Effect of Al Substitution of Si on the Microstructure, Retained Austenite Stability and Mechanical Properties of Low-Alloyed TRIP-Aided Steels

Abstract

In this work, the effect of partial to complete Al substitution of Si on the microstructure, retained austenite (RA) stability, and mechanical properties of cold-rolled TRIP-aided steels was investigated. Four experimental TRIP-aided steels (Fe-0.2C-1.5Mn-1.5/1.0/0.5/0Si-0/0.5/1.0/1.5Al-0.025Nb, wt.%) were designed. The results indicate that replacing Si with Al significantly increases the volume fraction of soft polygonal ferrite (from 52% to 73%) and decreases that of bainite. Although the volume fraction of RA decreases (from 15.6% to 12.4%), its average carbon content and, consequently, its mechanical stability are enhanced, which suppresses the strain-induced martensitic transformation. In terms of mechanical properties, the substitution leads to a monotonic decrease in both yield strength (from 573 MPa to 536 MPa) and ultimate tensile strength (UTS) (from 839 MPa to 648 MPa), primarily due to reduced solid-solution strengthening, coarsened ferrite grains, and a weakened TRIP effect. Conversely, the total elongation (TEL) increases from 28.3% to 32.4%, attributed to the higher fraction of ductile ferrite. Consequently, the product of tensile strength and total elongation (PSE) exhibits a slight decline. The 1.5Si-TRIP steel exhibited the most balanced mechanical properties, achieving the highest PSE of 23.7 GPa·%.

1. Introduction

As typical representatives of advanced high-strength steels, transformation-induced plasticity (TRIP) steels have a good strength–ductility combination, allowing them to exhibit extensive applications in the automotive industry [1,2,3,4]. It is well known that traditional TRIP steels are developed from Fe-C-Mn-Si alloy systems [5]. However, high Si addition will result in a strong oxide layer on the surface of hot-rolled TRIP steel plates and cause galvanizing problems [6,7]. Therefore, it is crucial to find elements that can replace Si in TRIP steels.

Al is considered an ideal substitution for Si due to their chemically similar behavior [6,7]. Its addition effectively suppresses cementite precipitation without deteriorating the coating ability of TRIP steel [8,9]. Consequently, the influence of partial or complete replacement of Si with Al in low-alloyed TRIP steels has been widely studied over the past two decades [7,8,9,10]. Research findings, however, have shown variations: El-Sherbiny et al. reported a decrease in ultimate tensile strength (UTS) and the product of tensile strength and total elongation (PSE) with the complete replacement of Si with Al [7], whereas Zhu et al. observed that TRIP steel with added Al achieved a higher PSE than its counterpart with Si added when the isothermal bainitic transformation (IBT) temperature exceeded 450 °C [8]. Similarly, Barbé et al. noted that Al substitution of Si leads to improved galvanizing properties but at the expense of some strength [9]. Conversely, Wang et al. demonstrated that the partial replacement of Si with Al significantly enhances total elongation (TEL) and PSE in Nb-Mo microalloyed TRIP steel [10]. These conflicting results indicate that the effect of partially or fully substituting Al for Si on the mechanical properties of TRIP steels remains unclear and necessitates further systematic investigation.

While understanding these variations in mechanical properties is important, the stability of retained austenite (RA)—a more fundamental aspect—has been largely overlooked in studies on Si/Al substitution in TRIP steels. Previous studies have primarily focused on the effect of this substitution on microstructure and mechanical properties, neglecting its critical effect on RA stability, which is fundamental to the TRIP effect [11,12,13]. To address this gap, four experimental TRIP steels (Fe-0.2C-1.5Mn-1.5/1.0/0.5/0Si-0/0.5/1.0/1.5Al-0.025Nb, wt.%) were designed to systematically study the effects of partial to complete Al substitution of Si on the microstructure, RA stability, and mechanical properties of TRIP steel.

2. Experimental

The chemical compositions of the experimental steels are given in

Table 1. Chemical composition (wt.%) of the experimental steels.

Steel	C	Mn	Si	Al	Nb	Fe
1.5Si Steel	0.21	1.48	1.40	0	0.025	Bal.
1.0Si0.5Al Steel	0.20	1.45	0.98	0.44	0.022	Bal.
0.5Si1.0Al Steel	0.19	1.43	0.53	0.94	0.022	Bal.
1.5Al Steel	0.20	1.50	0	1.36	0.028	Bal.

. First, 50 kg ingots were prepared by vacuum induction melting, followed by homogenization at 1200 °C for 2 h and hot forging into rectangular slabs. Subsequently, the slabs were reheated to 1200 °C for 2 h, hot-rolled to 3 mm sheets, and finally cold-rolled to a final thickness of 1 mm.

A two-stage salt bath heat treatment was applied. Samples were first austenitized by immersion in a salt bath at 825 °C for 3 min, then rapidly quenched by transfer to a second salt bath held at the IBT temperature of 410 °C for 3 min, followed by air cooling (AC) outside the furnace to room temperature (Figure 1). Based on their respective Si and Al contents, the resulting heat-treated steels are designated as the 1.5Si-TRIP, 1.0Si0.5Al-TRIP, 0.5Si1.0Al-TRIP, and 1.5Al-TRIP samples. Flat dog-bone tensile specimens with a gauge section of 25 mm × 6 mm were machined parallel to the rolling direction (Figure 2). Uniaxial tensile tests were performed at an initial strain rate of 1.3 × 10⁻³ s⁻¹ using a SANSCMT-5000 universal testing machine (MTS, Eden prairie, MN, USA).

Microstructural characterization was carried out employing optical micrograph (OM), field-emission scanning electron microscopy (FE-SEM), and transmission electron microscopy (TEM). For OM and SEM analysis, specimens were electropolished and then etched with 4% Nital or LePera reagent (prepared by mixing equal volumes of 1% aqueous Na₂S₂O₅ and 4% alcoholic picric acid). Eight representative micrographs were captured for each condition, and the volume fraction of ferrite (V_F) was quantified via a pixel-based method using Image-Pro Plus 6.0 software. Concave polygonal ferrite grains were manually segmented and colored in Photoshop, after which the total ferrite area (A_F) and the total micrograph area (A_T) were measured. This process was repeated eight times, and V_F was calculated as V_F = ΣA_Fi / ΣA_Ti (i = 1 − 8). The average ferrite grain size (D_F) was obtained from fifty randomly selected grains. Electron back-scatter diffraction (EBSD) was conducted at 20 kV with a step size of 50 nm. For TEM, foils were mechanically ground to approximately 50 µm, punched into 3 mm discs, and twin-jet-electropolished at −20 °C using a Struers Tenuol-5 unit with an electrolyte of 95% ethanol and 5% perchloric acid. The volume fraction (Vγ) and carbon content of RA were measured by X-ray diffraction (XRD, Rigaku D/Max2250/PC, Cu Kα radiation). Specifically, Vγ was determined via the direct comparison method [14], while the carbon (C) content was evaluated from the lattice parameter based on reference [15].

3. Results and Discussion

3.1. Heat Treatment Schedule

To select an appropriate intercritical annealing (IA) temperature, the equilibrium phase fractions of the experimental steels were simulated as a function of temperature using JMatPro. As shown in Figure 1, substituting Si with Al significantly shifted the critical transformation temperatures: the Ae1 temperature decreased from 707 °C to 691 °C, while the Ae3 temperature increased from 850 °C to 980 °C (Figure 3). This trend indicates a substantial expansion of the α + γ two-phase region with increasing Al content. To isolate the effect of this substitution on the resulting microstructure and properties, all steels were subjected to the same IA treatment at 825 °C.

The IBT temperature, along with the IA temperature, is a critical parameter for cold-rolled TRIP steels, as it dictates the grain size, morphology, carbon content, and volume fraction of RA. To clarify the microstructural evolution during bainitic transformation, time–temperature transformation (TTT) curves for the metastable austenite of the four experimental steels were calculated using JMatPro (Figure 4). Al substitution of Si increased the bainite start (B_s) temperature from 540 °C to 611 °C and shifted the nose of the bainitic bay from approximately 450 °C to about 550 °C. This substitution also elevated the martensite start (M_s) temperature from 376 °C to 402 °C (Table 2). Selecting an IBT temperature too close to this nose temperature accelerated bainite formation, which reduced the RA content at room temperature. Conversely, an IBT temperature below M_s led to the formation of martensite. Both conditions degraded the strength–ductility balance. Therefore, an IBT temperature of 410 °C was chosen in this study to optimize the microstructure and mechanical properties.

Table 2. B_s, M_s, and M_f temperature calculated by JMatPro for the experimental steels.

Steel	Bs/°C	Ms/°C	Mf/°C
1.5Si Steel	540	376	263
1.0Si0.5Al Steel	556	388	276
0.5Si1.0Al Steel	580	400	289
1.5Al Steel	611	402	292

Table 2. B_s, M_s, and M_f temperature calculated by JMatPro for the experimental steels.

Steel	Bs/°C	Ms/°C	Mf/°C
1.5Si Steel	540	376	263
1.0Si0.5Al Steel	556	388	276
0.5Si1.0Al Steel	580	400	289
1.5Al Steel	611	402	292

In addition to temperature, the holding times of IA and IBT significantly influence the final properties of TRIP steels. Wang et al. [16] demonstrated that an IA time of 1–3 min is sufficient to achieve an optimal strength–ductility balance. Similarly, Zhang et al. [17] reported that an IBT time of 3–4 min represents the most favorable processing window. A shorter IBT time leads to carbon-depleted RA with low mechanical stability, whereas prolonged holding promotes excessive bainite formation at the expense of austenite content—both scenarios suppress the TRIP effect. Accordingly, in this study, both the IA and IBT time were set to 3 min. The corresponding heat treatment schedule is illustrated in Figure 1.

3.2. Microstructure

Figure 5 shows the OM images of the four experimental steels after cold rolling. The microstructure of all samples is primarily composed of ferrite (the white phase) and pearlite (the black phase). Both constituents are markedly elongated along the rolling direction, forming a distinct banded structure. It is evident that substituting Si with Al leads to an increase in the volume fraction of ferrite and a corresponding decrease in pearlite.

Figure 6 shows the OM images of the four experimental steels after the two-stage heat treatment. The banded structure was eliminated in all samples, resulting in microstructures consisting of uniformly distributed white blocky polygonal ferrite, along with light gray bainite and RA. Both the volume fraction and size of the polygonal ferrite increase with Al content. This is attributed to the role of Al as a ferrite (α) stabilizer, which expands the α-phase field and contracts the austenite (γ) region (Figure 3). Therefore, during intercritical annealing at 825 °C for 3 min, a higher Al content promotes the formation of more ferrite and less initial austenite. This lower initial austenite content, in turn, results in reduced bainite content after the IBT transformation.

Figure 7 shows the SEM micrographs of the four experimental steels after the two-stage heat treatment. The microstructures primarily consist of polygonal ferrite and granular bainite, with RA (white phase) uniformly distributed both within the ferrite matrix and along the grain boundaries. The quantitative analysis shown in Figure 8 confirms that the volume fraction of ferrite increases systematically from 52% to 73% as Si is substituted with Al: specifically, 52% (1.5Si), 58% (1.0Si0.5Al), 68% (0.5Si1.0Al), and 73% (1.5Al).

To elucidate the effect of Al substitution of Si on the characteristics of RA, EBSD analysis was performed on four experimental steels subjected to the two-stage heat treatment. The corresponding phase maps and RA size distribution histograms are presented in Figure 9. In the phase maps, the α-bcc phase and RA are colored blue and red, respectively, while the white and black lines denote low-angle (2–15°) and high-angle (>15°) grain boundaries. As shown in Figure 9a,c,e,g, with Al substitution of Si, the RA in all steels primarily exhibits a granular or blocky morphology, with only a minor lath-like fraction. The RA is predominantly distributed along ferrite grain boundaries, with a small amount located within the grains, consistent with the SEM observations. As summarized in Figure 9b,d,f,h, the average RA grain size ranges from 0.36 to 0.45 μm.

To further elucidate RA morphology, TEM analysis was conducted. As shown in Figure 10, the RA in the four experimental steels predominantly exists as blocky or granular particles located at the ferrite grain boundaries, with a smaller fraction found within the grains, in good agreement with the SEM and EBSD results.

To further investigate the influence of Al substitution for Si on the volume fraction, carbon content, and mechanical stability of RA, XRD analyses were performed on the four experimental steels before and after tensile testing. The resulting data for the RA volume fraction and its average carbon content are presented in Figure 11. As Si is replaced by Al, the volume fraction of RA (Vγ) decreases from 15.6% to 12.4%, while its carbon content (Cγ) increases from 1.0 wt.% to 1.4 wt.% (Figure 11b,c). After tensile deformation, the RA contents in the four experimental steels drop to 5.4%, 5.0%, 4.7%, and 5.1%.

Austenite stability can be categorized into thermal stability and mechanical stability. This study focuses on the mechanical stability of RA, which reflects its tendency to transform into martensite under plastic deformation. The mechanical stability parameter k was quantitatively evaluated using the following equation [18]:

$$f_{\mathsf{\gamma }}=f_{{\mathsf{\gamma }}_0}\exp (-k\epsilon ),$$

where $f_{{\mathsf{\gamma }}_0}$ is the initial volume fraction of RA, $f_{\mathsf{\gamma }}$ is the volume fraction of RA at a true strain of ε, and k is the mechanical stability constant. A higher k value indicates a lower mechanical stability of RA [19]. Figure 12 shows the calculated k values for the four experimental steels after the two-stage heat treatment. As Si is replaced by Al, the k value decreases from approximately 4.0 to 3.2, indicating an enhancement in the mechanical stability of the RA. This enhanced stability is primarily attributed to the increased carbon content in the RA, which rises from 1.0 wt.% to 1.4 wt.% with Al addition (Figure 11c). As widely reported [20,21,22,23], the mechanical stability of RA is decided by its carbon content, grain size, and morphology. Specifically, a higher carbon content generally results in greater stability [24]. Consequently, the improved stability leads to a reduction in the amount of strain-induced martensitic transformation from 10.2% to 7.3% after tensile deformation (Figure 11b).

Figure 13 presents the engineering stress–strain curves of the four experimental steels after the two-stage heat treatment, and the corresponding mechanical properties are summarized in

Table 3. Mechanical properties of the experimental steels after two-stage heat treatment.

Steel	YS/MPa	UTS/MPa	EL/%	PSE/GPa·%
1.5Si-TRIP sample	573	839	28.3	23.7
1.0Si0.5Al-TRIP sample	543	764	30.4	23.2
0.5Si1.0Al-TRIP sample	541	693	32.3	22.4
1.5Al-TRIP sample	536	648	32.4	21.0

. As Si is replaced by Al, both the yield strength (YS) and UTS decrease monotonically from 573 MPa to 536 MPa and from 839 MPa to 648 MPa, respectively. In contrast, the TEL increases from 28.3% to 32.4%. Consequently, the product of UTS and TEL (PSE) shows a slight decline from 23.7 GPa·% to 21.0 GPa·%.

The balanced mechanical properties of TRIP steels arise from their three-phase microstructure of soft ferrite, hard bainite, and metastable RA. This strength–ductility synergy depends on the volume fractions of these phases and the mechanical stability of the RA. In the present experimental steels, the systematic substitution of Si with Al modifies both the phase fractions and the RA stability, thus controlling the overall mechanical response.

The YS of the experimental steels is described by the following equation [25]:

$$YS={\sigma }_0+{\sigma }_{\mathrm{s}}+{\sigma }_{\mathrm{g}}+{\sigma }_{\mathrm{p}}+{\sigma }_{\mathrm{d}},$$

where ${\sigma }_0$ is the lattice friction stress and ${\sigma }_{\mathrm{s}}$, ${\sigma }_{\mathrm{g}}$, ${\sigma }_{\mathrm{p}}$, and ${\sigma }_{\mathrm{d}}$ represent the strengthening increments from solid-solution, grain-refinement, precipitation, and dislocation hardening, respectively. The observed decrease in YS in the high-Al steels primarily stems from the concurrent weakening of two key strengthening mechanisms. First, the replacement of Si with Al reduces the solid-solution strengthening contribution (${\sigma }_{\mathrm{s}}$), as Al imparts a weaker strengthening effect compared to Si [8,26] Second, the ferrite grain size increases with Al addition, which diminishes the grain boundary strengthening contribution (${\sigma }_{\mathrm{g}}$). The combined reduction in both ${\sigma }_{\mathrm{s}}$ and ${\sigma }_{\mathrm{g}}$ thereby accounts for the overall lower YS of these steels.

The decline in UTS of the experimental steels is primarily due to the synergistic effect of three key microstructural and compositional changes. Firstly, the Al substitution for Si leads to a reduction in YS. Concurrently, a significant microstructural transformation occurs, characterized by an increase in the volume fraction of the soft ferrite phase from 52% to 73%, coupled with a decrease in the hard bainite phase from 32.4% to 14.6% (Figure 14). Furthermore, the stability of the RA is enhanced as its carbon content rises from 1.0 wt.% to 1.4 wt.%, which suppresses the strain-induced transformation from 10.2% to 7.3% during deformation. Consequently, the contribution of the TRIP effect to strengthening is significantly diminished, collectively accounting for the reduction in UTS.

Notably, despite the attenuated TRIP effect, the TEL of the experimental steels increases from 28.5% to 32.4% with Al’s substitution of Si. This enhancement in ductility is primarily attributed to a concurrent microstructural evolution characterized by a significant increase in the volume fraction of the soft, ductile ferrite phase (from 52% to 73%) and a corresponding decrease in the hard bainite phase (from 32.4% to 14.6%). The enlarged ferrite fraction provides an enhanced capacity for accommodating plastic strain, effectively compensating for the reduced TRIP effect and thereby leading to the improved TEL.

4. Conclusions

In this study, the effects of Al substitution of Si on the microstructure evolution, RA stability, and mechanical properties of low-alloy TRIP-aided steels were investigated. The primary conclusions are summarized as follows:

• With Al substituting for Si, the volume fraction of ferrite increased from 52% to 73%, while that of bainite decreased from 32.4% to 14.6%.
• The substitution of Si by Al reduced the volume fraction of RA from 15.6% to 12.4%, but increased its carbon content and mechanical stability.
• Al substitution in place of Si led to a decline in both YS (from 573 MPa to 536 MPa) and UTS (from 839 MPa to 648 MPa). In contrast, the TEL increased from 28.3% to 32.4%. Consequently, the 1.5Si-TRIP steel achieved an excellent combination of mechanical properties, characterized by a PSE of 23.7 GPa·%.