Introduction

High-strength steel is extensively utilized in automotive, aerospace and construction industries due to its exceptional strength and durability.1,2 High-strength steel achieves its remarkable properties primarily through complex microalloying with elements such as Nb, V and Ti, which are strong carbon (nitrogen) compound formers capable of forming stable precipitates. These precipitates refine the grain structure and impede dislocation movement, consequently enhancing steel yield and tensile strength.3 Among the various types of microalloyed high-strength steels, microalloying of titanium has shown the most significant effect on material strength. Also, it offers a distinct advantage in terms of raw material costs. Thus, microalloying of titanium is prevalently regarded as the principal approach for grain refinement and strength enhancement.4,5 However, large TiN particles (> 10 µm) precipitate quickly form in the center of the steel during the continuous casting process for a high Ti content in the steel. These large, hard TiN particles diminish the toughness, weldability, and fatigue limit of high-strength steel and lead to stress concentration, which accelerates the initiation and propagation of cracks, causing a notable reduction in the central region strength of the rolled product.6,7,8 Numerous studies have been conducted to investigate this issue through experiments and establish models for calculating TiN precipitation during the solidification process of molten steel. It has been found that a considerable amount of TiN precipitates within the solid-liquid two-phase zone due to the gradual enrichment and segregation of Ti and N atoms at the solidification front.9,10,11,12,13 Therefore, avoiding the formation of large TiN particles in the center of solidified steel is a critical issue that needs to be addressed urgently in the industrial production of Ti-microalloyed high-strength steel. Unfortunately, few reports exist on refining the large TiN particles in the center of high-titanium high-strength steel.

In recent years, extensive research has been conducted on adding trace amounts of magnesium to steel to manipulate the growth of inclusions. During the cooling and solidification process of molten steel, high-melting-point magnesium oxides can stably exist and provide heterogeneous nucleation sites for the secondary phase particles in molten steel, thereby refining the solidification structure. Jo14 and Ohta15 investigated the thermodynamic dominance region diagram of Mg–Al–O inclusions in molten steel containing 40%Ni and 20%Cr, revealing that only a small quantity of magnesium can modify Al2O3. Yang16 et al. investigated ladle furnace (LF) refining techniques applied to low carbon aluminum killed (LCAK) steel and calculated a thermodynamic diagram illustrating the relationship between Mg, Al and O in molten steel. Their findings indicated that when [Al] content reached 0.04%, approximately 0.5 ppm [Mg] was required to convert Al2O3 into MgAl2O4. Wang et al.,17 through their experimentation with H13 steel subjected to magnesium treatment, found that when the magnesium content was below 12 ppm, the inclusions present were primarily composed of Al2O3 and MgAl2O4. As the amount of added magnesium increased, so did the number of MgAl2O4 in the steels. Once reaching a concentration level at around 28 ppm, all existing inclusions transformed completely into MgAl2O4. Further increase in magnesium content led to precipitation of MgO. Additionally, it was observed that inclusion sizes larger than 2 μm decreased while those around 1 μm increased. Isobe18 and Kimura19 investigated the effect of Mg and Ti on the formation of equiaxed grains. They found that adding Mg promotes the formation of equiaxed grain, with a significant increase in the rate of equiaxed grain formation as Mg content increases. Qu et al.20 found that magnesium treatment in Ti-bearing peritectic steel consistently led to the presence of high-temperature oxides, such as MgAl2O4 or MgO, in molten steel. This treatment resulted in a 49% decrease in the average size of TiN particles post-treatment compared to pre-treatment. This indicated that the MgAl2O4 or MgO particles formed after magnesium treatment promote the heterogeneous nucleation of TiN, providing nucleation sites for the refinement of high-temperature δ-ferrite. Previous studies on homogeneous and heterogeneous nucleation behavior of TiN in steel mainly focused on medium to low titanium steels under fast cooling conditions such as air or water. Consequently, these findings cannot directly guide the control over the size of TiN refinement in the central region during continuous casting for refined high-titanium steels. Therefore, this study focuses on high-titanium high-strength steel as the research subject to investigate the influence of varying magnesium content on the type, quantity and size of TiN precipitation in the steel after slow solidification cooling process. The aim is to provide a scientific basis and practical guidance for precise control of TiN in the billet central region during continuous casting.

Experimental Materials and Methods

Four high-titanium high-strength steels with varying magnesium content were melted in a high-temperature muffle furnace at a temperature of 1600 °C and were cast into 1.2 kg ingots. The high-temperature muffle furnace (T-1700VCB) was shown in Fig. 1. When the temperature reached 1600 °C in the muffle furnace, the tested steel completely melted. The first step involved deoxygenating the molten steel by adding silicon and aluminum. The second step was the introducting magnesium–aluminum alloy into the molten steel, and the third step involved the addition of titanium to the molten steel. Then, the tested steels were cooled in the furnace to room temperature (cooling rate of 0.17 °C/s). The chemical composition of four tested steels was detected by an inductively coupled plasma optical emission spectrometer, and the content of N and O was detected by a TC500 nitrogen and oxygen analyzer. Swarf was cut from four tested steels, and the magnesium content was determined using chemical analysis method. Their chemical compositions were listed in Table 1. After the tested steel ingots were cooled to room temperature, the crucible was taken out from the high-temperature muffle furnace, then the tested steel ingots (Φ48 mm × 80 mm) were taken out from the crucible, as shown in Fig. 2. Wire electrical discharge machining was used to cut the bottom of the tested steel ingots at a height of 15 mm, then took out a Φ18-mm cylindrical sample in the middle. The Φ18-mm cylindrical sample was ground, polished and eroded with 4 vol.% nitric acid alcohol solution. The typical morphology and composition of TiN on the etched surface of each sample were observed and analyzed using a scanning electron microscope (JEM-6510). The etched surfaces were continuously observed and photographed using a metallurgical microscope (UCMOS03100KPA). Photoshop software was used to statistically analyze and measure the quantity and size of TiN within each field of view to calculate the TiN density in different size ranges. The TiN density is the total number of TiN divided by the total area. By utilizing the primary chemical reactions responsible for the formation of Mg inclusions and employing the principle of minimizing Gibbs free energy, this study calculates the correlation between Mg and Al content and the equilibrium products of deoxidation. Furthermore, it analyzes the association between Mg inclusions and heterogeneous nucleation of TiN in high-titanium high-strength steels with varying Mg contents.

Figure 1
figure 1

High-temperature Muffle furnace (T-1700VCB).

Full size image
Table 1 Chemical Composition of Tested Steels (wt.%)
Full size table
Figure 2
figure 2

Tested steel ingots: (a) 0 wt.% Mg; (b) 0.0007 wt.% Mg; (c) 0.0014 wt.% Mg; (d) 0.0023 wt.% Mg.

Full size image

Experimental Results and Analysis

Thermodynamic Analysis of MgO and MgAl2O4 Formation During Mg Deoxidation in the Tested Steels

After adding Mg to molten steel, the chemical reactions involved in the Fe–Mg–Al–O system at 1600 °C and their corresponding standard Gibbs free energy (\(\Delta G^{\theta }\)) of formation are expressed by Eqns. 1–5.20,21,22,23 Equations 4 and 5 are obtained through a linear combination of Eqns. 1–3. In molten steel, the content of elements other than Fe is relatively low, which conforms to Henry's Law, thus considering molten steel as an ideal solution.

$$ \begin{aligned} & {2}\left[ {{\text{Al}}} \right] + {3}\left[ {\text{O}} \right] = {\text{Al}}_{{2}} {\text{O}}_{{3}} \left( {\text{s}} \right) \\ & \Delta G_{1}^{\theta } = - 867500 + 222.5T \\ \end{aligned} $$
(1)
$$ \begin{aligned} & \left[ {{\text{Mg}}} \right] + \left[ {\text{O}} \right] = {\text{MgO}}\left( {\text{s}} \right) \\ & \Delta G_{2}^{\theta } = - 90000 - 82.0T \\ \end{aligned} $$
(2)
$$ \begin{aligned} & {\text{MgO}}\left( {\text{s}} \right) + {\text{Al}}_{{2}} {\text{O}}_{{3}} \left( {\text{s}} \right) = {\text{MgO}} \cdot {\text{Al}}_{{2}} {\text{O}}_{{3}} \left( {\text{s}} \right) \\ & \Delta G_{3}^{\theta } = - 20682 - 11.57T \\ \end{aligned} $$
(3)
$$ \begin{aligned} & {3}\left[ {{\text{Mg}}} \right] + {\text{4Al}}_{{2}} {\text{O}}_{{3}} \left( {\text{s}} \right) = {\text{3MgO}} \cdot {\text{Al}}_{{2}} {\text{O}}_{{3}} \left( {\text{s}} \right) + {2}\left[ {{\text{Al}}} \right] \\ & \Delta G_{4}^{\theta } = 535454 - 503.09T \\ \end{aligned} $$
(4)
$$ \begin{aligned} & {3}\left[ {{\text{Mg}}} \right] + {\text{MgO}} \cdot {\text{Al}}_{{2}} {\text{O}}_{{3}} \left( {\text{s}} \right) = {\text{4MgO}}\left( {\text{s}} \right) + {2}\left[ {{\text{Al}}} \right] \\ & \Delta G_{5}^{\theta } = 618182 - 456.93T \\ \end{aligned} $$
(5)

At 1600 °C equilibrium, the Gibbs free energy (\(\Delta G\)) for the formation of MgAl2O4 and MgO in molten steel within the Fe–Mg–Al–O system can be expressed by Eqns. 6 and 7:

$$ \Delta G = \Delta G_{4}^{\theta } + RT\ln \frac{{a_{{({\text{MgAl}}_{2} {\text{O}}_{4} )}}^{3} \cdot a_{{[{\text{Al}}]}}^{2} }}{{a_{{({\text{Al}}_{2} {\text{O}}_{3} )}}^{4} \cdot a_{{[{\text{Mg}}]}}^{3} }} $$
(6)
$$ \Delta G = \Delta G_{5}^{\theta } + RT\ln \frac{{a_{{({\text{MgO}})}}^{4} \cdot a_{{[{\text{Al}}]}}^{2} }}{{a_{{({\text{MgAl}}_{2} {\text{O}}_{4} )}} \cdot a_{{[{\text{Mg}}]}}^{3} }} $$
(7)

where \(a_{{({\text{MgAl}}_{{2}} {\text{O}}_{{4}} )}}\), \(a_{{({\text{Al}}_{{2}} {\text{O}}_{{3}} )}}\), a[Al] and a[Mg] denote the activities of MgAl2O4, Al2O3, MgO, Al and Mg in molten steel, respectively. Herein, \(a_{{({\text{MgAl}}_{{2}} {\text{O}}_{{4}} )}} = 1\), \(a_{{{\text{(Al}}_{{2}} {\text{O}}_{{3}} {)}}} = 1\) and a(MgO) = 1; w[Al] and w[Mg] denote the mass fractions of Al and Mg in molten steel, respectively; a[Al] and a[Mg] can be obtained by Eqns. 8 and 9:

$$ a_{i} = f_{i} [\% i] $$
(8)
$$ \lg f_{i} = \sum\limits_{j = 1}^{n} {e_{i}^{j} [\% i]} $$
(9)

where i are the elements of Al and Mg; j are the solute elements; fi denote the activity coefficients of Al and Mg; \(e_{i}^{j}\) represents the first-order interaction coefficients, the values of which are displayed in Table 2.

Table 2 \(e_{{{\text{Al}}}}^{j}\) and \(e_{{{\text{Mg}}}}^{j}\) in Molten Steel at 1600 °C24,25
Full size table

The relationship between the Al and Mg content during the formation of different deoxidation inclusions can be derived from Eqns. 4–9:

$$ {\text{MgAl}}_{2} {\text{O}}_{4} - {\text{Al}}_{2} {\text{O}}_{3} :\left[ {\% {\text{Mg}}} \right] = 3.347 \times 10^{ - 4} \left[ {\% {\text{Al}}} \right] $$
(10)
$$ {\text{MgO - MgOAl}}_{2} {\text{O}}_{4} :\left[ {\% {\text{Mg}}} \right] = 1.253 \times 10^{ - 2} \left[ {\% {\text{Al}}} \right] $$
(11)

Based on Eqns. 10 and 11, the relationship between Mg and Al contents and deoxidation inclusions in the tested steel at 1600 °C is shown in Fig. 3. As shown in Fig. 3, deoxidation products are categorized into three sections: Al2O3, MgAl2O4 and MgO. At 1600 °C, the composition of No. 2 steels (0.0007  wt.% Mg) and No. 3 (0.0014  wt.% Mg) is situated in the formation region of MgAl2O4, resulting in a large amount of MgAl2O4. The composition of No. 4 steel (0.0023  wt.% Mg) lies in the MgO region, resulting in a substantial formation of MgAl2O4 and a minor amount of MgO.

Figure 3
figure 3

Relationship between magnesium content and aluminum content in the tested steels at 1600°C.

Full size image

Solidification Structure of Tested Steels with Different Magnesium Contents

The statistical results of the solidification structure and average grain size of tested steels with varying Mg content are shown in Figs. 4 and 5. From Fig. 4, the solidification structures of tested steels with varying Mg contents are all equiaxed grains, and the size of these equiaxed grains diminishes progressively with magnesium content increasing. Specifically, the average size of the equiaxed grains in the tested steels decreases from 320 to 271 μm, 195 and 101 μm as the Mg content increases from 0 to 0.0007  wt.%, 0.0014 and 0.0023 wt.%, respectively. Therefore, it can be observed that in the tested steels with Mg content of 0 to 0.0023 wt.%, the solidification structure of high-titanium high-strength steel is significantly refined with a cooling rate of 0.17 °C/s. This phenomenon is due to the evident desulfurization and deoxidation effect of Mg. The introduction of even a small amount of Mg into the melt readily reacts with residual oxygen, leading to the formation of a large number of dispersed high-melting-point oxides and magnesium-aluminate spinel. Such inclusions can increase the proportion of equiaxed crystals and promote grain refinement,26 attributing to the decreasing critical undercooling degree and increasing nucleation-growth interval of δ-Fe after Mg treatment.27

Figure 4
figure 4

Macro-solidification structures of cylindrical samples with different contents of Mg: (a) 0 wt.% Mg; (b) 0.0007 wt.% Mg; (c) 0.0014 wt.% Mg; (d) 0.0023 wt.% Mg.

Full size image
Figure 5
figure 5

Statistical results of the average equiaxed grain size for tested steels with different Mg contents.

Full size image

TiN Morphology and Microstructure of Tested Steels with Different Magnesium Contents

The metallographic structure and TiN morphology of the central region of each tested steel with different magnesium contents are shown in Fig. 6. From Fig. 6a, the microstructure of the No. 1 steel consisted mainly of coarse-grained polygonal ferrite and pearlite with large polygonal TiN particles with sharp edges in the matrix. As the magnesium content increases from 0 to 0.0007 wt.%, 0.0014 and 0.0023 wt.%, there is a significant reduction in polygonal ferrite and pearlite size of the tested steel structure, while acicular ferrite (AF) emerges concurrently. This phenomenon is likely due to the following reasons: on the one hand, the high-titanium content in the steel, which readily binds with carbon and nitrogen to form carbonitrides, effectively impedes the diffusion of carbon and the precipitation of carbides; on the other hand, the addition of Mg leads to the formation of numerous fine acicular ferrites. An increasing Mg content results in more TiN precipitates while reducing their size, further increasing the acicular ferrite quantity. These fine ferrites can effectively slice through and refine the large ferritic grains.28,29 Additionally, the main reason for the apparent increase in the number of TiN precipitates and the reduction in their size in the central region of tested steels is likely due to the increased levels of microalloying solutes Ti and N in the steel with higher Mg contents. This increase creates a higher supersaturation and makes it easier to reach the nucleation undercooling of TiN, thus increasing the driving force for TiN nucleation, making precipitation easier.30

Figure 6
figure 6

Typical metallographic morphology of TiN in the tested steels with different Mg contents at 500× magnification: (a) 0 wt.% Mg; (b) 0.0007 wt.% Mg; (c) 0.0014 wt.% Mg; (d) 0.0023 wt.% Mg.

Full size image

The microstructure and TiN morphology in the central region of tested steels with different Mg contents were investigated. As shown in Figs. 7, 8 and 9, scanning electron microscopy revealed many Al2O3–TiN composite inclusions, MgO–TiN composite inclusions and MgAl2O4–TiN composite inclusions in the central region of tested steels. All three types of heterogeneous cores are less than 1 μm. In Fig. 7a, c, e, g, the sizes of heterogeneous nucleated cores such as Al2O3, MgO and MgAl2O4 are 0.68, 0.42, 0.34 and 0.44 μm, respectively. The single-TiN particles are defined as single TiN, and the TiN particles with other inclusions in the interior or edge are defined as composite TiN. The Al2O3 surfaces facilitate heterogeneous nucleation, leading to the formation of Al2O3-TiN composite inclusions.31 The MgO, MgAl2O4 and TiN all exhibit a face-centered hexagonal crystal system. The mismatch between them can be calculated using Eqn. (12), where s and n represent the substrate and nucleating phase, respectively.32 The (hkl), [uvw] and d[uvw] represent a low-index crystal plane in the substrate or nucleating phase, a low-index direction on the (hkl) crystal plane, and the interatomic spacing along the [uvw] direction, respectively. The θ represents the angle between [uvw]s and [uvw]n. According to Eqn. (12) and the relevant data in Table 3, the mismatch between MgO and TiN can be calculated as 0.02%, while the mismatch between MgAl2O4 and TiN is only 5.03%. Generally, when the mismatch between the two phases is less than 12%, the first precipitated compound phase can act as a non-spontaneous nucleus to promote nucleation. The high-melting-point MgO and MgAl2O4 preferentially precipitate and can serve as a heterogeneous nucleating core for TiN due to its low mismatch with TiN. During the solidification process, Ti segregation results in significantly higher Ti concentrations at the solidification front far exceeding the critical Ti concentration required for the heterogeneous nucleation of TiN.31 Consequently, TiN readily nucleates around MgO and MgAl2O4 cores in large quantities during solidification (Tables 4 and 5).

$$ \delta_{{(hkl)_{n} }}^{{(hkl)_{s} }} = \sum\limits_{i = 1}^{3} {\left[ {\frac{{\left| {d_{{\left[ {uvw} \right]_{s} }}^{i} \cos\uptheta - d_{{\left[ {uvw} \right]_{n} }}^{i} } \right|}}{{d_{{\left[ {uvw} \right]_{n} }}^{i} }}} \right]} \times \frac{1}{3} \times 100\% $$
(12)
Figure 7
figure 7

Morphologies and energy spectrum analysis of composite TiN in the tested steels with varying Mg contents: (a, b) 0 wt.% Mg; (c, d) 0.0007 wt.% Mg; (e, f) 0.0014 wt.% Mg; (g, h) 0.0023 wt.% Mg.

Full size image
Figure 8
figure 8

Morphology and line scanning analysis of MgAl2O4–TiN composite inclusions in 0.0023 wt.% Mg steel: (a) SEM morphology; (b) Ti spectrum; (c) N spectrum; (d) Fe spectrum; e Mg spectrum; (f) O spectrum; (g) Al spectrum.

Full size image
Figure 9
figure 9

Morphology and line scanning analysis of MgO–TiN composite inclusions in 0.0023 wt.% Mg steel: (a) SEM morphology; (b) Ti spectrum; (c) N spectrum; (d) Fe spectrum; € Mg spectrum; (f) O spectrum.

Full size image
Table 3 Crystallographic Parameters of MgO, MgAl2O4 and TiN33
Full size table
Table 4 Calculation Results of Mismatch Degree Between (111) Surface of MgO and (111) Surface of TiN
Full size table
Table 5 Calculation Results of Mismatch Degree Between (111) Surface of MgAl2O4 and (111) Surface of TiN
Full size table

Quantity, Size and Precipitation Location of TiN (≥ 1 μm) in the Central Region of Tested Steels with Different Magnesium Contents

The statistical results of the quantity, size and precipitation location of TiN (≥ 1 µm) in the central region of tested steels with different Mg contents are shown in Fig. 10. From Fig. 10a–d, in the central region of No. 1 steel (0 wt.% Mg), the size of TiN particles is primarily concentrated in the range of 5–15 μm, with a higher density in the 5–12 μm range (density of TiN 44.4 No./mm2). For the tested steels with Mg content ranging from 0.0007 to 0.0023 wt.%, there was a sharp increase in density of TiN (density of TiN 102.4–129.4 and 144.2 No./mm2). The size distribution mainly ranged between 1–8 μm, with higher densities observed within the range of 1–5 μm. No large-sized TiN particles exceeding 12 μm were found, and single TiN exhibited significantly higher densities than composite-TiN in four tested steels. As shown in Fig. 10e, the maximum size of TiN in the central region of tested steels changes from 16.5 to 10.4 μm, 9.8 and 9.3 μm as the Mg content increases from 0 to 0.0007 wt.%, 0.0014 and 0.0023 wt.%, respectively. The average size of TiN decreases from 9.1 to 4.2 μm, 3.3 and 3.1 μm, respectively. The precipitation of TiN in the central region of tested steel is mainly observed at grain interior, with a decrease in proportion precipitated at grain boundaries with increasing Mg content, ranging from 22.7 to 16.8%, 14.5 and 14.8%. The main reason is that larger grain interior area compare to the grain boundary area. Ti and N atoms diffuse, aggregate and continuously diffuse into MgO/MgAl2O4, creating more space and nucleation points at solidification defects. Thereby, TiN nucleation propensity is further heightened at grain interior.34

Figure 10
figure 10

Statistics of single-TiN and composite-TiN density and size in tested steels at different Mg contents: (a) 0 wt.% Mg; (b) 0.0007 wt.% Mg; (c) 0.0014 wt.% Mg; (d) 0.0023 wt.% Mg; (e) Maximum and average sizes of TiN.

Full size image

Nucleation and Growth Mechanisms of TiN in the Central Region of Tested steels During Solidification Cooling Process

The nucleation, growth and aggregation processes of TiN in the solid-liquid two-phase region of Nos. 1–4 steels with different Mg contents are shown in Fig. 11. Owing to the equilibrium partition coefficients of Ti and N atoms in solid and liquid phases being significantly less than 1, the solubility of Ti and N in the δ-phase is markedly lower than that in the liquid phase.35 Due to a consistently low cooling rate (0.17 °C/s) in four tested steels, heat dissipation in the central region of tested steels lacks directionality, allowing grain to grow freely within the liquid phase at similar rates on all sides. This results in the growth of δ equiaxed grains during solidification. In the nucleation and growth processes of the δ equiaxed grains at the solidification front, as both TiN and δ-phase belong to cubic crystal systems, δ equiaxial grains readily form on the surface of TiN particles due to their favorable lattice mismatch relationship with δ-Fe (3.9% at 1538 °C).19,36 The formation of these equiaxial grains contributes to refine grain size within the δ-phase. With the addition of Mg, the rapid vaporization of Mg in molten steel forms small bubbles that stir the molten steel during their rise and swiftly generate fine MgAl2O4 or MgO inclusions. Due to the favorable interface wetting characteristics of TiN and MgO/MgAl2O4, as well as the minimal lattice mismatch between TiN and MgO/MgAl2O4, the presence of MgO and MgAl2O4 in molten steel can enhance the nucleation rate of TiN. This leads to easier attainment of the critical concentration required for TiN nucleation around nucleating particles by facilitating a higher concentration of Ti and N atoms. Consequently, the growth driving force acting on individual TiN particles diminishes, resulting in an overall reduction in both average and maximum size of TiN growth. With the increase in Mg content in the tested steels, a higher amount of MgO and MgAl2O4 were introduced as nucleation particles, thereby inducing a significant number of spontaneous TiN nucleation and heterogeneous nucleation with MgO and MgAl2O4 as the core. It leads to the enhancement of the nucleation rate of δ equiaxed grains. When there is a more significant number of free δ equiaxed grains at the solidification front, more TiN inclusions form within δ equiaxed grains. During δ equiaxed grains growth process, they easily connect with other equiaxed grain arms, resulting in more closed liquid phase regions. This reduces mutual flow and mixing between equiaxed crystal melts, effectively hindering migration and enrichment of Ti and N atoms in the liquid phase. Consequently, in comparison to coarse equiaxed grains, fine equiaxed grains have the ability to generate segregation regions with lower concentrations of Ti and N atoms during growth. This minimizes solidification center Ti and N segregation while avoiding the formation of large-sized TiN particles. In subsequent δ → γ phase transition and slow cooling processes of γ-phase, an increased presence of small -sized MgAl2O4 and TiN particles hinders grain growth due to their strong pinning effect on grain boundaries. The resulting grain boundary area is also increased. The barrier effect of the grain boundary can effectively impede the diffusion of Ti and N atoms across boundaries, thereby inhibiting the continuous growth of TiN. Therefore, as the Mg content in steel changes from 0 to 0.0007 wt.%, 0.00014 and 0.0023 wt.%, density of single TiN and composite TiN gradually increase for tested steels, meanwhile maximum size of TiN and average size of TiN decrease.

Figure 11
figure 11

Schematic of equiaxed grains formation and TiN nucleation and growth during the solidification process in tested steels with different Mg contents: (a) 0 wt.% Mg; (b) 0.0007 wt.% Mg; (c) 0.0014 wt.% Mg; (d) 0.0023 wt.% Mg.

Full size image

Conclusion

  1. (1)

    According to the thermodynamic analysis of deoxidation products in molten steel, the deoxidation products of magnesium and aluminum can be categorized into three intervals: the Al2O3 interval, the MgAl2O4 interval and the MgO interval. Among these intervals, the formation range of MgAl2O4 is wider, and it can be formed even with a trace amount of magnesium present. In the tested steels containing 0.0007 wt.% Mg and 0.0014 wt.% Mg, MgAl2O4 forms at [%Mg] ≥ 3.347 × 10−4 [%Al]; in the tested steel with 0.0023 wt.% Mg content, MgAl2O4 and MgO form at [%Mg] ≥ 1.253 × 10−2 [%Al].

  2. (2)

    The tested steels exhibit equiaxed grains when the solidification rate is 0.17 °C/s. The average size of equiaxed grains in each tested steel varies from 320 to 271 μm, 195 and 101 μm as the Mg contents change from 0 to 0.0007 wt.%, 0.00014 and 0.0023 wt.%.

  3. (3)

    In No. 1 steel (0 wt.% Mg), composite-TiN particles are primarily Al2O3–TiN. As Mg content increases from 0.0007 to 0.0014 wt.% and 0.0023 wt.%, composite-TiN particles primarily consist of MgAl2O4–TiN and MgO–TiN. Calculations indicate that the small lattice mismatch between MgAl2O4 (5.03%) and MgO (0.02%) with TiN is the main reason for their role in heterogeneous nucleation of TiN. In the the No. 1 steel (0 wt.% Mg), TiN particle sizes are concentrated between 5–15 μm, with the highest precipitation density in the 5–12 μm range. In contrast, the No. 2 –No. 4 steels show a significantly higher total density of TiN precipitates, with sizes mainly ranging from 1–5 μm.

  4. (4)

    As Mg content in the tested steels increases from 0 to 0.0007 wt.%, 0.0014 and 0.0023 wt.%, the total areal density of TiN precipitates in the tested steels rises from 44.4 to 102.4 No./mm2, 129.4 and 144.2 No./mm2. The maximum size of TiN decreases from 16.5 to 10.4 μm, 9.8 and 9.3 μm, and the average size of TiN reduces from 9.1 to 4.2 μm, 3.3 and 3.1 μm.