Effect of High-pressure Quenching on Pure-iron Martensite Transformation and Its Strengthening Mechanism

Abstract

Industrial pure iron samples were austenitized and quenched (6°C/s cooling to room temperature) under hydrostatic pressure of 3–5 GPa. The morphology, phase transformation and strengthening mechanism of high-pressure quenched martensite are analyzed by the method of SEM, XRD, EBSD, and TEM. Lath martensite with hierarchical packet-block-lath structure is induced in industrial pure iron by high pressure, which keeps the Kurdjumov-Sachs (K-S) orientation relationship with the face-center-cubic (FCC) phase. Pressure refines the size of prior austenite by depressing the mobility of grain boundaries, leading to the decrease of the type of martensite variants. with the increment of pressure, the dislocation density increases gradually (2.04×10¹³ to 3.14×10¹⁴ m⁻²) and the martensite blocks are refined from 3.3 to 0.9 µm. In addition, an enormous number of twin boundaries and high-density dislocations are observed in 5 GPa-samples, which is fairly rare in lath martensite of low carbon steels. Superior tensile performances are obtained in industrial pure iron, especially in 5 GPa-sample with ultra-high yield strength of 700 MPa and excellent ductility of 27%. The strengthening mechanism is quantitatively analyzed by Olson’s strengthening model, and the results show that both of dislocation strengthening and Hall-Petch strengthening enhances with the increase of pressure. Based on the above findings, martensite transformation can be effectively controlled by hydrostatic pressure, which extends the knowledge into martensitic transformation mechanism and offers a new avenue for developing high performance metal materials.

1. Introduction

Lath martensite has always been research spot on fundamental material, depending on its excellent strength and processability.^1,2,3) Generally, the structure has distinct hierarchical structure, and a prior austenite is divided into packets, blocks and laths. Laths with a similar crystal orientation compose a block, while several blocks with the same habit plane form a packet.⁴⁾ Massive dislocations are accumulated in the interior of laths by the lattice deformation during martensite transformation. The extensively distributed grain boundaries and dislocations in martensitic matrix are important contributor for enhancing the mechanical properties of lath martensite.^5,6,7)

There is no denying that the role of lath martensite is irreplaceable in iron and steel. However, martensite transformation is difficult to take place in industrial pure iron, due to its ultra-low carbon content (~0.0029 wt.%).⁸⁾ As-quenched pure iron with martensite matrix can exist with the cooling rate of 10⁵°C/s under atmosphere pressure theoretically,⁹⁾ which requires extremely complicated technological process and expensive equipment. Prior studies have shown that high pressure would shrink the atom distance and change the electron distribution, result in an appreciable decrease in the transformation temperatures.^10,11) Martensite transformation may occur when transformation temperatures are very low with respect to Debye temperature.¹²⁾ Kang et al.¹³⁾ has achieved a rare fully martensitic structure of IF steel (heating to 1200°C and holding for 600 s followed by 8°C/s cooling to room temperature) by use of ultra-high hydrostatic pressure, which doubtlessly enhanced the strength of the lean-alloying steel to an unprecedented level. The calculated temperature-pressure phase diagram of iron provides the theoretical basis of martensite transformation in pure iron, from which austenite-martensite equilibrium temperature (T₀) for the A₃ transformation (FCC to BCC) decreases with increasing pressure.¹⁴⁾ With a further increase in pressure, ε martensite (HCP) appears with a triple point (11.5 GPa, 500°C),¹⁵⁾ and Lu et al.¹⁶⁾ illustrated the transformation mechanism by first-principle calculation.

Although many indications revealed the promising transformation behavior, in fact little attention has been devoted to elucidate the effect of high-pressure quenching on pure-iron martensite transformation and its strengthening mechanism systematically. Consequently, industrial pure iron was heated to 950°C and holding for 1200 s followed by a slow cooling of 6°C/s to the room temperature under constant pressure of 3, 4 and 5 GPa. Rare lath martensite and ultra-high strength was obtained in pure iron, and martensite transformation and strengthening mechanism were discussed in more depth. This research may shed new light on the intrinsic mechanism of high-pressure quenching, and the results demonstrate for the first-time evidence for strength enhancing and strengthening mechanism induced by pressure. As such, it presents a wide range of possibilities for the further innovation of processing technique.

2. Experimental Methods

An industrial pure iron with a composition Fe-0.0029 C wt.% was investigated. The iron was received after annealing stage, and specimens was cut into cylinders with a gauge section of Φ6 mm × 8 mm for GPa high-pressure quenching. The specimens were assembled into cubic pyrophyllite blocks,¹⁷⁾ then subjected to heating, homogenizing and cooling under high-pressure of 3, 4, 5 GPa on a hexahedral top ultrahigh pressure device (CS-1 V).¹⁸⁾ The high-pressure quenching used to investigate the martensite transformation consists of a heating to 950°C, followed by an isothermal holding during 1200 s and finally of a slow cooling at a rate of 6°C/s down to the room temperature under a constant pressure. The austenitizing temperature was set by temperature-pressure phase diagram of iron and measured with K-type thermocouples, and the pressure was calculated from calibration curves for the multi-anvil high pressure apparatus.¹⁹⁾

For phase identification, X-ray diffraction (XRD) experiments were performed at a X-Ray Diffractometer (Rigaku Corporation, Japan SMARTLAB 9) operating at 40 kV and 40 mA using Cu-K_α radiation (λ=0.15418 nm, 3°min⁻¹ from 30 to 90°). The patterns of electron backscattered diffraction (EBSD) were acquired by a SUPRA-55 scanning electron microscopy (SEM) equipped EDAX TSL (Mahwah, NJ) OIM system, and the orientation data were post-processed with Channel 5 software. Nanoscale lath observations were carried out by the Tecnai G2 F20 transmission electron microscope (TEM), and thin foils for that were prepared by mechanical grinding, twin-jet electrolytic polishing in a solution of 8% perchloric acid, 92% alcohol at 80 mA. The microstructure and morphology of high-pressure samples after tensile fracture was observed using SEM on the Zeiss Ultra 55 at 5 kV.

Vickers hardness was measured on a HV-1000A tester with a load of 5 N. Specimens with a gauge section of Φ1.8 mm x 6 mm were used for tensile testing at a strain rate of 0.2 mm/min at room temperature.

3. Results and Discussion

3.1. Microstructural Characterization

As shown by Fig. 1(a), the industrial pure iron annealing under atmospheric pressure has typical morphology of massive ferrite, and large size difference remained in grains. High pressure induces lath martensite in all high-pressure samples in Figs. 1(b)–1(d), of which the distinct hierarchical structure of packet-block-lath is observed in 3, 4 GPa-sample, highly similar to low carbon martensite.²⁰⁾ The typical lath martensite characteristics still remain when the quenching-pressure increases gradually, while the grain is remarkably refined in the function of pressure, especially in 5 GPa-sample.

Fig. 1.

Optical micrographs of industrial pure iron under different pressure. a) as-annealed at 10⁻⁴ GPa, b) quenched at 3 GPa, c) quenched at 4 GPa, d) quenched at 5 GPa.

Figure 2 shows the SEM microstructures of industrial pure iron austenitized at 950°C for 1200 s and cooled to room temperature at 6°C/s under different pressure. The distinct grain boundaries of prior austenite (PAG) for all high-pressure samples are observed, and a statistical size analysis was carried out using the mean value of the longest and shortest sides of the PAGs according to many such SEM micrographs. The statistical results exhibit that the average diameter of PAGs decreases with the increasing pressure, and the average sizes of those for 3–5 GPa samples are approximately 92, 67 and 10 μm, respectively. High pressure strongly refined PAGs of industrial pure iron, owning to the decrease of the mobility of grain boundary. It is generally accepted that low-angle grain boundaries move by the vacancy diffusion mechanism.²¹⁾ Consequently, the movement of low-angle grain boundaries is strongly slowed down due to the decrease in vacancy content with increasing pressure. Furthermore, Krawczynska²²⁾ confirmed that the migration of twin boundaries was hampered under high pressure in nano-twinned austenitic stainless steel, and it was found during annealing of cold-rolled copper under a pressure of 4 GPa that the pressure retarded recrystallization.²³⁾ The mobility of grain boundaries could be hampered by high pressure, causing the PAGs refinement of high-pressure samples.

Fig. 2.

SEM observations of high-pressure quenched martensite. a) 3 GPa, b) 4 GPa, c) 5 GPa. (Online version in color.)

The diffraction patterns in Fig. 3 for as-annealed and high-pressure samples have three broad peaks at about 45°, 65° and 82°, corresponding to (110), (200) and (211) of α-Fe, respectively. The result shows that the whole martensite matrix is obtained by high-pressure in industrial pure iron, since no retained austenite is detected by the method of XRD. Furthermore, the lattice constant and the full width at half maximum (FWHM) of (110)_α are calculated. Previous studies reported that the tetragonality of Fe–C quenched martensite was linearly dependent on the carbon content, hence pure iron after high-pressure quenching should be body centered cubic (BCC). The lattice constant of high-pressure samples is calculated by XRD. With a gradual increase in pressure from 3 GPa to 5 GPa, the a lattice parameter of martensite decreases from 0.2748 to 0.2737 nm and finally 0.2734 nm. Besides, the FWHM is reported a marked increase under high pressure, from 0.227° in 3 GPa to 0.355° in 5 GPa. Apparently, the pressure proliferates the extent of lattice distortion for industrial pure iron. To gain more insight, the dislocation density is estimated by Williamson-Hall (WH) analysis through the main three diffraction peaks. The estimation results showed that the dislocation density of high-pressure quenched samples for 3, 4 and 5 GPa is 2.04×10¹³ m⁻², 6.56×10¹³ m⁻² and 3.14×10¹⁴ m⁻², which increases continuously by pressure.

Fig. 3.

XRD profiles of quenched samples under different pressure. (Online version in color.)

In order to deeply investigate the high-pressure effect on grain refinement and orientation relationship in industrial pure iron, EBSD observations are taken as shown in Fig. 4. IPF images are applied for the direct observation of the hierarchical structure in Figs. 4(a)–4(c), of which each color represents a specific crystallographic orientation normal to the observed planes. The sizes of high-angle grain boundaries (blocks) are detected, and areas less than four pixels are removed for accuracy. The distribution frequency of the blocks is shown in Figs. 4(d)–4(f). The majority of martensite blocks in 3 GPa-samples are between 1–2 μm, with an average of 3.3 μm. With pressure increasing to 4–5 GPa, the average size of blocks decreases continuously to 2.6 μm and finally a surprising 0.9 μm. Therefore, the average equivalent grain size decreases with elevating pressure. As can be seen from (001)_γ pole figure (the area marked with a circle in the IPF images) in Figs. 4(g)–4(i), the experimental pole figure matches well with the theoretical one, indicating that the high-pressure martensite for 3–5 GPa is close to the K-S OR with respect to the prior austenite. However, the number of variant types decreases with the increasing pressure, which is related to the PAGs refinement by pressure. When the PAG is small in 5 GPa-sample, the strength of austenite during transformation is high. In order to reduce the elastic strain energy caused by phase transformation, the repeated nucleation of martensitic laths along the same crystal direction at α’/γ interface promotes the growth of nucleated martensitic variants, and finally inhibits the formation of other types of martensitic variants.²⁴⁾

Fig. 4.

EBSD inverse pole figure (IPF) images, block size distribution and (001)_γ pole figures of pure iron under high pressure. a, d, g) 3 GPa, d, e, h) 4 GPa, c, f, i) 5 GPa. (Online version in color.)

Furthermore, the frequencies of low angle boundaries (LAGs: 2°<θ<10°) and three groups of block boundaries are counted in Fig. 5. Generally, LAGs can be identified as a dislocation structure. Although the proportion of LAGs of 5 GPa is slightly smaller than 3 GPa, the dislocation density should be higher in 5 GPa-sample due to grain refinement. It is recognized that 60°<111> is twin boundary formed by twin-related variants.²⁵⁾ The frequency of twin boundaries is dominated in block boundaries for both high-pressure irons. As the pressure mounts, the proportion of 60°<111> twin boundaries increases from 23.0% (3 GPa) to 25.5% (5 GPa). Besides, the total grain boundaries rise as a result of the significant grain refinement in 5 GPa-sample, and thus the magnitude of the twin boundaries increases with increment of quenching pressure. Therefore, pressure promotes the formation of 60°<111> twin boundaries in martensite, the results in good with previous work in high-pressure quenching martensite of 0.2 mass% C steel.²⁶⁾

Fig. 5.

Observed ratio of low angle boundaries (LAGs: 2°<θ< 10°) and three groups of block boundaries. (Online version in color.)

Figures 6(a)–6(c) shows a typical low-magnification TEM images of high-pressure samples at 3–5 GPa, in which martensite laths tangled with high density of dislocations were observed. A width count taken from many such images, confirmed martensite laths were refined by high-pressure quenching. When the quenching-pressure increases from 3 GPa to 5 GPa, the mean width of laths significantly decreases from 540.5 to 204.3 nm. The selected area electron diffraction (SAED) patterns are indexed for the selected area in Fig. 6(d) as shown in Fig. 6(f). Figure 6(e) shows the dark field morphology of the twin, which is imaged by using the (1–10) twin diffraction spot in Fig. 6(f). It is found that the adjacent martensite laths mostly follow symmetric twin-related orientation relationships of {112} <111> at 5 GPa-sample. Such behavior implies that the generation of nano-twin lath can be attributed to the reduction of the total strain energy caused by martensite transformation, which result from volume expansion during transition constrained by pressure.

Fig. 6.

TEM images and SAED patterns of high-pressure quenched martensite. a–c) morphology of 3–5 GPa-samples; d–f) analysis of twin-related laths under 5 GPa-sample. (Online version in color.)

Fig. 7.

Mechanical property of high-pressure samples. a) engineering tensile strain-stress curve, b) histogram of tensile property and hardness. (Online version in color.)

3.2. Martensite Transformation Mechanism Induced by Pressure

It has been proved that, in the case of industrial pure iron quenched under high hydrostatic pressure of 3 to 5 GPa, the whole lath-martensite matrix is available even with ultra-low cooling rate of 6°C/s. However, it is commonly known that martensite is extremely hard to be obtained in pure iron at atmospheric pressure. From the results we have obtained, one can conclude that pressure is an effective means of microstructure regulation, which can cause martensite transformation of pure iron occur at very low cooling rate. In the process of metal solid-state transformation, martensite forms by a deformation of the austenite lattice without any diffusion of atoms, while diffusion transformations are accompanied by the decomposition of undercooled austenite. Nilan²⁷⁾ stated that high pressure enhanced the stabilization of austenite, and the decomposition of austenite is inhibited when cooling at GPa level high pressure. Hence, pressure can shift the Time-Temperature-Transformation to the right side and lower the critical cooling rate, so that diffusion transformations such as ferrite and pearlite are suppressed and martensitic transformation is able to occur in industrial pure iron with low cooling rate. Several excellent experiments have confirmed the credibility of this theory in Fe-15 wt.%Cr and IF steel,^28,29) thus pressure is critical for martensitic transformation at a low cooling rate in pure iron.

The martensite transformation temperature (Ms) plays vital roles in the morphology and performance of iron and steels, thus considerable research efforts have been devoted to the calculation of Ms by the difference in free energy of the parent and martensite phase. For thermodynamics, pressure exerts a significant influence on the difference in free energy between austenite and martensite. The change in free energy for martensite transformation in Fe–C steels was calculated by the formula³⁰⁾   

$$\mathrm{\Delta }{F}_P^{\gamma \to {\alpha }^{\prime }}=\mathrm{\Delta }{F}_{atm}^{\gamma \to {\alpha }^{\prime }}-24{\int }_{atm}^P\mathrm{\Delta }{V}^{\gamma \to {\alpha }^{\prime }}dP$$(1)

where $\mathrm{\Delta }{F}_P^{\gamma \to {\alpha }^{\prime }}$ and $\mathrm{\Delta }{F}_{atm}^{\gamma \to {\alpha }^{\prime }}$ are the changes in free energy for the reaction at atmospheric pressure and an applied hydrostatic pressure of P GPa (the unit ka in the original formula is uniformly converted to GPa for convenience), respectively, and $\mathrm{\Delta }{V}^{\gamma \to {\alpha }^{\prime }}$ is the change in the molar volume (in cm³/g mole of alloy).

According to the findings from Kaufman and Cohen,¹²⁾ $\mathrm{\Delta }{F}_P^{\gamma \to \alpha }$ is given as a function of the absolute temperature T. For industrial pure iron,   

$$\begin{array}{l}\mathrm{\Delta }{F}_{\text{atm}}^{\gamma \to {\alpha }^{\prime }}=\mathrm{\Delta }{F}_{\text{atm}}^{\gamma \to \alpha }=2.63\times {10}^{-3}{T}^2-1.54\times {10}^{-6}{T}^3\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}-1\mathrm{\hspace{0.17em} }202\mathrm{\hspace{0.17em} } \mathrm{c}\mathrm{a}\mathrm{l}/\mathrm{m}\mathrm{o}\mathrm{l}\\ \mathrm{\Delta }{V}_T{}^{\gamma \to {\alpha }^{\prime }}=\mathrm{\Delta }{V}_T{}^{\gamma \to \alpha }=\mathrm{\Delta }{V}_{293K}^{\gamma \to \alpha }-2\times {10}^{-4}\left(T-293\right)\end{array}$$(2)

The lattice parameters of austenite and martensite at atmospheric pressure for pure iron were reported as 0.35620 nm and 0.28650 nm.³¹⁾ Assuming that the effect of pressure on the lattice constant is ignored, the free energy change accompanying the martensite transformation can be calculated, as shown in Fig. 8. Patel and Cohen³⁰⁾ assume that the free energy change under the hydrostatic pressure remain the same, corresponding to the $\mathrm{\Delta }{F}_P^{\gamma \to {\alpha }^{\prime }}$ = −265 cal/mol, at ambient pressure at temperature Ms. Consequently, the Ms is calculated from the $\mathrm{\Delta }{F}_P^{\gamma \to {\alpha }^{\prime }}$ −T curves. The Ms-temperature is 562°C under atmospheric pressure, which is reduced to 376, 326 and 279°C under 3, 4 and 5 GPa for industrial pure iron. When GPa level high pressure is applied to the system, the free energies of the parent and martensite phases change results from the additional thermodynamical energy of the atomic volume compressed by pressure. At the temperature Ms, the free energy of the parent is smaller than that of the martensite phase since the value of $\mathrm{\Delta }{F}_P^{\gamma \to {\alpha }^{\prime }}$ is negative. This is consistent with the fact that the atomic volume of the martensite is larger than that of the parent phase in industrial pure iron. It is obvious that the Ms-temperature decreases with the increasing pressure. It is generally accepted that the decrease of the Ms-temperature is likely to inspire the martensite morphology from lath (dislocation substructure) to plate (twinning substructure) in Fe–C alloys at atmosphere. According to the microstructure observations, high-pressure quenching promotes the formation of 60°<111> twin boundaries and nano-twin laths in industrial pure iron. Such behaviors may be explained by the decrease of the Ms-temperature induced by high-pressure quenching.

Fig. 8.

Variation of free energy change from austenite to martensite. (Online version in color.)

Compared with the previous result on Fe-0.2C,²⁶⁾ new developments and findings are acquired by the study on pure-iron martensite induced by pressure. In Fe-0.2C alloy, martensite transformation can occur easily by normal quenching, so the effect of high pressure on the morphology and the C atom distribution of martensite is emphasized. In industrial pure iron, it is difficult to obtain martensite by conventional heat treatment and thus martensite formed by high-pressure quenching in pure iron is a novel phenomenon. Pure-iron martensite transformation have rarely been studied. In addition, pressure can influence the C atom distribution in Fe-0.2C alloy, which may change the Ms-temperature of the system. Analyzing the effect of pressure on the Ms-temperature temperature in pure iron, the result is more reasonable and convincing, given that it excludes the influence of carbon atom distribution.

To the best of my knowledge, the majority of studies on the properties of high-pressure quenched martensite focus on hardness, and no systematic study concerning the tensile strength and strengthening mechanism has been published yet.

3.3. Mechanical Properties and Strengthening Mechanism

The engineering stress-strain curves are displayed in Fig. 6(a), and the tensile properties and the harnesses of all high-pressure samples are shown in Fig. 6(b). The mechanical performances vary significantly with the pressure. The sample after 3 GPa high-pressure quenching shows an ultimate tensile strength (σ^b) of 425 MPa, yield strength (σ^0.2) of 365 MPa and break elongation (δ) of 36%. As is shown in Fig. 6(b), the σ^b and σ^0.2 of industrial pure iron increases remarkably with the increasing pressure, while the break elongation has a slight sacrifice. The increasing tendency of σ^b and YS basically keeps consistent, and the σ^0.2 is adopted to analyze the strengthening effect of pressure. It can be clearly observed that the strengthening effect is more striking when the pressure increased up to 5 GPa. The superior tensile performances in industrial pure iron are obtained after 5 GPa high-pressure quenching, with the ultra-high yield strength of 700 MPa and excellent ductility of 27%. In contrast to 3, 4 GPa-sample, the integrated mechanical properties of 5 GPa are better, as a result of comprehensive contributions on grain refinement, dislocation interaction and so on.

Due to high martensite start temperature (Ms), it’s hard to obtain martensite by quenching at atmosphere. Therefore, industrial pure iron is very soft with the hardness of 150 HV,⁹⁾ significantly less than that of high-pressure samples in Fig. 6(b). After high-pressure quenching under high pressure of 3 GPa, the hardness was increased to 223 HV. The hardness increases continuously with the increment of pressure, which exceeds 400 HV of 5 GPa-sample, even higher than 0.2 wt.% C steel after quenching at atmospheric pressure. Typically, the carbon content plays a crucial role in the hardness of martensite. However, high pressure provides an innovative means for the hardening of low-carbon or even carbon-free martensite.

In order to analyze the strengthening mechanism in high-pressure martensite, proper superposition laws (Olson’s strengthening model) are applied to predict the yield strength of the industrial pure iron by adding contributions from all strengthening constituents. Ignoring the effect of solid solution and precipitation in pure iron, the yield strength of steels with martensite matrix could be estimated by Eq. (3).³²⁾   

$${\sigma }_{\text{s}}={\sigma }_{P-N}+{\sigma }_{H-P}+M{\tau }_d$$(3)

where σ_P−N is the Peierls-Nabarro (P-N) strengthening, σ_H−P is the Hall-Petch strengthening, M = 3.2 is the Taylor factor, and τ_d are the critical-resolved shear stress (CRSS) for dislocation strengthening.

Based on the Peierls-Nabarro (P-N) stress model, σ_P−N of pure α-Fe could be expressed by Eq. (4).^33,34)   

$${\sigma }_{P-N}=M{\alpha }_{P-N}\frac{\mu }{1-v}exp\left[-\frac{2\pi d_{\text{s}}}{b(1-v)}\right]$$(4)

where α_P-N is the effective coefficient of P-N strengthening, μ = 77 GPa is the shear modulus, $d_{\text{s}}=\sqrt{2}/2a$ is the spacing of the slip planes (taking the slip system {110} <111>), $b=\sqrt{3}/2a$ is the Burgers vector for bcc iron, and ν=0.3 is Poisson’s ratio.

Grain refinement strengthening can be expressed by the Hall-Petch equation, as shown in Eq. (5).³⁵⁾   

$${\sigma }_{H-P}=K_{H-P}{d}^{-\frac{1}{2}}$$(5)

where K_H−P= 0.2 is the Hall-Petch constant for martensite packets and d is the block size of martensite.

The CRSS for dislocation interaction strengthening could be expressed by Eq. (6).³⁶⁾   

$${\tau }_d={\alpha }_d\mu b\sqrt{\rho }$$(6)

where α_d = 0.24 is the effective coefficient of dislocation interaction strengthening, and ρ is the dislocation density.

Figure 9 shows the simulation results of yield strength for high-pressure quenching martensite under 3,4 and 5 GPa. The errors are within ±15 MPa between simulation and experiment values, which could partly verify the rationality of the model. The simulation results of the strength contribution for each strengthening method in high-pressure samples are shown in Table 1. The result shows that the structural strengthening results from P-N strengthening, grain refinement strengthening and dislocation interaction strengthening. When industrial pure iron quenched at 3 GPa, P-N strengthening could represent two-third of the yield strength, and thus the resistance of crystal lattice to dislocation motion is critical for 3 GPa-sample. However, grain refinement and dislocation interaction have a little effect on strength at 3 GPa, and their potential has not been fully explored. As the most basic inherent resistance, the P-N stress will not change with pressure. The role of both grain refinement strengthening and dislocation strengthening enhances with the increase of pressure. The strengthening contributions of the grain refinement and dislocation mechanisms may be closely related to the applied pressure of quenching process. It should be pointed out that the grain refinement contribution reaches 209 MPa in 5 GPa-sample, almost three orders of magnitude higher than that of 3 GPa. Meanwhile, the dislocation interaction strengthening enhances by over four orders of magnitude relative to 3 GPa-sample (from 63 to 248 MPa). Consequently, the effect of pressure on dislocation strengthening is more remarkable. It is obvious that the enhancement of strength in industrial pure iron after high-pressure quenching caused by lifting pressure depends on grain refinement and dislocation interaction, and the dislocation interaction strengthening play a dominant role in 5 GPa-sample. These findings reveal an important new strategy for material strengthening without alloying.

Fig. 9.

Yield strength comparison of simulation results and experiment results. (Online version in color.)

Table 1. Simulation results of the strength contribution in high-pressure samples.

Strength contribution (MPa)	3 GPa	4 GPa	5 GPa
Peierls-Nabarro	231	231	231
Grain Refinement	84	124	209
Dislocation	63	114	248

4. Conclusions

In summary, a completed martensitic transformation was obtained by GPa high-pressure of industrial pure iron. The microstructural characterization, phase transformation and strengthening mechanism were carried out to study the high-pressure effect on martensite transformation, and the obtained results are as follows:

(1) Lath martensite matrix is acquired at ultra-low cooling rate of 6°C/s under 3–5 GPa pressure. High-pressure martensite has typical hierarchical structure of packet-block-lath, which is close to the K-S OR with respect to the prior austenite,

(2) High pressure causes the decrease of Ms, from 562°C at 10⁻⁴ GPa to 279°C at 5 GPa, inducing the formation of 60°<111> twin boundaries and nano-twin lath in industrial pure iron.

(3) The size of PAG strongly declined with the increasing pressure, and martensite blocks were also effectively refined. The dislocation density of 3, 4 and 5 GPa-samples is 2.04×10¹³, 6.56×10¹³ and 3.14×10¹⁴ m⁻², which increases continuously by pressure.

(4) Superior performance is exhibited in high-pressure quenched steel, especially at 5 GPa (700 MPa in yield strength and 27% in elongation). According to strengthening simulation, it is found that the strengthening contribution of the grain refinement and dislocation strengthening enhances with the increase of pressure. The effect of pressure on dislocation interaction is more remarkable, from 63 MPa at 3 GPa increasing to 248 MPa at 5 GPa.

Acknowledgements

This work was supported by the National Science Fund for Distinguished Young (Grant No. 51925105) and the National Natural Science Foundation of China (Grant No. 51675092).

References

• 1)   J. P.  Masse,  B.  Chéhab,  H.  Zurob,  D.  Embury,  X.  Wang and  O.  Bouaziz: ISIJ Int., 54 (2014), 235. https://doi.org/10.2355/isijinternational.54.235
• 2)   M.  Souissi and  H.  Numakura: ISIJ Int., 55 (2015), 1512. https://doi.org/10.2355/isijinternational.55.1512
• 3)   K.  Kwak,  T.  Mayama,  Y.  Mine and  K.  Takashima: Mater. Sci. Eng. A, 674 (2016), 104. https://doi.org/10.1016/j.msea.2016.07.047
• 4)   C.  Du,  J. P. M.  Hoefnagels,  S.  Kölling,  M. G. D.  Geers,  J.  Sietsma,  R.  Petrov,  V.  Bliznuk,  P. M.  Koenraad,  D.  Schryvers and  B.  Amin-Ahmadi: Mater. Charact., 139 (2018), 411. https://doi.org/10.1016/j.matchar.2018.03.011
• 5)   R.  Ueji,  N.  Tsuji,  Y.  Minamino and  Y.  Koizumi: Acta Mater., 50 (2002), 4177. https://doi.org/10.1016/S1359-6454(02)00260-4
• 6)   S.  Morito,  J.  Nishikawa and  T.  Maki: ISIJ Int., 43 (2003), 1475. https://doi.org/10.2355/isijinternational.43.1475
• 7)   B.  Hutchinson,  J.  Hagström,  O.  Karlsson,  D.  Lindell,  M.  Tornberg,  F.  Lindberg and  M.  Thuvander: Acta Mater., 59 (2011), 5845. https://doi.org/10.1016/j.actamat.2011.05.061
• 8)   B.  Liu and  Y. F.  Zheng: Acta Biomater., 7 (2011), 1407. https://doi.org/10.1016/j.actbio.2010.11.001
• 9)   H.  Zhang,  Y.  Wang,  Y.  Peng,  P.  Zhu,  J.  Liu,  Z.  Feng,  G.  Wu and  X.  Huang: Mater. Res. Lett., 7 (2019), 354. https://doi.org/10.1080/21663831.2019.1612792
• 10)   G.  Demazeau: J. Phys. Condens. Matter, 14 (2002), 11031. https://doi.org/10.1088/0953-8984/14/44/422
• 11)   V. A.  Chernenko: J. Phys. IV, 5 (1995), C2-77. https://doi.org/10.1051/jp4:1995211
• 12)   Z.  Nishiyama: Martensitic Transformation, Academic Press, New York, (1978), 263.
• 13)   J. Y.  Kang,  S. C.  Kim,  J. O.  Oh,  H. N.  Han,  K. H.  Oh and  H. C.  Lee: Scr. Mater., 66 (2012), 45. https://doi.org/10.1016/j.scriptamat.2011.10.002
• 14)   A. I.  Funtikov: High-Pressure Shock Compression of Solids VII, Springer, New York, (2004), 225.
• 15)   L. J.  Swartzendruber: Bull. Alloy Phase Diagr., 3 (1982), 161. https://doi.org/10.1007/BF02892374
• 16)   Z.  Lu,  W.  Zhu,  T.  Lu and  W.  Wang: Model. Simul. Mater. Sci. Eng., 22 (2014), 025007. https://doi.org/10.1088/0965-0393/22/2/025007
• 17)   L.  Wang,  X. P.  Lin,  C.  Xu,  S. S.  Zhao,  H.  Sun and  N.  Zhang: Mater. Sci. Technol., 35 (2019), 1690. https://doi.org/10.1080/02670836.2019.1639886
• 18)   X. P.  Lin,  Y.  Kuo,  L.  Wang,  J.  Ye,  C.  Zhang,  L.  Wang and  K. Y.  Guo: Trans. Nonferrous Met. Soc. China, 31 (2021), 1587. https://doi.org/10.1016/S1003-6326(21)65600-5
• 19)   B.  Yuan,  Q.  Tao,  X.  Zhao,  K.  Cao,  T.  Cui,  X.  Wang and  P.  Zhu: Rev. Sci. Instrum., 85 (2014), 013904. https://doi.org/10.1063/1.4862654
• 20)   B. Y.  Ma,  C.  Yan,  Y. S.  Ding,  Y.  Li and  Y. D.  Liu: Adv. Mater. Res., 750–752 (2013), 408. https://doi.org/10.4028/www.scientific.net/AMR.750-752.408
• 21)   V.  Sursaeva,  S.  Protasova,  W.  Lojkowski and  J.  Jun: Textures Microstruct., 32 (1999), 175. https://doi.org/10.1155/TSM.32.175
• 22)   A. T.  Krawczynska,  P.  Suchecki,  B.  Adamczyk-Cieslak,  B.  Romelczyk-Baishya and  M.  Lewandowska: Mater. Sci. Eng. A, 767 (2019), 138381. https://doi.org/10.1016/j.msea.2019.138381
• 23)   L. E.  Tanner and  S. V.  Radcliffe: Acta Metall., 10 (1962), 1161. https://doi.org/10.1016/0001-6160(62)90169-4
• 24)   C.  Celada-Casero,  J.  Sietsma and  M. J.  Santofimia: Mater. Des., 167 (2019), 107625. https://doi.org/10.1016/j.matdes.2019.107625
• 25)   S.  Morito,  H.  Tanaka,  R.  Konishi,  T.  Furuhara and  T.  Maki: Acta Mater., 51 (2003), 1789. https://doi.org/10.1016/S1359-6454(02)00577-3
• 26)   Q.  Cui,  J.  Liu,  C.  Xu,  Y. J.  Xu,  H.  Huang,  B.  Wen and  X. P.  Lin: ISIJ Int., 61 (2021), 2292. https://doi.org/10.2355/isijinternational.ISIJINT-2020-784
• 27)   T. G.  Nilan: Trans. Metall. Soc. AIME, 239 (1967), 898.
• 28)   Z.  Wang,  H.  Sun,  C.  Li,  N.  Liu,  S.  Zhang,  P.  Zhu,  D.  Yu and  H.  Zhang: Acta Mater., 214 (2021), 116978. https://doi.org/10.1016/j.actamat.2021.116978
• 29)   Y.  Li,  Z.  Wang,  S.  Zhang,  N.  Liu,  Y.  Wang,  P.  Zhu,  D.  Yu,  Y.  Peng and  H.  Zhang: Scr. Mater., 187 (2020), 163. https://doi.org/10.1016/j.scriptamat.2020.06.004
• 30)   S. V.  Radcliffe and  M.  Schatz: Acta Metall., 10 (1962), 201. https://doi.org/10.1016/0001-6160(62)90117-7
• 31)   X.  Liu,  F.  Zhong,  J. X.  Zhang,  M. X.  Zhang,  M.  Kang and  Z. Q.  Guo: Phys. Rev. B, 52 (1995), 9970. https://doi.org/10.1103/PhysRevB.52.9970
• 32)   C.  Wang,  Q.  Cui,  X.  Huo,  C.  Zhang and  W.  Xu: ISIJ Int., 59 (2019), 1715. https://doi.org/10.2355/isijinternational.ISIJINT-2019-143
• 33)   G.  Wang,  A.  Strachan,  T.  Çağin and  W. A.  Goddard III: Model. Simul. Mater. Sci. Eng., 12 (2004), S371. https://doi.org/10.1088/0965-0393/12/4/S06
• 34)   F. R. N.  Nabarro: Mater. Sci. Eng. A, 234–236 (1997), 67. https://doi.org/10.1016/S0921-5093(97)00184-6
• 35)   N.  Hansen: Scr. Mater., 51 (2004), 801. https://doi.org/10.1016/j.scriptamat.2004.06.002
• 36)   C.  Wang,  C.  Zhang,  Z.  Yang,  J.  Su and  Y.  Weng: Mater. Sci. Eng. A, 669 (2016), 312. https://doi.org/10.1016/j.msea.2016.05.069