2018 Volume 59 Issue 5 Pages 822-828
The carefully controlled forging of as-cast 30Cr2Ni4MoV steel was investigated as a mean to refine its coarse grain structure and crush the dendrites that cause segregation. This found that a fine-grained (∼50 um) structure could be achieved by deformation at 1000°C as a result of DRX. Grain boundary segregation of Mo caused a significant precipitation of MoC that could effectively pinned grain boundaries and suppressed the grain growth. In contrast, forging at temperatures above 1000°C led to a significant increase in grain size. Furthermore, the grain boundary segregation of sulfur resulted in the formation of low melting point sulfides near the grain boundaries that provided additional initiation points for intergranular cracking, thereby reducing the mechanical properties of the alloy. The results of this study therefore provide a basis for optimizing the forging process so as to improve the microstructure of heavy forgings, and sheds new light on the mechanism of grain refinement and crack initiation for 30Cr2Ni4MoV steel.
Although casting ingots are normally used as the initial blanks for heavy forgings, inherent defects in the form of their coarse grain structure, porosity, shrinkage and non-metallic inclusions become increasingly problematic as the weight of the ingot increases. Indeed, the microstructure of as-cast 30Cr2Ni4MoV steel, a material commonly used to fabricate low-pressure rotors in China, often manifests as excessively coarse grains, that fall well short of the quality standards (grain size ranging from 45 um to 65 um) required for such components.1) Given that the mechanical properties of heavy forgings are directly attributable to their microstructure, they can only be deemed suitable for use if the coarse grain structure of the steel can be suitably refined during forging.2–5) Consequently, enormous economic losses are suffered every year due to the need to scrap heavy forgings in which this cannot be achieved.
At present, most research into the microstructural evolution of 30Cr2Ni4MoV steel has been focused mainly on dynamic recrystallization (DRX), metadynamic recrystallization (MDRX) and static recrystallization (SRX) of small-sized grain structures during compression at constant temperature.6–9) From this, the kinetics of DRX, MDRX and SRX have been sufficiently established to allow the microstructural evolution during hot forging to be simulated.10) However, this overlooks the fact that the ingot is subjected to heating and cooling throughout the forging process, and that its as-cast microstructure can affect the microstructure of the final heavy forging, especially in the case of a coarse-grained cast structure. Furthermore, grain boundary segregation of alloying elements can occur during forging, which then leads to grain boundary precipitation; while some harmful elements, such as S, can even induce intergranular embrittlement.11–16) With this in mind, this paper explores the effect of temperature and reduction ratio on the microstructural evolution of coarse-grained as-cast 30Cr2Ni4MoV steel during FM (free from Mannesmann effect) forging (Fig. 1). The mechanism behind grain refinement and the generation of faults induced by grain boundary segregation is also investigated, with a view to optimizing the forging parameters so as to improve the mechanical properties of heavy forgings.
Sketch of FM forging method.
A 30Cr2Ni4MoV steel that is used in the commercial fabrication of rotors was selected for this study, with details of its composition given in Table 1. Before hot forging the grain structure in this steel was homogenized by heating to 1100°C and holding for 2 hours, as presented in Fig. 2. The alpha-gamma transformation temperature of 30Cr2Ni4MoV steel is about 750°C. It can be seen from Fig. 2 that although the microstructure in as-cast steel was austenitized and homogenized, rapid growth of the austenite grains occurred, leading to a coarse-grained structure with the average grain size of 130 um.
Homogenized grain structure by heating to 1100°C and holding for 2 hours.
An ingot is typically subjected to multiple upsetting and drawing deformations before it is formed into the final low-pressure rotor. After upsetting, the ingot is approximately shaped into a cylinder with a height-to-diameter ratio ranging from 1.2 to 1.5. Then the FM forging is used to draw the cylindrical shaped ingot. To ensure the consistency between experimental conditions and actual production, cylindrical specimens were therefore machined from this homogenized steel to a diameter of 30 mm and a height of 45 mm. These specimens were pre-treated by heating to 900, 1000, 1100 or 1200°C and holding for 15 min, and were then FM forged with a reduction ratio of 15% to ensure complete DRX.17)
All hot-deformed specimens were immediately quenched in water to preserve their austenite structure, during which a transformation from austenite to martensite occurred. After being sectioned longitudinally and mechanically polished, the specimens were etched in a C6H3N3O7–H2O solution. Metallographic observation and component analysis of the microstructure of each specimen was then performed using optical microscopy (OM), scanning electron microscopy (SEM) and energy-dispersive spectrometry (EDS).
To investigate the grain boundary segregation during forging, standard specimens were machined according to Fig. 3. These were then cooled by liquid nitrogen and held for 30 min. Immediately after the specimen was fractured, auger electron spectroscopy (AES) was used to observe the fracture surface and analyze micro areas of either the grain boundary or grain interior.
Standard specimen used for AES analysis.
Microstructures of the tested steel at the deformation temperatures of 900, 1000, 1100 and 1200°C and the reduction ratio of 15% are shown in Fig. 4. The 50 um average grain size obtained by forging at 1000°C is finer and more uniform than the 70, 110 and 150 um generated by forging at 900, 1100 and 1200°C, respectively. It can be seen from the micrographs that a large amount of serrated grain boundaries accompanied by several deformation twins have been obtained in the specimen treated at 900°C. Although the initial grains were deformed to elongate without any obvious appearance of dynamically recrystallized grains, their serrated and bulging grain boundaries could contribute to providing nucleation sites as long as DRX was initiated.18) When the deformation temperature increased to 1000°C, uniformly distributed fine grains could be observed, implying high degree of DRX was induced. A complete DRX was achieved at deformation temperatures of 1100 and 1200°C, and the growth of recrystallized grains was better than that at 1000°C, bring about a coarse grain structure in the forged specimen.
Microstructural evolution of 30Cr2Ni4MoV steel forged with a reduction ratio of 15%: (a) 900°C, (b) 1000°C, (c) 1100°C, (d) 1200°C.
Dendritic segregation in ingots is normally not permitted in heavy forgings due to concerns over embrittlement and crack initiation in areas of segregation. The morphology of dendrites in the specimens tested in this study was therefore observed in etched cross-sections, as shown in Fig. 5. It can be seen that the dendrites are more thoroughly crushed by forging at a higher temperature; and although a fine grain structure is formed at temperatures below 1000°C, the dendrites cannot be crushed completely unless the amount of deformation is continually increased. To ensure a final refined grain structure with the complete absence of dendrites in heavy forgings, it is suggested that FM forging should be conducted at a temperature above 1100°C and with a reduction ratio of 15%. The coarse austenite grain structure produced by this high temperature can then be refined by subsequent FM forging at 900°C using an appropriate reduction ratio to induce complete DRX.
Morphology of dendrites in different specimens.
During the grain growth that usually follows DRX, particles precipitating at grain boundaries can be formed that suppress the growth of dynamically recrystallized grains. This phenomenon is caused by the so-called Zener pinning effect. By equating the pinning force with the driving force for grain growth, Smith and Zener neatly derived an expression for the grain size of a pinned microstructure
| \begin{equation} D=Ad/f \end{equation} | (1) |
In the SEM image shown in Fig. 6(a), it is apparent that when the specimen was forged at 900°C some spherical particles ranging in size from 10 to 30 um are produced at the grain boundaries, which were identified as MoC through EDS analysis (Fig. 6(b)). The average d and f of these MoC particles measured on basis of the experimental data were 20 um and 0.076%, respectively. The expected grain size was then calculated as being 45 um, a size smaller than the actual grain size of 70 um. This can be explained by the assumption in the Smith-Zener analysis that each particle in contact with the grain boundary exerts maximum pinning force. Clearly though, the positions of the particles with respect to the grain boundaries vary continuously as the grain boundaries move. It follows that the pinning force exerted by each particle also depends on the position of the grain boundary. During SEM observation to the specimen treated at 900°C, it was found that some MoC particles were not at the grain boundaries (see Fig. 7), signifying they were less effective to retard the grain growth than those distributed exactly at the grain boundaries. However, the volume fraction of these particles was added into f, leading to a decreased grain size calculated by eq. (1).
Grain boundary precipitation of MoC: (a) SEM image of precipitates, (b) EDS analysis.
MoC distributed in the grain interiors.
When the deformation temperature increased above 1000°C, the MoC precipitates were absent, as shown in Fig. 8. Instead, only a few silicate and alumina inclusions generated by the metallurgical reaction were observed, as shown in Fig. 9 and Fig. 10. Since the absence of MoC particles precipitating at grain boundaries weakens the pinning force against the driving force for grain growth, a coarse-grained structure is produced.
Complete absence of precipitates at temperatures above 1000°C.
Silicate inclusion.
Alumina inclusion.
Past research has shown that grain boundary segregation of alloying elements such as B, Cr, Mn and Mo does certainly occur during the cooling of steels.20–22) Of these, S and P are considered particularly detrimental to the mechanical properties of processed steels, as they can weaken their stress rupture property point and induce temper brittleness. Meanwhile, Mo can actually improve the fracture resistance of steels by combining with C.23–25)
Figure 11 shows the results of AES analysis of the deformed specimens, with the Mo content of the grain boundary and grain interior regions shown in Fig. 12(a). It can be seen from this the Mo content of both the grain boundary and grain interior is much lower after forging at 900°C than at any other temperature, which is related to the aforementioned precipitation of MoC. Conversely, the absence of MoC at temperatures above 1000°C indicates that it is dissolved in the metal matrix, as evidenced by the increase in Mo content observed by AES. Figure 12(a) also shows that the Mo content of the grain boundary region is higher than the grain interior for all tested temperatures other than 900°C. This can be explained by the precipitation of MoC near grain boundaries at 900°C, with its subsequent dissolution at higher temperatures leading to Mo-depleted microzones in the grain interior.
AES analysis of the deformed specimens.
Element content of the grain boundary and grain interior: (a) Mo content, (b) S content.
As shown in Fig. 12(b), the S content was significantly higher at the grain boundaries than the grain interiors after forging at 1000°C, with both regions having a higher S content than those achieved at other temperatures. This suggests grain boundary segregation of S occurred. Grain boundaries are the most common defects in a solid containing sites for which solute atoms (e.g. S, P, etc.) generally have a lower free energy as compared to the sites in the bulk.26) The occupation of these sites at grain boundaries by solute atoms in order to minimize the total free energy of the material system is known as grain boundary segregation. Basic theoretical treatments, such as the Langmuir-McLean theory, can be used to discuss this phenomenon quantificationlly. Plenty of analytical studies focusing on S segregation have revealed the tendency of such atoms enriched towards the grain boundaries at a specific temperature. A detailed investigation of grain boundary segregation has been reported by Seah and Lea.27) Their data obtained from AES measurements were compared to earlier results from Seah and Hondros, who employed the same technique to study grain boundary segregation of S in the same system.28) The complex temperature dependence of enrichment factors was attributed to temperature dependent enthalpies and different entropies of segregation for grain boundary. Larere et al. calculated the S segregation free energy for grain boundary to be 18 kJ/mol and reported the competitive segregation of C and Ca with either repulsive (C) or attractive (Ca) interaction with S.29) The grain boundary segregation behavior of S was examined by Boutassouna et al.30) They pointed out that the main aspect caused by S segregation was the failure to achieve intergranular fracture of samples annealed at temperatures higher than 650°C, and for the sample annealed at 650°C, AES quantitative analysis gave an estimation of an average grain boundary composition that was comprised of about 15.08% of atomic S.
The Langmuir-McLean theory predicts that the magnitude of segregation decreases as temperature increases. However, the tendency for the present results seems to be opposite, i.e., the S segregation is higher at 1000°C than that at 900°C. This can be explained by the physical characteristics of the sulfides precipitating at 900°C, as shown in Fig. 13. Since these sulfides have a melting point of about 1000°C, which means when the 30Cr2Ni4MoV steel is forged at 900°C, low S content can be tested by AES due to their precipitation in the form of sulfides. However, as the temperature increases to 1000°C, these sulfides are melted and then dissolved in the metal matrix, leading to an increase in S content. It is noteworthy that forging at temperatures higher than 1000°C can usually cause sulfides to melt and further weaken the bond strength of the grain boundary. The fracture strength of the grain boundary is therefore also greatly reduced, with Fig. 14 showing the initiation of a crack by sulfides, and its subsequent propagation along the grain boundary to a void, where it then combines to form a much bigger fault.
Sulfides precipitating at 900°C.
Propagation of cracks initiated by sulfides along the grain boundary to a void: (a) OM image of cracks, (b) EDS analysis.
In order to ensure the quality of 30Cr2Ni4MoV steel heavy forgings, it is suggested that a temperature above 1100°C should be used to ensure the complete absence of dendrites and the elimination of any voids. More attention should also be given to limiting deformation at this temperature to prevent intergranular cracking caused by grain boundary segregation of S. Subsequent forging at 1000°C with a reduction ratio of 15% can then be used to achieve a homogenous distribution of fine grains in the final structure.
Experiments during the plastic deformation of 30Cr2Ni4MoV forged steel were conducted to reveal the effects of grain boundary segregation and precipitation on the microstructural evolution. From this study, the following conclusions are drawn:
