracture surfaces fatigued in high-cycle region for coarse-grain alloy A286 (Stress amplitude: 200 MPa, Nf: 5.48 106 cycles, stress ratio: 0.01).
Fig. 5. Fracture surfaces fatigued in low-cycle region for coarse-grain alloy A286 (Stress amplitude: 500 MPa, Nf: 1.11 104 cycles, stress ratio: 1).
and it is irrespective of the grain size. Fig. 5 is a case of coarse-grain alloy A286. The arrow in the figure indicates the initiation site of the fatigue fracture. It is clear that the initiation site is a surface of the specimen.
In high-cycle fatigue regions, on the other hand, the initiation site was an interior of the specimen as shown in Fig. 6. In this case the material is coarse-grain alloy A286. Dotted circle in the figure indicates a fish-eye pattern [11]. The fish-eye pattern is discriminated from an oxidation color and a roughness of the surface. High magnification reveals that there are many crystallographic facets inside the dotted circle as shown in Fig. 6b. The size of the facets is almost the same as the grain size. As for fine-grain alloys A286, the initiation site was also interior. Fig. 7 is a case of fine-grain alloy A286. Dotted circle in Fig. 7a indicates also the fish-eye pattern [11]. Many crystallographic facets are also observed inside the dotted circle as shown in Fig. 7b. In a case of alloy 718 almost the same fracture surfaces as alloy A286 were observed [10]. In a case of 304L steel, the initiation site of the fracture was a surface of the specimen. It is considered that the 304L steel is not strengthened like the superalloys, and the fracture mode is not interior even in high-cycle region.
K. Kobayashi et al. / International Journal of Fatigue 30 (2008) 1978–1984
1983
The estimated ratio of the fatigue limit at 107 cycles, rcoarse-grain/rfine-grain is 0.86 for alloy A286, and 0.68 for alloy 718. On the other hand, actual ratio of the fatigue strength, rcoarse-grain/rfine-grain is 0.66 for alloy A286 and 0.63 for alloy 718. The value for alloy 718 is normalized by the tensile strength at each temperature. The estimated ratio values are not equal completely to the actual ratio values. The reason is not clear at present, but it seems that this is one of the physical meanings to the effect of austenitic grain size on the fatigue strength. 4.2. Fatigue fracture map for austenitic materials Fatigue limit or fatigue strength at high-cycle region is known to correlate with tensile strength. The results of this study will be compared with reference data for various steels at high and room temperatures [10,12,13]. In this study there are two test conditions for stress ratio R. So the converted fatigue strength derived by the following equation [12] was used. rwðR¼1Þ ¼ 1:317 rwðR¼0:01Þ 17:0;
Fig. 7. Fracture surfaces fatigued in high-cycle region for fine-grain alloy A286 (Stress amplitude: 280 MPa, Nf: 1.57 106 cycles, stress ratio: 0.01).
4. Discussion
ð2Þ
where rw(R=1) and rw(R=0.01) are the fatigue strength for each stress ratio. In this study the fatigue strength is a value at 107 cycles. Eq. (2) is proved to be valid for various carbon steels and austenitic steels [12]. Using this equation makes it possible to evaluate the fatigue strength based on the common test condition of R = 1. The relationship between the fatigue strength and the tensile strength is shown in Fig. 8. The plotted data are the values under R = 1 or converted values from Eq. (2). In Fig. 8 the box shows the results of rotating bending fatigue tests at high-temperatures for various high-temperature materials [9,13]. In this case, the stress ratio of R is 1 and the fatigue strength is a value at 108 cycles. The dotted box shows the results of axial load-controlled fatigue tests for many kinds of steels with a stress ratio of R = 1 at room temperature [12]. In this case the fatigue
In high-strength steels it is well known that fish-eye fatigue fracture occurs in high-cycle region and a non-metallic inclusion exists at a center of fish-eye pattern [11]. Murakami proposed a method for estimating the fatigue limit from a size of the inclusion. The equation is expressed as follows, p 1=6 rw ¼ 1:56ðH v þ 120Þ=ð areaÞ ð1Þ where rwpis an estimated fatigue limit, Hv is Vickers hardness and area is a square root of the defect projected vertically to the load axis. In this study this equation seems applicable to the crystallographic facets fracture. However, it is difficult to determine a location of a center of the fracture inside the fish-eye pattern, because there are many facets everywhere inside the fish-eye pattern. So the maximal grain diameter in Table 1 is used as substitute for the defect area in Eq. (1).
7 Fatigue strength at 10 cycles σ w / MPa
4.1. Effect of grain size 800 700 600
All the data are under R=-1. * means the data converted from Equation (2).
500
a at ld ia x A
300
ta Ro
200
tin
d en g-b
304L
718
ta
400
600
ain se-gr Coar
A286
: Fine-grain A286 (R=-1) : Fine-grain A286 (R=0.01)* : Coarse-grain A286 (R=-1) : Coarse-grain A286 (R=0.01)*
100 200
in
a gd
ain
Equation (3)[12] ( σ w = 0.5 × σ B )
] 13 [9,
400
0 0
r e-g Fin
2] [1
800
: Fine-grain 718 (R=-1) : Coarse-grain 718 (R=-1) : Coarse-grain 304L (R=0.01)*
1000
1200
1400
Tensile strength σ B / MPa Fig. 8. Relationships between fatigue strength and tensile strength.
1984
K. Kobayashi et al. / International Journal of Fatigue 30 (2008) 1978–1984
strength is a value at 107 cycles. These boxes indicate the 95% confidence region. In general, the relationship between the fatigue strength and the tensile strength is expressed as follows [12]. rwðR¼1Þ ¼ 0:5 rB
ð3Þ
where rw(R=1) is the fatigue strength in high-cycle region and rB is the tensile strength at each temperature. The data points such as circles, triangles and squares show the results of this study. The fatigue strength of the 304L steel tested at 500 °C is located in the box. The data of fine-grain alloys 718 and A286 are located near the boxes or near the line of Eq. (3). However, the data of coarse-grain alloys 718 and A286 deviate far from the boxes or the line of Eq. (3). These large deviations are discussed as follows. The reason is thought to be that the fracture site is interior, and the fracture is initiated by the large facets of the crystallographic features. For very fine-grain materials, the fatigue strength is controlled by Eq. (3). Between the fine and the coarse highstrength materials,the grain size effect may appear in the fatigue strength. On the other hand, a low-strength material such as 304L steel shows no grain size effect, because the crystallographic facet fracture does not occur. Therefore, Fig. 8 also shows the fatigue fracture map for austenitic materials as well as a relationship between the fatigue strength and the tensile strength for all kinds of materials. 5. Conclusions Fatigue strength for austenitic superalloys was discussed from the viewpoint of austenitic grain size and the tensile strength. Our conclusions are as follows. High-strength austenitic materials with a coarse-grain showed interior fractures in high-cycle fatigue region. The initiation site was composed of various large facets of the crystallographic features and the facets were as large as the grain size. So the fatigue strength deviated from the equation line of rw(R=1) = 0.5 rB in the relationship between rw(R=1) and rB, where rw (R=1) was the fatigue strength under the stress ratio R of 1, and rB was the tensile strength. On the other hand, high-strength austenitic materials with a fine-grain also showed interior fractures in high-
cycle fatigue region. However, the facets at the initiation sites were as small as the fine-grain size. The fatigue strength was obeyed the equation of rw(R=1) = 0.5 rB. From the above results, Fatigue fracture map for austenitic materials are proposed. This map shows that high-cycle fatigue strength for high-strength austenitic steel or superalloys is determined by the austenitic gain size and the tensile strength, since the high-cycle fatigue strength is controlled by the facet size of the crystallographic features at the initiation area on the fracture surface. References [1] Cozar R, Pineau A. Precipitates and thermal-stability of inconel 718 type alloys. Metall Trans 1973;4(1):47–59. [2] Aerospace Material Specification AMS 5663J. The engineering society for advancing mobility land sea air and space. Warrendale; 1997. p. 1. [3] Bornstein L. Numerical Data and Functional Relationships in Science and Technology. Adv Mater Technol 2006;2:319–21. [4] NIMS. Space Use Materials Strength Data Sheet No.0. Tsukuba: National Institute for Materials Science; 2003. [5] NIMS. Space Use Materials Strength Data Sheet No.4. Tsukuba: National Institute for Materials Science; 2004. [6] NIMS. Space Use Materials Strength Data Sheet No.6. Tsukuba: National Institute for Materials Science; 2005. [7] NIMS. Space Use Materials Strength Data Sheet No.7. Tsukuba: National Institute for Materials Science; 2006. [8] Kitagawa M, Yamaguchi K, Hukuda Y, Komine Y, Hirata H. Analysis on VAMAS low cycle fatigue round robin test in Japan. ISIJ Int 1993;33:817–24. [9] Yoshida S, Kanazawa K, Yamaguchi K, Sasaki M, Kobayashi K, Sato M. Elevated-temperature fatigue properties of 15Cr-26Ni-MoTi-Al-V heat-resisting steel bars for general application. Trans Natl Res Inst Metal 1978;20(6):392–400. [10] Kobayashi K, Yamaguchi K, Hayakawa M, Kimura M. Grain size effect on high-temperature fatigue properties of alloy718. Mater Lett 2005;59:383–6. [11] Murakami Y. Metal fatigue: effects of small defects and nonmetallic inclusions. Oxford: Elsevier; 2002. p. 75. [12] Nishijima S, Ishii A, Kanazawa K, Matsuoka S, Masuda C. Fundamental fatigue properties of JIS steel for machine structural use. National Research Institute for Metals, NRIM Material Strength Data Sheet Technical Document, No.5, 1989. [13] Kanazawa K, Yamaguchi K, Sato S. Elevated temperature, highcycle fatigue properties of engineering materials for high-temperature components. National Research Institute for Metals, NRIM Material Strength Data Sheet Technical Document, No. 6, 1990.