Elsevier

Acta Materialia

Volume 91, 1 June 2015, Pages 10-18
Acta Materialia

Direct measurement of carbon enrichment in the incomplete bainite transformation in Mo added low carbon steels

https://doi.org/10.1016/j.actamat.2015.03.021Get rights and content

Abstract

The overall kinetics and carbon enrichment in austenite during the incomplete isothermal bainite transformation in Fe–0.1C–1.5Mn–(0, 0.03, 0.3, 0.5, 1)Mo (mass%) alloys were investigated with quantitative metallography and Electron Probe Microanalysis in the transformation temperature range of 773–873 K. The incomplete transformation appears at 823–873 K when Mo addition exceeds 0.3 mass%; at 773 K substantial carbide precipitation accompanies bainite transformation and no transformation stasis is observed. Transformed fractions in the stasis stage are hardly affected by prior austenite grain size. Carbon concentrations in austenite in the stasis stage are lower than T0 line and decrease with the increase of Mo addition and temperature. T0′ limit, solute drag and WBs limit theories are used to examine the experimentally measured carbon concentration limits in austenite, and their respective flaws are pointed out.

Introduction

Low carbon bainitic steels have found a wide application in many industries, e.g. shipbuilding, automobile, architecture, etc., due to their good combination of high strength and high toughness. In order to achieve enough hardenability to obtain bainite microstructures during continuous cooling in industrial production, several alloying elements like B, Mo and Nb are usually added into steels in appropriate amounts. However, such microalloying also increases the fraction of untransformed austenite, most of which finally transforms into martensite/austenite constituent (M/A) when cooled to room temperature, leading to a decrease in toughness [1]. Similar effects are also responsible for degradation of toughness in heat affected zones (HAZ) after welding, causing failure at welding joints of structural steels [2]. Since the amount of M/A is mainly determined by the degree to which bainite transformation proceeds, this might relate to a phenomenon called “incomplete transformation” (ICT) [3], in which bainite formation stops prematurely before the equilibrium amount is attained. However, despite intensive studies [3], [4], [5], [6], [7], [8], [9] over decades, the nature of ICT remains in dispute, with three main hypotheses being the T0′ limit, solute drag and WBs limit theories.

When Zener [10] first clearly defined the diffusionless mechanism of bainite, he believed that unlike martensite, there is no strain energy attending the growth of bainite. Thus Zener predicted that the upper temperature limit of bainite formation, Bs, should coincide with T0 temperature, where austenite and ferrite of the same composition have the same Gibbs free energy. The ICT phenomenon is also readily accounted for in a way that during the growth of bainite, untransformed austenite continuously absorbs carbon rejected from supersaturated bainitic ferrite and its actual T0 temperature is decreasing, thus eventually a certain point would be reached, beyond which diffusionless formation of bainite is thermodynamically impossible. Bhadeshia and Edmonds [11] inherited the diffusionless view of bainite formation and demonstrated Zener’s theory in an Fe–C–Mn–Si alloy, showing that the carbon concentration limits in austenite measured by X-ray diffraction were in good agreement with T0 line, although some data exceeded T0 by 0.2–0.4 mass% below 623 K. In a subsequent article, after carefully examining experimental data in an Fe–C–Ni–Si alloy, Bhadeshia [12] estimated that there is a strain energy of ∼400 J/mol associated with diffusionless growth of bainite, thus proposed a new limit T0′ that is in essence similar to T0 but takes into account the strain energy. Recently Caballero et al. [13] studied the ICT phenomenon in a series of sophisticated alloy steels. The measured carbon concentrations in austenite in the stasis stage were again proved to be close to T0 or T0′ limits thus were taken as new experimental evidence that ICT is a manifestation of diffusionless formation of bainite.

In contrast, Aaronson proposes that bainite forms in a reconstructive way, as allotriomorphic ferrite or pearlite does [5], and ICT results from the solute drag effect (SDE) [14] which shifts the upper limit for unpartitioned growth of bainite to much lower than Para-Ae3 [15] line. In this mechanism, substitutional solute atoms that have slow diffusion rates would segregate at migrating interfaces due to negative interaction energy with these interfaces, exerting a drag force on them thus causing free energy dissipation, a concept originally developed for grain boundary motion [16]. When this drag force is powerful enough, the migration of interfaces could be entirely stopped, leading to the ICT phenomenon. Subsequently, after realizing that the solute diffusion is too sluggish at temperatures of interest to exert a substantial drag force, Bradley and Aaronson [17] modified the original SDE and put forward the solute drag-like effect (SDLE), which states that substitutional alloying elements that have a strong attractive interaction with carbon, when absorbed at migrating interfaces, may decrease the activity of carbon in the austenite in immediate contact with the interfaces, which in turn decrease the activity gradient of carbon in front of the interface and cause sluggish kinetics. This SDLE theory was adopted by Reynolds et al. [7] and Goldenstein and Aaronson [18] to explain the ICT phenomena observed in several high-purity ternary alloys, e.g. Fe–C–Mo, Fe–C–Mn and Fe–C–Cr. Later, Aaronson et al. [19] integrated SDE and SDLE and developed the term coupled solute drag effect (CSDE) to describe these drag forces. When the chemical potential of carbon in austenite at mobile areas of interfaces is decreased to the same level as that in regions remote from these interfaces, growth of bainite would stop. Although Aaronson’s opinion experienced the gradual evolution from SDE to CSDE, hereafter this theory is still referred to as the solute drag theory for conciseness.

Unlike T0′ limit and solute drag theories, WBs limit theory proposed by Hillert [9] does not start from assuming the mechanism of bainite transformation (i.e. diffusionless or reconstructive) but takes a completely different approach. When applying Zener–Hillert equation [20] to examine experimental lengthening rates of acicular ferrite in Fe–C binary alloys, Hillert found that the extrapolation of measured growth rates to zero velocity did not yield the equilibrium carbon concentration. He directly related this effect to what was later described as the ICT phenomenon, proposing that there is a thermodynamic barrier to the growth of acicular ferrite. The experimentally determined value for this barrier indicated no sharp change between Widmanstatten ferrite at higher temperatures and bainitic ferrite at lower. Therefore, Hillert held that there is no essential difference between Widmanstatten ferrite and bainite in terms of growth mechanism, and suggested using a common WBs to refer to the limits for carbon diffusion controlled growth of acicular ferrite [9]. Hillert tested the validity of WBs by examining published data in some alloy steels showing a transformation stasis and found that these data could be fitted equally well with the WBs and T0′ lines. In most alloys investigated, the shift of WBs by addition of substitutional elements can be well explained purely by thermodynamic effect; in Mo or Cr added alloys, however, an additional amount of energy needs to be added to the growth barrier obtained in Fe–C systems, which is suggested to originate from solute drag effect.

As can be seen from the brief review above, considerable controversies still remain upon the mechanism of ICT and further study is required. Recently, the ICT phenomena in Nb, Mo or B + Mo added low carbon steels have been preliminarily investigated by Furuhara et al. [21], [22]. However, effects of alloy contents on ICT have yet to be revealed in detail. In the present study, different amounts of Mo, which is a potent hardenability enhancing element, were added into an Fe–0.1C–1.5Mn (mass%) alloy and the incomplete bainite transformation was more systematically investigated. In particular, direct measurement of carbon concentration in austenite was conducted by Electron Probe Microanalysis (EPMA) to clarify the role of carbon enrichment in the ICT phenomenon, based on which the three theories mentioned above were examined. In addition, effects of prior austenite grain size on ICT were also briefly studied.

Section snippets

Experimental procedure

A series of Fe–0.1C–1.5Mn–(0, 0.03, 0.3, 0.5, 1)Mo (mass%) alloys were used in this study. Hereafter they are referred to as 0Mo, 0.03Mo, 0.3Mo, 0.5Mo and 1Mo alloys, respectively. The alloy ingots were prepared by vacuum melting and casting, followed by hot rolling to ∼15 mm. 1Mo alloy was homogenized at 1473 K in Ar atmosphere for 172.8 ks. Samples of 0Mo–0.5Mo alloys were encapsulated into Ar-filled silica tubes, first homogenized at 1423 K for 86.4 ks, cold rolled to a final thickness of ∼5 mm,

Results

Fig. 1(a) shows the variations in volume fractions of transformed regions with time in 0Mo and 0.5Mo alloys isothermally held at 848 K. It is seen that transformation proceeds relatively quickly and soon approaches completion in 0Mo alloy. In contrast, the transformation kinetics curve of 0.5Mo alloy exhibits a distinct transformation stasis stage, during which the volume fraction of transformed regions almost maintains constant, i.e. ∼48 vol.%. After prolonged holding, reaction resumes in 0.5Mo

Discussion

It was seen in Fig. 2(g)–(i) that substantial carbide precipitation accompanies bainitic ferrite formation and no transformation stasis is observed at 773 K in all the five alloys. Since carbide precipitation is unfavorable to carbon enrichment, absence of transformation stasis at 773 K implies that carbon enrichment might play an important role in the premature cessation of bainite transformation. In addition, Fig. 6(a) demonstrated that prior austenite grain size hardly influence transformed

Conclusions

The present study investigated the incomplete isothermal bainite transformation at 773–873 K in a series of Mo added Fe–0.1C–1.5Mn alloys. Emphases were placed on clarifying the effects of Mo content and temperature on bainite transformation kinetics and carbon enrichment in austenite. The following results were obtained:

  • 1.

    Addition of over 0.3 mass% Mo could induce a transformation stasis at 823–873 K, while at 773 K substantial carbide precipitation accompanies bainite formation and no stasis is

Acknowledgements

A part of this research was supported by “Creation of Innovative Functions of Intelligent Materials on the Basis of Element Strategy” by CREST Basic Research Program. T.F. gratefully acknowledges the financial support provided by the Ministry of Education, Culture, Sports, Science, and Technology through a Grant-in-Aid for Scientific Research (B) No. 23360316 (2011–2013). X.Y. gratefully acknowledges the financial support from the Joint Education Program between Tohoku University and Tsinghua

References (33)

  • H. Bhadeshia et al.

    Acta Metall.

    (1980)
  • H. Bhadeshia

    Acta Metall.

    (1981)
  • F.G. Caballero et al.

    Acta Mater.

    (2009)
  • M. Hillert et al.

    Acta Metall.

    (1976)
  • Z.Q. Liu et al.

    Acta Mater.

    (2013)
  • S.B. Singh et al.

    Mater. Sci. Eng. A

    (1998)
  • M. Enomoto

    Acta Mater.

    (1999)
  • H. Chen et al.

    Acta Mater.

    (2013)
  • V. Biss et al.

    Metall. Mater. Trans. B

    (1971)
  • H. Ikawa et al.

    Trans. Jpn. Weld. Soc.

    (1980)
  • H.I. Aaronson et al.

    Metall. Mater. Trans. A

    (2006)
  • R.F. Hehemann et al.

    Metall. Trans.

    (1972)
  • H.I. Aaronson et al.

    Metall. Trans. A

    (1990)
  • H. Bhadeshia et al.

    Metall. Trans. A

    (1990)
  • W.T. Reynolds et al.

    Metall. Trans. A

    (1990)
  • W.T. Reynolds et al.

    Metall. Trans. A

    (1990)
  • Cited by (0)

    View full text