2016 Volume 57 Issue 11 Pages 1945-1951
The effect of nitrogen addition on the microstructure formation and hardness during solidification and heat treatment was investigated and the possibility of nitrogen as an alloying element was discussed in terms of alloy chemistry for high-carbon high-speed tool steel type cast alloys. Nitrogen with a concentration from 48 ppm to 1542 ppm was successfully introduced by mixing Cr2N into a molten alloy. Analysis of the diffraction pattern revealed that the primary V2CN carbonitride crystallized upon the addition of nitrogen, whereas eutectic carbides mainly formed in N-free specimens. The chemical composition of the carbonitride is also affected by the addition of nitrogen. With increasing quenching temperature, the Vickers hardness gradually increased to a peak and then decreased. Nitrogen addition helps to increase the hardness similarly to carbon. A N-containing specimen also exhibited superior secondary hardening after tempering. It is known that a large amount of residual austenite finally transforms to a hard martensite phase after tempering. According to the results of XRD analysis, nitrogen addition increases the volume fraction of retained austenite in the matrix at a higher holding temperature. Furthermore, the precipitation of nanosize carbonitride was observed around the primary V2CN carbonitride in addition to the standard precipitation. Therefore, this carbonitride precipitation may induce the superior secondary hardening and ultimately increase the macrohardness of N-containing specimens.
This Paper was Originally Published in Japanese in J. Japan Inst. Met. Mater. 79 (2015) 169–175.
In the last few decades, the conditions required for hot rolling have become increasingly severe because of the increasing strength required for mill products. The abrasion speed and exchange cycle of rolls have also accelerated because the quantity and speed of the rolling process have increased. Therefore, higher quality has been required for hot-rolling mill materials in terms of hardness, wear resistance, heat resistance, and lifetime, as well as a reduction in the production cost1–4). High-carbon high-speed steel type cast alloys (HCHSSs) have been developed as work roll materials for hot rolling. HCHSSs have similar chemical compositions to high-speed steel and include hard eutectic MC- and M2C-type carbides (M is a carbide-generating element such as Cr, Mo, or V) between primary austenite dendrites5–10). Furthermore, austenite dendrites transform to hard martensite, in which fine secondary carbides precipitate after quenching and tempering, and exhibit superior hardening11–15).
To improve the wear resistance of HCHSSs, it is necessary to control the species, volume fraction, and distribution of carbides crystallized in the solidification process and to optimize the heat treatment conditions16–20). We here focused on the addition of nitrogen as an alloying element, which forms hard intermetallic compounds with Al, Cr, Nb, Ti, or V and may be as effective as carbon during the solidification and heat treatment processes21). In addition, it has been reported that considerable secondary hardening was induced by the precipitation of fine CrN in martensite stainless steel including nitrogen22). HCHSSs contain many alloying elements, which can easily react to form nitrides. Thus, the solidification and heat treatment processes may be affected by the addition of nitrogen. However, there has been little investigation of the effect of nitrogen on the characteristics of HCHSSs.
Therefore, in this study, nitrogen-containing specimens were produced by a conventional casting process with chrome nitride (Cr2N), and the microstructural formation and the variation of the Vickers hardness were investigated in relation to the composition of nitrogen and the amount of retained austenite during the solidification and heat treatment, and the advantage of adding nitrogen to HCHSSs is discussed.
High-purity iron, pure carbon, pure chromium, ferromolybdenum, and ferrovanadium were measured in accordance with a standard chemical composition (Fe-1.7-2.3mass%C (abbreviated to %)-5%Cr-5%Mo-5%V). The materials were melted in an alumina crucible at about 1573 K in an induction furnace then cast into permanent molds to form as-cast cylindrical specimens of 20 mm diameter and 200 mm length. Nitrogen (N)-containing specimens were fabricated by the addition of powdered Cr2N to the same raw materials. The N-free and N-containing specimens were cast in Ar and N2 atmospheres, respectively. The chemical compositions of the specimens are listed in Table 1.
| Alloy | Chemical compositions | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| (mass%) | (ppm) | Fe | |||||||
| C | Cr | Mo | V | N | O | ||||
| 1.7%C | 1.76 | 4.94 | 4.95 | 5.04 | 78 | 24 | Bal. | ||
| N+1.7%C | 1.75 | 4.97 | 4.89 | 5.04 | 1542 | 24 | Bal. | ||
| 2.0%C | 2.04 | 4.96 | 5.00 | 5.07 | 57 | 49 | Bal. | ||
| N+2.0%C | 2.06 | 5.01 | 4.87 | 5.12 | 1319 | 69 | Bal. | ||
| 2.3%C | 2.35 | 4.96 | 4.87 | 5.10 | 48 | 39 | Bal. | ||
| N+2.3%C | 2.33 | 5.07 | 5.01 | 5.08 | 1173 | 21 | Bal. | ||
To evaluate the optimum heat treatment conditions, specimens were subjected to holding, quenching, and tempering processes. Specimens were placed in a SiC electric-resistance furnace for 3.6 ks with the temperature set at intervals of 50 K in the range from 1123 to 1373 K. After solution treatment, the specimens were quenched in an oil bath. Finally, the quenched specimens were tempered two or three times from 673 to 873 K for 3.6 ks and subsequently underwent oil quenching.
After the heat treatment, the specimens were sectioned and polished by a standard metallographic procedure, etched by Murakami reagent (30 g KOH, 30 g K3Fe(CN)6, 30 mL water) at 333 K for 60 s, and observed with an optical microscope. The macroscopic hardnesses of the as-cast and heat-treated specimens were measured by Vickers hardness testing equipment (Mitutoyo, HV-115) under a load of 294.2 N for 10 s. Hardness test was performed 10 times per specimen, and the standard deviation of hardness was calculated less than 4 HV. Specimens were also scanned by XRD (Rigaku, RINT-2100) at an accelerating voltage of 40 kV, a current of 40 mA, a scan speed of 2°/min, and a scan step of 0.02 mm in the range from 30 to 100° using a Co target. The amount of retained austenite was also measured by XRD using a Mo target. Compositional analysis was carried out on the as-polished specimens by SEM (Zeiss, ULTRA55) with EDS (SII NanoTechnology) at an acceleration voltage of 8–10 kV and the ZAF calibration method. The cross-sectional TEM specimens were prepared by lift-out in dual-beam scanning electron microscope/focused ion beam (DB-SEM/FIB) systems (FEI, Quanta 3D 200i). Then, the TEM specimens were irradiated by an Ar+ ion beam to remove the damaged layer on the surface. A selected area diffraction pattern of carbonitride was analyzed by TEM (JEOL, JEM-2010HCKM) at an accelerating voltage of 200 kV. The STEM observations and EDS analysis were carried out at an accelerating as above voltage of 200 kV using a scanning TEM (JEOL JEM-ARM200F) equipped with double Cs-correctors, and an SDD with an enlarged sensing area of 100 mm2.
Nitrogen with a concentration of more than 1000 ppm was successfully incorporated in the as-cast specimen, however, the N composition decreased with increasing C composition as shown in Table 1. The solubility of nitrogen in a molten Fe-Ni-Cr steel has been reported to satisfy the following equation proposed by Japan Society for the Promotion of Science23)
| \[\begin{split} \log[N] = &-\frac{518}{T} - 1.063 + 0.046[Cr] - 0.00028[Cr]^2 \\ & + 0.02[Mn] - 0.007[Ni] - 0.048[Si] \\ & + 0.12[O] - 0.13[C] + 0.011[Mo] \\ & - 0.059[P] - 0.007[S] \end{split}\] | (1) |
Figure 1 presents the microstructures of the as-cast specimens etched by Murakami reagent. In the N-free specimen, eutectic structures including colony-formed MC- and platelike M2C-type carbides were observed between well-developed primary austenite dendrites in the 1.7%C and 2.0%C specimens. A eutectic structure including an M7C3-type carbide was also observed in the 2.3%C specimen. It was found that the area fractions of MC and M2C carbides are approximately 8–12% and 4–7%, respectively, and tend to increase with increasing C composition. On the other hand, in the N-containing specimens, a primary dendritic phase develops with an area fraction of around 0.6–1.1% inside the austenite dendrites as well as eutectic MC and M2C carbides. The composition and lattice parameters of the primary dendritic phase were identified by EDS and TEM. The results are shown in Fig. 2 along with a chemical analysis of the MC and M2C phases in the N+2.0%C specimen. Lattice spacings of 0.2075 nm in the $(\bar{2}00)$ direction and 0.1467 nm in the $(\bar{2}20)$ direction were estimated from the diffraction pattern in Fig. 2(b), and the composition ratio of V+Mo, C, and N is approximately 2:1:1 from Fig. 2(c), where the X-ray diffraction spectra of M2CN was obtained. These results indicate that the primary dendritic M2CN-type carbonitride formed during the solidification. The eutectic MC carbide was coarsened and the area fractions of MC and M2C were reduced by the formation of M2CN carbonitride as shown in Figs. 1(d)–(f). From Fig. 2(c), nitrogen preferentially accumulated in primary M2CN than in the eutectic carbides. The eutectic MC carbide consisting of 10% N with C, V, Mo, and Fe may have a solid solubility limit of N in the N+2.0%C specimen. On the other hand, N did not dissolve in the eutectic M2C carbide.
Solidification microstructures of Fe-(1.7-2.3)C-5Cr-5Mo-5V alloys and the change upon the addition of nitrogen. (a) 1.7%C, (b) 2.0%C, (c) 2.3%C, (d) N+1.7%C, (e) N+2.0%C, and (f) N+2.3%C specimens.
(a) SEM image of carbonitride and carbides, (b) diffraction pattern of carbonitride, and (c) contents of alloying elements in each phase for as-cast specimen with added nitrogen and 2.0% carbon.
To find the optimum heat treatment conditions, the as-cast specimens were held at an elevated temperature then quenched in an oil bath. The variation of the Vickers hardness with the quenching temperature of the HCHSSs is summarized in Fig. 3. With increasing quenching temperature, the Vickers hardness gradually increased to a peak value and then decreased for each specimen. The N-free 1.7%C specimen had the lowest hardness of 635 HV at 1123 K. On the other hand, the N+1.7%C specimen had a hardness of 706 HV at 1123 K, higher than that of the N-free specimen. The 2.0%C and 2.3%C specimens had similar hardnesses at 1123 K. Figure 4 shows the effect of the quenching temperature on the ratio of retained austenite. Less than 10 vol% of austenite remained, which indicates that 90% of the matrix transformed to martensite at 1123 K. To evaluate the effect of the chemical composition on Vickers hardness, the hardnesses at 1123 K were arranged in order of total composition of C and N, and a linear fit was made to the data as shown in Fig. 5. However, the ability of N to increase the hardness of carbides is almost half that of C according to Takano25). Thus, it is unlikely that the increase in the hardness is due to the incorporation of N. Therefore, the increased hardness with nitrogen addition may be because the C content in the martensite matrix is increased by the reduction of C from the M2CN or MC carbide due to the substitution of N with C.
Effect of quenching temperature and the addition of carbon and nitrogen on the Vickers hardness of specimens.
Relationship between ratio of retained austenite and quenching temperature.
Effect of total carbon and nitrogen content on Vickers hardness of specimens quenched at 1123 K.
The Vickers hardness gradually decreased from the temperature of peak hardness to 1373 K. In particular, the hardness of the N+2.3%C specimen decreased from 945 HV at 1173 K to 603 HV at 1373 K. Figure 4 indicates that 60% of the soft austenite remained at 1373 K. This is because a large amount of the alloying elements dissolved in the austenite matrix, which may have decreased the martensitic transformation temperature (Ms temperature). Moreover, the hardness may be affected by the C and N compositions, similarly to other alloying elements. Thus, the compositions of C and N in the primary M2CN phase were measured as a function of the holding temperature and are plotted in Fig. 6. The composition of C was one-third lower than that of N at 1123 K, but it increased with increasing holding temperature, and the composition of C was almost twice that of N at 1373 K. From Fig. 3, the difference in hardness between the N-containing and N-free specimens became small as the holding temperature increased. It can be seen that the occupation of C and N in the primary M2CN phase affected the C composition in martensite and the hardness of the specimen even when the area ratio of M2CN was about 1%.
Redistribution of carbon and nitrogen between carbonitride and austenite during quenching.
To optimize the heat-treatment process of the high-carbon high-speed steel alloys suitable for hot-rolling materials, the effect of the tempering temperature on the microstructure and macrohardness was evaluated. Prior to the tempering heat treatment, two series of representative specimens were selected. The first series comprised specimens with a peak hardness; these composed the 2.0%C and N+2.0%C specimens quenched from 1223 K and the 2.3%C and N+2.3%C specimens quenched from 1173 K. The second series comprised specimens quenched from 1373 K. First, the specimens with a peak hardness were tempered twice and subjected to macrohardness analysis. It was found that the hardness gradually decreased in all specimens at an elevated tempering temperature, while resistance to temper-softening appeared around 800 K as shown in Fig. 7(a). Little retained austenite was detected in these tempered specimens as shown in Fig. 7(b), which indicates that the treatment temperature of 1273 K was too low to incorporate the alloying elements to induce secondary hardening. Then, the specimens quenched from 1373 K were tempered three times in the temperature range from 673 to 873 K, and the obtained relationship between the hardness and tempering temperature is summarized in Fig. 8. All the specimens exhibited secondary hardening and had a peak hardness of more than 800 HV at a tempering temperature of 823 K. The comparison between the curves for the tempered N-containing and N-free specimens suggests that the peak hardness was enhanced by N addition. To consider the effect of the microstructure, the amount of the retained austenite in the tempered specimens was measured by XRD and is shown in Fig. 9. The XRD analysis indicated that a relatively large amount of the austenite phase was retained in the N-containing and high-carbon specimens compared with the N-free and low-carbon specimens. The amount of the retained austenite did not change in the range from 673 to 773 K, which suggests that the retained austenite was stable even after three tempering treatments and that it was difficult for the secondary carbide to precipitate in this temperature range. On the other hand, the amounts of austenite in all specimens abruptly decreased upon tempering at 823 K and above, indicating that the martensite transformation occurred at these temperatures. It is known that heat treatment at a higher temperature accelerates the precipitation of carbides in the austenite matrix13). This precipitation reduces the composition of the alloying elements, destabilizes the retained austenite, and finally induces the martensite transformation. From the experimental results, the greater the incorporation of alloying elements, the greater the secondary hardening that occurred at 823 K and above. The macrohardness may be related to the microstructure and the distribution of alloying elements. Thus, we performed microstructural observation and chemical analysis of the carbide or carbonitride in the N+2.0%C specimen tempered at 823 K for 3.6 ks after quenching from 1373 K (Fig. 10). Lattice spacings of 0.2084 nm in the $(\bar{2}00)$ direction and 0.1475 nm in the $(\bar{2}20)$ direction were measured from the diffraction pattern of M2CN, as shown in Fig. 10(b), which are slightly larger than the values for the as-cast specimen. Since the lattice spacing of VC is larger than that of VN, M2CN may transform to the MC-type carbide during quenching and tempering as shown in Fig. 10(c). The change in the N composition from 36.3% to 10% may have been because of the balance of the free energy between the M2CN and the austenite phase or between the M2CN and the liquid phase, but further study is necessary to conclude this. Finally, because the change in the composition of the primary M2CN or eutectic MC and M2C phases may affect the distribution of N and the matrix microstructure, TEM and EDS analyses were carried out on the tempered specimen. Figures 11 and 12 show the elemental mappings of the N+2.0%C specimen in the vicinity of M2CN and far from the M2CN carbonitride, respectively. The V Kα, C Kα, Al Kα, N Kα, Mo Lα, Cr Lα, and Fe Lα peaks were used for the Cs-corrected STEM-EDS mappings. The segregation of Mo and AlN precipitates with a size of nm to 100 nm order was observed at the interface of the primary M2CN phase as shown in Fig. 11. Furthermore, nanosize carbides containing V or Mo precipitated on the left side of the V and Mo mapping images. Secondary carbides containing V, W, Mo or Ta are utilized for secondary hardening at high tempering temperatures11–15). Although we could not clarify the composition of the V- or Mo-containing carbides under these experimental conditions, these phases may be general secondary carbides. Therefore, the increase in the Vickers hardness particularly depends on these precipitates. In contrast, an AlN phase precipitated in the N-containing specimens. Since Al was not used as a raw material in this experiment, it might have been included as an impurity in the ferrovanadium alloy or originated from a reaction with the Al2O3 crucible in the casting process. In addition to the large AlN precipitates under these experiment conditions, a few nanosize phases may also have precipitated in the matrix. Fine carbides containing V or Mo and the AlN phase were observed in the matrix microstructure in the STEM-SEM image as shown in Fig. 12. Furthermore, more fine V- or Mo-containing carbides were observed in in Fig. 12 than in Fig. 11. Since Mo diffused from the M2C carbide during tempering, as shown in Fig. 10(c), more Mo-containing carbides can be precipitated. From the above results, the increase in the macrohardness is predominantly caused by the incorporation of C in the matrix phase due to the substitution of N with C in M2CN or MC carbides. Furthermore, secondary carbides precipitated by the diffusion of V and Mo increase the hardness, similarly to in standard precipitation hardening, and fine AlN and other carbonitrides may partially contribute to the secondary hardening during the tempering process.
Effect of tempering temperature on Vickers hardness (a) and ratio of retained austenite (b) for specimens quenched at peak hardness.
Relationship between Vickers hardness and tempering temperature for specimens quenched from 1373 K.
Relationship between ratio of retained austenite and quenching temperature for specimens quenched from 1373 K.
(a) SEM image of carbonitride and carbides, (b) diffraction pattern of carbonitride, and (c) contents of alloying elements in each phase of tempered specimen with added nitrogen and 2.0% carbon.
STEM-EDS mapping at phase boundary of tempered specimen with added nitrogen and 2.0% carbon.
STEM-EDS mapping of matrix phase of tempered specimen with added nitrogen and 2.0% carbon.
To develop a work roll material with superior wear resistance for hot rolling, the effect of nitrogen addition on microstructure development and heat-treatment characteristics was investigated for high-carbon high-speed tool type alloys. The obtained results are as follows.
