According to the Gibbs free energy diagram of martensite (α) and austenite (γ) phases at a fixed partitioning temperature , the Gibbs energy gap between two phases (∆Gα→γ) is the driving force for carbon partitioning from martensite into austenite during Q&P treatment, as shown in Fig. 10(a). In contrast to that in the low carbon and medium carbon steels, the martensite in high carbon steel has a higher carbon content, thereby shifting the carbon content rightward and leading to a larger ∆Gα→γ in the diagram. Consequently, the driving force for carbon partitioning from martensite into austenite increases with the increase in carbon content of steel. The paraequilibrium carbon content of two phases shown in Fig. 10(a) and the schematic diagram of carbon distribution at α/γ interface shown in Fig. 10(b) indicate the existence of a carbon content gradient at the interface, illustrating the carbon partitioning direction from martensite into austenite. Figs. 10(c)–10(f) show the microstructures and corresponding line scanning of carbon on the retained austenite of the as-quenched and PT-250 samples, respectively. The carbon distribution of the PT-250 sample has a parabolic-shaped gradient in blocky austenite with the carbon content increasing progressively with line scanning from the center to the edge (Fig. 10(f)). This finding is in sharp contrast to that of austenite in the quenched state, in which the carbon content in austenite grains remains nearly constant with only slight fluctuations (Fig. 10(d)). With further partitioning at 250°C for 600 s, the carbon content of supersaturated martensite decreases from (2.3 ± 0.1)wt% to (1.5 ± 0.1)wt% and the carbon content of RA increases from 0.65wt% to 1.1wt% simultaneously; thereby corroborating the intrinsic mechanism of carbon diffusion and kinetics at the martensite–austenite interface during partitioning [24–25].
Figure 10. (a) Gibbs free energy diagram showing the effect of carbon content on ∆Gα→γ; (b) carbon distribution diagram illustrating the carbon partitioning mechanism (∆Gα→γ—Gibbs energy gap between α and γ phases; Cα and Cγ—Carbon contents of α and γ phases away from α/γ interface; Cα/γ and Cγ/α—Paraequilibrium carbon contents of α and γ phases at α/γ interface); microstructures and corresponding line scanning of carbon on the retained austenite phases of the (c, d) as-quenched and (e, f) PT-250 samples, respectively.
On the basis of the carbon partitioning mechanism, the evolution of RA is analyzed as a function of partitioning time. In the initial partitioning stage, because of the large carbon content gradient at the martensite–austenite interface, carbon is transported rapidly from the supersaturated martensite to adjacent unstable austenite, increasing the carbon content of RA and thus the stability of RA remarkably. As the partitioning process continues, the carbon content of martensite diminishes gradually, and the driving force for carbon partitioning decreases progressively. When the carbon that flows into RA through the partitioning process is less than that exhausted by the competing reactions, such as carbide precipitation, the carbon content of RA decreases inevitably. As the carbon content of RA decreases below a critical point, the decomposition of RA occurs, leading to the decrease in the fraction of RA.
To summarize the results obtained and the arguments previously reported, the formation of RA during Q&P treatment is closely associated with carbon diffusion and the amount and stability of RA depend on the degree of how much carbon is involved in the diffusion from supersaturated martensite into RA. Because of the stronger driving force for carbon partitioning in high carbon steel, the maximum fraction of RA is higher than that obtained in low carbon and medium carbon steels [13,26]. With the increase in partitioning temperature from 250 to 400°C, the carbide precipitation effect that exhausts the carbon source becomes more pronounced, and the residual amount of carbon available to participate in the formation of RA decreases, thereby leading to the decrease in the fraction and size of RA.
Another phenomenon is worth noting. The decomposition of quenched untransformed austenite into martensite occurs in the initial partitioning stage, which is not observed in low carbon steel . Because the quenching process occurs from the fully austenitic temperature region, the microstructure is in a high energy state and relatively metastable. The carbon content in quenched austenite is determined to be approximately 0.65wt%, as shown in Fig. 10(d). By substituting the composition of quenched austenite into the following equation , Ms (°C) is estimated to be 229°C.
where WC, WMn, WSi, and WAl are the contents (wt%) of C, Mn, Si, and Al, respectively.
Upon heating to a partitioning temperature above Ms, the decomposition of quenched austenite is expected to occur. In addition to the carbon content, another factor should be taken into consideration in the decomposition of quenched austenite. Previous studies reported that the stability of austenite was affected by the constraining effect from phases surrounding austenite, i.e., an increase in stress concentration surrounding metastable austenite decreases its stability [27–29]. The level of stress concentration at the interface between quenched martensite and austenite is assumed to be high in high carbon steel because the volume dilatation effect resulting from austenite- martensitic transformation during quenching is prominent in high carbon steel due to its high carbon content. Upon heating to the partitioning temperature, the stress concentration tends to be relieved instantaneously through the decomposition of RA into martensite.
In the microstructural design of structural steel, the characteristics of RA (i.e., morphology, stability, and volume fraction) are of significance in determining the toughness of high carbon steel. However, determining which characteristic, i.e., volume fraction or stability, has a dominant role in affecting the toughness of high carbon steel is difficult. In this work, different characteristics of RA, obtained through different processing parameters, are analyzed to determine which characteristic of RA affects the toughness of high carbon steel. The results shown in Figs. 2 and 7 illustrate that the partitioning time at which the maximum toughness is achieved is identical to the time at which the peak volume fraction of RA is obtained for both partitioning temperatures, indicating the important role of RA fraction in determining the toughness of high carbon steel.
Another interesting feature is noted by comparing the toughness and volume fraction of RA between PT-250 and PT-400 samples. Despite the higher fraction of RA transformed during the impact toughness test, a lower toughness is observed for the PT-250 sample than that of the PT-400 sample, which seems to show a paradox based on the positive correlation between the fraction of RA and toughness (Table 1). Previous studies reported that, apart from the fraction of RA, the toughness of high carbon steel was closely correlated with the mechanical stability of RA, which was affected by the carbon content of austenite [30–31], the size of austenite grains , the morphology , and the constraining effect [27–28]. Generally, the film-like austenite is more preferable than blocky austenite in microstructural design of high carbon steel because of its higher stability . When subjected to the impact toughness test, the large blocky austenite grains in PT-250 sample trigger strain-induced martensitic transformation at a relatively small strain level because of its low stability and transform into coarse plate martensite [26,33–34], which is prone to stress concentration at the interface and promotes crack propagation during deformation . In contrast, finer blocky austenite grains accompanied with a larger amount of film-like austenite in PT-400 sample help in achieving a higher toughness by accommodating a larger strain. Moreover, the surrounding tempered martensites with a lower hardness in PT-400 sample can accommodate the strain more effectively than the harder martensite in PT-250 sample and lead to a lower stress concentration at the austenite–martensite interface, which, in turn, delays the strain-induced martensitic transformation during impact toughness test. Generalizing from the previous analysis, compared with the volume fraction of RA, the stability of RA in the microstructure plays a more significant role in determining impact toughness of high carbon steel.
Sample Heat treatment After the impact toughness test After the abrasion test PT-250 32 ± 1.5 16 ± 1 0 PT-400 23 ± 1 12 ± 1 0
Table 1. Changes in the volume fractions of retained austenite under different conditions
Further analysis of Fig. 7(a) revealed a second increase in impact toughness at longer partitioning times (i.e., 1800–5400 s) for both temperatures. Given the small amount of RA at this stage (Fig. 2), the contribution of TRIP effect of RA to impact toughness of high carbon steel is less significant, and the residual tempered martensites in the microstructure begin to play an important role. It was reported that the tempered martensite with a low hardness was ductile, whereas tempered martensite with a high hardness was brittle . This finding is supported by the fact that the sample partitioned at 400°C for 5400 s with a hardness of HV 530 has a higher impact toughness than the sample partitioned at 250°C with a hardness of HV 590. Given the progressive decrease in hardness at this stage, the ductility of tempered martensites is expected to increase gradually, leading to the second increase in impact toughness.
The fracture feature was characterized to some degree by its parent microstructure, in which the transformation of austenite grains to martensite occurred during impact toughness tests . The microstructure of PT-250 sample, which is dominated by large blocky austenite grains, is expected to exhibit a low resistance to crack propagation and leads to martensitic transformation at low strains, thus leading to the resulting quasi-cleavage fracture. Meanwhile, the small austenite grains show a high capability to accommodate strains and contribute to a high degree of ductile fracture features , in which the fine dimples may be associated with the coalescence of voids resulting from secondary carbides in tempered martensitic matrix [14,38].
The wear resistance of a material is governed by hardness, and a higher hardness usually corresponds to a better wear resistance. This rule holds in this work, in which the PT-250 sample with the highest hardness of HV 610 exhibits the best wear resistance among three samples, as evidenced by the role of hardness in enhancing wear resistance. Moreover, the contribution of RA to the wear resistance is identified by comparing the wear resistance and hardness of the QT-250 and PT-400 samples. Despite the similar hardness, the PT-400 samples exhibits a higher wear resistance than QT-250 sample, suggesting the role of microstructural characteristic in affecting wear performance. It was reported that abrasion-induced martensitic transformation of RA contributed significantly to wear resistance during abrasion [9,39–40]. The comparison of the microstructural difference between QT-250 and PT-400 samples shows that the presence of a considerable amount of RA in PT-400 sample is responsible for its higher wear resistance.
An increase in hardness could enhance the resistance of steel to penetration by abrasives, whereas an increase in toughness could hinder the nucleation and propagation of microcracks during wear, thus increasing the wear resistance of materials [39,41–42]. Both hardness and toughness contribute to the wear resistance of high carbon steel; however, it is unclear which factor has a dominant role. Despite its lower toughness, the PT-250 sample with a higher hardness showed a higher wear resistance than PT-400 sample, indicating the more significant role of hardness in determining wear resistance of high carbon steel. In conclusion, under the low-impact abrasive conditions in this work, hardness is the primary factor affecting the wear resistance of high carbon steel, followed by toughness.
Because of the beneficial role of RA in affecting the mechanical and wear properties of high carbon steel, these properties can be adjusted as needed to best fit the application conditions by purposely modifying the characteristics of RA in multiphase microstructure by controlling the processing parameters.
Effect of quenching-partitioning treatment on the microstructure, mechanical and abrasive properties of high carbon steel
8 June 2020
Revised: 3 August 2020
Accepted: 5 August 2020
Available online: 10 August 2020
Abstract: The present work employed the X-ray diffraction, scanning electron microscopy, electron backscattered diffraction, and electron probe microanalysis techniques to identify the microstructural evolution and mechanical and abrasive behavior of high carbon steel during quenching-partitioning treatment with an aim to enhance the toughness and wear resistance of high carbon steel. Results showed that, with the increase in partitioning temperature from 250 to 400°C, the amount of retained austenite (RA) decreased resulting from the carbide precipitation effect after longer partitioning times. Moreover, the stability of RA generally increased because of the enhanced degree of carbon enrichment in RA. Given the factors affecting the toughness of high carbon steel, the stability of RA associated with size, carbon content, and morphology plays a significant role in determining the toughness of high carbon steel. The analysis of the wear resistance of samples with different mechanical properties shows that hardness is the primary factor affecting the wear resistance of high carbon steel, and the toughness is the secondary one.