The effect of grain size on fracture behaviour in tempered martensite embrittlement for AISI 4340 steel

The effect of grain size on fracture behaviour in tempered martensite embrittlement for AISI 4340 steel

~laterials Acieme aml Engineering, 100 ( 19881121 12S 121 The Effect of Grain Size on Fracture Behaviour in Tempered Martensite Embrittlement for AI...

1MB Sizes 0 Downloads 10 Views

~laterials Acieme aml Engineering, 100 ( 19881121 12S

121

The Effect of Grain Size on Fracture Behaviour in Tempered Martensite Embrittlement for AIS14340 Steel HOON KWON and J1N C. CHA

Department of Metallurgical Engineering, Kookmin Univer~'io,,861-1 Jeongneung-1)ong. Seongbuk-Ku. Ak,ou1136 (Korea) CHONG H. KIM

l)epartment ~['Materials Science and Engineering, Korea A dvam'ed Institute of Science and ?~'c/mology, I'. O. Box 131, ('hongrvang, Seou1131 (Korea)

(Received March 10. 1987; in rcvised form August 31. 19871

Abstract

1. I n t r o d u c t i o n

Tempered martensite embrittlement (TME) in AISI 4340 steel was studied Jor how variations in the test temperature and grain size affect the plastic flow. The grain size was changed by varying the austenitizing temperature in the range of 870-1200°C. f o r the evahmtion of TAlE with test temperature, ('harpy impact testing was performed in the range of - 196-23 °C. TME occurs because o/" an effective activation of intergranular brittle fracture in the 300 °C tempered condition where grain boundary carbides are present, the ductile-brittle transition temperature (I)BT~I) increases with itu'reasing grain size attd the transition to brittle .[}'actttre is attributed to the occttrrence of intergranttlar brittle fracture. This effect o[grain size on the ,fiacture behaviour indicates that the intergruntdar brittle Jracture is controlled by the stress concentration susceptibilio,, i.e. the extent of dislocation pile-ttp at the grain boundaries, which increases with increasing groin size. In the 3(f(1°C tempered condition (in the presettce of grain boundary carbides), the D B T T is higher by 70-150 °C, compared with the 200 °C tempered condition (near O, devoid of grain boundary carbide# where the transition to brittle fracture results from transgmnular brittle fracture. A critical test temperalure below which intergranular TME can occur is reduced with decreasing groin size. Therefore, intergranular TME can be produced by the occurrence of intergranular brittle fracture in the presence of grain bounda O, carbides, which can be more eff~etively activated as the stress concentration susceptibili O, increases with increasing groin size or with decreasing test temperature.

A loss in toughness (i.e. tempered martensite embrittlement, TME) when medium-carbon martensitic steels are tempered in the range of 250-450 °C occurs in spite of a decrease in strength with increasing tempering temperature. TME can be classified into two types according to fracture mode: intergranular and transgranular TME. Transgranular TME has been generally observed in high-purity steels and is associated with the formation of interlath carbides [1-5]. Intergranular TME examined in commercial-purity steels and recently in some high-purity steels is correlated with the combined action of impurities (phosphorus and/or sulphur) segregated during austenitizing and carbides formed during tempering at the prior austenite grain boundaries [4-12j. Recent studies suggest that whether intergranular TME or transgranular TME occurs is determined by the plastic flow (strain hardening) behaviour of the matrix [13] or by the intrinsic matrix toughness directly affecting the dislocation motion [14, 15]. Of course, the fracture behaviour in TME will be influenced by the variation of impurity content, relative to matrix toughness. The occurrence of TME is affected by test temperature influencing the matrix toughness [2, 7, 10, 14] and also by the grain size influencing the plastic flow [4, 6, 7, 16, 17]. However, few studies have been conducted on the combined effects of test temperature and grain size on TME. Therefore, in this study the intergranular tempered martensite embrittlement in 4340 steel was systematically analysed by variations in the test temperature and grain size.

{1025-5416/88/$3.50

© Elsevier Sequoia/Printed in The Netherlands

122

2. Experimental procedure AISI 4340 steel used in this study was made by air-melting and hot-forging. The chemical composition was 0.38 wt.% C, 1.69 wt.% Ni, 0.72 wt.% Cr, 0.68 wt.% Mn, 0.27 wt.% Si, 0.24 wt.% Mo, 0.011 wt.% P and 0.014 wt.% S. Charpy V-notch impact specimens 10 mm x 10 mm in cross-section were made from the hot-forged bar. The notch root radius of the impact specimen was modified to 0.5 mm. This modification maximized the tendency for grain size to affect the existence and/or the depth of TME trough developed in room-temperature testing (from no trough in the fine-grained condition to a deep TME trough in the coarse-grained condition). The impact specimens were austenitized for 1 h in a flowing argon atmosphere at 870°C, 1030°C and 1200°C respectively, and then oil-quenched. Decarburization was avoided and a fully martensitic structure was obtained. Austenitized specimens were tempered for 1 h at 200-500°C and then water-quenched. Impact testing was performed in the range of - 196-23 °C. The fracture surfaces were examined in a J E O L scanning electron microscope (SEM) operated at 20 kV. Thin foils were prepared from the fractured specimens and examined by transmission electron microscopy (TEM). Rockwell C hardness was measured at the specimen centres. The specimens tempered at 500°C were etched in a boiling solution of saturated picric acid and the prior austenite grain size was measured by the linear intercept method. The prior austenite grain sizes of the specimens austenitized at 870°C, 1030°C and 1200°C were about 20 /am, 70 /am and 150/am respectively.

tests conducted at - 1 9 6 ° C , the trough becomes shallow. Each - 196 °C datum point represents the average value of two or three tests with scatter bands of about _+ 1.5 J. Although the trough is shallow at - 1 9 6 ° C , the fracture behaviour observed therein is quite different and significant. The impact toughness changes with tempering temperature at various test temperatures in the specimens austenitized at 1030°C are presented in Fig. 3. For room-temperature testing, the impact toughness slightly decreased in the 300 °C tempered condition. For tests at - 9 5 ° C and - 1 9 6 ° C , the data present similar tendencies to those shown in Fig. 2. Figure 4 shows the impact toughness changes with tempering temperature at various test temperatures in the specimens austenitized at 1200 °C. Even for room-temperature testing, a TME trough is explicitly developed with the minimum energy in the 300°C tempered condition. With decreasing test temperature, the depth of the TME trough tends to be decreased. The deep trough is observed at test temperatures where the impact toughness in the 200 °C tempered specimen remains near the upper shelf energy, compared with that in the 300°C tempered specimen which declines toward the lower shelf energy. In

60

-.._..

50

40 r'r-r-

3. Results 3.1. Mechanical properties Room-temperature hardness changes with tempering temperature at various austenitizing temperatures are shown in Fig. 1. Hardness values with tempering temperature are little affected by the variation of austenitizing temperature. Figure 2 shows the impact toughness changes with tempering temperature at various test temperatures in the specimens austenitized at 870°C. For room-temperature testing, the impact toughness continually increases with increasing tempering temperature, i.e. TME is not observed. For a test at - 95 °C, however, a trough, which is deepest for the 300°C tempered condition, is exhibited. For the

3O

O o I

870 °C

Z~ 1030 °C 2O

[]

1200 °C

IO

0

!

I

I

I

200

300

400

500

Tempering temp, °C Fig. l. Fig. l. Room-temperature hardness changes with tempering temperature in the specimens austenitized at 870°C, 1030°C or 1200°C.

123 60

80

70 - Test temp, or-

870 °C

Test temp,

0 2s

~

60

/

50

°C

1200 °C

0 25 A -95

~ -95 [] --196

~

[] -196 :£ 40

o~

5O c @

£z

b~. 50 E

40

>,

o (D

20

t.)

2O

I0

0

I

1

I

200 300 400 Tempering temp, °C

i

500

0

200

300

400

500

Tempering ternp, °C

Fig. 2. Impact toughness changes with tempering temperature in the specimens austenitized at 870 °C and tested at various temperatures,

Fig. 4. Impact toughness changes with tempering temperature in the specimens austenitized at 1200 °C and tested at various temperatures.

60

Test temp, °C 50

1030 °C

the specimens austenitized at 870°C, 1030°C and 1200 °C respectively. From these results, we see that TME can occur at lower test temperatures in the fine-grained specimen than in the coarse-grained specimen. Hence the critical test temperature below which the TME trough is observed decreases with decreasing grain size.

0 25

A -95 []

196

40 @ c @

o

50

el_

E >,

/

()

o

r~

I

l

200

300

Ternoermg

!

400 temp, °C

I

500

Fig. 3, Impact toughness changes with tempering temperature in the specimens austenitized at 1030 °C and tested at various temperatures.

contrast, the shallow trough at - 196°C is caused by the reduction in impact toughness of the 200 °C tempered specimen to the lower shelf energy. For room-temperature testing, in addition, no trough, a shallow trough and a deep trough are observed in

3.2. Microstructural and fractographic observations The transmission electron micrographs in the 200°C and 300°C tempered conditions of the specimens austenitized at 870°C are shown in Fig. 5. In the 200°C tempered condition, the grain boundary carbides are almost absent whereas they are present in the 300 °C tempered condition. SEM fractographs, taken at the centres of the specimens tempered at 200 °C and 300°C, were analysed and related to the existence of grain boundary carbides. Figure 6 shows the fractographs of the specimens austenitized at 870°C. As shown in Figs. 6(a) and (b), the specimens tested at room temperature contain mostly dimples (ductile fracture) in both tempered conditions. Thus, TME is not observed at room temperature since the impact toughness is controlled by the ductile fracture in the 870°C austenitizing condition where the grain size is small. In Figs. 6(c) and (d), however, the fractographs of

124

Fig. 5. Transmission electron micrographs of the specimens austenitized at 870°C: (a) tempered at 200°C; (b) tempered at 300 °C.

the specimens tested at - 196 °C represent the large area of intergranular brittle fracture in the 300°C tempered condition, compared with the mostly transgranular brittle fracture in the 200°C tempered condition. Hence TME troughs observed at low test temperatures are attributed to the intergranular brittle fracture in the 300°C tempered condition with grain boundary carbides (Fig. 5(b)). The fractographs of the specimens austenitized at 1030°C are shown in Fig. 7. The fractographs at room temperature show mixtures of ductile (dimples) and some intergranular brittle fracture (Figs. 7(a) and (b)). The amount of intergranular brittle fracture is greater in the 300°C tempered condition than it is in the 200°C tempered condition. This means that a small TME trough at room temperature is caused by the relatively easy occurrence of intergranular brittle fracture in the 300 °C tempered condition with the grain boundary

carbides. At a test temperature of - 196 °C, a much larger area of intergranular brittle fracture was obtained in the 300 °C tempered condition, whereas the mostly transgranular brittle fracture was obtained in the 200 °C tempered condition owing to the near-absence of grain boundary carbides (Figs. 7(a) and (b)). Thus, TME results from the activation of intergranular brittle fracture in the presence of grain boundary carbides. Figure 8 represents the fractographs of the specimens austenitized at 1200°C. From the fractographs of the specimens tested at room temperature in Figs. 8(a) and (b), a mostly intergranular brittle fracture was obtained in the 300 °C tempered condition whereas a mixture of ductile (dimples) and some intergranular brittle fracture was obtained in the 200 °C tempered condition. The fact that almost all intergranular brittle fracture occurs in the 300 °C tempered condition at room temperature explains the large drop in impact toughness, i.e. a deep TME trough. The fractographs at a test temperature of - 1 9 6 ° C in Figs. 8(c) and (d) show that mostly intergranular brittle fracture was observed in the 300 °C tempered condition at this test temperature whereas mostly transgranular brittle fracture was observed in the 200 °C tempered condition. In spite of a similarly shallow TME trough, the fracture behaviour in both conditions is quite different at - 1 9 6 °C. In other words, the intergranular brittle fracture can be effectively activated by the grain boundary carbides in the 300°C tempered condition, compared with that in the 200°C tempered condition with the near-absence of grain boundary carbides, in the coarse-grained condition. As seen from the fractographic observations, in the 300°C tempered condition in which the grain boundary carbides are present, the intergranular brittle fracture can be more effectively activated with decreasing test temperature or with increasing grain size, and TME is thus caused by the activation of intergranular brittle fracture in the presence of grain boundary carbides. 4. Discussion

In the 4340 steel examined in this study, TME results from the activation of intergranular brittle fracture in the 300 °C tempered condition with the grain boundary carbides. It is suggested that the intergranular TME occurs in the condition where the matrix toughness is relatively high for the intergranular brittle fracture to be activated by grain boundary carbides. This activation of intergranular brittle fracture can be influenced by the test temper-

125

Fig. 6. Fractographs of the specimens austenitized at 870 °C: (a) tempered at 200 °C and tested at 23 °C, (b) tempered at 3 0 0 ° C and tested at 23 °C; (c) t e m p e r e d at 2 0 0 ° C and tested at - 196 °C; (d) t e m p e r e d at 300 °C and tested at - 196 °C.

!

m

Fig. 7. Fractographs of the specimens austenitized at 1030 °C: (a) tempered at 200 °C and tested at 23 °C; (b) t e m p e r e d at 300 °C and tested at 23 °C; (c) t e m p e r e d at 200 °C and tested at - 196 °C; (d) t e m p e r e d at 300 °C and tested at - 196 °C.

126

Fig. 8. Fractographs of the specimens austenitized at 12(I0 °C: (a) tempered at 200 °C and tested at 23 °C; (b) tempered at 300 °C and tested at 23 °C; (c) tempered at 200 °C and tested at - 196 °C; (d) tempered at 300 °C and tested at - 196 °C.

ature and grain size which affect the plastic flow behaviour. In this study, the fracture behaviour in the 200 °C tempered condition in which the grain boundary carbides are nearly absent, in the 300 °C tempered condition where they are present, and finally in TME, has been systematically analysed by varying the test temperature and grain size. For this purpose, the impact toughness changes with test temperature in the specimens austenitized at 870°C, 1030°C or 1200°C and then tempered at 200 °C or 300 °C are shown in Fig. 9.

5O

]Tempering temp, °C

]

I 200 soo I -o--~-

I I

87o

] [.'~.] :-~.:. 1030 Austerqtlzingtemp, °C I ,

/

/f,'" /

///,,"/ / #/;'/

,o"

I

/ ,"./ //

/

;"

//

,'"

I

,,,;I / I I

,"

'

4.1. 200 °C tempered condition (in the near-absence of grain boundary carbides') The impact toughness at room temperature approaches the upper shelf energy (USE) level whereas the impact toughness at - 196 °C does not seem to approach the lower shelf energy (LSE) level in the 200°C tempered condition. When it is assumed that USE and LSE are 22 J and 3 J respectively, and that the ductile-brittle transition temperature (DBTT) is defined as a temperature at which the impact toughness has a midway value

0 1

I

I

I

I

--200

-I00

0

I00

I

Test temperature, °C Fig. 9. Impact toughness changes with test temperature in the specimens austenitized at 870°C, I03/)°C or 12(/(/°C and then tempered at 200 °C or 300 °C.

127 between the USE and LSE, the DBTT in the 870°C, 1030°C and 1200°C austenitized conditions are about - 1 7 0 ° C , - 1 2 0 ° C and - 6 0 ° C respectively. The D B T T increases with increasing grain size. From the fractographic observations, the resistance to transgranular brittle fracture, i.e. transgranular toughness Y-I,decreases with increasing grain size since the transition to brittle fracture takes place by transgranular fracture. In other words, as the grain size becomes larger, the transgranular brittle fracture can occur at a higher test temperature where the matrix toughness has a higher value. In the near-absence of grain boundary carbides, the decrease in y~ with increasing grain size seems to be attributed to the increase in stress concentration susceptibility. At present, however, it is uncertain how the initiation of transgranular brittle fracture is associated with the stress concentration susceptibility increasing with increasing grain size.

4.2. 300 °C tempered condition (in the presence of grain boundary carbides) In the 300°C tempered condition in Fig. 8, if USE and LSE are approximately 24 J and 2 J respectively, the DBTT in the 870°C, 1030°C and 1200°C austenitized conditions are about - 8 5 °C, -50°C and 80°C respectively. The D B T T increases with increasing grain size. In the 300 °C tempered condition where the grain boundary carbides are present, the transition to brittle fracture is produced by the occurrence of intergranular fracture as seen from the fractographs. As the grain size increases, the resistance to intergranular brittle fracture, i.e. intergranular toughness y~, decreases. This cannot be explained by the variation of impurity content at the grain boundaries segregated during austenitizing because the impurity content at the grain boundaries decreases inversely with increasing austenitizing temperature [71. Hence, the stress concentration susceptibility at the grain boundaries in the presence of carbides increases with increasing grain size or with decreasing test temperature, and the intergranular brittle fracture is more effectively activated. If the test temperature is lowered, i.e. the matrix toughness decreases, the relaxation of stress concentrations (e.g. by cross-slip) becomes more difficult. Also, in the coarse-grained condition, the dislocation pile-up at the grain boundaries becomes relatively large and thus the stress concentration susceptibility is enhanced, compared with thai in the fine-grained condition.

4.3. Tempered martensite embrittlement The DBTT in each austenitized condition is higher by 7(I- 150 °C in the 300 °C tempered condition than that in the 200°C tempered condition. Hence, TME results from the activation of intergranular brittle fracture. Even at relatively high temperatures where the matrix toughness is high, the stress concentration at the grain boundaries with carbides are developed with relative ease and intergranular brittle fracture can occur. A critical test temperature below which the intergranular TME can occur is about - 2 0 ° C , 75°C and 170°C in the 870°C, 1030°C and 1200°C austenitizing conditions respectively (Fig. 8 ). As the grain size becomes smaller, TME caused by the effective activation of intergranular brittle fracture can occur at test temperatures below a lower critical temperature. At these lower temperatures where the matrix toughness has a lower value, i.e. the plastic deformation within the grain is more constrained, the stress relaxation at the grain boundary becomes more difficult. Hence, the larger stress concentration susceptibility at the grain boundary can lead to an effective activation of intergranular brittle fracture below a lower critical lest temperature. For the occurrence of intergranular TME, it is necessary that 7-i in an unembrittled condition in the near-absence of grain boundary carbides is larger than )q in an embrittled condition in the presence of grain boundary carbides and that Yl > }'~ in an embrittled condition. In other words, intergranular TME occurs at test temperatures below a critical temperature where the intergranular brittle fracture can be effectively activated, in alloy steels with relatively high intrinsic matrix toughness, which leads to the condition that 71 > 7~. If the impurity content becomes large, ~,~becomes low and thus intergranular TME can occur in a condition with relatively low ),~.. This means that intergranular TME can be produced in an alloy steel with relatively low intrinsic matrix toughness if the impurily content is high 118]. 5. Conclusions

(i) In 4340 steel examined in this study, TME occurs because of an effective activation of intergranular brittle fracture in the 300°C tempered condition. (it) In the 200°C tempered condition in the nearabsence of grain boundary carbides, the DBTT increases with increasing grain size and the transition to brittle fracture is attributed to the occur-

128 rence of transgranular brittle fracture. This effect of grain size on the fracture behaviour seems to indicate that the transgranular brittle fracture is correlated with the variation of stress concentration susceptibility with grain size. (iii) In the 300°C tempered condition in the presence of grain b o u n d a r y carbides, D B T T increases with increasing grain size and the transition to brittle fracture is attributed to the occurrence of intergranular brittle fracture. This effect of grain size on the fracture behaviour means that the intergranular brittle fracture is controlled by the stress concentration susceptibility, i.e. by the extent of dislocation pile-up at the grain boundaries, which increases with increasing grain size. (iv) In the 300°C tempered condition in the presence of grain b o u n d a r y carbides, the D B T T is higher by 7 0 - 1 5 0 ° C , c o m p a r e d with the 200°C tempered condition where grain b o u n d a r y carbides are nearly absent. Thus, T M E occurs because of the intergranular brittle fracture activated by the grain b o u n d a r y carbides. In addition, the critical test temperature below which intergranular T M E occurs is reduced with decreasing grain size.

Acknowledgment T h e authors gratefully acknowledge Mr. N. K. Cho of the K A I S T for assistance with the T E M .

References 1 G. Thomas, Metall. Trans. A, 9 (1978) 439. 2 H. Kwon and C. H. Kim, Metall. Trans. A, 14 (1983) 1389. 3 H. K. D. H. Bhadeshia and D. V. Edmonds, Met. Sci., 13 (1979) 325. 4 J. P. Materkowski and G. Krauss, Metall. Trans. A, 10 (1979) 1643. 5 G. Krauss, Principles" of Heat Treatment of Steel, Am. Soc. Met., Ohio, 1980, p. 212. 6 S.K. Banerji, C. J. McMahon, Jr., and H. C. Feng, Metall. Trans. A,, 9(1978) 237. 7 C. L. Briant and S. K. Banerii, Metall. Trans. A, 10 (1979) 1729. 8 C. L. Briant and S. K. Banerii, Metall. Trans. A, 10(1979) 1151. 9 C. L. Bfiant and S. K. Banerii, Metall. Trans. A, 13(1982) 827. 10 H. Kwon and C. H. Kim, J. Mater. Sci., 18 (1983) 3671. 11 H. Kwon and C. H. Kim, Metall. Trans. A, 15 (1984) 393. 12 N. Bandyopadhyay and C. J. McMahon, Jr., Metall. Trans. A, 14(1983) 1313. 13 E Zia-Ebrahimi and G. Krauss, Acta Metall., 32 (1984) 1767. 14 H. Kwon and C. H. Kim, Metall. Trans. A, 17 (1986) 1173. 15 H. Kwon and C. H. Kim, Metall. Trans. A, 17(1986) 745. 16 J. L. Youngblood and M. Raghavan, Metall. Trans. A, 8 (1977) 1439. 17 D.S. McDarmaid, Met. Tech., 5(1978) 7. 18 K.S. Lee, H. Kwon, Y. C. Lee, J. H. Lee and N. D. Cho, J. Mater. Sci., in the press.