Effect of intercritical quenching soaking time on cryogenic toughness and internal friction characterization of 9Ni steel

Quenching, intercritical quenching and high-temperature tempering of 9Ni steel is widely used in large-size and high-capacity liquefied natural gas storage tanks due to its excellent mechanical properties at low temperatures. This paper studied the microstructure, internal friction and mechanical properties of different intercritical quenching heat soaking times on the cryogenic toughness of 9Ni steel. The effect of martensitic strip thickness change on thermal stability and morphology of reversed austenite after intercritical quenching heat soaking time of 9Ni steel was revealed. The findings show that the thickness of the martensitic slats increases by 4.5 μm when the intercritical quenching of steel is conducted for 50 min, and the maximum concentration of Ni and Mn reaches 13.5% and 10.25%, respectively. The volume fraction of thin film reversed austenite is about 5%. Different frequency under the condition of isothermal Snoek-Kê-Köster peak changes shows there will be a loss of mechanical energy in the process of phase transition. They are important factors for the maximum volume fraction of 9Ni steel film reversed austenite and the significant improvement of cryogenic toughness.

Because of its safety, reliability, clean environmental protection, and low cost, liquefied natural gas (LNG) is frequently used in various countries [1].With the advancement of the LNG industry, the research and development of high-performance materials used in LNG storage and transportation have been widely discussed worldwide.9Ni steel has good mechanical properties in ultra-low temperature environments.It has become the preferred choice for fabricating LNG storage tank materials.With the increasing demand for LNG storage and transportation, the research and development of 9Ni steel with large size and good low-temperature performance have become thriving research [2][3][4].
Developing low carbon, high nickel steel [5] can be summarized as double normalizing and tempering (NNT), quenching and tempering (QT), and quenching, intercritical quenching and high temperature tempering (QLT) processes.Many studies have shown [6,7] that the low-temperature mechanical properties of low-carbon, high-nickel steels directly depend on the reversed austenite (RA) precipitated along the grain boundary of martensite and prior austenite after tempering.Notably, adding intercritical quenching treatment (quenching between A c1 and A c3 temperatures, referred to as L) to the QT process, intercritical quenching heat soaking time (IQST) is vital in regulating reversed austenite.Nickel-based low-temperature steel treated by the QLT has excellent mechanical properties.Hence, the QLT has gradually become essential in preparing 9Ni steel.Literature [7] shows that 16% Cr-5% Ni super martensitic stainless steel can obtain film and block reversed austenite after tempering at 620 °C for different times, and only part of block reversed austenite is transformed into martensite during the tensile process, and the mechanical stability of steel is poor.Liu et al [8] designed a new process for regulating retained austenite, which produced a high content of retained austenite and improved the strength and toughness of maraging steel.Thin film austenite is the main reason for the unique mechanical properties of steel.Wu et al [9,10] studied the effect of grain size of martensite and retained austenite at different intercritical quenching temperatures, significantly impacting the mechanical properties of 5Ni steel.The published studies have focused on the formation of reversed austenite and the impact of changing the grain on the reversed austenite.Still, the stability attained after the formation of reversed austenite and its related mechanism need investigation.This is conducive to the diversification of tank performance optimization research in the future.
In this study, the quenching, intercritical quenching (QL) process is designed to study the evolution of fresh martensite (FM) in 9Ni steel and explore the influence of IQST on the amount of FM and RA formation.In addition, the morphological characteristics are investigated to reveal the relationship between the thermal stability of reversed austenite and the cryogenic toughness of 9Ni steel under different intercritical quenching holding times.The study provides theoretical guidance for large-scale production of 9Ni steel for LNG storage tanks with improved strength.

Materials and methods
The chemical composition (wt%) of 9Ni steel used in this study was 0.043% C, 0.43% Si, 0.64% Mn, 9.35% Ni, and Fe allowance.The ingot was smelted using a 200 kg vacuum induction furnace.The bar was hot forged into a 125 mm (height) × 200 mm (length) ×150 mm (width) billet.j4 mm × 10 mm specimens were cut from the billet.The phase transition point of the tested steel was determined using the German DIL 805A thermal dilatometer following YB T 5127-1993.The A c1 and A c3 of the tested steel were 538 °C and 685 °C, respectively, with M s at 325 °C.The billet with a thickness of 125 mm was heated at 1250 °C for 1 h using the Φ550 mm test mill.After 10 rolling passes, it was transformed into a 16 mm thick steel plate.The initial and final rolling temperatures were 1200 °C and 1000 °C, respectively.Following the final rolling, a laminar water-cooling treatment was applied.150 mm × 80 mm × 16 mm size samples were taken from the surface of the rolled steel plate, following the rolling direction (RD).These samples experienced the QLT heat treatments in a boxresistant furnace.IQST for 30 min, 50 min, 100 min and 150 min, as shown in figure 1.
The standard tensile (j10 mm × 120 mm) and V-notch impact (10 mm × 10 mm × 55 mm) specimens were extracted from the test steel after heat treatment.The tensile properties at room temperature and lowtemperature impact properties were tested.The tensile test was conducted on a Zwick-Z100 Universal tensile testing machine at 2 mm/min.Subsequently, a standard V-notch impact sample of 10 mm × 10 mm × 55 mm was cut and soaked at −196 °C in liquid nitrogen.Then, the impact test was performed on the JBN-500 impact testing machine.A cross cross-section perpendicular to the RD was cut to prepare the sample.Subsequently, the sample was ground and mechanically polished.The samples were etched using 4% nitric acid alcohol corrosion fluid.The microstructure morphology of the fractured sample was observed after the impact testing using the Quanta400HV field-emission scanning electron microscope.The fine microstructure of tested steel was observed using 10 mm×10 mm×300 μm thin slices by employing a JEOL-2100F transmission electron microscope.The x-ray diffraction spectra were collected using an X'pert PRO x-ray diffractometer.The measurement parameters included a Co target, tube voltage of 35 kV, and tube current of 40 mA.The integrated intensities of the (211) α, (200) γ, (220) γ, and (311) γ crystal planes were compared to calculate the volume fraction of reversed austenite [10].
where V γ is the volume fraction of reversed austenite.I α and I γ are the cumulative intensity of diffraction peaks corresponding to martensite and austenite, respectively.G is the ratio of the integrated intensity factor corresponding to the fcc crystal plane (hkl) γ to the bcc crystal plane (hkl) α , where (hkl) is the corresponding crystal plane index.The sample, 1 mm × 2 mm × 50 mm, completed QLT heat treatment states.The experiment was conducted on the MFP-1000 multifunctional internal friction instrument.The forced vibration frequency is 0.5Hz, 1Hz, 2Hz and 4Hz.The internal friction and modulus variation rules of test steel with different frequencies were obtained.The effect of interstitial C element diffusion on the formation of reversed austenite was further studied.The activation energy of the internal friction peak and the activation energy (H) were calculated using the following expression [11]: where R is the ideal gas constant, T P (K) and f (s −1 ) are the temperature and frequency corresponding to internal friction peaks, respectively.k B is Boltzmann constant, h is Plank constant, and ∆S is entropy change (1.1 × 10 -4 eV K −1 ).

Effect of intercritical quenching heat soaking time on mechanical properties of 9Ni steel
Figure 2 shows the tensile properties of steel at room and low temperatures with the intercritical quenching heat soaking time.From figure 2 (a), it can be observed that with the increase of the heating preservation time, the steel's yield strength first increases and then remains constant, with a maximum value of 649 MPa.The tensile strength increases first and then decreases slightly.When the intercritical quenching heat soaking time is 100 min, the tensile strength is minimum (700 MPa).The elongation of the test steel after the intercritical quenching heat soaking time for 30 to 100 min remains unchanged, and the elongation decreases to 26.5% when the holding time is 150 min.In figure 2(b), at −196 °C, the Charpy impact energy of the test steel first increases and then decreases with the rise of heating preservation time.Although the tensile properties of the test steel are stable at the beginning of the heating preservation, the low-temperature impact work value changes significantly.
Especially when the intercritical quenching heat soaking time is 50 min, the low-temperature impact work is 268 J. Overall, the test steel demonstrates good comprehensive mechanical properties when the holding time is 50 min.Therefore, it is necessary to investigate the mechanical properties of the test steel at the beginning of the intercritical quenching heat soaking time.
Figure 3 shows different intercritical quenching heat soaking times after processing at −196 °C and the morphology of impact fracture.From the figure, it can be observed that when the heat soaking times are 30 min and 150 min (figure 3(a) and (d)), the impact fracture morphology of the test steel contains apparent river patterns.This behavior follows the cleavage fracture characteristics; when the heat soaking time is 50 min (figure 3(b)), the impact fracture morphology of the test steel contains several small and deep dimples.The shallow dimples are diffused, better inhibiting crack propagation and improving the cryogenic toughness of the test steel, thus indicating that the fracture mode is a ductile fracture.When the heat soaking time is 100 min (figure 3(c)), the impact fracture morphology contains tearing ridges and a few deep dimples.The low-temperature impact work sustains only for 50 min, which better stabilizes the cryogenic toughness of the test steel [11].According to the published reports [12], when the intercritical quenching heat soaking time is 50 min, the metastable austenite of the test steel supports the TIRP effect, and more external energy is absorbed in the transition of FCC austenite to BCC martensitic phase, ensuring the toughness and plasticity of the test steel.Excellent mechanical properties are observed during the intercritical quenching heat soaking time of 50 min.

Microstructure evolution at different intercritical quenching heat soaking times
Figure 4 shows the SEM image of the test steel microstructure after intercritical quenching treatment with different heat soaking times.After different heat soaking time treatments, the microstructure of the tested steel comprises polygonal ferrite and lath martensite, and the volume fraction of ferrite increases significantly with the increase of heating preservation time.When the IQST is 30 min, the microstructure transformation of ferrite is insufficient, indicating a minimum volume fraction of about 10%.The average length of the lath clusters and lath spacing are 3.9 μm and 1.7 μm, respectively, as shown in figure 4(a).When the IQST is 50 min, the average size of martensitic slat clusters increases to 4.5 μm, the slat spacing is about 1.3 μm, and soft phase massive ferrite is uniformly distributed around the slat clusters (volume fraction of about 14%), as shown in figure 4(b).When the tested steel experiences an external force, the soft phase massive ferrite effectively alleviates the stress concentration around the hard phase structure, cuts off the elongation of the lath martensite, and effectively hinders crack propagation.Therefore, the tested steel demonstrates excellent cryogenic toughness during heating preservation.As shown in figure 4(c) and (d), with the continuous increase of IQST, the bulk ferrite phase transformation is sufficient.The volume fraction of the ferrite structure increases (The final volume fraction is about 20%), and it no longer disperses evenly and joins together.Therefore, the tensile strength and yield strength are reduced in terms of macroscopic mechanical properties, and the change in elongation after fracture can be related to the coarsening of martensitic lath clusters [9], as shown in figure 5.
Figure 6 shows the SEM images of the QLT for intercritical quenching with different heating soaking times.It can be observed that after tempering treatment, the slat structure of martensite disappears to varying degrees and is broken into a small bulk structure.This structure is accompanied by the formation of reversed austenite, tempered martensite, and cementite, among which the bright white area is the alloyed element enrichment at the grain boundary [13].When the intercritical quenching heat soaking time is 30 min, the volume fraction of the martensite structure is the maximum in the test steel, so under the same tempering condition, the martensite lath structure dominates, as shown in figure 6(a).When the intercritical quenching heat soaking time reaches 50 min, the microstructure (in the bright uniform distribution area) accounts for about 85% of the area.This  shows that reversed austenite volume fraction formation is the largest when the C, Ni, and Mn alloy elements partially gather to martensite lath boundaries [14], as shown in figure 6 (b).The ferrite volume fraction is larger when the intercritical quenching heat soaking time exceeds 100 min.However, the distribution is uneven, and the amount of retained austenite at room temperature is less.Moreover, during the tempering process, cementite precipitation decreases the test steel's impact performance.Figure 7 is a schematic diagram of the evolution of martensite and reversed austenite after the QLT treatment of test steel.When the intercritical quenching heat soaking time is 50 min, polygonal block ferrite blocking martensitic lath and the number of reversed austenite at the interface of the lath increase simultaneously.In contrast, the number of ferrite decreases with the increase of soaking time, and the number of reversed austenite decreases accordingly.The behavior shows that the comprehensive mechanical properties of the test steel are excellent when the intercritical quenching heat soaking time reaches 50 min, and the change in the stability of the reversed austenite after tempering causes a difference in the mechanical properties under the intercritical soaking state.

Reversed austenite content and its stability
Figure 8 shows the XRD characterization of the test steel after QLT treatment and the changing trend of reversed austenite in the test steel before and after soaking in liquid nitrogen.It can be observed that the intercritical quenching soaking at 670 °C (for 30 min) develops a small amount of austenite structure.With increased intercritical quenching soaking time, considerable martensite appears, as shown in figure 8(a).Figure 8(b) shows that before and after immersion in liquid nitrogen, the content of reversed austenite in the test steel after intercritical quenching soaking time for 50 min and 100 min is almost similar, indicating that the reversed austenite after intercritical quenching maintains high stability at low temperatures [15].Literature [16] describes that the change of critical quenching temperature can affect the morphology of slat martensite, and the increase in the number of martensite leads to the formation of fine-grained needle-like reverse austenite, which is the reason for the stability of reverse austenite in 7Ni steel.Considering the use conditions of low-temperature steel, soaking liquid nitrogen can better judge the stability of reversed austenite.Before soaking in liquid nitrogen, the volume fraction of reversed austenite after intercritical quenching soaking for 100 min and 150 min is roughly equal.Still, the austenite in the test steel is significantly reduced after soaking in liquid nitrogen.No change occurs in the reversed austenite during the 50∼100 min soaking, and the alloying elements are enriched to the maximum extent.The degree of alloying elements is reduced after increasing the heat soaking time, and insufficient Ni and Mn elements ensure the stability of the reversed austenite.In addition, the reversed austenite maintains a stable volume fraction after 100 min.After soaking in liquid nitrogen for 100 min, the test steel becomes unstable with reversed austenite.A few reversed austenites suffer martensite transformation at low temperatures, which reduces the stability of reversed austenite after 100 min, consistent with the published literature [16].Finally, the degree of segregation of alloying elements in the L-stage affects the nucleation of reversed austenite in the two-phase region (α + γ), and the content of reversed austenite determines the stability.

Partition mechanism of Ni and Mn elements
The earlier studies have shown that the volume fraction of reversed austenite determines the cryogenic toughness of the test steel after QLT treatment, and the stability of reversed austenite is estimated by the partition of Ni and Mn elements in the intercritical quenching process [17][18][19][20].Figure 9 shows test steel microstructure analysis after different intercritical quenching soaking times.Literature [20] shows that the increase of intercritical quenching temperature is conducive to the reversed austenite segregation of Ni and Mn elements.The reversed austenite in tested steel preferentially nucleates and grows along the boundary of high dislocation density lath martensite, where Ni and Mn elements continuously diffuse from BCC structure ferrite to FCC structure austenite.An element enrichment area with relatively high concentrations of Ni and Mn is formed through short-distance diffusion.Thus, a short rod and blocky shape reversed austenite is formed at the  boundary of martensitic slats.This paper shows that intercritical quenching soaking time can also affect the composition of Ni and Mn elements.
When intercritical quenching soaking times are 30 min, 100 min, and 150 min, respectively, the steel block experiences reversed austenite, as shown in figures 9(a), (c), and (d).However, when the soaking time is 50 min, in addition to the bulk austenite, the film austenite is also formed in the test steel (figure 9(b)), demonstrating continuity.Continuous thin-film austenite has higher carbon content, stability, and hardness than bulk austenite [13].Therefore, shear strain does not occur in thin-film reversed austenite.Under low temperature and external loading, the more driving force transforms relatively stable thin-film reversed austenite into martensite [21], which corresponds to the low-temperature properties of test steels treated with different soaking times after intercritical quenching (figure 2(b)).
The changes of Ni and Mn elements in the reversed austenite and SEM structures of the test steel were analyzed by EDS.During the intercritical quenching soaking time (from 30 to 150 min), the Ni and Mn contents in reversed austenite changed by 8.52% to 9.18% and 0.74% to 0.9%, respectively, and the contents of Ni and Mn in the matrix were changed by 9.64% to 8.26% and 0.96% to 0.75%, respectively, as shown in figure 10.During the intercritical quenching, Ni and Mn elements were redivided between ferrite and austenite and finally enriched into austenite [22].As shown in figure 10(a), with the heating preservation time increasing to 50 min, the Ni element in reversed austenite increased to the peak value of 13.46%, then decreased with the soaking time increasing.Mn element showed the same trend, when the soaking time was 50 min, Mn element content reached its peak (10.5%), as shown in figure 10(b).In contrast, the degree of change of the Mn element is greater than the Ni element, which indicates that the Mn component is more sensitive in the high-temperature stage.It can be concluded that the enrichment of Ni and Mn elements during intercritical quenching soaking time is a crucial factor affecting the stability of reversed austenite in test steel [23].This phenomenon is related to the changing trend of the volume fraction of reversed austenite (figure 8).According to the selected area electron diffraction (SAED) and EDS analysis of the TEM fine structure of the test steel, SAED results show the presence of austenite (points a to d).EDS results showed that the Ni and Mn contents in reversed austenite are much higher than those in steel components.The Mn content ranges from 7.95% to 8.46% at points a to d, respectively, as shown in figure 11.Therefore, the degree of nickel enrichment of reversed austenite depends on the intercritical quenching holding time of the test steel precision, which can prove that Ni and Mn elements ensuring the stability of reversed austenite.However, it is essential to study the influence of the interstitial C element on the reversed austenite phase transition.

Internal friction characterization
Literature [22,24] shows that the stability of reversed austenite is affected by the diffusion of Ni and Mn elements.C also affects the formation of reversed austenite by influencing the movement of the Cottrell atmosphere.Internal friction experiments were conducted under different forced vibration frequencies to study the composition of interstitial C elements during the formation of reversed austenite.The internal friction study can accurately represent the influence of the atomic diffusion process on the material's internal structure.In the internal friction experiment, the internal structure defects caused by the vibration of the material produce inner friction peaks, resulting in internal friction caused by the dislocation movement in the crystal.Internal friction peaks due to dislocation motion include the Snoek-Kê-Köster (SKK) peak, the Snoek peak, and the Kê peak.The SKK peak is the internal friction peak caused by different dislocations and the precipitated interaction between the dislocation and carbide.Snoek peak is the internal friction peak caused by the interaction between solid solution atoms and dislocation.Kê peak is the internal friction peak caused by grain boundary structure, grain boundary area, and grain boundary movement [25].
Figure 12 shows the internal friction peak curve of the test steel under different forced vibration frequencies.According to equation (2), the activation energies of different frequencies are 1.3003 ev, 1.2720 ev, 1.2438 ev and  1.2156 ev, respectively.It can be observed that the internal friction of the test steel under different forcing frequencies reaches its peak at 225 min (figure 12(a)).Since the peak temperature does not change with the change of frequency, it implies that the internal friction peak at 225 min is the SKK peak [25].The change curve of the internal friction shows that when the test steel experienced phase change, the internal friction decreased sharply and gradually compressed with the extension of holding time.The internal friction can be measured by the following expression using the original definition of internal friction [26]: where ∆W is the mechanical energy consumed per unit volume of the sample in a vibration period, and W is the maximum potential energy that the sample stores per unit volume during the vibration period.The partition of alloy elements also consumes a certain amount of mechanical energy.During the generation of the phase transition peak, the C atom promotes the reversed austenite, which maximizes the phase transition driving force.However, the sharp decline in the internal friction peak indicates that the element partitioning consumes much mechanical energy during the phase transformation.With prolonged soaking time, the loss of mechanical energy decreases, and the element partitioning gradually reaches homogenization [27].In this case, the 9Ni steel exhibits the most reversed austenite content under 50 min holding time.Therefore, the synergistic action of C, Ni, and Mn elements in the L stage supports the formation of stable austenite and causes the change of the austenite content with the intercritical quenching soaking time.

Conclusions
(1) After 50 min of the IQST, the volume fraction of reversed austenite in the test steel is the largest, mostly a thin film shape.It reduces crack growth and improves cryogenic toughness with a low-temperature impact energy of 268 J.When IQST from 100 to 150 min, 20% ferrite structure and cementite precipitation, reversed austenite content reduction.This is the cause of the decrease in the low-temperature impact performance of the tested steel.
(2) The cryogenic toughness of the test steel is figured out by the stability of the reversed austenite.When the IQST is 50 min, the volume fraction of reversed austenite in the test steel is up to 5.4%, increasing its stability and ensuring the cryogenic toughness of the test steel.When the IQST is 50∼100 min, the diffusion and enrichment behavior of Ni and Mn elements affect the reversed austenite content of the test steel.When the IQST exceeds 100 min, the unstable reversed austenite is transformed into martensite, resulting in the loss of cryogenic toughness of the test steel.
(3) The change of SKK peak at different vibration frequencies shows that a large amount of mechanical energy is consumed in the partition of C element into reverse austenite during the phase transition process after critical quenching, and the driving force of phase transition is the largest.The disappearance of the internal friction peak indicates that the element partitioning gradually reaches homogenization.

Figure 2 .
Figure 2. (a) tensile properties at room temperature and (b) low-temperature impact energy with IQST of the test steel.

Figure 3 .
Figure 3. Fracture morphology of the test steel under different soaking times.

Figure 7 .
Figure 7. Reversed austenite evolution of 9Ni steel under the QLT.

Figure 8 .
Figure 8.(a) XRD characterization of test steel after QLT treatment, (b) reversed austenite content before and after soaking in liquid nitrogen.

Figure 9 .
Figure 9. Microstructure and EDS analysis of test steel after different IQSTs.

Figure 10 .
Figure 10.Changes of Ni and Mn elements in SEM structures at different soaking times under EDS analysis.

Figure 11 .
Figure 11.Changes of Ni and Mn elements after EDS analysis of TEM images of reversed austenite with different soaking times.

Figure 12 .
Figure 12.Internal friction peak curve of test steel under different forced vibration frequencies.