Annealing Time Effects on Mechanically Long Fatigue Crack Growth of TRIP-maraging Steels

Abstract

Compact tension tests for fatigue crack growth were conducted on transformation-induced plasticity (TRIP) maraging steel with two different annealing times (1 and 8 h). Interestingly, resistance to the long crack growth increased with an increasing annealing time at ΔK ranging from 33 to 50 MPa∙m^1/2, whereas a short crack growth resistance, for example, crack growth in a smooth specimen, was reported to show an inverse trend. It is also noteworthy that increasing the annealing time in TRIP-maraging steel decreases both the yield and tensile strengths. Namely, the resistance to the long crack growth showed an inverse trend in the tensile properties, in terms of annealing time. The major microstructural change caused by increasing annealing time was the retained austenite fraction. Specifically, increasing the annealing time increases the austenite fraction, which may have assisted TRIP-related phenomena and associated resistance to long crack growth, for example, transformation-induced crack closure.

1. Introduction

A ductility-strength balance is important for the machining and design of structural components, and fatigue strength/life is crucial for determining the period of use. Transformation-induced plasticity maraging steel (TRIP-M steel) was developed to achieve simultaneous improvements in the ductility-strength balance and fatigue properties.^1,2,3,4,5,6) The fatigue strength is nearly proportional to the tensile strength in smooth specimens,⁷⁾ whereas appropriate control of the microstructures can achieve a better fatigue strength than that predicted from the tensile strength. For example, the fracture surface roughness formation associated with the lamellar structures and hardness increase and/or volume expansion owing to a deformation-induced martensitic transformation improve the fatigue crack growth resistance.^{8,9,10,11,12)} TRIP-M steel intrinsically has a high tensile strength because of precipitation hardening and its extraordinary work-hardening capacity owing to the TRIP effect.¹⁾ In addition, TRIP-M steel has a high fatigue strength and life because of crack deflection and volume expansion, owing to crack propagation along the lamellar structure and a deformation-induced martensitic transformation.^2,3,4,5,6)

Based on the above background, the high fatigue tolerance of TRIP-M steel and its microstructural mechanisms have attracted significant attention. However, to determine the fatigue strength and the life or periodical inspection period of actual structures, an evaluation of the fatigue crack growth characteristics under a wide range of test conditions is required because the fatigue crack growth behavior changes under various conditions (e.g., crack length and stress intensity factor range).¹³⁾ However, the fatigue crack growth characteristics of TRIP-M steel have been evaluated only for mechanically short cracks.^2,3,4,5) Therefore, it is necessary to evaluate the characteristics of mechanically long fatigue crack growth. In this study, we evaluated the growth behavior of mechanically long fatigue cracks in TRIP-M steel under two different heat-treatment conditions. As a result of the evaluation, we report that long-term heat treatment decreases the tensile strength and decreases the resistance to mechanically short fatigue crack growth, but increases mechanically long fatigue crack growth resistance in TRIP-M steel.

2. Experimental Procedure

2.1. Material

Fe-9.3Mn-3.1Ni-1.4Al-0.002C (mass%) steel was used in this study. The steel ingots were cast in a vacuum induction furnace. The ingot was heated to 1200°C, and then hot-rolled at 1000°C or above. Subsequently, the hot-rolled bar was solution-treated at 1100°C and then water-quenched. To obtain a retained austenite, the solution-treated bar was annealed at 600°C for 1 or 8 h and subsequently water-quenched. Consequently, an austenite/martensite laminated multi-phase microstructure was obtained. Hereafter, 1h annealed TRIP-M steel is referred to as TRIP-M 1h, and 8h annealed TRIP-M steel is referred to as TRIP-M 8h. The average lamellar thicknesses of TRIP-M 1h and TRIP-M 8h are 0.21 and 0.30 μm, respectively.

The initial austenite fraction was determined through X-ray diffraction (XRD) measurements using Co radiation using the reference intensity ratio method. The XRD measurement conditions were 2θ = 40–56° and a scan speed of 1 deg/min, sampling interval of 0.05°, voltage of 40 kV, and current of 20 mA. The initial austenite fractions of TRIP-M 1h and TRIP-M 8h were 8% and 34%, respectively.

2.2. Mechanical Tests

Uniaxial tensile tests and fatigue crack growth tests using compact tension (CT) specimens were conducted in this study. Tensile tests were applied at room temperature (23°C) with an initial strain rate of 1.0 × 10⁻⁴ s⁻¹. The strain was measured using a video extensometer. Tensile test specimens had grip sections and a gauge section with dimensions of 30 mm in length, 4 mm in width, and 1 mm in thickness.

Fatigue crack growth tests were conducted using CT specimens and test methods in accordance with ASTM-E647.¹⁴⁾ The test conditions were as follows: For the ΔK decreasing tests, ΔK ≈ 10–35 MPa∙m^1/2, stress ratio R = 0.1, frequency f = 10 Hz, ΔK gradient C = −0.08 mm⁻¹, and for the ΔK increasing tests, ΔK ≈ 30–50 MPa∙m^1/2, stress ratio R = 0.1, and frequency f = 10 Hz. After the ΔK increasing tests were completed, the CT specimens were cut at the center of the thickness using an electrical discharge machining to measure the crack shape and observe the microstructure inside the CT specimens.

Austenite fractions near the crack tip and the crack shapes at the center of the CT specimen thicknesses were investigated for the TRIP-M 1h and TRIP-M 8h specimens after the ΔK increasing tests. The local γ fraction around the cracks was determined using the electron backscatter diffraction (EBSD) method. EBSD measurements were carried out under an acceleration voltage of 20 kV with a step size of 0.15 μm. For both materials after fatigue crack growth tests, the deformation-induced martensite fraction was calculated at ΔK ≈ 30 MPa∙m^1/2 at the center of the thickness of the CT specimens. To determine the observation area for measuring the deformation-induced martensite fraction, the plastic zone size r_p(θ) was calculated using the following equation with the Irwin¹⁵⁾ approach:   

$$\begin{array}{l}{r}_{\text{P}}\left(\theta \right)=2\times r_{\text{y}}\left(\theta \right)\\ \hspace{1em}\hspace{1em}\hspace{0.5em}=\frac{1}{2\pi }{\left(\frac{K_{\text{I}}}{{\sigma }_{\text{YS}}}\right)}^2\left[\left(1-2\nu \right){\left(1+\text{cos}\theta \right)}^2+\frac{3}{2}{\text{sin}}^2\theta \right]\end{array}$$(1)

Although Eq. (1), which is based on an elastic solution, overestimates the plastic zone size r_p(θ), it was used in this study to determine the assumed maximum plastic zone size. To calculate the plastic zone size in the loading direction, θ was set to 90°. The yield stress σ_YS was assumed to be a 0.2% proof stress. Equation (1) shows that the higher the yield stress is, the smaller the plastic zone size. The plastic zone size r_p(θ) of TRIP-M 1h, which has a higher yield strength, is 729 μm according to Eq. (1).

In this study, austenite fractions were plotted against the distance from the fracture surface to the loading direction in the plastic zone at the same level of ΔK. Therefore, based on the plastic zone of TRIP-M 1h, the observation area for both TRIP-M 1h and TRIP-M 8h is approximately 750 μm from the crack tip. The measurement of the austenite fraction by the EBSD method has errors depending on the observation area, beam step size, and beam penetration depth. In this study, the measurement of the γ fraction using the EBSD method needed to be performed with a step size of 0.15 μm because of the wide measurement area, and the measurement error was large owing to the large step size.

Therefore, the ratio of the austenite fraction of the initial microstructure measured using the EBSD method to that of the initial microstructure measured using the XRD was obtained, which was multiplied by the γ fraction measured for the specimen after the fatigue crack growth test, and subtracted from the γ fraction of the initial microstructure measured using the XRD to obtain deformation-induced martensite fractions.

3. Results and Discussion

3.1. Results of Uniaxial Tensile Tests and Fatigue Crack Growth Tests

Figure 1(a) shows the relationship between nominal stress and nominal strain. The 0.2% proof stress and tensile strength of TRIP-M 1h were 719 and 844 MPa, respectively. The 0.2% proof stress and tensile strength of TRIP-M 8h were 465 and 779 MPa, respectively. With an increasing annealing time, both the 0.2% proof stress and tensile strength decreased. It is assumed that the reason for the decrease in the 0.2% proof stress and tensile strength is the thickening of the austenite film, which is preferentially deformed. In addition, not only the thickening of the γ film but also the increase in the initial γ fraction may have contributed to the decrease in the 0.2% proof stress. However, as shown in Fig. 1(b), the work hardening rate of TRIP-M 8h is higher than that of TRIP-M 1h, indicating the enhancement of the TRIP effect owing to the increase in the initial γ fraction.

Fig. 1.

Stress–strain response in the TRIP-M 1h and TRIP-M 8h.: (a) Engineering stress–strain curves and (b) work hardening rate curves. (Online version in color.)

Figure 2(a) shows the results of ΔK decreasing tests in TRIP-M steels, which are compared with the results of conventional steels reported in previous papers.^16,17,18,19) The fatigue crack growth rates of both TRIP-M steels are comparable to those of conventional steels when ΔK is above 20 MPa∙m^1/2. By contrast, when ΔK is below 20 MPa∙m^1/2, the fatigue crack growth rates are lower than those of the conventional cases. In addition, fatigue crack growth behaviors at ΔK > 20 MPa∙m^1/2 show a peculiar relationship between TRIP-M 1h and TRIP-M 8h. To examine the details of the fatigue crack growth behaviors within the high ΔK region, the results of ΔK increasing tests for TRIP-M steels were also applied, as shown in Fig. 2(b). Notably, within the high ΔK region, the fatigue crack growth rates of TRIP-M 8h were lower than those of TRIP-M 1h, despite the 0.2% proof stress and tensile strength of TRIP-M 8h being lower than those of TRIP-M 1h. The resistance to small fatigue crack propagation in TRIP-M steels increases with a decrease in heat treatment time.^2,3,4,5,6) That is, the inverse dependence of fatigue crack growth resistance on the tensile strength is a unique feature of mechanically long cracks. The results shown in Fig. 2(b) are summarized in terms of the Paris law (Eq. (2)), which describes the propagation behavior of mechanically long cracks:   

$$\frac{da}{dN}=C{\left(\mathrm{\Delta }K\right)}^m$$(2)

where C and m are constants. Because C in the Paris law is 1.03 × 10⁻¹⁰ for TRIP-M 1h and 2.73 × 10⁻¹¹ for TRIP-M 8h, the difference was 3.8 fold. However, the value of m was 2.47 for TRIP-M 1h and 2.59 for TRIP-M 8h, which do not show a significant difference. In other words, the crack growth rate of TRIP-M 8h was approximately 1/4 that of TRIP-M 1h within the range of ΔK ≈ 30–50 MPa∙m^1/2. This fatigue crack growth resistance cannot be explained by the tensile strength, and it is necessary to consider the microstructural factors.

Fig. 2.

(a) Comparison of fatigue-crack growth behavior of TRIP-M and conventional steels (dual phase steel,¹⁶⁾ pearlitic steels,¹⁷⁾ SUS3040,¹⁸⁾ AISI 4340,¹⁹⁾ and 300 M.¹⁹⁾ (b) Details of the fatigue-crack growth behavior of TRIP-M 1h and TRIP-M 8h in ΔK increasing test. (Online version in color.)

3.2. Martensitic Transformation

As mentioned in the introduction, the fracture surface roughness and martensitic transformation are the microscopic factors affecting the fatigue crack growth resistance. In this section, we focus on the martensitic transformation behavior, which showed the most significant difference between TRIP-M 1h and TRIP-M 8h. Figure 3 shows the relationship between the deformation-induced martensite fraction (the transformed austenite fraction) and the distance from the crack surface to the loading direction for TRIP-M 1h and TRIP-M 8h. In both TRIP-M 1h and TRIP-M 8h, almost all of the austenite was transformed into martensite near the crack surface. The amount of deformation-induced martensite in TRIP-M 8h was consistently more than twice that in TRIP-M 1h, even in the area far from the crack. In addition, the amount of martensite transformation near the crack surface was measured to be more than 3-times larger. In other words, there is a large difference in the work hardening rate originating from the TRIP effect between TRIP-M 1h and TRIP-M 8h within the large plastic deformation region near the crack. Such a relationship between local plastic deformation and martensitic transformation, which is difficult to evaluate using tensile tests, may have caused the resistance to fatigue crack propagation, which was contrary to the tensile strength shown in Fig. 2(b).

Fig. 3.

Relationship between the transformed austenite fraction after the fatigue crack growth test and the distance from crack surface in TRIP-M 1h at ΔK = 35 MPa·m^1/2 and TRIP-M 8h at ΔK = 33 MPa·m.^1/2 (Online version in color.)

In addition, the effect of martensitic transformation on the fatigue properties, apart from the strength, is known to be a reduction of the crack opening displacement owing to transformation-induced volume expansion. When a martensitic transformation with a volume expansion occurs during fatigue crack growth, the crack closing point is increased by a mechanism similar to a plasticity-induced crack closure, which is called a transformation-induced crack closure (TICC).¹¹⁾ The important factor for TICC is the amount of martensitic transformation in the plastic zone. This is because, when austenite in the plastic zone transforms into martensite, the amount of plastic deformation in the tensile direction that exists around the crack increases because of the expansion caused by the martensitic transformation, not only in the case of propagation through the transformed region, but also in the case of propagation outside the transformed region, which increases the crack closure point. The transformed γ fractions in the plastic zone of TRIP-M 1h and TRIP-M 8h shown in Fig. 3 indicate that a martensitic transformation occurred in the plastic zone for both TRIP-M 1h and TRIP-M 8h. In a future study, we plan to quantitatively analyze the effect of this difference in transformation on the fatigue crack growth rate.

4. Conclusions

Fatigue crack growth tests with an increase in ΔK were conducted on TRIP-M 1h and TRIP-M 8h. Furthermore, the fatigue crack growth resistance of long cracks with different austenite fractions was evaluated. The following conclusions were drawn:

(1) The propagation resistance of mechanically long cracks in TRIP-M steels is superior to that of conventional steels within the range of ΔK below 20 MPa∙m^1/2.

(2) As the heat treatment time increased, the 0.2% proof stress and tensile strength decreased, whereas the growth resistance of the long crack increased. This trend is opposite to that of the resistance to short fatigue crack growth, which increases with the tensile strength. This trend suggests the presence of important factors other than the 0.2% proof stress and tensile strength near the crack, for example, the crack closure and work hardening for a large deformation (post-uniform elongation stage), which cannot be evaluated through tensile testing.

Acknowledgement

This work was financially supported by the 28th ISIJ Research Promotion Grant, and JSPS KAKENHI (20H02457), the Super Research Assistant program at Kyushu University International Research Center for Hydrogen Energy, and the Robert T. Huang Entrepreneurship Center of Kyushu University.

References

• 1)   M. M.  Wang,  C. C.  Tasan,  D.  Ponge,  A. Ch.  Dippel and  D.  Raabe: Acta Mater., 85 (2015), 216.
• 2)   Z.  Zhang,  M.  Koyama,  M. M.  Wang,  K.  Tsuzaki,  C. C.  Tasan and  H.  Noguchi: Int. J. Fatigue, 100 (2017), 176.
• 3)   M.  Koyama,  Z.  Zhang,  M.  Wang,  D.  Ponge,  D.  Raabe,  K.  Tsuzaki,  H.  Noguchi and  C. C.  Tasan: Science, 355 (2017), 1055.
• 4)   Z.  Zhang,  M.  Koyama,  M. M.  Wang,  K.  Tsuzaki,  C. C.  Tasan and  H.  Noguchi: Int. J. Fatigue, 113 (2018), 126.
• 5)   Z.  Zhang,  M.  Koyama,  M.  Wang,  C. C.  Tasan and  H.  Noguchi: Metall. Mater. Trans. A, 50 (2019), 1142.
• 6)   M.  Koyama,  T.  Mizoguchi,  Z.  Zhang and  H.  Noguchi: CAMP-ISIJ, 33 (2020), 283, CD-ROM (in Japanese).
• 7)   S.  Suresh: Fatigue of Materials, Cambridge University Press, Cambridge, UK, (1998), 222.
• 8)   G. T.  Gray,  J. C.  Williams and  A. W.  Thompson: Metall. Trans. A, 14 (1983), 421.
• 9)   S.  Suresh and  R. O.  Ritchie: Metall. Trans. A, 13 (1982), 1627.
• 10)   A. G.  Pineau and  R. M.  Pelloux: Metall. Trans., 5 (1974), 1103.
• 11)   Z.  Mei and  J. W.  Morris: Metall. Trans. A, 21 (1990), 3137.
• 12)   Z. G.  Wang and  S. H.  Ai: ISIJ Int., 39 (1999), 747.
• 13)  The Society of Materials Science, Japan: Hirou Sekkei Binran (Fatigue Design Handbook), Yokendo, Tokyo, (1995), 158 (in Japanese).
• 14)  ASTM E647-13: 2013, Standard Test Method for Measurement of Fatigue Crack Growth Rates.
• 15)   T. L.  Anderson: Fracture Mechanics (3rd ed.), CRC Press, Boca Raton, FL, (2005), 61.
• 16)   S.  Li,  Y.  Kang and  S.  Kuang: Mater. Sci. Eng. A, 612 (2014), 153.
• 17)   A. A.  Korda,  Y.  Mutoh,  Y.  Miyashita and  T.  Sadasue: Mater. Sci. Eng. A, 428 (2006), 262.
• 18)   N.  Nagaishi,  M.  Yoshikawa,  S.  Okazaki,  J.  Yamabe,  F.  Yoshida and  H.  Matsunaga: Eng. Fract. Mech., 215 (2019), 164.
• 19)   R. O.  Ritchie: Int. Met. Rev., 24 (1979), 205.