Effect of Molybdenum Content on the Hardenability and Precipitation Behaviors of Boron Steel Austenitized at High Temperatures

Abstract

The effect of molybdenum (Mo) contents on hardenability and precipitation behaviors in Mo–B simultaneously added steels were investigated placing a focus on high austenitizing temperature. The hardenability of 0.5% Mo - 11 ppm B steel austenitized at 1150°C was decreased compared with that austenitized at 950°C, whereas 1.0% Mo - 10 ppm B and 1.5% Mo - 9 ppm B steels were less affected by high austenitizing temperature than 0.5% Mo - 11 ppm B steel. The Fe₂₃(C, B)₆ precipitation by increasing austenitizing temperature was also revealed to be suppressed in 1.5% Mo – 9 ppm B steel. These results indicate that the improved effect of the Mo addition on hardenability by retarding the precipitation of Fe₂₃(C, B)₆ still appear in B-added steels austenitized at high temperature. Furthermore, Fe₂₃(C, B)₆ precipitation start temperature was increased in Mo–B added steels austenitized at 1150°C. This result implies that non-equilibrium B segregation mechanism during cooling from high austenitizing temperature enhances the amount of segregated B on grain boundaries leading to the promotion of the borides precipitation at high austenitizing temperature region. However, Mo is presumed to fix a part of thermal vacancies as Mo–V complex resulting in the suppression of non-equilibrium B segregation to grain boundaries during cooling, which is speculated to inhibit the Fe₂₃(C, B)₆ precipitation. Thus, the effect of Mo–B combined addition on hardenability was presumably maintained even in high austenitizing temperature region.

1. Introduction

Boron (B) is commonly used to manufacture high-strength steel because the addition of small amounts of B (about 10 mass ppm; ppm hereafter) markedly increases steel hardenability.^1,2) The increase in hardenability is attributed to B in solid-solution segregating at pre-transformation austenite grain boundaries.^3,4,5) Moreover, the hardenability-increasing effect of B decreased when borides, such as Fe₂₃(C, B)₆, precipitate before transformation. Therefore, Mo can be added to steel to promote the hardenability-increasing effect of B at low cooling rates during quenching, especially in the center of thick steel sheets. Mo suppresses Fe₂₃(C, B)₆ precipitation during cooling, and this method is known as combined Mo–B addition.^2,6,7)

Conversely, at high austenitizing temperatures, the effect of Mo tends to decrease. For example, Asahi⁶⁾ reported the effect of austenitizing temperature on the hardenability of steel containing 0.5 mass% Mo (mass% is hereinafter abbreviated %) and 13 ppm B. Furthermore, although combined Mo–B addition increased the hardenability of steel austenitized at temperatures of up to 1000°C, the effects of B addition and combined Mo–B addition on hardenability were not apparent at austenitizing temperature range of 1100–1250°C. In addition, at an austenitizing temperature of 1250°C, Fe₂₃(C, B)₆ precipitates were observed although at an austenitizing temperature of 950°C, no precipitates were present after cooling.

The accelerated precipitation of Fe₂₃(C, B)₆ with increasing austenitizing temperature was attributed to the amount of B segregated at grain boundaries increasing above the solid-solution limit during cooling. Furthermore, B combined with excess vacancies, which was explained using a non-equilibrium segregation model termed “rapid diffusion”.^{8,9,10,11,12,13)} Quantification of the amount of B segregated at grain boundaries has been challenging; therefore, the applicability of a non-equilibrium segregation model is still uncertain. However, Miyamoto et al.¹⁴⁾ have recently measured the amount of B segregated at grain boundaries using a three-dimensional atom probe (3DAP) and demonstrated that, at quenching temperatures in the range of 1050–1150°C, the amount of B segregated at grain boundaries was higher than the calculated equilibrium segregation amount with limited diffusion. These findings confirmed that the amount of segregated B increased during cooling from high austenitizing temperatures.

Consequently, it can be postulated that Mo may affect the promotion of Fe₂₃(C, B)₆ precipitation in B-added steels austenitized at high temperatures, which is attributed to the non-equilibrium segregation of B. However, the effect of Mo content on the hardenability and precipitation behavior of B-added steels austenitized at high temperatures has not yet been elucidated. Asahi⁶⁾ investigated the effect of austenitizing temperature on the hardenability of combined Mo–B added steels with Mo contents ≤ 0.5%. However, the precipitation-promoting effect of Fe₂₃(C, B)₆ by high austenitization temperature has not yet been demonstrated at Mo contents of ≥ 0.5%. In addition, we analyzed the hardenability and precipitation behavior of combined Mo–B added steels with Mo contents in the range of 0–1.5% Mo austenitized at 950°C. Our results indicate that, for the steel samples with high Mo (≥ 0.75%) and B (12 ppm) contents, Mo₂FeB₂ precipitated, thereby demonstrating that the effect of combined Mo–B addition was saturated.¹⁵⁾ However, the changes in hardenability and the dissolution of Mo₂FeB₂ with austenitizing temperature have not yet been elucidated, and it is unclear whether Fe₂₃(C, B)₆ precipitates during cooling after the dissolution of Mo₂FeB₂.

In this study, we evaluated the effect of Mo content on the hardenability of combined Mo–B added steels austenitized at high temperatures, and elucidated the mechanisms of the effect behind. Furthermore, the relationships between the hardenability and precipitation behaviors of steel samples with various Mo and B contents were investigated.

2. Methods

2.1. Steel Samples

The tested steel samples were the same as those used in previous studies on hardenability, precipitation behavior, and 3DAP analysis.^15,16,17,18) Test samples were prepared by vacuum melting, and the chemical compositions of all steel samples are shown in Table 1. The basic composition of the tested steel was Fe–0.15C–1.3Mn (%), and the B and Mo contents of the steel samples ranged between 0 ppm and 25 ppm and 0% and 0.5%, respectively. The steel containing 0.5% Mo and 0 ppm B is denoted as M5B0 steel, and the steel containing 0.5% Mo and 5 ppm B is denoted as M5B5. Ti (0.02%) was added to each steel sample to suppress boron nitride precipitation, such that almost the entire amount of N in steel precipitated as titanium nitride. After molding them into 50 kg ingots followed by reheating at 1250°C for 3600 s, the steel samples were hot-rolled into plates with thicknesses of up to 35 mm at a finishing temperature of 950°C. Thereafter, the samples were air-cooled and used to evaluate their hardenability and precipitation behavior.

Table 1. Chemical compositions of test steels.

(mass%,*ppm)
No.	C	Si	Mn	P*	S*	Ti	Al	Mo	B*	N*	O*
(0, 10, 20)B	0.15	0.27	1.31	<20	19	0.020	0.020	<0.01	<3, 10, 20	7	<10
M5(0, 5, 9, 11, 14, 16, 20, 25)B	0.14	0.27	1.29	<20	20	0.020	0.017	0.50	<3, 5, 9, 11, 14, 16, 20, 22	8	<10
M7(0, 10, 20)B	0.14	0.27	1.28	<20	20	0.020	0.017	0.76	<3, 10, 20	8	<10
M10(0, 5, 10, 12, 14, 17, 25)B	0.14	0.28	1.27	<20	20	0.020	0.017	1.01	<3, 5, 10, 12, 14, 17, 25	13	<10
M15(0, 4, 9, 11, 14, 16, 20, 23)B	0.14	0.28	1.26	<20	20	0.020	0.017	1.52	<3, 4, 9, 11, 14, 16, 20, 23	18	<10

2.2. Hardenability and Transformation Start Temperature

Hardenability was evaluated using the Jominy test. The test samples (25 mm in diameter and 100 mm in length) were heated to 1150°C in a furnace under an Ar atmosphere until one end became hardened. Subsequently, a 1 mm layer was ground off the lateral side of each test sample, and the changes in hardness of the hardened end were evaluated using the Rockwell method. As previously reported,¹⁵⁾ the critical cooling rate (V_c-90 (°C/s)) at which the samples contained 90% martensite, was considered to be the hardenability index, as proposed by Ueno et al.²⁾ Figure 1 shows the effect of Mo content on V_c-90 for a B-free steel sample and B-added steel samples containing 10 and 20 ppm B, which were maintained at 950 and 1150°C for 2700 s. Figure 2 shows the effects of B content on V_c-90 for steel samples with various Mo contents. To minimize the effect of deboronization in the steels with small amount of B addtion,¹⁹⁾ the steel samples were austenitized at 950 and 1150°C for 600 s. Under these conditions, the deboronization effect on hardenability was negligible.

Fig. 1. Effect of Mo content on hardenability (Vc_-90) of B-free steels and B-added steels, which are austenitized at (a) 950°C and (b) 1150°C. Dashed line was calculated by using Ueno’s prediction formula of hardenability.²⁾ (Online version in color.)

Fig. 2. Effect of B content on hardenability (V_c-90) influenced by Mo content, which are austenitized at (a) 950°C and (b) 1150°C. (Online version in color.)

The transformation temperature was measured using test samples with diameters and lengths of 3 and 10 mm, respectively. A Formaster test device (Fuji Electronic Industrial Co., Ltd.) was used to austenitize the test samples at 1150°C for 20 s, followed by cooling below 50°C at various cooling rates of 0.5–30°C/s. The changes in linear expansion in the longitudinal direction with temperature were measured. Moreover, the temperature of a 10% fraction of transformed was calculated from the change in transformation expansion, and it was defined as the transformation temperature.

2.3. Precipitation State of B

Rolled steel test samples (3 mm in diameter and 10 mm in length) were austenitized at 1150°C for 20 s using a Formaster tester (Fuji Electronic Industrial Co.), followed by cooling at cooling rates of 1–5°C/s. Once the temperature reached 650°C, each steel sample was quenched using He gas, and the as-obtained samples were used to evaluate the precipitation behavior of B. If bainite transformation occurs at a temperature higher than the He gas quenching temperature, it can affect the segregation and precipitation behavior of B at the austenite grain boundaries, thereby affecting hardenability. Moreover, the bainite start temperature Bs of M5Bx, M10Bx, and M15Bx steels were calculated to be 631, 590, and 548°C, respectively, using the empirical formula proposed by Steven et al.²⁰⁾ Therefore, the Bs points of all steel samples were designed to be lower than 650°C. The cooling rate during He gas quenching was ≥ 100°C/s. It was considered that this type of heat treatment can quench the segregated and precipitated B at austenite grain boundaries, thereby affecting hardenability. To investigate the B distribution after heat treatment, specimens with surface area of 200 × 200 μm, which were cut from the central parts of the test samples, were analyzed using time of flight–secondary ion mass spectrometry (ToF–SIMS). As previously reported,¹⁷⁾ ToF–SIMS was performed using a TOF.SIMS5 instrument (ION-TOFGmbH). Bi⁺ was the primary ion and BO₂⁻ was the secondary ion (mass-to-charge ratio (m/z) = 43).^21,22,23) In addition, the heat-treated samples were subjected to selective potentiostatic etching via electrolytic dissolution to prepare specimens for extraction replica observation. Subsequently, a 200 kV field-emission transmission electron microscopy (TEM; JEOL, Ltd.) instrument was used to observe and identify B precipitates.

To determine the Fe₂₃(C, B)₆ precipitation start temperature, we used M7B20 steel sample (8 mm in diameter and 12 mm in length) that was subjected to the solution heat-treatment at 1200°C for 600 s. Following the complete dissolution of Fe₂₃(C, B)₆, the samples were subjected to austenitization at 1150°C for 20 s, followed by He gas quenching to 850 or 650°C (cooling rate of 10°C/s). The changes in precipitation behavior of the austenitized and quenched steel samples with quenching start temperature (T_Q) were investigated using TEM.

3. Results

3.1. Effect of Mo Content on the Hardenability of B-added Steels

3.1.1. Jominy Test

Figure 1 shows the effect of Mo content on the V_c-90 values of B-free steel and steels containing 10 and 20 ppm B (MxB0, MxB10, and MxB20 steels, respectively, where x denotes the Mo content) for test samples austenitized at 950°C (hereafter 950°C materials) and 1150°C (hereafter 1150°C materials). In the previous reports, the B contents of 950°C and 1150°C materials were 0 and 20 ppm, respectively,^15,18) and in the present paper the results using these specimens are cited for the sake of comparison. The effect of Mo content on the V_c-90 values of 950°C materials is shown in Fig. 1(a). For MxB10 steel, V_c-90 gradually decreased with increasing Mo content at Mo contents ≤ 1.5%. In contrast, for MxB20 steel V_c-90 remained constant at Mo contents ≥ 0.75%. According to a previous report,¹⁵⁾ for MxB20 steels with Mo contents ≥ 0.75% (saturated V_c-90), Mo₂FeB₂ was confirmed to precipitate at the heating temperature of 950°C. It was hypothesized that V_c-90 did not change because the following two factors were balanced: 1) the increase in hardenability, which was attributed to the increase in Mo in solid-solution with increasing Mo content, and 2) decrease in hardenability, which was ascribed to the decrease in the amount of B in solid-solution due to the increasing amount of Mo₂FeB₂ precipitate. The effect of Mo content on the V_c-90 values of 1150°C materials are shown in Fig. 1(b). The V_c-90 values of MxB20 steels with Mo contents ≥ 0.75% austenitized at 950°C were constant. In contrast, the V_c-90 values of M7B20 and M15B20 steels austenitized at 1150°C were considerably different. In other words, hardenability increased with increasing Mo content. In the 950°C materials, the formation of Mo₂FeB₂ was a main factor to saturate the Mo effect on hardenability in B-added steels; however, owing to the dissolution of Mo₂FeB₂ at high temperature (1150°C), V_c-90 decreased approximately linearly at Mo content ≤ 1.5%. Nevertheless, the V_c-90 values of 1150°C materials were still higher than those of 950°C materials, and upon increasing austenitizing temperature, the hardenability of all MxB20 steel samples decreased. This indicates that hardenability no longer increased even if Mo₂FeB₂ was dissolved at high temperature (1150°C). In addition, the hardenabilities of the MxB10 and MxB20 steel samples austenitized at 1150°C were lower than those of the corresponding steel samples austenitized at 950°C. These results suggest that, even for steel samples with Mo contents ≥ 0.5%, rapid B diffusion during cooling from high austenitizing temperature promotes B segregation at grain boundaries. This favors the rapid Fe₂₃(C, B)₆ precipitation, which causes a decrease in hardenability owing to the non-equilibrium segregation. This is discussed in more detail in Section 4.

Figure 2 shows the effect of B content on the V_c-90 values of B-added steels with various Mo contents. Figure 2(a) shows the results for 950°C materials, and previously reported results are included for comparison.¹⁵⁾ For the 950°C materials, the effect of B on hardenability was the highest at 11 ppm B in the M5Bx steel. Ueno et al.²⁾ reported that for Mo-free steel with B contents of ≥ 5 ppm, Fe₂₃(C, B)₆ precipitation prevented the further improvement of in V_c-90. Therefore, as reported by Asahi,⁶⁾ for M5Bx steels, the Mo addition was suggested to suppress Fe₂₃(C, B)₆ precipitation, thereby increasing hardenability. For the 950°C materials with B contents of ≥ 12 ppm, the increase in Mo addition by 1.0% or 1.5% did not increase hardenability markedly. This is consistent with the V_c-90 values of MxB20 steels remaining constant at Mo contents ≥ 0.75% (Fig. 1(a)). Figure 2(b) shows the results for 1150°C materials. For M5Bx steels, the effect of B on hardenability was the highest at abount5 ppm B for the1150°C materials, while that was about 10 ppm for the 950°C materials. These results were similar to those reported previously by Asahi⁶⁾ for 1250°C materials and are consistent with the present finding that Fe₂₃(C, B)₆ precipitation occurred for steel samples with low B contents. In contrast, for M10Bx and M15Bx steels, V_c-90 decreased, and hardenability increased at B contents ≤ 10 ppm. Moreover, at higher B contents (≥ 10 ppm) hardenability did not change considerably and V_c-90 remained constant or decreased with increasing B content. These results suggest that the amounts of Fe₂₃(C, B)₆ precipitate for M10Bx and M15Bx steels were comparable to that for M5Bx steels. In addition, the difference was noticed in hardenability between M10Bx and M15Bx steels with B contents ≥ 12 ppm austenitized at 1150°C. This behavior is different from those austenitized at 950°C. These findings indicate that Mo₂FeB₂ precipitates did not form in M10Bx steel when austenitized at 1150°C.

3.1.2. Changes in Transformation Behavior

To further analyze the decrease in hardenability with increasing austenitizing temperature described in Section 3.1.1, we analyzed the effect of austenitizing temperature on Fe₂₃(C, B)₆ precipitation and transformation behaviors at each cooling rate for three 950°C materials: M5B11, M10B10, and M15B9 steels with about 10 ppm B. This B content was selected because it should not cause B precipitation. Figures 3(a)–3(c) show the transformation start temperature for each type of steel measured with a Formaster equipment. To predict the type of transformation, the Bs and Ms points were calculated using the empirical formula proposed by Steven et al.,²⁰⁾ and the results are shown in Figs. 3(a)–3(c). The effect of quenching temperature on transformation temperature was considered negligible, when transformation temperature is lower than the Ms point. In contrast, the effect of quenching temperature considerably affected transformation temperature, when that is between the Bs and Ms points. These results suggest that the difference in behavior was not caused by a thermodynamic effect but by a kinetic effect caused by the decrease in the amount of B in solid-solution at the austenite grain boundaries. That is, the nucleation frequency for bainitic transformation increased due to the change in hardenability. Next, the effect of austenitizing temperature on transformation start temperature is discussed. For M5B11 steel, at cooling rates ≤ 10°C/s, the transformation temperature increased markedly as the austenitizing temperature increased. M10B10 and M15B9 steels exhibited a similar increase in transformation temperature at low cooling rates. However, the cooling rates at which transformation temperature increased markedly with increasing austenitizing temperature for these steels were about 5°C/s and 3°C/s, respectively, which were lower than the corresponding rate for M5B11 steel. Such an increase in transformation temperature with increasing austenitizing temperature is presumed to be associated with a decrease in hardenability, which was consistent with the results of the Jominy test. In addition, the changes in transformation temperature were estimated using the decrease in amount of segregated B caused by Fe₂₃(C, B)₆ precipitation, and the precipitation states for each type of steel are described in Section 3.2.

Fig. 3. Effect of austenitizing temperature on transformation start temperature. (a) 05Mo12B steel, (b) 10Mo10B steel, and (c) 15Mo10B steel.

3.2. B Precipitation State

3.2.1. Precipitation States of M5B11, M10B(5, 10), and M15B(4, 9) Steels Austenitized at 1150°C

To assess the pre-transformation precipitation states over a range of cooling rates close to the V_c-90 value of each type of steel, we used SIMS. M10B10 and M15B9 steels were cooled from 1150 to 650°C at cooling rates of 3 and 1°C/s, respectively, followed by He-gas quench below 650°C. Figures 4(a) and 4(b) show the SIMS BO₂⁻ ion maps of M10B10 and M15B9 steels, respectively. The linear areas for which luminance increases correspond to B in solid-solution segregated at the austenite grain boundaries, and the areas with granular high luminance correspond to B precipitates. Segregated solute B was present at the austenite grain boundaries of M10B10 and M15B9 steels. Granular areas corresponding to precipitates, are indicated with arrows, and can only be observed in the SIMS BO₂⁻ ion map of M10B10 steel (Fig. 4(a)). The TEM-identified precipitates present in M5B11, M10B(5, 10), and M15B(4, 9) steels are summarized in Table 2. The test samples used for TEM were subjected to a similar heat treatment as that for the samples used for SIMS, and the rates used to cool the samples from 1150°C to the quenching temperature are presented in Table 2. Fe₂₃(C, B)₆ was confirmed to precipitate in M5B11 and M10B10 steels. Figure 5 shows the representative TEM image of a Fe₂₃(C, B)₆ precipitate for M10B10 steel. In contrast, no precipitates were present in M15B9 steel. In other words, in the steels containing about 10 ppm B, the addition of 1.5% Mo suppressed Fe₂₃(C, B)₆ precipitation with increasing austenitizing temperature. In addition, no B precipitates were present in M15B4 or M10B5 steels, which means that the addition of a low Mo content (1.0%) suppressed Fe₂₃(C, B)₆ precipitation with increasing austenitizing temperature in the steels containing about 5 ppm B.

Fig. 4. SIMS BO₂^- ion maps. Arrows indicate precipitates. (a) M10B10 steel cooled from 1150°C to 650°C at 3°C/s. (b) M15B9 steel cooled from 1150°C to 650°C at 1°C/s. (Online version in color.)

Table 2. Observation results of precipitated borides in M5B11, M10B(5, 10), M15B(4, 9) steels austenitized at 1150°C.

Material	Mo (%)	B (ppm)	Cooling rate from 1150°C to quenching temperature (°C/s)	Precipitated borides
M10B5	1.0	5	3	No boride
M15B4	1.5	4	1	No boride
M5B11	0.5	11	5	Fe23(C,B)6
M10B10	1.0	10	3	Fe23(C,B)6
M15B9	1.5	9	1	No boride

Fig. 5. Extraction replica TEM photograph, electron diffraction pattern and EDS spectrum of borides in M10B10 steel cooled from 1150°C to 650°C at 3°C/s. (Online version in color.)

3.2.2. Precipitation States of M7B20, M10B25, and M15B(14, 23) Steels Austenitized at 1150°C

To investigate the effect of austenitizing at 1150°C on B precipitates using the steels in which Mo₂FeB₂ precipitated when austenitized at 950°C, the B precipitates in M7B20, M10B25, and M15B(14, 23) steels were investigated using TEM. The TEM-identified precipitates present in M7B20, M10B25, and M15B(14, 23) steels are summarized in Table 3. Mo₂FeB₂ precipitate was not present in M15B4 steel. However, Mo₂FeB₂ precipitates were present in M15B14 and M15B23 steels. These findings were similar to those for the 950°C materials. Figure 6 shows the TEM of a Mo₂FeB₂ precipitate for M15B23 steel. Conversely, Fe₂₃(C, B)₆ precipitates were present in M7B20 and M10B25 steels austenitized at 1150°C, similar to those for 950°C materials. This suggests that Fe₂₃(C, B)₆ easily precipitates during cooling after the dissolution of Mo₂FeB₂ during heating at 1150°C, even for steel samples in which Mo₂FeB₂ precipitates during austenitization at 950°C. These results are discussed in detail in Section 4.

Table 3. Observation results of precipitated borides in M7B20, M10B25, M15B(14, 23) steels austenitized at 1150°C.

Material	Mo (%)	B (ppm)	Cooling rate from 1150°C to quenching temperature (°C/s)	Precipitated borides
M7B20	0.75	20	10	Fe23(C,B)6
M10B25	1.0	25	3	Fe23(C,B)6
M15B14	1.5	14	1	Mo2FeB2
M15B23	1.5	23	1	Mo2FeB2

Fig. 6. Extraction replica TEM photograph, electron diffraction pattern and EDS spectrum of borides of M15B23 steel cooled from 1150°C to 650°C at 1°C/s. (Online version in color.)

To estimate the precipitation behavior of Fe₂₃(C, B)₆ during high-temperature heating, M7B20 steel was austenitized at 1150°C followed by continuous cooling. Thereafter, the TEM images of test samples with different T_Q values were obtained. Our results indicate that even at T_Q = 1150°C (materials were quenched immediately after austenitizing), Fe₂₃(C, B)₆ precipitated (Table 4 and Fig. 7). In addition, Fe₂₃(C, B)₆ coarsened with decreasing T_Q to 850 and 650°C. This indicates that, for the samples austenitized at 1150°C, Fe₂₃(C, B)₆ precipitates formed and grew over a wide temperature range (1150°C (cooling start temperature) to 650°C (approximately temperature at which bainite transformation starts).

Table 4. Observation results of borides in M7B20 steel austenitized at 1150°C.

Quenching temperature (°C), TQ	Cooling rate from 1150°C to TQ (°C/s)	Precipitated borides
1150	–	Fe23(C, B)6 Size: 100 nm
850	10	Fe23(C, B)6 Size: 100 nm −1 μm
650	10	Fe23(C, B)6 Size: 200 nm −1.8 μm

Fig. 7. Extraction replica TEM photograph of borides in M7B20 steel quenched from various temperature T_Q after austenitized at 1150°C. Arrows indicate Fe₂₃(C, B)₆. Cooling rate from 1150°C to T_Q was 10°C/s in (b) and (c).

4. Discussion

4.1. Relationship between Hardenability and B Precipitation State

We compared the B precipitation state that affected hardenability for samples austenitized at 1150°C with that of previously reported samples austenitized at 950°C.¹⁵⁾ The experimental results for each austenitizing temperature are summarized as below. The effect of B content on V_c-90 together with experimentally demonstrated precipitation states of B for steel samples added with 0.5%, 1.0% and 1.5% Mo are shown in Fig. 8. The steel samples in which the precipitates were analyzed are indicated using arrows. To estimate the boundaries between B precipitates and no B precipitates in steels samples, the B contents at which the V_c-90 values were the lowest are indicated by open symbols in Fig. 8. Figure 8 also represents the schematic image of the effect of heating temperature on the B precipitation states of combined Mo–B added steels with various Mo contents. This schematic image is drawn assuming the precipitation state of B at a cooling rate of V_c-90 for the 15 ppm B steel sample. No B precipitates were present in M5B9 steel austenitized at 950°C; however, Fe₂₃(C, B)₆ precipitate was present in M5B20 steel austenitized at 950°C (Fig. 8(a)). In addition, because M5B11 steel presented the lowest V_c-90 value, it was expected that Fe₂₃(C, B)₆ precipitate would form in M5Bx steels with B contents ≥ 11 ppm. Similarly, for the 1150°C materials, Fe₂₃(C, B)₆ precipitate was present in M5B11 steel, and the V_c-90 value of M5Bx (x ≈ 5) was the lowest. Therefore, Fe₂₃(C, B)₆ precipitates was expected to form in M5Bx (x ≥ 5) steels. Similar estimates were made using the experimental results of precipitates and B contents at the lowest V_c-90 values, for M10Bx and M15Bx steels (Figs. 8(b) and 8(c), respectively). With increasing austenitization temperature for M10Bx (x ≥ 10) steel, Mo₂FeB₂ precipitates, which were present after cooling the steel samples at a rate of approximately V_c-90, was expected to change to Fe₂₃(C, B)₆ precipitates (Fig. 9). Furthermore, even for M15Bx steel austenitized at 1150°C, Mo₂FeB₂ precipitated at the austenitizing temperature, and the precipitate type did not change with austenitization temperature. Moreover, within the B content ranges in which B precipitates were not present in M10Bx and M15Bx steels, hardenability decreased slightly with increasing heating temperature (Figs. 8(b) and 8(c)). The test samples used to evaluate the B precipitation states were quenched at 650°C; however, precipitates could have formed at lower temperatures. Nevertheless, this was a hypothesis, and further studies should be performed to investigate the B precipitation states at lower temperatures.

Fig. 8. Relationship between V_c-90 and B content taking into account precipitation state in Mo-B combined added steels. Schematic image of effect of reheating temperature on precipitation behavior of Mo₂FeB₂ and Fe₂₃(C, B)₆ is shown at each Mo contents. (a) 0.5%Mo, (b) 1.0%Mo and (c) 1.5%Mo. (Online version in color.)

Fig. 9. Precipitation map of borides in Mo–B combined added steels at cooling rates corresponding to the V_c-90, austenitized at (a) 950°C¹⁵⁾ and (b) 1150°C. (Online version in color.)

These results for steel samples with various Mo and B contents are summarized in the precipitation maps of borides in Mo–B combined steels (Fig. 9). The B content range in which Mo suppressed Fe₂₃(C, B)₆ precipitation are labeled as “no boride” in Fig. 9, where this range shifted toward lower B contents with increasing austenitizing temperature. However, even for the samples austenitized at 1150°C, for Mo15Bx (x ≈ 10) steel, Fe₂₃(C, B)₆ precipitation was suppressed. This indicates that the effects of Mo addition were not eliminated, and the interactions between Mo and B were not negligible even for high austenitizing temperature. In addition, considering the data in Table 3 and Fig. 7 together with previously reported B range in which Mo₂FeB₂ precipitated,¹⁵⁾ we assumed that for Mo–B added steels with Mo contents in the range of 0.75–1% and B content of approximately 20 ppm, Mo₂FeB₂ precipitate can be dissolved during heating at 1150°C, where Mo₂FeB₂ was a factor affecting the saturated effect of combined Mo–B addition for 950°C materials. However, even though Mo₂FeB₂ precipitate was dissolved, Fe₂₃(C, B)₆ would easily precipitate during cooling. The B content in which Mo₂FeB₂ precipitated, is illustrated in the precipitation map of combined Mo–B added steels when austenitized at 950°C (Fig. 9(a)). However, for the combined Mo–B added steels austenitized at 1150°C, a part of the region of Mo₂FeB₂ precipitate was converted to the region of Fe₂₃(C, B)₆ precipitate. In this study, Mo₂FeB₂ precipitates were present in Mo15Bx (x ≥ 14) steels at an austenitizing temperature of 1150°C. However, upon further increasing austenitizing temperatures, Mo₂FeB₂ dissolution is expected to occur; however, Mo₂FeB₂ converts to Fe₂₃(C, B)₆. These changes in B precipitation behavior with increasing austenitizing temperature are discussed in Sections 4.2.1. and 4.2.2. below from thermodynamic and kinetic perspectives, respectively.

4.2. Interactions between B and Mo during High Austenitizing Temperature

4.2.1. Thermodynamic Effects

To discuss the competition between Fe₂₃(C, B)₆ and Mo₂FeB₂ precipitation, and the influence of Mo on the promotion of Fe₂₃(C, B)₆ precipitation via high austenitizing temperature, thermodynamic calculation was performed in relation to B precipitation, as investigated by Takahashi et al.²⁴⁾ and Otani et al.²⁵⁾ Figures 10 and 11 show the effects of temperature and addition of Mo and B on the precipitation of face centered cubic (FCC) M₃B₂ (equivalent to Mo₂FeB₂) and M₂₃C₆ (equivalent to Fe₂₃(C, B)₆) phases calculated using the Thermo-Calc software. The calculations were performed for the Fe–0.15%C–B–Mo quaternary system and considering only the FCC phases of M₃B₂ and M₂₃C₆, and TCFE10 was used as the database. Figure 10 shows the effects of B and Mo contents on the M₂₃C₆ and M₃B₂ precipitation ranges at 950 and 1150°C. Figure 10(a) shows the calculation results at 950°C. The calculated results matched the experimentally demonstrated precipitation behavior; however, they were not completely quantitatively consistent. A M₂₃C₆ phase forms in the region with a low Mo content and high B content, whereas a M₃B₂ phase forms in the region with high Mo and B contents (Fig. 10(a)). Figure 10(b) shows the calculation results at 1150°C. M₃B₂ precipitates in a narrow region with Mo and B contents higher than those calculated for the austenitizing temperature of 950°C, and these findings matched the experimentally determined tendency. Meanwhile, with respect to the M₂₃C₆ phase, no equilibrium phase was present at 1150°C, and precipitation promotion of the M₂₃C₆ phase by the addition of Mo during high austenitizing temperature cannot be explained solely using of an equilibrium diagram. Therefore, the rapid diffusion of B during high austenitizing temperature should be considered. In other words, a mechanism such as the non-equilibrium segregation is presumably necessary.

Fig. 10. Effect of B and Mo contents on calculated phase diagram. (Online version in color.)

Fig. 11. Effect of B and Mo contents on equilibrium precipitation starting temperature of (a) M₃B₂ and (b) M₂₃C₆ phases. (Online version in color.)

Figure 11 shows the effects of B and Mo additions on the precipitation start temperatures of the M₃B₂ and M₂₃C₆ phases when the amount of B segregated at grain boundaries increased during cooling. For the samples with high B and Mo contents, the M₃B₂ precipitation start temperature was high. Moreover, for Mo10Bx (x ≤ 30) and Mo15Bx (x ≤ 23) steels, M₃B₂ is predicted to be dissolved at 1150°C. In this study, Mo₂FeB₂ precipitates were not present in Mo10Bx (x ≤ 25), M15B4, and M15B9 steels austenitized at 1150°C; however, Mo₂FeB₂ precipitates were present in Mo15B14, and Mo15B23 steels austenitized at 1150°C. Although the experimental and calculated results did not match 100%, the calculated results shown in Fig. 11(a) replicated the aforementioned experimental results. Moreover, even for B-added steels austenitized at high temperature, the Mo₂FeB₂ precipitation range can be predicted using thermodynamic calculations. Figure 11(b) shows the effects of B and Mo contents on the precipitation start temperature of the M₂₃C₆ phase. The M₂₃C₆ precipitation start temperature increased with increasing B content. However, unlike for the M₃B₂ precipitation start temperature, that of M₂₃C₆ remained almost constant with increasing Mo content. At 1150°C, the precipitation of M₂₃C₆ (molar ratio) is predicted to be achieved at a B content of 60 ppm. The corresponding Fe:C:B:Mo molar ratio was 0.79:0.02:0.19:4×10⁻⁸, and under these conditions, the amount of Mo in solid-solution in M₂₃C₆ is negligible. Mo was not observed in the energy-dispersive X-ray spectroscopy images of Fe₂₃(C, B)₆ precipitates (Fig. 5); therefore, the results of the thermodynamic calculations were consistent with the experimental results. Such a relationship can be interpreted as a repulsive interaction between M₂₃C₆ and Mo, where the replacement of Fe by Mo in the Fe₂₃C₆ composition causes an increase in free energy of the M₂₃C₆ phase. Consequently, the M₂₃C₆ precipitation start temperature remains unchanged with increasing Mo content. Therefore, even for the Mo added steel with the B content higher than 10 ppm, where Mo₂FeB₂ is dissolved during austenitization at 1150°C, B segregation at grain boundaries progresses during cooling. Since Mo thermodynamics does not affect the Fe₂₃(C, B)₆ precipitation start temperature, the precipitation rate of Fe₂₃(C, B)₆ is considered high, resulting in the precipitation at grain boundaries. Therefore, we focused on the local B content to induce the M₂₃C₆ precipitation (Fig. 11(b)). We hypothesized that if B in solid-solution segregates and Fe₂₃(C, B)₆ precipitates at grain boundaries, it would be expected that, within the temperature range of 1150–950°C, M₂₃C₆ would precipitate when the concentration of B in solid-solution segregated at grain boundaries increases from 20 to 60 ppm during cooling. It corresponds to the boron concentration only several times higher than the matrix concentration of test steels. The non-equilibrium segregation was presumed to affect the increase in concentration of B in solid-solution segregated at grain boundaries in the temperature range of 1150–950°C, and this contribution is discussed further in Section 4.2.2.

4.2.2. Effect of Mo on the Non-equilibrium Segregation Behavior of B

For the combined Mo–B added steels with B contents ≤ 10 ppm, Fe₂₃(C, B)₆ precipitation was suppressed with increasing Mo content even at an austenitized temperature of 1150°C (Fig. 9). This has not been reported previously, and it suggests that Mo affected the rapid diffusion of B during cooling. The effect of Mo is discussed further based on the non-equilibrium segregation model used to explain the rapid diffusion of B during conventional high austenitizing temperature. The Karlsson model is typically used to analyze the non-equilibrium segregation.¹⁰⁾ The equilibrium vacancy concentration (C_V) at each temperature (T) can be calculated as follows:

  

$$C_{\text{V}}=K_{\text{V}0}\cdot \text{exp}\left(\frac{-E_{\text{VF}}}{kT}\right),$$(1)

where K_V0 is a constant associated with configurational entropy, E_VF is the vacancy formation energy, and k is the Boltzmann’s constant.

Considering the B–V complexes formed by pairing B with vacancies (V), which promote the non-equilibrium segregation of B, the concentration of B–V complexes (C_VB) can be calculated as follows:

  

$$\frac{C_{\text{VB}}}{C_{\text{V}}\cdot C_B}=K_{\text{VB}0}\cdot exp\left(\frac{B_{\text{VB}}}{kT}\right)$$(2)

where C_B is the B content, K_VB0 is a constant associated with configurational entropy, E_VF is the vacancy formation energy, and B_VB is the binding energy of B and V.

Next, we attempted to evaluate the effect of Mo using the method of Kimura et al.,²⁶⁾ who evaluated the effect of Sn on the precipitation behavior in Al–Cu alloys. Kimura et al.²⁶⁾ considered that the delay in the formation of the Guinier–Preston zone in Al–Cu alloys induced by the Sn addition was attributed to Cu diffusion being delayed by Sn which traps V. Using the standardized formula proposed by Kimura et al.,²⁶⁾ we calculated the concentration of V bound by B (C_1V) as follows:

  

$$C_{1\text{V}}=\frac{\frac{C_{\text{B}}}{C_{\text{Mo}}}{C}_{\text{V}}\cdot exp\left(-\frac{\mathrm{\Delta }B}{k{T}_s}\right)}{1+\frac{C_{\text{B}}}{C_{\text{Mo}}}{C}_{\text{V}}\cdot exp\left(-\frac{\mathrm{\Delta }B}{k{T}_s}\right)},$$(3)

where C_V is the equilibrium vacancy concentration at the austenitizing temperature. C_B and C_Mo are the B and Mo contents, respectively, C_1V is the vacancy concentration when quenching was conducted in the temperature range between the austenitizing temperature and cooling stop temperature (T_s), and excess vacancies at the austenitizing temperature are present at T_S. These calculations were performed assuming that the binding energy (B_VMo) between an atomic vacancy (V) and a ternary element (Mo in this study) is higher than the binding energy (B_VB) between B and V, where the difference between B_VMo and B_VB is defined as ΔB. Moreover, it was hypothesized that no free atomic vacancies existed (i.e., C_V = C_1V + C_2V, where C_2V is the concentration of V bound to Mo).

The C_V, C_VB, and C_1V values calculated using Eqs. (1), (2) and (3) are shown in Fig. 12, and parameters necessary for calculation are summarized in Table 5. Figures 12(a) and 12(b) show, the C_V and C_VB values, respectively. As anticipated, C_VB increased with increasing austenitizing temperature. For example, upon cooling from 1150°C, the B–V complex of C_VB ≈ 2 ppm is expected to promote non-equilibrium segregation. The B segregation coefficient at 950°C at a B content of 10 ppm was of the order of several tens to several hundreds.¹⁶⁾ Therefore, it was assumed that B of 2 ppm may rapidly diffuse, and contribute to the increase in such an amount of segregated B at grain boundaries. Moreover, when the precipitation start temperature line for the M₂₃C₆ phase is crossed, Fe₂₃(C, B)₆ is able to precipitate (Fig. 11(b)). Figure 12(c) shows the C_1V values. The C_V value calculated at 1150°C was used in Eq. (3), and C_1V was calculated when the vacancy concentration at 1150°C was present at temperature T_S. As the Mo content is increased to 1% and 1.5%, C_1V decreases. This indicates that the increase in Mo content causes an increase in C_2V, leading to the decrease in C_VB, which promotes the non-equilibrium segregation of B at the temperature T_S. In other words, suppression of Fe₂₃(C, B)₆ precipitation during cooling owing to an increase in Mo content even at an austenitizing temperature of 1150°C (Fig. 9), was attributed to the decrease in the C_VB value for the B–V complexes formed during high austenitizing temperature and suppression in the increase in segregation at grain boundaries owing to non-equilibrium segregation during cooling. This discussion was based on the hypothesis that B diffuses rapidly owing to B–V complex formation, because the amount of segregated B increased during cooling. However, the rapid diffusion of B has not been sufficiently confirmed experimentally or theoretically, which requires further research in future. In addition to rapid B diffusion induced by B–V complex formation, the grain boundary phase model proposed by Takahashi et al.,²⁴⁾ which considers precipitate formation in grain boundary,²⁸⁾ can be used to explain the increase in the amount of segregated B during cooling by calculating the equilibrium segregation amount of B. In addition, even if no precipitates form, segregation can be attributed to clustering of solute elements. This indicates that more studies should be conducted, and researchers should not only observe precipitates using TEM, as we did herein, but also perform microstructure analysis using 3DAP and other suitable techniques to quantify the amount of segregated B. Furthermore, previous research suggested that B atoms in γ-Fe occupy substitutional sites;²⁹⁾ however, a fraction of the B atoms in γ-Fe may change its stable site from the substitutional to the interstitial site during high austenitizing temperature, thereby contributing to the significant segregation. This is because B atoms in α-Fe can occupy both in interstitial and substitutional sites in energetic terms.³⁰⁾ Future experimental and theoretical validation of the site of B in γ-Fe is important. In addition, the precipitation suppression mechanism by Mo should be further investigated. In terms of the effect of Mo on conventional 950°C materials, Hara et al.³¹⁾ proposed that suppression of Fe₂₃(C, B)₆ precipitation was caused by Mo retarding the diffusion of C toward grain boundaries.³¹⁾ The number of B–V pairs formed in γ-Fe at 950°C is lower than that formed at 1150°C. However, the effect of Mo suppressing B diffusion, as discussed herein, can contribute to suppression of Fe₂₃(C, B)₆ precipitation. Therefore, future studies should develop models that can comprehensively explain the effects of quenching temperature and element content.

Fig. 12. Calculated contents of (a) equilibrium vacancy (C_V), (b) equilibrium boron-vacancy pair (C_VB) as a function of temperature, (c) vacancy trapped as B–V pair (C_1V) in Mo added steels as a function of Ts. (Online version in color.)

Table 5. Thermodynamic parameter values used in the calculation of C_V, C_VB, C_1V.

CB (mass ppm)	CMo (mass ppm)	KV0	KVB0	EVF10)	BVB10)	BVMo27)
10	1, 1.5	12	12	1.4	0.5	3.44

5. Conclusions

In this study, we used Fe–0.15C–1.3Mn (%) steel as the fundamental composition, and the B-added steel and combined Mo–B added steel with various B and Mo contents to investigate the effect of austenitizing temperature on hardenability and B precipitation state over a wider range of Mo and B contents. The principal findings of our study are as follows:

(1) For the combined Mo–B added steel with about 10 ppm B, the decrease in the effect of combined Mo–B addition with increasing the austenitized temperature was suppressed by increasing Mo content. Fe₂₃(C, B)₆ precipitates formed upon austenitizing Mo5Bx steels at 1150°C. However, no Fe₂₃(C, B)₆ precipitates were present after austenitizing Mo15Bx steels at 1150°C. Moreover, it was determined that even during high austenitizing temperature, the effects of combined Mo and B addition were achieved over a specific B content range.

(2) Mo₂FeB₂ precipitated in combined Mo–B added steels with about 20 ppm B austenitized at 950°C can be dissolved during austenitization at 1150°C. However, even if Mo₂FeB₂ was dissolved at 1150°C, Fe₂₃(C, B)₆ precipitated immediately after cooling started, leading to poorer hardenability than the case of austenitization at 950°C. In addition, for Mo15Bx steels, Mo₂FeB₂ dissolution does not occur even at 1150°C, and it was demonstrated that for Mo15Bx steels with about 20 ppm B, B precipitation was not entirely suppressed by the Mo addition.

(3) Calculations assuming that B and Mo combines with vacancies were proposed and demonstrated that an increase in Mo content caused a decrease in the concentration of B–V complexes. This contributed to suppress the formation of Fe₂₃(C, B)₆ due to the reduced B non-equilibrium segregation to grain boundaries, which supports the mechanism for the result that the effect of combined Mo–B addition owing to an increase in Mo content was confirmed even at high austenitizing temperatures.

References

• 1)   M. A.  Grossman: Trans. AIME., 150 (1942), 227.
• 2)   M.  Ueno and  K.  Itoh: Tetsu-to-Hagané, 74 (1988), 1073 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.74.5_910
• 3)   R. A.  Grange and  T. M.  Garvey: Trans. ASM, 37 (1946), 136.
• 4)   C. R.  Simcoe,  A. R.  Elsea and  G. K.  Manning: Trans. AIME, J. Metals, 8 (1956), 984.
• 5)   S.  Yoshida,  K.  Ushioda,  Y.  Abe and  J.  Ågren: ISIJ Int., 54 (2014), 685. https://doi.org/10.2355/tetsutohagane.98.482
• 6)   H.  Asahi: ISIJ Int., 42 (2002), 1150. https://doi.org/10.2355/isijinternational.42.1150
• 7)   T.  Fujishiro,  T.  Hara and  G.  Shigesato: Tetsu-to-Hagané, 101 (2015), 300 (in Japanese). https://doi.org/10.2355/tetsutohagane.101.300
• 8)   L.  Karlsson,  H.  Norden and  H.  Odelius: Acta Metall., 36 (1988), 1. https://doi.org/10.1016/0001-6160(88)90023-5
• 9)   L.  Karlsson,  H.  Norden and  H.  Odelius: Acta Metall., 36 (1988), 13. https://doi.org/10.1016/0001-6160(88)90024-7
• 10)   L.  Karlsson,  H.  Norden and  H.  Odelius: Acta Metall., 36 (1988), 25. https://doi.org/10.1016/0001-6160(88)90025-9
• 11)   L.  Karlsson and  H.  Norden: Acta Metall., 36 (1988), 35. https://doi.org/10.1016/0001-6160(88)90026-0
• 12)   X. L.  He,  Y. Y.  Chu and  J. J Jonas: Acta Metal., 37 (1989), 2905. https://doi.org/10.1016/0001-6160(89)90325-8
• 13)   G.  Shigesato,  T.  Fujishiro and  T.  Hara: Metall. Mater. Trans. A, 45 (2014), 1876. https://doi.org/10.1007/s11661-013-2155-3
• 14)   G.  Miyamoto,  A.  Goto,  N.  Takayama and  T.  Furuhara: Scr. Mater., 154 (2018), 168. https://doi.org/10.1016/j.scriptamat.2018.05.046
• 15)   K.  Ishikawa,  H.  Nakamura,  R.  Homma,  M.  Fujioka and  M.  Hoshino: ISIJ Int., 58 (2018), 551. https://doi.org/10.2355/isijinternational.ISIJINT-2017-579
• 16)   J.  Takahashi,  K.  Ishikawa,  K.  Kawakami,  M.  Fujioka and  N.  Kubota: Acta Mater., 133 (2017), 41. https://doi.org/10.1016/j.actamat.2017.05.021
• 17)   K.  Ishikawa,  H.  Nakamura,  M.  Fujioka,  M.  Hoshino,  J.  Takahashi and  K.  Kawakami: Tetsu-to-Hagané, 106 (2020), 321(in Japanese). https://doi.org/10.2355/tetsutohagane.TETSU-2019-099
• 18)   K.  Ishikawa,  H.  Nakamura,  R.  Homma,  M.  Fujioka and  M.  Hoshino: Proc. Asia Steel International Conf. (2015), (Yokohama), The Iron and Steel Institute of Japan, 624, CD-ROM.
• 19)   J. C.  Shyne and  E. R.  Morgan: Metal Progress, 65 (1954), 88.
• 20)   W.  Steven and  A. G.  Haynes: J. Iron Steel Inst., 183 (1956), 349.
• 21)   S.  Hashimoto,  S.  Doi,  K.  Takahashi,  M.  Terasaka and  M.  Iwaki: Mater. Sci. Eng., 90 (1987), 119. https://doi.org/10.1016/0025-5416(87)90202-3
• 22)   S.  Lozano-Perez,  M.  Schroder,  T.  Yamada,  T.  Terachi,  C. A.  English and  C. R. M. Grovenor: Appl. Surf. Sci., 255 (2008), 1541.
• 23)   T.  Tanaka,  T.  Fujishiro,  G.  Shigesato and  S.  Hayashi: Surf. Int. Anal., 46 (2014), 297. https://doi.org/10.1002/sia.5414
• 24)   K.  Takahashi,  K.  Ishikawa,  M.  Fujioka,  M.  Enoki and  H.  Ohtani: Tetsu-to-Hagané, 106 (2020), 310 (in Japanese). https://doi.org/10.2355/tetsutohagane.TETSU-2019-097
• 25)   H.  Ohtani,  M.  Hasebe,  K.  Ishida and  T.  Nishizawa: Trans. Iron Steel Inst. Jpn., 28 (1988), 1043. https://doi.org/10.2355/isijinternational1966.28.1043
• 26)   H.  Kimura and  R. R.  Hasiguti: Acta Met., 9 (1961), 1076. https://doi.org/10.1016/0001-6160(62)90050-0
• 27)   T.  Tsuru and  Y.  Kaji: Journal of Nuclear Materials, 442 (2013), 684. https://doi.org/10.1016/j.jnucmat.2013.03.051
• 28)   M.  Hillert: Lectures on the Theory of Phase Transformations, AIME, NY, (1975), 36.
• 29)   P. E.  Busby,  M. E.  Warge and  C.  Wells: Trans. Met. Soc. Amer. Inst. Min. Met. & Pet. Eng., 197 (1953), 1463.
• 30)   D. H. R.  Fors and  G.  Wahnström: Phys. Rev. B, 77 (2008), 132102. https://doi.org/10.1103/PhysRevB.77.132102
• 31)   T.  Hara,  H.  Asahi,  R.  Uemori and  H.  Tamehiro: ISIJ Int., 44 (2004), 1431. https://doi.org/10.2355/isijinternational.44.1431