Achieving Superior Ductility at High Strain Rate in a 1.5 GPa Ultrahigh-Strength Steel without Obvious Transformation-Induced Plasticity Effect

Abstract

The development of ultrahigh-strength steels with good ductility is crucial for improving the crashworthiness of automobiles. In the present work, the mechanical responses and deformation behaviors of 1.5 GPa ultrahigh-strength steel were systematically investigated over a wide range of strain rates, from 10⁻³ s⁻¹ to 10³ s⁻¹. The yield strength and tensile elongation at quasi-static strain rate (10⁻³ s⁻¹) were 1548 MPa and 20%, respectively. The yield strength increased to 1930 MPa at an extremely high strain rate (10³ s⁻¹), and the steel maintained excellent ductility, with values as high as 17%. It was found that the prevailing of the transformation-induced plasticity (TRIP) effect at quasi-static condition resulted in the formation of fresh martensite. This produced strong hetero-deformation-induced (HDI) stress and strain partitioning, contributing to the enhancement of strain hardening. The TRIP effect is remarkably suppressed under high strain rates, and thus the retained austenite with excellent deformation ability sustains the subsequent deformation, leading to superior ductility when the TRIP effect and HDI strengthening are retarded. Ultrahigh-strength steel with great strength–ductility combination over a wide range of strain rates has great potential in improving component performance while reducing vehicle weight.

1. Introduction

Driven by the demands to improve crashworthiness and reduce weight in the automobile industry, numerous efforts have been made to develop steels with high strength and good formability at low cost [1,2]. Advanced high-strength steels (AHSSs) are desirable for the automobile manufacturing of critical anti-collision components such as A- and B-pillars and crash-boxes [3,4]. The excellent strength–ductility combination of third-generation AHSSs mainly depends on the transformation-induced plasticity (TRIP) effect of retained austenite [5,6]. In the event of a catastrophic accident, automobile collision is a complex and highly non-linear deformation process. When a vehicle collides at a velocity of 60–80 km/h, these crashworthy components are subjected to large impact loads at high strain rates (10² s⁻¹–10³ s⁻¹) [7,8]. Thus, it is of critical significance in providing anti-intrusion barriers to achieve superior comprehensive mechanical properties over a broad range of strain rates.

Extensive research on strain-rate-dependent deformation behavior is mainly focused on the effects of strain rates on overall mechanical properties. In general, high-strain-rate (>10³ s⁻¹) deformation induces both the strain rate hardening effect and the adiabatic softening effect [9]. The microstructural evolution and corresponding mechanical properties depend on the competition between both effects. For example, the yield strength of PHS2000 increases by ~149 MPa over strain rates from 470 to 1450 s⁻¹ due to larger friction stress for dislocation slip [1]. V-containing TRIP steels [10] and austenitic stainless steels [11] show a sudden reduction in tensile elongation under dynamic tensile loading (~2000 s⁻¹) compared to quasi-static tension. Wang et al. [12] reported that increasing the strain rate to 10³ s⁻¹ significantly lowered the strain-hardening ability of quenching–partitioning steel owing to the inhibited dislocation multiplication in the martensite matrix. In terms of medium Mn steels with a yield strength of 1500 MPa, there is limited literature associated with high-strain-rate deformation. This lack can be attributed to the complexity of microstructures and intricate deformation mechanisms. Sevsek et al. [13] conducted tensile tests over strain rates from 10⁻⁵ s⁻¹ to 230 s⁻¹ and found that the TRIP effect is partially suppressed at very low and very high strain rates in medium Mn steels. The effect of strain rate on the plastic anisotropy of Fe-0.16C-10Mn-1.56 Al steel has been studied, while the microstructural evolution at various strain rates has not been discussed [10]. Thus, the in-depth understanding of high-strain-rate deformation behaviors in ultrahigh-strength steel is needed.

Investigations have been conducted to optimize the comprehensive mechanical properties of high-strength steel at high strain rates, such as grain refinement [14] and the addition of Si [15]. Also, it has been demonstrated that introducing heterogeneous microstructures like microstructural heterogeneity [16] and chemical heterogeneity [17] is an effective approach to improve strength without significantly sacrificing ductility [18]. Mechanical incompatibility between adjacent domains with different strengths leads to the generation of geometrically necessary dislocations, resulting in strong heterogeneous deformation-induced (HDI) strengthening [19,20]. Dynamic compressive experiments with a wide range of strain rates (10⁻³ s⁻¹ to 5 × 10³ s⁻¹) were undertaken in multiple heterostructures (grain and component heterogeneities); however, the relationship between HDI strengthening and mechanical properties was not discussed [20]. Thus, the effect of designing heterogeneous microstructure on improving the strength–ductility combination at high strain rates is still not clear. The deformed and partitioned (D&P) steel in our previous study exhibited extremely high strength and ductility as well as fracture toughness [21,22] at a quasi-static strain rate. It is of great importance for improving the impact resistance of automotive steel and passenger protection to explore the mechanical response at high strain rates in this ultrahigh-strength steel.

In this work, taking D&P steel with multiphase microstructure as a model material, the strain-rate-dependent mechanical properties and deformation mechanisms over a wide range of strain rates from 10⁻³ s⁻¹ to 10³ s⁻¹ were investigated. The effect of strain rates on microstructural evolution during deformation as well as the relationship between heterostructures and comprehensive mechanical properties are discussed in details. The present D&P steel exhibits great strength–ductility synergy over a wide range of strain rates from 10⁻³ s⁻¹ to 10³ s⁻¹.

2. Materials and Methods

The chemical composition of the investigated D&P steel is shown in

Table 1. Chemical composition (wt. %).

C	Mn	Al	V	Si	Fe
0.47	10	2.0	0.7	0.05	Bal.

. The investigated steel was cast in a vacuum furnace and forged into a 30 kg ingot. The as-received ingot was homogenized at 1150 °C for 2 h, followed by cooling, to obtain a 4 mm-thick hot-rolled plate with a fully austenitic microstructure. This is due to the fact that the material composition of this work possesses a low finish temperature for martensite transformation. Subsequently, multistep treatment was carefully designed to introduce heterogeneous multiphase in ultrahigh-strength steel. The hot-rolled plate was reheated to 750 °C, held for 30 min, and warm-rolled with a reduction ratio of 4 mm to 2 mm. The warm-rolled sheet was subjected to intercritical annealing at 625 °C for 5 h, followed by water-quenching to ambient temperature. After that, the cold-rolling process (15% thickness reduction) introduced defects in austenite to further improve yield strength. The final sheet thickness was 1.67 mm. Next, the cold-rolled sheet was tempered at 350 °C for 6 min to promote the carbon enrichment of austenite during the partitioning step.

The initial and deformed microstructures were observed by electron backscattering diffraction (EBSD) characterization, performed on a Crossbeam 350 scanning electron microscope equipped with an Oxford detector. EBSD scanning was performed under an operating voltage of 20 kV with a step size of 0.05 μm. The phase fraction was determined by X-ray diffraction (XRD) using Cu-Kα radiation (λ = 1.54 $Å$), and scanning was performed over the 2θ range from 30° to 10° with a scanning speed of 1° min⁻¹ and a step size of 0.05°. The samples for EBSD and XRD were firstly mechanically polished and electro-polished in a solution of 6 vol% perchloric and 94 vol% glacial acetic acid. The volume fraction of austenite was measured using the intergrade intensities of the (200)_γ, (222)_γ and (311)_γ peaks of austenite and (211)_α and (200)_α peaks of ferrite or martensite. The austenite volume fraction, V_γ, is calculated using the following equation [23]:

$$V_{\gamma }={\displaystyle \frac{I_{\gamma }/R_{\gamma }}{I_{\gamma }/R_{\gamma }+I_{\alpha }/R_{\alpha }}}$$

(1)

where R is the material scattering factor depending on the scattering angle (θ), and I_α and I_γ are the average intensity of ferrite and austenite diffraction peaks, respectively.

To symmetrically study the mechanical properties of the investigated steel, macroscopic mechanical tests were performed within the strain rate range 10⁻³ s⁻¹ ≤ $\dot{\epsilon }$≤ 10³ s⁻¹. Both quasi-static and intermediate-strain-rate tensile tests were conducted in a high-performance test system (MTS) at strain rates from 10⁻³ s⁻¹ to 1 s⁻¹, and a mechanical extensometer was used. Tensile dog-bone-shaped samples were cut along the rolling direction with the gauge dimensions of 12 mm × 4 mm. An extensometer was used to measure the strain accurately during the tensile deformation. An improved split Hopkinson tension bar was carried out with dynamic tensile tests with gauge dimensions of 10 mm × 4 mm; technical details about this apparatus can be found in Refs. [24,25]. The engineering stress (σ_e), engineering strain (ε_e) and strain rate (${\dot{\epsilon }}_e$) were determined by the elastic waves recorded by strain gauges mounted on the incident and transmitted bars as follows:

$${\mathsf{\sigma }}_{\mathrm{e}}=\left({\mathrm{A}}_{\mathrm{b}}/{\mathrm{A}}_{\mathrm{s}}\right)\cdot {\mathrm{E}}_{\mathrm{b}}\cdot {\mathsf{\epsilon }}_{\mathrm{b},\mathrm{t}}\left(\mathrm{t}\right)$$

(2)

$${\mathsf{\epsilon }}_{\mathrm{e}}=\left(-2{\mathrm{C}}_{\mathrm{b}}/{\mathrm{L}}_{\mathrm{s}}\right)\cdot {\int }_0^{\mathrm{t}}{\mathsf{\epsilon }}_{\mathrm{b},\mathrm{r}}\left(\mathrm{t}\right)\mathrm{dt}$$

(3)

$${\dot{\mathsf{\epsilon }}}_{\mathrm{e}}=\left(-2{\mathrm{C}}_{\mathrm{b}}/{\mathrm{L}}_{\mathrm{s}}\right)\cdot {\mathsf{\epsilon }}_{\mathrm{b},\mathrm{r}}\left(\mathrm{t}\right)$$

(4)

where ε_b,t and ε_b,r are the transmitted and reflected strain pulses, respectively. A_b and A_s are the cross-sectional areas of the bars and sample, respectively. E_b and C_b are the elastic modulus and velocity of stress wave of the bar material, respectively. L_s is the initial gauge length of the specimen. t is testing duration. The data acquisition rate was 107 Hz, and the bandwidth of filter was 105 Hz. From various wave form signals obtained from the Hopkinson Tension bar, the stress–strain relationship under high-strain-rate test conditions can be derived. All experiments were repeated at least three times at room temperature, and their mean values were taken as the experimental data.

Several loading–unloading–reloading (LUR) cycles were conducted for different strain rates to evaluate the evolution of back stress and effective stress components. In each cycle, the specimens were firstly stretched to a designed strain (e.g., 5%) and then unloaded to 20 N at an unloading rate of 200 N min⁻¹, followed by reloading to the same applied stress before the next unloading.

3. Results and Discussion

3.1. Initial Microstructure

The initial microstructure of the investigated steel is shown in Figure 1. Approximately 42 vol% austenite was retained in the present D&P sample according to the XRD results (Figure 1a). Also, a three-dimensional reconstructed phase map shows that the microstructure was primarily composed of austenite, tempered α′-martensite and a small amount of ultrafine-grained (UFG) ferrite (Figure 1b). And the austenite phase had both coarse grains elongated along the rolling direction and reverted grains derived from intercritical annealing (Figure 1c). Such a heterogeneous microstructure was attributed to the irregular austenite grains during intercritical annealing and the introduction of defects during multistep treatments, as well as ferrite transformation from austenite. Thus, a triplex-phase heterogeneous microstructure after partitioning treatment was successfully introduced to the investigated steel.

3.2. Strain Rate Dependent Mechanical Properties

The engineering stress–strain curves of the D&P steel deformed at strain rates ranging from 10⁻³ s⁻¹ to 10³ s⁻¹ are shown in Figure 2. The yield stress (YS), ultimate tensile stress (UTS) and total elongation (TE) at different strain rates are summarized in

Table 2. Mechanical properties of the D&P steel obtained at different strain rates.

Strain Rates	YS/MPa	UTS/MPa	TE/%	PSE/GPa·%
10−3 s−1	1548	1804	20	36.1
10−1 s−1	1603	1724	26	44.8
1 s−1	1574	1691	26	44.0
103 s−1	1930	1732	17	29.5

. At a quasi-static strain rate of 10⁻³ s⁻¹, the D&P sample exhibits ultrahigh strength in both YS (1548 MPa) and UTS (1804 MPa), as well as excellent uniform elongation (20%). When the strain rate is increased to the intermediate strain rates (0.1 s⁻¹ and 1 s⁻¹), the ductility is surprisingly improved, and the tensile strength is maintained at an ultrahigh level (1724 MPa and 1691 MPa, respectively). In particular, a necking stage occurs, and the tensile elongation obtained at 0.1 s⁻¹ increases from 20% at 10⁻³ s⁻¹ to 26% and thus achieves a strength and ductility combination of ~45 GPa% (product of strength and ductility, PSE). At the extremely high strain rate of 10³ s⁻¹, a significant uptrend of YS (1930 MPa) and superior comprehensive mechanical properties are obtained, and good ductility can be maintained at 17%. This phenomenon of increased YS at a high strain rate is consistent with the observations in the PHS2000 [1] and quenching–partitioning steels [26,27].

Temperature rises due to adiabatic heating are commonly observed under high-strain-rate deformation. In certain cases, adiabatic heating can cause substantial thermal softening, resulting in the formation of adiabatic shear bands, which has been observed in various metals [4,28]. It is critical for plastic strain to convert into heat during the deformation process of AHSS steels [29]. The softening is superimposed by the geometric size softening and adiabatic temperature rise softening due to the presence of the adiabatic temperature rise effect. The hardening effect is smaller than the soft effect at high strain rates for D&P steel so that the tensile strength at high strain rates is lower than that in a quasi-static condition.

According to the Corporate Average Fuel Economy standard, all original equipment manufacturers in the automotive industry are required to meet the fuel economy target by the average vehicle weight. The international Iron and Steel Institute proposed a method [30] to calculate the lightweight coefficient of the whole vehicle [31] in the Ultra-Light Steel Auto Body project with the following equation:

$${\mathrm{L}}_{\mathrm{v}}={\mathrm{M}}^2\mathrm{Q}/\mathrm{VPA}$$

(5)

where L_v is the lightweight coefficient of whole vehicle, and M and P are car mass and engine power, respectively. V is volume, and Q is combined fuel consumption per 100 km. Compared to QP980, with a yield strength of ~698 MPa and tensile strength of ~1054 MPa, the ultrahigh-strength steel sheet with 1500 MPa yield strength in this work allowed a reduction in vehicle weight by ~6%. In this case, the improvement in the lightweight coefficient resulted in a reduction in the inertia and friction resistances of the car body, with an expected increase in fuel efficiency of ~4.0% and a reduction in vehicle exhaust emissions of ~5%. The results indicate that the ultrahigh-strength steel in the present work meets the multiple demands of both low emissions, high-performance structural components and easy recyclability. This further promotes the development of lighter, stronger and greener automobiles.

The corresponding strain hardening curves indicate that the strain-hardening ability appears to be sensitive to strain rates (Figure 2b). The strain hardening rate of the D&P sample deformed at quasi-static strain rate (10⁻³ s⁻¹) increases steadily with strain until the true strain of 0.14 and then decreases, while its counterpart at 0.1 s⁻¹ decreases rapidly until facture. The D&P sample drops sharply in the small strain range at extremely high strain rates, although there exists an increase in the strain hardening rate. The superior comprehensive properties at high-strain-rate deformation stand out among materials such as AHSSs and stainless steel [4,32,33,34,35,36] together with other Al/Mg alloys [37,38], as shown in Figure 2c. For example, dual-phase (DP) steel with different strengths does have enough time to undergo plastic deformation at 10³ s⁻¹, resulting in the elongation of ~7% in DP600 and ~12% in DP1000, respectively. Also, the yield strength of high-Mn steel is seriously insufficient at high strain rates, even if it can maintain a ductility within the range of ~20–25%, which limits the application of high-Mn steel sheet for automotive vehicles. As a result, this outstanding combination of high yield strength and superior ductility in this work exhibits great potential to fulfill the requirements of mass reduction and driving economy improvement.

3.3. Microstructure Evolution during Deformation at Various Strain Rates

The evolution of microstructures with respect to strain rates is characterized to illustrate the deformation behavior of the D&P steel. Figure 3 shows the EBSD phase maps and XRD patterns of the deformed specimens obtained at different strain rates. Regarding the quasi-static condition, austenite is almost fully exhausted after finishing L$\ddot{\mathrm{u}}$ders deformation (~10% strain), which can be further confirmed by XRD results (Figure 3a–c). The microstructure of the D&P steel consists of tempered α′ martensite, UFG ferrite and fresh α′-martensite originating from the TRIP effect, whereas the D&P sample did not immediately facture but continued to undergo plastic deformation, although the microstructure was fully in b.c.c phases. This result demonstrates that the TRIP effect is not the key mechanism responsible for enhanced strain hardening at high strain levels. The exact mechanism of high-strain hardening under quasi-static conditions will be discussed later. As the strain rate increases to 1 s⁻¹, the kinetics of the TRIP effect is sluggish. Approximately 16% austenite is retained upon straining to 10%. After that, the martensitic transformation further proceeds until fracture (Figure 3e,f). This indicates the success of phase-transformation kinetics at the intermediate strain rate of 1 s⁻¹. The TRIP effect is remarkably inhibited throughout the deformation process under the high strain rate of 10³ s⁻¹ (Figure 3g,h). Accordingly, ~31% austenite is retained till fracture (Figure 3i).

Under high-strain-rate deformation, a large amount of plastic deformation energy is converted into heat, which cannot be fully diffused into the atmosphere in an extremely short deformation time (~10⁻⁴ s) and can cause a temperature rise phenomenon [7,8,9]. This temperature rise (∆T) can be calculated by the following expression [9]:

$$∆\mathrm{T}={\displaystyle \frac{\mathsf{\beta }}{\mathsf{\rho }{\mathrm{C}}_{\mathrm{P}}}}\int \mathsf{\sigma }\left(\mathsf{\epsilon }\right)\mathrm{d}\mathsf{\epsilon }$$

(6)

where Cv is the heat capacity (taken as 0.46 kJ/kgK at room temperature), ρ is the density (taken to be 7.8 g cm⁻³), σ is the flow stress, ε is the plastic strain, and β is the fraction of the heat generated by the plastic deformation of the specimen. Lots of experiments under high-strain-rate loading have been conducted to determine the value of β experimentally, and the results suggested that it was suitable to taken the value of β to be 0.9 [8]. According to the above equation, the temperature rise of D&P steel at the strain rates of 10³ s⁻¹ is estimated to be 34.10 K. Most of the mechanical work is converted to heat, and a rapidly evolving microstructure can consume more of the energy as permanent changes in the microstructure. As a result, the initial microstructure and its evolution rate during deformation have a dramatic influence on the final released heat during high-strain-rate deformation. The heating further affects the austenite stability and the overall mechanical performance.

To illustrate the mechanism responsible for the high-strain hardening ability at higher strength levels in the D&P steel, LUR tests under the strain rates of 10⁻³ s⁻¹ and 1 s⁻¹ were performed (Figure 4). Both curves for the D&P sample deformed at 10⁻³ s⁻¹ and 1 s⁻¹ exhibit an obvious hysteresis loop in each LUR cycle, as shown in Figure 4a, indicating strong Bauschinger effect associated with mechanical incompatibility and strain gradients [39,40]. The HDI stress is calculated using the following equation, as proposed by Yang et al. [41]:

$${\mathsf{\sigma }}_{\mathrm{HDI}}={\displaystyle \frac{{\mathsf{\sigma }}_{\mathrm{u}}+{\mathsf{\sigma }}_{\mathrm{r}}}{2}}$$

(7)

where σ_u and σ_r are the yield stress during the unloading and reloading process, respectively. And the flow stress can be regarded as the sum of back stress induced by HDI strengthening and effective stress derived from dislocation slip and multiplication [19,42] (Figure 4b).

The D&P sample deformed at the quasi-static condition of 10⁻³ s⁻¹ possesses an overall higher HDI stress than that deformed at 10⁻¹ s⁻¹. Unlike the slight fluctuation with a small stress range obtained at 10⁻¹ s⁻¹, the HDI stress increases to 997 MPa when the true strain reaches 10% at 10⁻³ s⁻¹. After that, the HDI stress considerably increases from 997 MPa to 1415 MPa, and HDI stress accounts for a much higher proportion (~64%) of the overall flow stress during quasi-static tests (Figure 4c). These results indicate that the Bauschinger effect is stronger with strain at high strain levels. This can be ascribed to the fact that the martensite phase contains tempered α′-martensite and fresh martensite with different strengths in the present study. Almost all martensite phases in AHSS undergo tempering or auto-tempering during the partitioning process. However, there are differences in facilitating or resisting the onset of the local plasticity ability between tempered martensite and martensite transformed from austenite phase, as reported by Wang et al. [43]. In other words, the mechanical incompatibility derived from different martensite constituents results in a stronger strain gradient and thus promotes higher HDI stress in the later deformation stage [13,43]. On the other hand, the sluggish TRIP effect results in a progressive austenite transformation for the D&P sample deformed at 0.1 s⁻¹ and 10³ s⁻¹. Thus, the subsequent deformation process is mainly borne by austenite and accompanied by the increase in uniform elongation without obvious HDI strengthening.

To sum up, the strain-rate-dependent microstructural evolution is sketched in Figure 5. The triplex-phase heterogeneous microstructure consists of UFG ferrite, coarse and reverted austenite grains and tempered α′-martensite (α′_temp). In the case of quasi-static deformation (10⁻³ s⁻¹), the prevailing TRIP effect causes the complete depletion of austenite grains as early as in the strain of 10%. The microstructure is therefore a mixture of different martensitic phases, i.e., α′_temp and fresh α′-martensite (α′_TRIP) resulting from the TRIP effect. This more significant strain gradient between phase constituents with the deformation process increases to 20%, resulting in higher HDI stress and strain hardening ability. Intense plastic deformation promotes the rapid attainment of fracture strength and subsequent fracture of materials under quasi-static conditions. Regarding the high-strain-rate condition (10³ s⁻¹), the instinct deformation ability of TRIP effect is suppressed obviously due to both hindered dislocation movement and increased stacking fault energy induced by adiabatic heating [13,44]. As a result, the large amount of retained austenite further sustains tensile deformation and thus obtains good ductility.

4. Conclusions

In the present work, the mechanical properties and associated deformation mechanisms in a wide strain-rate regime (10⁻³ s⁻¹ to 10³ s⁻¹) of an ultrahigh-strength D&P steel were investigated. Based on the experimental results, the key findings are summarized as follows:

• The present D&P steel with a heterogeneous microstructure consisting of 42 vol% austenite, tempered α′-martensite and UFG ferrite exhibits superior strength–ductility combination over a wide range of strain rates from 10⁻³ s⁻¹ to 10³ s⁻¹, which is rarely achieved in other metallic materials. Particularly, the yield strength is 1930 MPa and the tensile elongation is 17% under the extremely high strain rate of 10³ s⁻¹.
• Strain rate exerts a major effect on the deformation behaviors of the D&P steel. The austenite is entirely exhausted after deformation to the strain of 10% at the quasit-static strain rate. The fresh martensite induced by the TRIP effect promotes significant HDI hardening at high strains. The sequential occurrence of the TRIP effect and HDI hardening contributes to enhanced strain hardening at the strain rate of 10⁻³ s⁻¹.
• The TRIP effect is remarkably suppressed by high-strain-rate deformation. As the strain rate increased to 10³ s⁻¹, ~31% austenite was retained in the D&P steel until fracture. Austenite phase with excellent deformation ability sustains the subsequent deformation and thus achieves superior ductility without obvious TRIP effect and HDI strengthening.