Gigacycle Fatigue Fracture of Low Strength Carbon Steel, Tested using a Simulated Heat Affected Zone Microstructure

Abstract

In this study, gigacycle fatigue properties were investigated for several microstructures prepared by heat treatment designed to simulate the heat-affected zone (HAZ) that results from welding. The results showed that internal matrix crack origin gigacycle fatigue becomes dominant in coarse-grained microstructures in spite of low tensile strength of only about 600 MPa. It was found that, high material strength is not always necessary and that microstructure plays an important role in the development of internal-origin gigacycle fatigue fractures.

1. Introduction

Gigacycle fatigue is a phenomenon in which the fatigue limit disappears after 10⁷ fatigue cycle region; this is related to internal-origin fatigue fracture, and is commonest in high-strength materials such as spring steel,¹⁾ high-strength steels,²⁾ and titanium alloys.³⁾ Internal cracks are often initiated from inclusions, so the effects of inclusion configuration have been investigated in detail in the past.^2,4) In addition, effect of microstructures has also been pointed out before.^5,6)

Since the heat-affected zone (HAZ) of welded joints shows a relatively wide range of microstructures, the fatigue behaviors of this area are likely to differ from common fatigue data obtained for everyday materials. For example, weld joints sometimes show no clear fatigue limit in spite of their low tensile strength.^7,8) These fatigue behaviors, seen in very high cycle regimes, may be related to the microstructure of the HAZ area. This suggests that clarifying the gigacycle fatigue behavior of HAZs will lead to a better understanding of the role of the microstructure in gigacycle fatigue.

In this study, to clarify the effect of HAZ microstructure on gigacycle fatigue properties, fatigue behavior of up to 10¹⁰ cycles was investigated using ultrasonic fatigue testing on simulated HAZ microstructures that were obtained by heat treatment. There are numerous reports that the ultrasonic loading frequency effect on fatigue life is generally negligible in the internal-origin fatigue fracture mode.^1,2,3)

2. Material and Experimental Procedure

The materials used in this study were low carbon steel, subjected to several simulated HAZ heat treatments. Table 1 shows chemical compositions. The maximum heat treatment temperatures were 1673 K for coarse microstructure materials, and 1273 K for relatively fine-grain microstructure materials. After holding at a maximum temperature for 5 seconds, the cooling rates were controlled as follows. When cooling from 1673 K, the cooling rate was 50 K/s until 1273 K and 30 K/s or 3 K/s after 1273 K. When cooling from 1273 K, the cooling rates were 30 K/s and 10 K/s respectively. Table 2 shows heat treatment conditions and tensile strength of different microstructures A to D. Figure 1 shows the microstructure images of the simulated HAZs. A and B show a relatively fine-grained microstructure, and C and D show a coarse-grained microstructure. Prior austenite grain sizes were 30–50 μm for A and B, and 300–500 μm for C and D.

Table 1. Chemical compositions (mass%).

C	Si	Mn	P	S	Cr	Al	N	O
0.158	0.015	1.52	0.007	0.0012	0.541	0.029	0.0020	0.0026

Table 2. Heat treatment conditions and tensile strength of microstructure A to D.

	Maximum temperature	Cooling rate	Tensile strength
A	1273 K	30 K/s	997 MPa
B	1273 K	10 K/s	812 MPa
C	1673 K	30 K/s	1112 MPa
D	1673 K	3 K/s	663 MPa

Fig. 1.

Simulated HAZ Microstructure images.

Fatigue tests were carried out using ultrasonic-type fatigue testing at 20 kHz and servo hydraulic-type fatigue testing at 20 Hz. The specimens were air-cooled during ultrasonic fatigue testing to prevent any temperature increase. The specimens had an hourglass-type configuration with a diameter of 3 mm.

3. Results and Discussions

Figure 2(a) shows the resultant S–N curves. Since the ultrasonic fatigue testing data showed no discontinuity with the conventional servo hydraulic fatigue testing data, the effect of loading frequency appears to be negligible. Microstructures A and B, with their relatively fine-grained structure, did not break at more than 10⁶ fatigue cycles, and a clear fatigue limit appeared on the S–N curves. In contrast, fatigue fracture occurred at more than 10⁷ fatigue cycles and the fatigue limit disappeared in the coarse-grained microstructures C and D. All fracture origins for microstructure A and B appeared at the specimens’ surfaces, while for C and D, internal matrix origin fractures appeared. These are represented by dashed marks in Fig. 2(a). The tens of microns facets appeared at fracture origins were smaller than the prior austenite grain size of about 500 micrometers.

Fig. 2.

(a) S–N curve and (b) Correlation between fatigue strength and tensile strength. (Online version in color.)

Figure 2(b) shows the relationship between fatigue limit and tensile strength as compared with conventional data.¹⁾ Normally, the fatigue limit is near to being proportional to tensile strength within the range where tensile strength is below 1200 MPa. When the tensile strength exceeds 1200 MPa, the fatigue strength does not rise any further due to internal-origin gigacycle fatigue fracture. The fatigue strengths of microstructure A and B, which had a relatively fine-grain structure, were comparable to those of other structural steels. In contrast, for coarse-grained microstructures C and D, internal-origin fracture appeared in the very low tensile strength region, and the fatigue limit seems to reach an upper limit. The boundary of tensile strength where internal-origin fractures appear, is only about 600 MPa. These results are possibly important to understand the long term fatigue strength of actual weld joint.

Figure 3 is a cross-sectional SEM image of the internal fracture origin facet of microstructure D. As shown in Fig. 3(c), the facet does not seem to correspond to any prior austenite grain boundary and crosses a martensite block boundary. The facet that acted as the gigacycle fatigue origin, therefore appears to be related to a martensite packet boundary.

Fig. 3.

Cross-sectional observation of internal fracture origin facet (Microstructure C, σ_a = 360 MPa, N_f =1.2×10⁸). (a) Cutting line of fracture surface, (b) Higher magnification image of (a) and (c) Cross-sectional SEM image including fracture origin facet.

Tokaji et al., reported similar grain size-related internal-origin fatigue fractures for beta-type titanium alloy.⁹⁾ They reported that the internal-origin fatigue fractures appeared only for the alloys with coarse-grained microstructures in spite of their tensile strength being comparable to that of other materials. Grain size thus appears to play important role in internal-origin fatigue fracture, regardless of tensile strength.

The fracture origin of coarse grain microstructure appears to be related to packet size, as shown in Fig. 3. Even for internal-origin gigacycle fatigue, fatigue life is related to internal crack growth behavior.¹⁰⁾ Hence, the effect of grain size on the fatigue strength is discussed by considering the packet size as initial crack size.

Since packet size is usually proportional to prior austenite grain size,¹¹⁾ the packet size of fine-grained microstructure A, or the initial crack size, is predicted to be one tenth of that of coarse-grained microstructure due to the prior austenite grain size of each microstructure. Since the fatigue limit is inversely proportional to the 6th root of √area, the square root of the projected area of defect as proposed by Murakami.¹²⁾ An explanation of the difference in fatigue strength between microstructures A and C of 520 MPa and 320 MPa respectively, requires an 18-fold difference in grain size. The experimental results of this study cannot therefore be fully explained by the effects of initial crack size, such as packet size. To elucidate the effect of grain size on gigacycle fatigue, it seems necessary to consider the peculiarly of internal small fatigue crack growth behavior.

4. Summary

In this research, ultrasonic fatigue tests were conducted to study the giga-cycle fatigue properties of a carbon steel with different microstructures prepared by various heat treatment process that were designed to stimulate the HAZ’s. We concluded that, even if the tensile strength is low, internal-origin gigacycle fatigue can become dominant depending on the microstructure. In addition, we have shown that whether internal-origin gigacycle fatigue becomes dominant depends on whether the microstructure is coarse-grained.

Acknowledgement

This work was supported by the Council for Science, Technology and Innovation (CSTI), the Cross-ministerial Strategic Innovation Promotion Program (SIP), “Structural Materials for Innovation” (Funding agency: JST). Material preparation and heat treatment was supported by Professor Tadashi Kasuya of Tokyo University.

References

• 1)   T.  Abe,  Y.  Furuya and  S.  Matsuoka: Fatigue Fract. Eng. Mater. Struct., 27 (2004), 159.
• 2)   Y.  Furuya,  H.  Hirukawa,  T.  Kimura and  M.  Hayaishi: Metall. Mater. Trans. A, 38 (2007), 1722.
• 3)   Y.  Furuya and  E.  Takeuchi: Mater. Sci. Eng. A, 598 (2014), 135.
• 4)   Y.  Murakami,  N. N.  Yokoyama and  J.  Nagata: Fatigue Fract. Eng. Mater. Struct., 25 (2002), 735.
• 5)   H.  Oguma and  T.  Nakamura: Scr. Mater., 63 (2010), 32.
• 6)   P.  Zhao,  G.  Gao,  R. D. K.  Misra and  B.  Bai: Mater. Sci. Eng. A, 630 (2015), 1.
• 7)  NIMS Fatigue Data Sheet No. 114, National Institute for Materials Science, Ibaraki, (2011).
• 8)   Y.  Liu,  C.  He,  C.  Huang,  M. K.  Khan and  Q.  Wang: Struct. Eng. Mech., 52 (2014), 889.
• 9)   K.  Tokaji,  J. C.  Bian,  T.  Ogawa and  M.  Nakajima: Mater. Sci. Eng. A, 213 (1996), 86.
• 10)   Y.  Furuya: Mater. Lett., 112 (2013), 139.
• 11)   T.  Maki,  K.  Tsuzaki and  I.  Tamura: Trans. Iron Steel Inst. Jpn., 20 (1980), 207.
• 12)   Y.  Murakami and  M.  Endo: Int. J. Fatigue, 16 (1994), 163.