1 Introduction

To avoid serious environmental pollution, the International Maritime Organization (IMO) has recently reinforced regulations on nitrogen oxide (NOx) and sulfur oxides (SOx). In accordance with MARPOL Annex VI, the NOx limit in Tier III and the global SOx limit will be reduced from 3.5 to 0.5%, which will be effective in January 2020 [1]. This has led to increasing demands for liquefied natural gas (LNG) in China, India, and the Middle East [2]. As the demand for LNG increases, more LNG carriers (LNGCs), floating liquefied natural gas (FLNGs), and very large gas carriers (VLGCs) are required.

The structural integrity of storage tanks is one of the most important issues in the construction of LNG carriers. In order to establish the structural integrity of storage tanks, the primary barrier is typically manufactured with low-temperature metals, as shown in Table 1. In general, commonly employed low-temperature metals such as Invar alloys, Al 5083, SUS304L, 9 wt.% nickel alloy, and high-manganese steels exhibit excellent fatigue performance at cryogenic temperature. Many studies have been carried out to investigate the fatigue and fracture performance of such metals.

Table 1 Operation temperature of low-temperature metals [3]
Full size table

Oh et al. [3] investigated the fatigue performance of low-temperature metals such as Invar alloy, Al 5038-O, 9 wt.% nickel alloy steel, and SUS304L. In particular, they quantitatively evaluated the improvement in fatigue performance of each low-temperature metal at cryogenic temperatures [3]. Hwang et al. [4] evaluated the low-cycle fatigue (LCF) performance of 304-L stainless steel weldments to assess the fatigue life in the plastic strain region. The LCF results of weldments were found to exhibit similar tendencies in terms of the thickness [4].

Hyundai Heavy Industry (HHI) [5] assessed the fatigue and fracture performance of 9 wt.% nickel alloy steel. The fatigue performance satisfied the requirements of the fatigue design curves suggested by the IIW recommendations [6]. In terms of fracture toughness, the heat-affected zone (HAZ) has the lowest crack tip opening displacement (CTOD) value among 3 locations (HAZ, parent, and weld metals). Park et al. [7] investigated the fatigue crack growth rates (FCGRs) of nickel alloy steels while considering the nickel content. The material constant C in the Paris law associated with the FCGR of 9 wt.% nickel alloy steel was the highest among various nickel alloy steels. In addition, the slope of the relationship between the fatigue crack growth exponent, m, and the nickel content is increased as the temperature decreases.

Jeong et al. [8] evaluated the near-threshold FCGR of low-temperature metals such as high-manganese, SUS304, and 9 wt.% and 5 wt.% nickel alloy steels. High-manganese steels showed excellent near-threshold FCGRs at 25 °C and − 100 °C compared to other low-temperature metals [8]. However, only a limited number of studies discuss the effects of alloying components on the fatigue and fracture behavior of low-temperature metals. Therefore, this study addresses these effects on the fatigue and fracture performance of the low-temperature metals such as SUS304L, 9 wt.% nickel alloy, and high-manganese steels. In addition, we investigate the fatigue and fracture performance with various welding processes.

2 Low-temperature metals

The chemical composition is very important for the fatigue and fracture performance of low-temperature metals. The mechanical properties and the fatigue and fracture performance are strongly affected by the contents of P, S, Cr, and Si. Toughness is adversely affected by high contents of P and S, and high contents of Si and Cr are known to decrease the grain size while increasing the amount of retained austenite [9]. This study considers SUS304L, 9 wt.% nickel alloy (9% Ni), and high-manganese (high Mn) steels in tensile, CTOD, and FCGR tests. The chemical compositions of the metals are summarized in Table 2.

Table 2 Chemical composition of low-temperature metals
Full size table

3 Test preparation

Five welding consumables manufactured with various chemical compositions were compared to evaluate the fatigue and fracture performance in welds. Welding consumables are denoted by the acronyms in Table 3. In addition, Tables 4 and 5 present the welding parameters related to the heat input of each process and information of welding procedure, respectively.

Table 3 Applied welding processes and chemical composition of welding consumables
Full size table
Table 4 Welding parameters
Full size table
Table 5 Information of welding procedure
Full size table

A servo-hydraulic testing machine (Instron Model 8803) was used for the tensile, fatigue, and fracture tests. The machine has a maximum load capacity of ± 500 kN. The COD gauge length is 10 mm, and the travel length is 4 mm. The test machine and COD gauge are presented in Fig. 1.

Fig. 1
figure 1

Test equipment. a Servo-hydraulic test machine. b COD gauge

Full size image

Tensile tests were conducted according to ASTM E8 [10]. The mechanical properties were measured, including the yield and tensile strengths, elongation, and reduction of area [10]. The tensile test specimens were manufactured according to Part 6.4 in ASTM E8. The geometry of a standard round bar specimen is illustrated in Fig. 2.

Fig. 2
figure 2

Standard round tension test specimen in ASTM E8 [10]

Full size image

The CTOD test is a common fracture mechanics test that measures the resistance of a material to fracture. The CTOD test was conducted with CT specimens according to BS 7448, as illustrated in Fig. 3 [11]. The thickness of test specimens is 10–20 mm, the width is 40–60 mm, and the height is 48–72 mm, depending on the thickness of the materials.

Fig. 3
figure 3

Dimensions of CT specimen (9% Ni)

Full size image

The FCGR test was conducted according to ASTM E647 [12]. The test specimens for SUS304L, 9% Ni steel, and high Mn-steel are 10–20 mm in thickness, 50–80 mm in width, and 60–96 mm in height. Figures 3 and 4 present the typical geometry of a 9% Ni test specimen. The test specimens for the other metals have similar geometry except for the dimensions.

Fig. 4
figure 4

Dimensions of FCGR specimen (9% Ni)

Full size image

4 Results and discussion

4.1 Parent metals

4.1.1 Tensile properties

Figure 5 shows the yield strength (σYS) and tensile strength (σTS) of the parent metals from the tensile test results [13]. In addition, Table 6 summarizes the requirement for mechanical properties of parent metals. In comparison of Table 6, the mechanical properties of three parent metals satisfy the steel grade requirements. High-Mn steel exhibits excellent tensile strength at room temperature. It is well known that the tensile strength increases with the manganese content [14]. The high-Mn steel had the highest carbon content among the three parent metals considered. Therefore, the increase in tensile strength can be explained by the high carbon content [9]. Despite the excellent tensile strength, the weakness of high-Mn steel is the relatively low yield to tensile strength ratio [15]. There have been some studies to increase the yield strength in high-Mn steel [15]. The tensile properties of SUS304L exhibit a noticeable variation at cryogenic temperature compared to other metals.

Fig. 5
figure 5

Comparison of tensile properties of parent metals. a Yield strength. b Tensile strength

Full size image
Table 6 Requirement of mechanical properties
Full size table

4.1.2 Fracture toughness

CTOD tests for parent metals were carried out three times for the parent metals at both room and cryogenic temperatures, and the results are shown in Fig. 6. The CTOD of SUS304L is the highest at room temperature compared to the other two metals. In addition, the CTOD values of SUS304L and high manganese steel are similar. According to Mertinger et al. [16], increasing the chromium contents is known to increase the fraction of the retained austenite.

Fig. 6
figure 6

CTOD values of parent metals: a 25 °C and b − 163 °C

Full size image

Based on the microstructure analysis, the dimple structure of each low-temperature metal was examined, as shown in Fig. 7. At cryogenic temperature, the dimple size was the smallest for SUS304L, followed by high-Mn and 9% Ni steels. The dimple fracture mode of SUS304L can explain the higher toughness, while the dimple structure of 9% Ni exhibits a quasi-cleavage fracture mode.

Fig. 7
figure 7

Fracture surface in low-temperature metals after CTOD test at − 163 °C: a 9% Ni, b SUS304L, and c high Mn. The SEM images were obtained from the front of the fatigue pre-crack tip

Full size image
Fig. 8
figure 8

da/dN–ΔK curves of parent metals

Full size image

4.1.3 Fatigue crack propagation

Figure 8 compares the FCGRs of the parent metals. The material constant m in the Paris equation is fixed at 3.0 to compare it with the FCGR value in BS7910 [17]. All FCG curves are presented in terms of the stress intensity factor (SIF) range of 20 and 50 \( \mathrm{MPa}\sqrt{m} \).

FCGR showed a different tendency from the yield strength. The FCGR of all parent metals tended to decrease as the temperature decreased. In addition, the FCGR of 9% Ni steel was greater than that of SUS304L and high-Mn steels. SUS304L exhibits the slowest FCGR behavior among the metals, which is attributed to the higher silicon content decreasing the grain size [18].

Figure 9 presents the fatigue striation observed in the parent metals at room temperature. The striation spacing is one of the criteria used for the assessment of fatigue crack retardation. The 9% Ni steel had the largest striation spacing among the metals, which suggests that faster fatigue crack growth will occur.

Fig. 9
figure 9

Fracture surface of low-temperature metals after FCGR test at 25 °C: a 9% Ni, b SUS304L, and c high Mn. Red arrows: direction of fatigue crack. The SEM images were obtained from the front of the fatigue pre-crack tip

Full size image

4.2 Weld metal

4.2.1 Tensile properties

Figure 10 shows the mechanical properties of weld metals [19]. At room temperature, the tensile strength of the Mn-SA is the highest among the five weld metals. The increased tensile strength is attributed to the higher chromium content of up to 3.0 wt.% [20]. The tensile strengths of SUS-FC and SUS-TI are higher than those of the other weld metals at cryogenic temperature. Moreover, SUS-FC and SUS-TI exhibit drastic increases in the tensile strength at cryogenic temperature. It is estimated that the mechanical properties of SUS304L are substantially influenced by the secondary hardening [21].

Fig. 10
figure 10

Comparison of mechanical properties for weld metals. a Yield strength. b Tensile strength

Full size image

The secondary dendrite arm spacing (SDAS) was investigated to provide further insight into the tensile behavior. Figure 11 compares the SDAS of the five weld metals, and the values are summarized in Table 7. The SDAS of SUS-TI is the largest, while Mn-SA has the smallest value. The SDAS tends to decrease as the tensile strength increases [22]. Therefore, the strong increase in tensile strength can be explained by the low SDAS.

Fig. 11
figure 11

Secondary dendrite arm spacing of five weld metals: a 9Ni-FC, b 9Ni-SM, c SUS-FC, d SUS-TI, and e Mn-SA. Yellow arrows: secondary dendrite arm

Full size image
Table 7 Secondary dendrite arm spacings of five weld metals
Full size table

4.2.2 Fracture toughness

The CTOD values of the weld metals were measured three times for each metal with various welding processes and consumables, and the results are compared in Fig. 12. All five weld metals satisfy the minimum CTOD requirement by DNVGL at both room and cryogenic temperatures [23]. In addition, there was a large variation of CTOD values in SUS-FC and SUS-TI at cryogenic temperature. According to the reference paper, the increase in sigma phase has an adverse effect on fracture toughness [24]. In addition, Gow and Harder [25] suggested that if ratio factor > 1.7, the sigma phase can precipitate in stainless steels.

$$ \mathrm{Ratio}\ \mathrm{factor}=\frac{\% Cr-16\ast \%C}{\% Ni} $$
Fig. 12
figure 12

CTOD values of weld metals: a 25 °C and b − 163 °C

Full size image

In cases of SUS-FC and SUS-TI, the ratio factors are 1.84 and 1.93, respectively. Therefore, it is estimated that the fracture toughness of SUS-FC and SUS-TI at cryogenic temperature is affected by sigma phase.

The Mn-SA has the highest CTOD value at room temperature. According to Havel [20], the content of molybdenum is typically increased to stabilize the austenite in high-Mn steel. In this study, the welding consumable used for high-Mn steel includes molybdenum. As the temperature decreases, the CTOD values of weld metals in 9% Ni and high-Mn steel become higher than that of SUS304L. The contents of sulfur and phosphorus were negligible in the welding consumables of 9% Ni and high-Mn steel.

The dimple structures of weld metals are illustrated in Fig. 13. The dimple size of SUS-TI is the smallest, and that of SUS-FC was sensitive to the temperature changes. Therefore, the variation of the dimple size with temperature can account for the CTOD values at cryogenic temperature.

Fig. 13
figure 13

Fracture surface in weld metals after CTOD test at − 163 °C (d): a 9Ni-FC, b 9Ni-SM, c SUS-FC, d SUS-FC at 25 °C, e SUS-TI, and f Mn-SA. The SEM images were obtained from the front of the fatigue pre-crack tip

Full size image
Fig. 14
figure 14

da/dN–ΔK curves in weld metals.

Full size image

4.2.3 Fatigue crack propagation

Figure 14 compares the FCGRs in weld metals. SUS-FC and SUS-TI exhibit slower fatigue crack growth rates at both room and cryogenic temperatures. The silicon content influences the decrease in FCGR due to the presence of martensite [26]. One major observation is that the FCGRs of SUS-FC and SUS-TI exhibit significant variation with temperature compared to the other weld metals. In the case of the weld metal in SUS304L, the austenite changes into martensite when the temperature decreases [27]. The martensitic transformation that occurs at the tip of fatigue cracks in SUS304L appears to reduce the FCGR significantly [28].

We observed the striation in weld metals at room temperature to investigate the microstructure effects on FCGR, as shown in Fig. 15. SUS-TI has the lowest striation spacing. FCGR tends to decrease as the striation spacing decreases. According to the SEM images, the FCGR results can be reasonably explained by the observed striation spacing, indicating slower fatigue crack growth with lower striation spacing.

Fig. 15
figure 15

Fracture surface of weld metals after FCGR test at 25 °C: a 9Ni-FC, b 9Ni-SA, c SUS-FC, d SUS-TI, and e Mn-SA. The SEM images were obtained from the front of the fatigue pre-crack tip

Full size image

5 Summary and conclusion

We investigated the fatigue and fracture performance of the low-temperature metals SUS304L, 9% Ni, and high-Mn steel while considering the effects of different alloying components. In addition, microstructure analysis was performed on five weld metals to investigate the fatigue and fracture performance. The important observations are summarized as follows:

  • The tensile strength of high-Mn steel exhibits excellent tensile strength at room temperature in the parent and weld metals. The tensile strength is affected by the amounts of manganese, carbon, and chromium. In the case of the weld metal, the low SDAS can explain the substantial increase in tensile strength. In addition, the tensile strength of SUS304L has significant variation with the temperature decrease as shown in Fig. 5b. The increase in tensile strength of 9% Ni, high Mn, and SUS304 are 21%, 42%, and 20%, respectively. Therefore, the tensile strength of SUS304L is strongly increased at cryogenic temperatures compared to the other two steels. It is estimated that the mechanical properties of SUS304L are substantially influenced by secondary hardening.

  • The CTOD values of SUS304L are the highest at room temperature compared to the other two metals. At cryogenic temperature, CTOD values of SUS304L and high manganese steel are similar. In case of FCC austenite steels, it can be explained that higher chromium content is known to increase the fraction of the retained austenite. The dimple fracture mode of SUS304L can explain the higher toughness than that of 9% Ni steel, where the dimple structure showed a quasi-cleavage fracture mode.

  • The minimum CTOD values of SUS-FC and SUS-TI weld metal were lower than those of other three weld metals as the temperature decreased. The decrease in fracture toughness at low temperature is influenced by the contents of sulfur and phosphorus. Mn-SA has the highest CTOD value at room temperature. In high-Mn steel, the amount of molybdenum is increased to stabilize the austenite. As the temperature decreases, the CTOD values of weld metals in 9% Ni and high-Mn steel became higher than that of SUS304L. The contents of sulfur and phosphorus were negligible in the welding consumables of 9% Ni and high-Mn steel. The dimple size of SUS-TI is the smallest among weld metals. Therefore, the variation in the dimple size with temperature can account for the CTOD values at cryogenic temperature.

  • SUS304L exhibits the slowest FCGR behavior among the low-temperature metals, which is attributed to the higher silicon content, which in turn plays a role in reducing the grain size. Furthermore, 9% Ni steel had the largest striation spacing compared to SUS304L and high-Mn, which suggests that faster fatigue crack growth will occur.

  • The fatigue crack growth rates of SUS-FC and SUS-TI weld metals was the slowest at both room and cryogenic temperatures. The silicon content influences the decrease in FCGR due to the presence of martensite. FCGRs of SUS-FC and SUS-TI exhibit strong variation with temperature. The martensitic transformation that occurs at the tip of a fatigue crack in SUS304L significantly reduces the FCGR. SUS-TI exhibits the lowest striation spacing among the five weld metals. According to the SEM images, FCGR results can be reasonably explained by the striation spacing, indicating slower fatigue crack growth with lower striation spacing.