1 Introduction

316LN austenitic stainless steel has been considered a promising candidate material for the Generation IV reactor due to their excellent high temperature mechanical properties, corrosion resistance and adequate weldability [1, 2]. With the advancement of nuclear industry and the pursuit of higher power generation efficiency, structural components have to suffer higher service temperature [3, 4]. Therefore, it is imperative to improve the mechanical properties of 316LN steel at elevated temperatures, particularly creep properties.

During long-term exposure, precipitation strengthening by alloying is usually supposed as an effective approach to improve the creep strength of austenitic steels [5, 6]. The creep rupture life of 316LN steel at 600 °C/200–300 MPa was increased by six times with 0.1 wt% N, inhibiting the precipitation and coarsening of Cr23C6 on grain boundaries [79]. Small addition of Nb could promote fine and stable Z-phase precipitation in 316LN steel, which remarkably reduced the minimum creep rate and prolonged the primary and secondary stages during creep at 650 °C/100 MPa [10, 11]. Additionally, dispersive nanosize Cu-rich particles arose from Cu addition in 316LN and 347 austenitic steels significantly increased the creep resistance under low stress by considerable precipitation strengthening effect because of the coherency strain and partial stacking-fault strengthening [12, 13]. Recently, it was concerned that rare earth (RE) elements could improve creep performance by changing the precipitation behaviors [14, 15]. Yttrium in HP heat-resistant steels increased the fragmentation of primary carbides obstructing the continued growth of creep cracks [16, 17]. Cerium in 27Cr-7Ni duplex steel retards the precipitation of sigma phase during annealing at the temperature of 600–1000 °C after solid-solution heat treatment, which inhibited the degradation of mechanical properties [18]. As a result, it is considered to be feasible to improve the creep properties of 316LN steel by RE addition. However, there is little work about the effect of Ce on creep properties and their mechanism in 316LN steel.

In present study, effect of Ce on creep properties was investigated in 316LN steel. The precipitation behaviors of 316LN steel with 320 ppm Ce have been analyzed in comparison with that of such steel without Ce crept at 700 °C/150 MPa under various strains. The dependence of precipitation behaviors on Ce in 316LN steel was elaborately discussed, and the mechanism of Ce to enhance the creep strength of 316LN steel has been deliberately elucidated.

2 Experimental

Two pilot ingots of 316LN steel without and with Ce additions were smelted in vacuum induction melting furnace. The chemical compositions of these two studied steels designated LN0 and LN1 are listed in Table 1, respectively. The ingots were forged into square bars in section dimension of 45 mm × 45 mm during 1150–900 °C, and then annealed at 1150 °C for 2 h followed by water cooling. The creep specimens were machined in standard round forms with a diameter of 5 mm and a gauge length of 25 mm. Creep tests were conducted at 700 °C with the stress of 150 MPa to simulate the service environments of next generation nuclear power. To analyze the precipitate evolution during creep, the interrupted creep tests were performed at creep strains of 5%, 10% and 20%. The microstructure after interrupt and rupture creep tests was analyzed by scanning electron microscopy (SEM, FEI Apreo). Further observation of precipitates in two studied steels was conducted in scanning transmission electron microscopy (STEM, Talos F200X) equipped with energy dispersive spectroscopy (EDS). Precipitation strengthening was quantitatively determined by the creep threshold stress and Orowan dislocation bypass stress [19]. Additionally, the activity and diffusion coefficient of Mo dependent on Ce content and its influence on chemical driving force of Laves phase in 316LN was calculated by Thermo-Calc with TCFE 11.0 and MOBFE 6.0 databases.

Table 1 Compositions of two studied 316LN steels (wt%)

3 Results and Discussion

Typical creep rupture curves of two studied steels tested at 700 °C/150 MPa are shown in Fig. 1a, while the creep rate as a function of time is shown in Fig. 1b. The creep curves typically consist of secondary, tertiary and negligible primary creep regimes, as shown in Fig. 1a. It is worth noting that Ce addition significantly improved the creep life of 316LN steel from 313 to 556 h at 700 °C/150 MPa. In comparison with the Ce-free 316LN steel, Ce addition increased the creep rupture life by 77.6%. Moreover, the creep rate of LN1 steel is obviously smaller than that of LN0 steel among all creep stages, as revealed in Fig. 1b. The minimum creep rate of LN0 steel is 7.08 × 10–4 h−1, while that of LN1 steel is 2.98 × 10–4 h−1, suggesting that Ce addition effectively decreased the minimum creep rate and substantially increased the creep resistance of 316LN steel.

Fig. 1
figure 1

Creep strain versus time curves a and creep rate versus time curves b at 700 °C/150 MPa. [Inset in a is the enlarged view of the black box showing the negligible primary creep stage]

Figure 2 shows the microstructure of two studied steels after interrupt or rupture creep tests. The intergranular precipitates in the shape of continuous films and isolated particles were found both in two studied steels at 5% creep strain, as shown in Fig. 2a, b. Subsequently, at 10% creep strain, the coarsening of intergranular precipitates and fragmentation of film precipitates also occurred in both studied steels, as demonstrated in Fig. 2c, d. As creep strain is increased to 20%, the continuous film precipitates on grain boundaries are fully replaced by granular precipitation, as revealed in Fig. 2e, f. Similar intergranular precipitation characteristics were observed both in LN0 and LN1 steel at different creep stages, indicating that Ce addition showed insignificant effect on the intergranular precipitation. Furthermore, it is worthwhile noting that great many fine and dense precipitates with an average size of 100.3 nm appeared within matrix of LN1 steel at 5% creep strain (Fig. 2b). Subsequently, the intragranular precipitates in LN1 steel grew gradually as creep strain increased, and their average size coarsened to 166.8 nm after creep rupture. Differently, no intragranular precipitates occurred in LN0 steel throughout entire creep tests, implying that Ce addition induced the formation of intragranular precipitation in 316LN steel during creep at 700 °C.

Fig. 2
figure 2

SEM images of the microstructure in LN0 steel after 5% a, 10% c, 20% e interrupt crept and rupture crept g, and that in LN1 steel after 5% b, 10% d, 20% f interrupt crept and rupture crept h

Figure 3a displays the representative STEM image of the intragranular precipitates in LN1 steel after interrupt crept at 20%. A number of granular precipitates are uniformly distributed in LN1 steel. The structural and chemical analysis of a precipitate inside the grain is shown in Fig. 3b–e. Such intragranular particle was identified as Laves phase with hexagonal structure on the basis of the selected area diffraction pattern (SADP) with [10–1–2] zone axis. Compared to the austenitic matrix, the Laves phase shows Mo (37.9 ± 4.9 wt%) and Si (2.4 ± 0.5 wt%) enrichment and Ni (5.7 ± 1.0 wt%) depletion. Hence, these intragranular particles in LN1 steel during creep deformation were identified as Fe2Mo-Laves phase [20]. Especially, it is worth to note that a nanoscale nucleus with an average diameter of less than 50 nm was found inside the Laves phase, as indicated by yellow arrows in Fig. 3a. These individual nanoscale nuclei about 30 nm in diameter in LN1 steel after creep rupture were observed by STEM, as seen in Fig. 3f. The STEM-EDS line analyses for one of these nuclei are shown in Fig. 3g, indicating that these nanoscale nuclei enrich in Ti, N and Ce. According to the atom ratio of the Ti (0.160 ± 0.044 at.%) and N (0.178 ± 0.012 at.%), these nanoscale nuclei were supposed as TiN particles. Moreover, such TiN particles originally existed in solid-solution LN1 steel but were absent in LN0 steel, indicating that its formation is related to Ce addition.

Fig. 3
figure 3

STEM micrographs of the intragranular precipitates in LN1 steel after interrupt crept at 20% strain a and the enlarged view of an intragranular precipitate b corresponding with selected area diffraction pattern (SADP) c, elemental concentration distribution of Mo d and Si e; STEM images of an individual nanoparticle within matrix in LN1 steel after crept rupture at 700 °C/150 MPa f together with EDS line scanning analyses g. [Yellow arrows in a, b represent nucleus of intragranular particles, while black arrow in f marks the position of an EDS scanning line]

The creep threshold stresses for LN0 and LN1 steel were experimentally determined as 61.7 MPa and 76.6 MPa, respectively. The threshold stress deviation about 14.9 MPa between these two studied steels could be corresponding to intragranular Laves phase [21]. Meanwhile, Orowan dislocation bypass stress for intragranular Laves phase has been also evaluated to be 16.2 MPa close to the threshold stress deviation of 14.9 MPa. It is further confirmed that intragranular Laves phase plays an important role in enhancement of the creep resistance by strengthening the matrix at 700 °C.

Figure 4a describes the relationship between the Fe2Mo Laves nucleation driving force and the Mo content in 316LN steel by thermodynamic calculation at 700 °C. The nucleation chemical driving force of Laves phase significantly increases with the amount of Mo. For instance, raising the Mo contents from 2.5 to 3.0% leads to remarkable increase in driving force from 1080.7 to 1128.9 J/mol. In addition, the dependence of the activity as well as the diffusion coefficient of Mo on the content of Ce was also calculated and plotted in Fig. 4b. It can be seen that the activity and diffusion coefficient of Mo slightly increases in linear relationship with Ce content, suggesting that Ce addition could favor the substitutional diffusion of Mo in 316LN steel at 700 °C. Therefore, Ce enrichment in TiN nanoscale particles shown in Fig. 3g could facilitate Laves precipitating around these particles by heterogeneous nucleation in LN1 steel during creep (seen Fig. 3a). It is consistent with Michael’s research that TiN particles provide nucleation sites for Laves phase during annealing at 600 °C in 316L steel [22]. Besides, Ce addition in 316LN steel might also accelerate Laves phase precipitation by homogeneous nucleation in the matrix pointed by the red arrow in Fig. 3a, because of the slight increase in Mo activity and diffusion coefficient which could reduce the nucleation activation energy of Laves phase. Therefore, the intragranular Laves phase precipitated during creep through heterogeneous or homogeneous nucleation. The former predominated in the early stages of creep, while the latter prevailed after long-term creep (about 150 h in 10% creep strain), both improve creep strength.

Fig. 4
figure 4

Chemical driving force of Laves phase variation with Mo fraction a and the dependence of Ce content on the activity and diffusion coefficient of Mo b, calculated by Thermo-Calc software at 700 °C

4 Conclusion

A small amount of Ce in 316LN steel could remarkably improve the rupture life from 313 to 556 h during crept at 700 °C/150 MPa. Ce has enriched in TiN nanoscale particles as nuclei for intragranular Laves phase precipitates, which facilitates forming fine and dense Laves phase particles. Besides, Ce could increase the activity and diffusion coefficient of Mo in 316LN steel, which might accelerate the precipitation of intragranular Laves phase particles by homogeneous nucleation. In the result, 0.032 wt% Ce addition in 316LN steel has enhanced its creep resistance by the creep threshold stress about 14.9 MPa or Orowan dislocation bypass stress of 16.2 MPa, attributing to the intragranular Laves phase precipitation strengthening effect.