1 Introduction

The generation IV nuclear reactors are characterized by economics, safety, reliability, proliferation resistance, physical protection and sustainability (Ref 1). However, the generation IV nuclear reactors have higher outlet temperature, higher radiation dose and extreme corrosion environment compared with the previous generation reactors (Ref 1). Therefore, structure materials with higher performance are needed for the generation IV nuclear reactors. Reduced activation ferritic/martensitic (RAFM) steels are considered as the candidate material for the generation IV nuclear reactors due to their low coefficient of thermal expansion, good thermal conductivity, excellent creep resistance and oxidation resistance (Ref 2). The microstructure of RAFM steels at room temperature is fully martensite or martensite with a small number of ferrite. RAFM steels also contain a large number of precipitates, such as M23C6 carbide (M = Cr, Fe, W), MX phase (M = V, Ta; X = C, N) and Z phase (Ref 3,4,5). The M23C6 carbide and MX phase impede dislocations movement and play a role in precipitation strengthening, while MX phase is more stable than M23C6 carbide at high temperature (Ref 6, 7). In addition to precipitation strengthening, the strengthening mechanisms of RAFM steels also include solid solution strengthening, dislocation strengthening and grain boundary strengthening (Ref 8, 9).

Mn is one of the usually used alloying elements in steels, which has a strong influence on the inclusions, mechanical properties and microstructure of steels. Mn has a strong ability of deoxidation and desulfurization in steels. It can combine with S to form MnS and eliminate the harmful effects of S (Ref 10). And it can combine with O to form MnO and reduce the O content in steels (Ref 11). Besides, Mn can improve the hardness, strength and wear resistance of steels (Ref 12, 13). The increased Mn content in high-C and high-Mn austenitic steels enhances the stacking fault energy and promotes the formation of deformation twins and work hardening characteristic, thus improving the strength and plasticity of the steels (Ref 13). Furthermore, Mn can also affect the microstructure of steels. The average grain size of high silicon alloyed non-oriented electrical steels decreases as Mn content increases (Ref 14). The addition of Mn in Fe-NiAl steels accelerates the precipitation of NiAl nanoparticles, which increases the hardness and shortens the time to reach the peak hardness (Ref 15). The addition of Mn in Fe-Cu-Ni-Al steels enhanced the effect of precipitation hardening at early aging and accelerated the whole process of precipitation hardening (Ref 16).

Mn in RAFM steels is often used to control the Cr equivalent (Ref 17). It can expand austenite area and avoid the formation of δ-ferrite during the austenitization heat treatment (Ref 18). The δ-ferrite is harmful for tensile and Charpy impact properties of RAFM steels (Ref 19,20,21,22). Furthermore, Mn is considered to reduce the MS temperature, thus reducing the residual tensile stress in steels after thermal cycle processing (welding, additive manufacturing, etc.) (Ref 23). There are only a few studies on the effect of Mn on the properties of RAFM steels. Chen et al. (Ref 24, 25) found that Mn plays a key role in the oxide scale formation of RAFM steels, which promotes the formation of Mn oxides on the surface of steels and improves the compactness of oxide scale. Jin et al. (Ref 26) found that the high temperature oxidation resistance of RAFM steels is improved as the content of Mn increased from 0.04-0.93 wt.%. And the mechanical properties of RAFM steels with different Mn content have no obvious difference (Ref 26). Dong et al. (Ref 27) found that increasing the Mn content in ferritic/martensitic steels reduced the corrosion resistance of ferritic/martensitic steels in supercritical water (SCW). As can be seen, quantitative and systematic studies on the effect of Mn on the microstructure, mechanical properties and high temperature oxidation resistance of RAFM steels are still rare.

In this work, RAFM steels with Mn contents in the range of 0.5-2.0 wt.% were designed. A systematic study of Mn on the microstructure, mechanical properties and high temperature oxidation resistance of RAFM steels were conducted. The correlation between properties and microstructure were presented.

2 Experimental

The chemical compositions of RAFM steels with various Mn contents are given in Table 1. The RAFM steels with 0.5, 1.2 and 2.0 wt.% Mn are named 0.5Mn, 1.2Mn and 2.0Mn, respectively. The RAFM steels were prepared by the vacuum induction melting method, followed by hot-forging into a 45 × 60 mm plate. Subsequently, the hot-forged plates were hot-rolled to 8-mm-thick plate at temperatures between 950 and 1100 °C. In order to obtain the steels with the best performance, normalizing experiments at different normalizing temperatures (950, 980, 1010, 1040, 1070 °C) were carried out on the hot-rolled 0.5Mn steel. After normalizing, the steels were tempered at 760 °C for 90 min.

Table 1 Chemical compositions of the investigated RAFM steels (wt.%)
Full size table

Cubic samples with the dimension of 2.5 × 10 × 30 mm3 were used for high temperature oxidation resistance test. The surfaces of the rectangular samples were ground to 2000 grit using SiC and polished using diamond polishing paste. The polished samples were cleaned by ultrasonic cleaning in ethanol and then drying. High temperature oxidation is carried out in a muffle furnace with air atmosphere, and the oxidation temperature is 650 °C. The samples were taken out at 50, 100, 200, 300 and 400 h, respectively, and weighed discontinuously to obtain the oxidation weight gain curves. The total oxidation time is 400 h.

In terms of the mechanical properties, tensile and hardness tests at room temperature were performed. For tensile test, samples with a gauge section of 5 mm in diameter and 25 mm in length were tested at a crosshead rate of 1 mm/min at room temperature using a universal tensile test machine (Shimadzu AG-X plus Series, Japan) for tensile teste. Standard V-notch Charpy impact specimens with the size of 10 × 5 × 55 mm were subjected to the impact test at room temperature. Three samples of each type RAFM steels were evaluated in tensile test and Charpy impact test. The hardness test was performed by a Vickers hardness tester (HVS-1000Z) with loading force of 500 g for 10 s, and the obtained hardness values were the average of 10 measurements.

The analysis of the microstructure and precipitates was performed using scanning electron microscope (SEM, HITACHI S-3700 N, Japan) and transmission electron microscope (TEM, TECNAI G2 F20, USA). The SEM samples were mechanically polished and etched with 1% solution (5 g FeCl3 + 50 mL HCl + 100 mL H2O) for 1 min. The TEM samples were obtained by jet polishing in a solution of 90% C2H5OH and 10% HClO4 at − 30 °C and 20 V. The phase compositions of oxides on the surface of RAFM steels after oxidation were analyzed with an x-ray diffraction (XRD, Rigaku smartLab-9, Japan) by Cu Kα radiation (45 kV, 200 mA).

3 Results and Discussion

3.1 Optimization of Normalizing Temperature

Figure 1 shows the hardness of 0.5Mn steel after normalized at different temperatures. It is obvious that the hardness of 0.5Mn steel normalized at 950-1010 °C is lower than 200 HV. The hardness of 0.5Mn steel normalized at 1040 °C is the highest (225 ± 2 HV), and the hardness decreased as the normalizing temperature increased to 1070 °C.

Fig. 1
figure 1

Vickers hardness of 0.5Mn steel after normalized at different temperature

Full size image

In order to analyze the reason for the evolution of hardness with normalizing temperature, the microstructure of 0.5Mn steel with different normalizing temperature was investigated by SEM. Figure 2 shows the microstructure of 0.5Mn steel after normalized at different temperature. It can be found that the microstructure of 0.5Mn steels normalized at 950-1010 °C is composed of martensite and ferrite, while the 0.5Mn steels normalized at 1040 and 1070 °C is a fully martensite structure. It is known that the hardness of martensite is higher than that of ferrite. Therefore, the hardness of 0.5Mn steels normalized at 1040 and 1070 °C is higher than 0.5Mn steels normalized at 950-1010 °C. Furthermore, the hardness decreased as normalizing temperature increased from 1040 to 1070 °C. This is mainly due to the coarsening of martensite lath with increasing austenitizing temperature (Ref 28). Therefore, in order to get a fully martensitic structure and optimum hardness, the normalizing temperature of 0.5Mn steel is selected as 1040 °C.

Fig. 2
figure 2

SEM images of 0.5Mn steel after normalized at (a) 950 °C, (b) 980 °C, (c) 1010 °C, (d) 1040 °C and (e) 1070 °C

Full size image

The addition of Mn will reduce the Ac3 temperature of ferritic/martensitic steels (Ref 26). Hence, 1.2Mn and 2.0Mn steels can obtain a fully martensite structure after normalizing at 1040 °C. In order to study the effect of Mn content on the microstructure, mechanical properties and high temperature oxidation resistance of ferritic/martensitic steels, the normalizing temperature of 1.2Mn and 2.0Mn steels is also selected as 1040 °C.

3.2 Microstructure Characterization

Figure 3 shows the phase diagrams of 0.5, 1.2 and 2.0Mn steels obtained using Thermo-Calc with TCFE10 database. There are five types of precipitates in the steels, i.e., M23C6 carbides, Zr(C, N) carbonitrides, Laves phase, Z phase and σ phase. In this work, only M23C6 carbides and Zr(C, N) carbonitrides were present after heat treatment. Other precipitates have slow formation kinetics and usually require extended thermal aging. The Laves phase usually formed after long-term thermal aging, which will lead to creep strength failure (Ref 29). The Z phase also formed after long-term thermal aging, which does not contribute to strengthening and consume the MX particles during their growth, causing a breakdown in creep strength (Ref 30, 31). The σ phase is rare in RAFM steels, which typically formed at boundaries between δ-ferrite and martensite (Ref 32). The effect of σ phase on the mechanical properties of RAFM steels is generally considered to be deleterious (Ref 32). The fraction of M23C6 carbides and Zr(C, N) carbonitrides in 0.5Mn, 1.2Mn and 2.0Mn steels is predicted to be -4 and -0.3%, respectively.

Fig. 3
figure 3

Phase diagrams of (a) 0.5, (b) 1.2 and (c) 2.0Mn steels calculated by Thermo-Calc with TCFE10 database.

Full size image

The SEM microstructures of 0.5Mn, 1.2Mn and 2.0Mn steels are shown in Fig. 4. It is clear that the microstructures of the steels show a fully martensite structure. There are two sizes of precipitates. The larger-size precipitates are rare in steels, while the dimeter can be several microns. The smaller-size precipitates distribute along martensitic lath boundaries and prior austenite grain boundaries (PAGBs), which are probably the M23C6 carbides. The identification of smaller-size precipitates will be conducted by the following TEM analysis. The compositions of larger-size precipitates were analyzed by EDS. As shown in Fig. 4(d), the larger-size precipitates composed of C, N, Zr, Cr and Fe, which corresponding to the Zr(C, N) carbonitrides in previous phases analysis. The Zr(C, N) carbonitrides is formed by ZrC and ZrN, which will precipitate when steels contains C and N (Ref 33).

Fig. 4
figure 4

SEM images of (a), (d) 0.5Mn steel, (b), (e) 1.2Mn steel and (c), (f) 2.0Mn steel normalized at 1040 °C and tempered at 760 °C.

Full size image

In order to further analyze the martensitic lath and precipitates in the steels, TEM analysis was performed. The TEM microstructures of 0.5Mn, 1.2Mn and 2.0Mn steels are shown in Fig. 5. It is clear that the microstructure of the steels is composed of martensitic lath and precipitates. The precipitates distributed along martensitic lath boundaries, which have long strip and ellipse shape, as shown in Fig. 5(a), (c) and (e). In order to identify the precipitates, EDS analysis and selected area diffraction (SAD) analysis were employed. Figure 6 illustrates the EDS maps and mainly element compositions of the precipitates in the steels. It can be found that the precipitates are enriched in Cr, followed by Fe and W. Furthermore, the SAD patterns of precipitates in Fig. 5(b), (d) and (e) are matched well with the diffraction patterns of Cr23C6. Therefore, it can be proved that the precipitates distributed along martensitic lath boundaries are M23C6 (M = Cr, Fe, W) carbides.

Fig. 5
figure 5

TEM images of (a, b) 0.5Mn steel, (c, d) 1.2Mn steel and (e, f) 2.0Mn steel.

Full size image
Fig. 6
figure 6

The EDS analyses of precipitates in Fig. 5

Full size image

In order to uncover the effect of Mn on the tempered martensite and Cr-rich M23C6 precipitates, martensitic lath width and precipitates diameter in the RAFM steels were statistically analyzed, the results are shown in Fig. 7 and  8, respectively. The results of martensitic lath width and precipitates diameter were analyzed from five TEM photographs. It can be seen in Fig. 7 that the martensitic lath width in the steels is 100-500 nm. The average lath width (Lave) of martensite in 0.5Mn, 1.2Mn and 2.0Mn steels is 296 ± 69, 300 ± 79 and 243 ± 69 nm, respectively. It can be found that the Lave of 0.5Mn and 1.2Mn steel is similar, while the Lave of 1.2Mn steel is significantly lower than that of other steels. However, there is no obvious difference in precipitates diameter between the steels, as depicted in Fig. 8. The diameter of precipitates in steels is mainly 50-200 nm. The average diameter (Dave) of precipitates in 0.5Mn, 1.2Mn and 2.0Mn steels is 101 ± 45 nm, 122 ± 49 nm and 111 ± 48 nm, respectively. It can be concluded that the increase of Mn content can refine the martensitic lath of steels, but has little effect on the precipitation behavior. This was mainly because Mn tends to segregate to grain boundaries to lower its energy and induced the solute drag effect (Ref 14). The solute drag effect will confine the grain growth (Ref 34, 35).

Fig. 7
figure 7

The lath width distribution of martensite in (a) 0.5, (b) 1.2 and (c) 2.0Mn steels.

Full size image
Fig. 8
figure 8

The diameter distribution of precipitates in (a) 0.5, (b) 1.2 and (c) 2.0Mn steels.

Full size image

3.3 Mechanical Properties

The mechanical properties of 0.5Mn, 1.2Mn and 2.0Mn steels are listed in Table 2. It is clear that the yield strength (YS, σs) and ultimate tensile strength (UTS, σb) of the steels increase with the increase of Mn content. The YS and UTS of 1.2Mn steel have a little increase compared with 0.5Mn steel, while 2.0Mn steel exhibits the highest YS (625.3 ± 14.0 MPa) and UTS (812.9 ± 4.9 MPa). The strength and toughness of steels are generally opposite. The total elongation (TE, E) of steels decreases with the increase of Mn content. Table 2 also lists the Vickers hardness and impact energy of the steels. The hardness of the steels has a little increase with increasing Mn content. Compared with the little variation of hardness, the impact property exhibits a significant variation. The impact energy of 0.5Mn, 1.2Mn and 2.0Mn steels at room temperature is 43.7 ± 1.7 J, 37.3 ± 0.9 J and 30.6 ± 1.2 J, respectively. The impact toughness of the steels decreases dramatically with the increase of Mn content.

Table 2 Tensile properties, Vickers hardness and impact energy of 0.5, 1.2 and 2.0Mn steels
Full size table

The fracture surfaces of 0.5Mn, 1.2Mn and 2.0Mn steels after tensile test are shown in Fig. 9. It shows that the steels are fractured in a ductile manner and have a large number of dimples. It can be seen in Fig. 9(d) that there are larger-size particles in dimples, which are Zr(C, N) carbonitrides as analyzed previously. The fracture surfaces of steels after Charpy impact test are shown in Fig. 10. It can be found that the fracture surface of 2.0Mn steel is smooth compared with 0.5Mn and 1.2Mn steels. And there are cracks in the fracture surface of 2.0Mn steel, as depicted in Fig. 10(f), which indicates the lower impact toughness. The results of fracture morphologies are in agreement with the mechanical properties of the steels.

Fig. 9
figure 9

SEM images of fracture surfaces of (a), (d) 0.5Mn, (b), (e) 1.2Mn, (c), (f) 2.0Mn steels after tensile tests.

Full size image
Fig. 10
figure 10

SEM images of fracture surfaces of (a), (d) 0.5Mn, (b), (e) 1.2Mn, (c), (f) 2.0Mn steels after impact tests.

Full size image

The results show that the strength increases and the toughness decreases with the increase of Mn content. The yield strength (σs) of steels with primarily martensite and bainite can be estimated by (Ref 36):

$$ \sigma_{s} = \sigma_{Fe} + \sigma_{ss} + \sigma_{P} + \sigma_{L} + \sigma_{D} $$
(1)

where σFe represents the strength of annealed pure Fe, σss represents the solid solution strengthening, σP represents the precipitation strengthening, σL represents the grain size strengthening (or grain boundary strengthening), here we use the lath size instead of grain size, σD represents the dislocation strengthening,

The solid solution hardening σss can be estimated by (Ref 36):

$$ \sigma_{{{\text{ss}}}} = 1723\left[ {\text{C}} \right]^{1/2} + 105\left[ {{\text{Si}}} \right] + 45\left[ {{\text{Mn}}} \right] + 37\left[ {{\text{Ni}}} \right] + 18\left[ {{\text{Mo}}} \right] + 5.8\left[ {{\text{Cr}}} \right] + 4.8 \, \left[ {\text{V}} \right] $$
(2)

where σss in unit of MPa, all the elements [Z] are in weight percent.

The grain size strengthening σL can be estimated by (Ref 36):

$$ \sigma_{L} = K_{L} \cdot L_{ave}^{ - 1 } $$
(3)

where KL represents the coefficient for strengthening due to lath size, ~ 115 MPa, and Lave represents the average lath width with units of μm.

As previous microstructure analysis, Mn has no significant effect on the precipitation behavior. Therefore, the effect of Mn content on the precipitation strengthening (σP) can be neglected. In addition, dislocations in RAFM steels are generally related to the produced process and some alloying elements (C and N) (Ref 37). As the same preparation process and alloying elements (except for Mn) among the RAFM steels, we consider that the dislocation strengthening σD of the RAFM steels is equivalent. Therefore, the strength variation of RAFM steels caused by Mn content is mainly attributed to solution strengthening and grain size strengthening (lath size strengthening). The toughness of the RAFM steels decreases with the increase of Mn content. This is mainly due to the effect of solution strengthening, which improves the strength and reduces the toughness of steels (Ref 38).

3.4 Oxidation Behavior

Figure 11 shows the weight gain curves of 0.5Mn, 1.2Mn and 2.0Mn steels oxidized at 650 °C up to 400 h. It is clear that 1.2Mn steel shows the lowest weight gain, followed by 0.5Mn steel and 2.0Mn steel. In other words, 1.2Mn steel shows the optimum high temperature oxidation resistance among the RAFM steels. Furthermore, the oxidation rate at 200-300 h is obviously higher than that of other times.

Fig. 11
figure 11

The plot of weight gain vs. oxidation time of 0.5Mn, 1.2Mn and 2.0Mn steels oxidized at 650 °C for oxidation times ranging from 0 to 400 h.

Full size image

In order to determine the oxides on the surface of steels after oxidized at 650 °C, the XRD experiment has been carried out and the results are shown in Fig. 12. Two types of oxides were identified except for Fe-Cr matrix, namely Cr2O3 oxides and MnCr2O4 oxides. It is known that the intensity of the XRD peak can indicate the phase content. The higher the peak intensity is, the higher the phase content is. The intensity of MnCr2O4 phase in 0.5Mn steel is higher than that of Cr2O3 phase. Therefore, it can be concluded that the oxides on the surface of 0.5Mn steel is composed of a large number of MnCr2O4 oxides and a small number of Cr2O3 oxides. However, the intensity of MnCr2O4 phase and Cr2O3 phase of 2.0Mn steel is equivalent nearly. This indicates that the content of MnCr2O4 oxides and Cr2O3 oxides on the surface of 2.0Mn steel is equivalent nearly.

Fig. 12
figure 12

XRD analysis of the RAFM steels after oxidation at 650 °C for 400 h.

Full size image

The morphologies of oxides layers of 0.5Mn, 1.2Mn and 2.0Mn steels oxidized at 650 °C for 400 h are shown in Fig. 13. It can be found that only granular oxides are formed on the surface of 0.5Mn steel, while there are both granular oxides and lamellar oxides on the surface of 1.2Mn steel and 2.0Mn steel. In addition, the number and size of lamellar oxides on the surface of 1.2Mn steel are smaller than that of 2.0Mn steel. The element compositions of lamellar oxides and granular oxides are analyzed by EDS, and the results are listed in Table 3. It can be found that the A, B and D particles are granular oxides, which have a similar elemental composition of ~ 50 at.% Cr, ~ 28 at.% Cr and ~ 20 at.% Mn. And the C and E particles are lamellar oxides, which have a similar elemental composition of ~ 50 at.% Cr, ~ 40 at.% Cr and ~ 10 at.% Mn. It should be noted that the compositions of lamellar oxides have a higher Cr content and a lower Mn content than that of granular oxides. There are Cr2O3 oxides and MnCr2O4 oxides on the surface of steels as previous XRD analysis. Therefore, granular oxides and lamellar oxides can be inferred as MnCr2O4 oxides and Cr2O3 oxides, respectively. In addition, there are both Cr2O3 oxides and MnCr2O4 oxides on the surface of 0.5Mn steel according to the previous XRD analysis, while no Cr2O3 oxides is observed. This was mainly due to the oxide layers on the steels have shown a two-layer structure: outer layer and inner layer (Ref 39). The MnCr2O4 oxides are distributed on the outer layer and the Cr2O3 oxides are distributed on the inner layer.

Fig. 13
figure 13

SEM images showing the surface morphologies of (a, d) 0.5Mn, (b, e) 1.2Mn, (c, f) 2.0Mn steels after oxidation at 650 °C for 400 h.

Full size image
Table 3 The element content of oxide particles in Fig. 13 by EDS analyses (at.%)
Full size table

The size of granular oxides and lamellar oxides on the surface of the steels were statistically analyzed by image analysis, and the results are listed in Table 4. The results of average diameter of granular oxides and average length of lamellar oxides were calculated from more than one hundred particles and more than fifty particles, respectively. It can be found that the 1.2Mn steel has the minimum granular oxides with an average diameter of 0.9 ± 0.3 μm. And the 2.0Mn steel has the maximum granular oxides with an average diameter of 1.4 ± 0.4 μm. For lamellar oxides, the average length of 2.0Mn steel is 10.4 ± 2.3 μm, which is ~ 7 times higher that of 1.2Mn steel. Therefore, the reason for 1.2Mn steel exhibits the excellent oxidation resistance could be ascribed to the finer oxides formed on the surface. The finer oxides contribute to form a compactness oxide layer, which improving the oxidation resistance of the steel.

Table 4 The size of oxides on the surface of the RAFM steels oxidized at 650 °C for 400 h
Full size table

It is found that the oxidation resistance of the steels increases with the Mn content increases from 0.5 to 1.2 wt.%. The reason for this increase could be ascribed to the addition of Mn promotes the formation of Mn oxides (Ref 26), which refined the diameter of granular MnCr2O4 oxides. However, the oxidation resistance of steels dramatically decreased as the Mn content further increased to 2.0 wt.%. This was mainly because the presence of higher Mn content reduces the critical concentration of Cr to form Cr oxides (Cr2O3) (Ref 26). Therefore, the Cr2O3 oxides are easier to form in 2.0Mn steel. Finally, the size of lamellar Cr2O3 oxides on the surface of 2.0Mn steel is much higher than that of 1.2Mn steel.

4 Conclusions

The effect of Mn contents (0.5, 1.2 and 2 wt.%) on the microstructure, mechanical properties and high temperature oxidation resistance of the RAFM steels was systematically investigated. In addition, different normalizing temperatures were carried out in order to optimize its heat treatment process. The major results are summarized as follows:

  1. (1)

    The optimum normalizing temperature of the RAFM steels is 1040 °C, at which the structure of steels is fully martensite and the hardness of steels is the highest. Except for lath martensite, a large number of M23C6 carbides and a small number of large-sized Zr(C, N) carbonitrides were detected in the steels. The martensitic lath width was refined as the Mn content reaches 2 wt.%. However, the Mn content has little effect on the precipitation behavior of steels.

  2. (2)

    The strength increases and the toughness decreases with the increase of Mn content. The increase of strength can be ascribed to the solution strengthening by Mn and the gain size strengthening by refining martensitic lath width.

  3. (3)

    The oxides on the surface of the RAFM steels after oxidation at 650 °C for 400 h are composed of MnCr2O4 oxides and Cr2O3 oxides. The steels with 1.2 wt.% Mn exhibit the optimum high temperature oxidation resistance, followed by steels with 0.5 wt.% Mn. The suitable Mn content is contributed to form fine granular MnCr2O4 oxides and improve the oxidation resistance of steels. However, the higher Mn content reduces the critical concentration of Cr to form large-sized lamellar Cr2O3 oxides and reduces the oxidation resistance of steels.