Study on thermal deformation behavior and microstructure evolution of P550 high nitrogen austenitic stainless steel

The high nitrogen austenitic stainless steel is widely used as power generation and geological exploration equipment materials because of its excellent strength, corrosion resistance and non-magnetic. In this paper, the mechanical behavior and microstructure evolution of P550 steel in the range of 900 °C–1200 °C and 0.001–10 s−1 deformation conditions were studied by physical and heat treatment simulations, metallographic observations and thermal processing maps. The results showed that the flow curves quickly reach the peak and then soften to a steady state, which indicates dynamic recrystallization (DRX) behavior. DRX is easy to occur when the deformation temperature is above 1080 °C. The activation energy of the forged P550 stainless steel was calculated as 519 kJ mol−1. There is a positive correlation between the peak stress, DRX critical stress, strain and Z value of the tested steel. The instability of the tested steel is easy to produce in the high strain rate region and low temperature region during hot working. Crack germinates and expands preferentially at the ‘necklace structure’ of inadequate dynamic recrystallization. Under the deformation state of 0.001 s−1, coarse crystals and mixed crystals are easily emerged during subsequent heat treatment. Combining the hot working map, the maximum deformation resistance and the grain evolution behavior during hot working and heat treatment, the suggested working window is T = 1020 °C–1200 °C and έ = 0.01–1 s−1.


Introduction
The Cr-Mn(-Ni)-N series high nitrogen austenitic stainless steels are extensively used as equipment materials for retaining ring for thermal power generator, geological exploration and oil and gas development [1,2]. With the depletion of near surface oil resources, deep and ultra deep layers will be the main positions for major oil and gas discovery in the future. The harsh environment of deep drilling puts forward more stringent requirements for the comprehensive performance of high-end non-magnetic steel forging materials used in the above mentioned equipments. At present, the annual demand for high-strength non-magnetic steel in the world market is more than 10000 tons and increases year by year. In addition, this kind of forging is a high value-added product, with a price of nearly 14000 dollars/ton. And they are consumed greatly in the process of oil drilling, with a high replacement rate which brings significant economic benefits. Accordingly, it is very meaningful to develop such high-end materials and research the matching manufacturing processes.
The typical production process of high nitrogen stainless steel forgings is: electric arc furnace steelmakingmultistage refining-(electroslag remelting) -open forging (fast forging machine) -rough forging (diameter forging machine) -precision forging (diameter forging machine) [1,[3][4][5]. The matching and control of key alloy elements such as C, N, Mn, Cr, Mo, especially the control of the lower limit of N element [6] is the key to obtain excellent strength, toughness, high corrosion resistance and ultra-low magnetic permeability of high nitrogen austenitic stainless steel [7,8]. The high nitrogen steel has high deformation resistance during hot forging because of its large number of alloy elements and high N content, which results in a narrow hot processing working window [8][9][10]. In the actual forging process, forging cracks, coarse grains, mixed grains and other problems are easy to occur [9,11,12], and even the forgings will be scrapped in serious cases. This not only reduces the production efficiency and increases the production cost, but also poses a great threat to the service safety of equipment. Poor hot workability and difficulties in controlling the microstructure and property are the main subjects in the production of high nitrogen steel heavy forgings. Therefore, it is necessary to study the thermal processing parameters through physical simulation and heat treatment simulation tests to optimize the forging process, so as to accurately control the grain size and mechanical properties of the forgings, and prevent forging cracks to ensure its integrity. In addition, there are few systematic reports on the grain evolution behavior of high nitrogen steel in various hot deformation states after solution treatment, which is critical to control the uniformity of grain size in engineering. Based on the above considerations, this paper takes P550 product forging material (N content is 0.68 wt%) as the research object, studies its thermal deformation behavior and microstructure evolution law, and then obtains the best hot working process window, in order to provide theoretical guidance for formulating reasonable hot working process parameters to improve the quality of such forgings.

Materials and methods
The P550 high nitrogen steel used in the test is a solid forging with a diameter of 190 mm and a length of 7500 mm. And the chemical composition measured using chemical analysis and X-ray fluorescence spectroscopy analysis methods is as follows (wt%): 0.05 C, 0.15 Si, 18.60 Cr, 20.71 Mn, 2.01 Ni, 0.57 Mo and 0.68 N. Compression test specimens (Φ10 × 15 mm) were processed axially along the near surface of the P550 forging. Before processed, the large cutting sample shall be kept at 1170°C for 2 h in a resistance furnace to coarsen grains. Hot compression tests were carried out on a Gleeble 3800 thermal/mechanical simulation testing machine. High temperature tensile test was conducted on a DDL 100 testing machine. Firstly, the specimens were quickly heated to 1200°C for 60 s to preheat, then cooled to each deformation temperature (900°C∼1200°C) at 10°C s −1 , then compressed to the strain of 0.7 at the strain rate of 0.001∼10 s −1 , and quenched immediately. The compressed Gleeble specimens were sectioned along the deformation direction by wire cutting, and one of the faces with the largest deformation area was taken for metallographic observations. The area calculation software was used to calculate the recrystallization percentage of each deformed specimen. The other half of the sample was subjected to simulated solution treatment at 1080°C for 1 h in a resistance furnace, and water cooling after heat treatment. Also, conducting metallographic observations on the maximum deformation area of the profile, and the grain size was measured using the line intercept method. The polished surface was etched with 1.5 g KMnO 4 + 10 ml H 2 SO 4 + 90 ml H 2 O solutions (60°C) for 0.5 h, and immersed in 10% oxalic acid solution to remove the surface corrosion residue. The microstructures were observed with Axiovert 200 MAT optical microscope.

Results and discussion
3.1. The stress-strain curves It is well known that the microstructure evolution of metal materials during deformation can be reflected by the changes of stress and strain to a certain extent. The true stress-strain curves of the tested steel under different deformation conditions of 900°C∼1200°C and 0.001∼10 s −1 are shown in figure 1. As we can see, these curves exhibit peaks and softening to a steady state, which is consistent with the plastic strain behavior of low stacking fault energy metals. The evolution characteristics of these flow stresses indicate that DRX occurs in P550 stainless steel during hot compression. Temperature and strain rate play a decisive role in the evolution of flow stress during hot working. The curves in figure 1 show that, within the range of test conditions, when the deformation temperature is from 1200°C to 900°C and the strain rate is from 0.001 s −1 to 10 s −1 , the peak stress of the tested steel is gradually increased. At the initial stage of plastic deformation, the changes of stress and strain are positively correlated, and the stress increases rapidly. At this initial stage, work hardening plays the most important role in the rapid increase of stress [13]. In the next stage, with the further increase of the strain, the stress growth slows down and the work hardening rate decreases. This is mainly because the softening effect of DRX partially offsets the hardening effect caused by hot working. Then, with the further increase of the strain, the dislocation density continues to rise, and the softening effect of DRX becomes more and more significant until it is equal to the work hardening effect. At this time, the flow stress tends to a relatively stable value, indicating that the curve enters the steady state stage. On the other hand, under the same strain rate, the work hardening rate is negatively related to the deformation temperature, that is, the higher the temperature is, the lower the hardening rate is, and the greater the material softening degree is. It can also be seen from the rheological curve that the deformation rate has a significant effect on the amount of deformation required to reach each stage. The increase of the strain rate makes it more difficult to achieve the peak stress, and the material will enter the steady rheological stage later. In addition, it is worth mentioning that compared with the 0.001∼1 s −1 deformation state, the flow stress instability is more sensitive to the 10 s −1 deformation condition. This is mainly due to the adiabatic temperature rise effect of metal materials during thermal deformation [14]. It is apparent that, with the increase of strain rate, the influence of adiabatic temperature rise on the rheological curve will be more remarkable.

Hot deformation equation
According to the stress-strain curves, the thermal deformation process is notablely affected by deformation temperature and strain rate. Since the high temperature plastic deformation process of P550 stainless steel is a heat activated process, the evolution law of its flow stress can be described by a modified Arrhenius function containing various deformation parameters. During forging of P550 steel at all stress levels, the relationship between σ, T and έ can be described by the classical hyperbolic sine formula [15]: In formula (1), the Q is the activation energy of thermal deformation (J mol −1 ), and the σ is the flow stress (MPa). A, α and n is the material constant independent of deformation temperature, and R is the gas constant (8.314 J·mol −1 ·K −1 ). The σ can be peak stress, steady flow stress or flow stress corresponding to a specified strain. The peak stress value is selected as the stress value in this study. According to literature [16], the relationships between σ p -lnέ and lnσ p -lnέ can be achieved by simplifying the hyperbolic sine function. According to the peak stress data of P550 steel under various conditions of hot compression, to draw σ p -lnέ and lnσ p -lnέ relationship diagrams respectively, as shown in figures 2(a) and (b). The graph results show that σ p -lnέ and lnσ p -lnέ are approximately linear. Carrying out linear regression on the data with the least square method, and then determining the material constant through calculation. The α parameter value is 0.005819831 (MPa −1 ).
According to the rheological curve datas in figure 1, natural logarithms are taken on both sides of formula (1) to obtain ln[sinh(ασ p )]-lnέ and ln[sinh(ασ p )]−1/T relationship curves respectively, as shown in figures 2(c) and (d). The value of each parameter is obtained by linear regression with the least square method: A is 1.38 × 10 19 , n is 5.535597747, Q is 519008 J mol −1 . Substituting the obtained values into Formula (1) Q in the formula is a critical parameter reflecting the thermal deformation ability of materials. The Q value is negatively related to the thermal deformability of the material, which means the smaller the Q value is, the better the deformability of material is. The Q value of P550 steel measured in this study is 519008 J mol −1 , which is higher than the hot deformation activation energy of the traditional hard deformation 321 austenitic stainless steel containing Ti and the nitrogen controlled 316LN austenitic stainless steel (N content is 0.15 wt%). The Q values of the latter two are about 470000 J mol −1 [17,18]. However, compared with the high nitrogen Cr-Mn austenitic stainless steel containing V and Nb, the activation energy of P550 hot deformation is significantly lower than that of its 767000 J mol −1 [19]. From the perspective of alloy composition, this is mainly due to the high N content of P550 stainless steel used in the test, but there is no precipitation during high-temperature deformation. Therefore, in terms of hot deformation ability, the P550 steel is inferior to the 321Ti and 316LN austenitic stainless steels. But compared with high nitrogen austenitic stainless steel containing V and Nb, the P550 steel has stronger high-temperature deformation ability and is easier to be formed by hot working.

Z parameter and its relationship with peak stress
The temperature compensated strain rate factor (Zener-Hollomon parameter), namely Z parameter, is introduced to characterize the comprehensive influence of various deformation parameters on deformation resistance. According to the above test datas, the Z parameter expression of the tested steel is: The peak stress σ p , one of the primary reference indexes for metal hot working, directly determines the selection of forging equipment and the establishment of relevant force and energy parameters to a large extent. The Z value of P550 steel in different deformation states can be calculated by equation (3), and then the correspondence between lnZ and σ p will be identified, as is shown in figure 3. The correlation coefficient (R 2 ) of the linear fitting data is 0.978. The relationship between the peak stress and lnZ of the P550 stainless steel as forged is as follows: The Z parameter comprehensively reflects the deformation behavior of P550 steel under different deformation conditions. It can be seen intuitively from figure 3 that, there is a good linear relationship between σ p and lnZ. In other words, there is a positive correlation between σ p and Z values. This is mainly because the smaller the Z value is, the higher the mobility of dislocations and grain boundaries in the material is, which is more conducive to the occurrence of DRX. The more developed the DRX structure is, the better the softening effect is. On the contrary, the larger the Z value is, the smaller the driving force of DRX occurs until DRX does not occur, so the higher the peak stress is.

The critical stress, critical strain of DRX and their relationship with Z parameter
Poliak [20] and Abbas [21] believed that the critical stress (σ c ) of DRX is related to the inflection point of the relationship curve of work hardening rate (θ) and flow stress (σ). The relationship between θ and σ can be expressed by equation (5): Here d d , q s = έ / and A, B, C and D are constants related to the deformation conditions. According to the partial derivative of equation (5), the relationship curve between work hardening rate and flow stress of P550 stainless steel can be obtained. The curve obtained is a quadratic parabola, which corresponds to a specific quadratic equation. The minimum value of the quadratic equation is the critical stress of DRX, i.e.
The relationship between σ c and lnZ of P550 stainless steel can be obtained by the above mentioned approaches, as is shown in figure 4(a). After polynomial fitting, the R 2 is 0.874, and the relationship between σ c and lnZ is as follows: It can be seen from figure 4(a) that, with the increase of Z value, the critical stress of DRX of the tested steel enlarges accordingly. In addition, it is worth noting that the greater the Z value, the larger the growth trend. This means that DRX is more hard to occur at higher strain rates and lower deformation temperatures, which has been verified by the flow stress curves in figure 1.
According to Formula (6), calculating the critical stress of DRX under each deformation condition, and then determining the corresponding critical strain of DRX through the stress-strain curves. As shown in figure 4(b), the R 2 after polynomial fitting is 0.743, and the relationship between ε c and lnZ is as follows: It is easy to see from the figure that the critical strain of DRX of P550 steel is also positively correlated with the Z value. That is, when the deformation temperature is lower or the strain rate is higher, the critical strain required for the occurrence of DRX is larger. Furthermore, combined with the stress-strain curves in figure 1, it can be found that the actual starting point of DRX of the tested steel is earlier than the peak point determined from the stress-strain curve, which further proves that the peak point cannot be used as the starting point of DRX.

Hot processing maps
The microstructure evolution rule in the thermo-mechanical processing of materials is directly reflected by the hot working map. It is an effective tool to study the relationship between the hot deformation behavior of materials and various process parameters, and is used to assist in selecting technological parameters and avoiding hot working defects. At present, the hot working map [22] based on the dynamic material model is widely used in stainless steel, nonferrous alloys, superalloys and other materials. According to this theory, the strain rate sensitivity index (m) of P550 steel under a certain strain can be obtained, and the hot working map was achieved from the power dissipation ratio (η) and plastic flow instability criteria (ξ). The η and ξ can be expressed by the following formulas: Selecting the test datas when the strain is 0.6 to make the hot working map of P550 steel, as shown in figure 5. The shaded area corresponds to the rheological instability scope, and the numbers on the contour represent the percentage of dissipation efficiency. It can be seen that there was a high power dissipation efficiency area in the lower part of the hot working map, and there were two rheological instability areas in the upper part. The deformation condition corresponding to the high power dissipation efficiency area was about 950°C∼1150°C, 0.001∼0.01 s −1 , and the maximum η value was about 46%. Generally speaking, the DRX area is considered as a safe domain for hot processing. The higher the η value is, the more likely the DRX occurs in the material. Therefore, the alloy possesses a good machinability in the area with larger η value. But it doesn't mean that the higher the η value is, the better the thermal deformation capacity of the material is. The η value may also be high in the instable zone. It is clear from figure 5 that there is a certain negative correlation between the power dissipation efficiency and the strain rate, and the sensitivity of power dissipation to the change of strain rate increases with the decrease of temperature. Rheological instability is mainly related to the low power consumption. The two deformation ranges corresponding to the rheological instability zone of the tested steel were 900°C∼1020°C, 0.1∼10 s −1 and 1075°C∼1180°C, 1∼10 s −1 . This high strain rate area shall be avoided during hot working of the tested steel. Therefore, the optimal hot working range of the tested steel is 920°C∼1020°C, 0.001∼0.1 s −1 and 1020°C∼1200°C, 0.001∼1 s −1 from the view of hot working map alone. In addition, the rheological curves ( figure 1) show that when the temperature exceeds 1100°C and the strain rate is Figure 5. Hot working map of tested steel at a strain of 0.6. 0.001 s −1 , the maximum value of deformation resistance is lower than 50 MPa, under which the risk of cracking in the actual forging process is higher. Considering the maximum deformation resistance, combined with the rheological instability range under different deformation conditions, the recommended hot working range of the tested steel is T = 1020°C∼1200°C, έ = 0.01∼1 s −1 .

Microstructure observation
The thermal deformation process plays an important role in the grain evolution of materials. Figure 6 shows the metallographic microstructure of the maximum deformation zones of the tested steel at 950°C or 1150°C under different strain rates. It is apparent that the microstructure varies greatly under different deformation conditions. As is well known, when DRX occurs in materials, the deformed original coarse grains will be replaced by fine equiaxed grains after recrystallization. Figure 6(a)-(c) (950°C) show some fine DRX grains produced at the original austenite grain boundary, presenting typical necklace like DRX, which is the structural feature of rheological instability [10,17]. In figure 6(d)-(f) (1150°C), when the deformation temperature is high, the proportion of fine equiaxed grains and the average grain size increase significantly, and the original coarse grains have been basically disappeared, which indicates that DRX is more sufficient in the high-temperature deformation state of the tested steel.
When the strain rate is constant, the deformation temperature has a great influence on the grain evolution behavior of P550 steel: the deformation temperature is positively related to the content and grain size of DRX, that is, the higher the temperature is, the more the content is, and the larger the grain size is. Under the low strain rate, the higher the temperature is, the less the original grains remain. At this time, due to the occurrence of complete DRX, the tested steel can obtain a more ideal grain structure. When the deformation temperature is constant, the strain rate also has a significant effect on the microstructure of the tested steel: at the same deformation temperature, there is a positive correlation between the strain rate and the dislocation accumulation rate of the material. In other words, the higher the rate is, the faster the accumulation rate is, thus providing more nucleation sites for DRX. Simultaneously, DRX grains do not have enough time to grow under the high strain rates. Therefore, the higher the strain rate, the smaller the DRX grain size. At the low temperature, under the high strain rate, there are more elongated primary austenite grains without DRX in the metallographic structure of the stainless steel, and only a few fine DRX grains are observed at the grain boundary, forming a 'necklace structure' [23]; At the high temperature, under the high strain rate, there is still a 'sub necklace structure' in the microstructure of the tested steel, i.e., a ring of fine equiaxed DRX grains are formed around the original austenite grain boundary, which is in sharp contrast to the larger grains formed inside the original crystals. The higher the strain rate is, the greater the difference of grain structural is. This grain size inhomogeneity between the intragranular and grain boundary of the original austenite grain after deformation will also significantly expand the sensitivity of forging cracking of the tested steel. As shown in figure 7, the microstructures of the fracture and its vicinity of the tested steel under 950°C/1 s −1 tensile condition were examined. There are many dimples with different sizes on the macro fracture surface, and the fracture mechanism presents the typical characteristics of microporous aggregation fracture. The metallographic photos of longitudinal section of fracture show that some twins are appeared in the matrix. Many cracks are emerged in the structure near the fracture, distributed along the 'necklace structure'. During the deformation process, inadequate dynamic recrystallization occurs, easily lead to stress concentration and strength and plasticity reduction in the fine recrystallized grains. These micropores germinate at the mixed grain boundary and expand along the grain boundary during high temperature tension, causing intergranular or transgranular fracture of the tested steel.
In order to more intuitively display the DRX situation of the tested steel under various deformation conditions, the DRX percentage under each state is enumerated by calculating the DRX grain area. The statistical results are shown in figure 8. The contour line in the figure represents the percentage of DRX. The light gray area corresponds to the area without DRX, and the dark gray area corresponds to the area with complete DRX. The results show that the percentage of DRX increases with the deformation temperature raised and the strain rate decreased. When the deformation temperature is lower than 950°C and the strain rate is higher than 0.56 s −1 , DRX is difficult to occur in the tested steel. When the deformation temperature is above 1080°C, DRX is easy to occur, and complete DRX is more likely to happen with the diminishing strain rate.
Furthermore, it is very essential to investigate the precipitation behavior of the second phase during the deformation process of high nitrogen steel. Figure 9 shows the microstructures of tested steel under 950, 900°C/0.001 s −1 deformation conditions. As can be seen, no obvious second phase precipitate was found in the tested steel under 950°C state, while many precipitated phase particles were observed along grain boundaries under 900°C deformation condition. EDS analysis shows that the precipitated phase was Cr 2 N. Due to the  presence of these nitrides, stress concentration occurs at grain boundaries. The forging process of large high nitrogen steel forgings often takes a long time. With the increase of the residence time of the low-temperature section on the surface of the forging, the number and size of the second phase enlarge. The presence of a large number of larger and unevenly distributed Cr 2 N phases will intensify the proliferation, aggregation, and expansion of micropores at grain boundaries, leading to faster crack initiation rate and ultimately leading to intergranular cracking. It is necessary to avoid forging at this lower temperature during the forging process.
The rheological instability zone in the low temperature section shown in figure 5 overlaps with the zone without DRX or the zone with incomplete DRX of the tested steel. That is, the flow instability zones are located in the high strain rate and relatively low temperature sections. This is mainly because DRX is a thermal activation process. The energy storage in the crystal is low at the low temperature, which is not enough to produce complete DRX. At the same time, the reduction of the grain boundary migration ability also has a negative impact on the recrystallization expansion. Therefore, there will be a large structural divergence in different parts of the forging, leading to the appearance of uneven deformation in the actual forging process. At the low temperature, the uneven surface state of the forging further expands this kind of deformation structural divergence, which eventually brings about the rheological instability. But, instead, under the high temperature and high strain rate, due to the high temperature, short deformation time, and poor heat conduction performance of the high nitrogen steel, a large amount of heat is not dissipated in time, which results in the increment of the temperature of the core of forgings and the reduction of deformation resistance. In addition, the high strain rate would increase the incongruity of deformation in the incomplete DRX region with 'sub necklace structure' characteristics, which ultimately prompts a higher risk of forging cracking of P550 steel under the high temperature and high strain rate deformation conditions. The solution treatment after hot working is considered to be the key process for controlling the grain size of tested steel forgings, since the subsequent cold-worked will not alter the grain size of forgings. Therefore, the compression samples under different deformation conditions were subjected to simulated heat treatment at 1080°C for 1 h to investigate the grain evolution behavior after solution treatment. The average grain size after simulated solution treatment measured by the intercept method is shown in figure 10. It can be seen that after solution treatment, the average grain sizes of the hot deformed samples, except 0.001 s −1 deformation state, are roughly the same. And the average grain sizes are 76∼107 μm, the discrepancy of grain size grade is within 1, which can meet the requirements of the grain size grade 3∼5 of forging products. Under the 0.001 s −1 condition, when the deformation temperature exceeds 1050°C, the grains of the tested steel grow rapidly after the solution treatment. It is mainly related to the low strain rate resulting in sufficient time for DRX grains to grow during the deformation process, which gives rise to the rapid growth of the original DRX grains with larger size and less quantity during the solution treatment process. This discrepancy in grain growth trend in the subsequent heat treatment process caused by deformation conditions is easy to cause problems of coarse grains and mixed grains. In the actual hot working process, the deformation resistance under this low strain rate condition is also small, and hot deformation is prone to instability. Therefore, comprehensive consideration should be taken to avoid forging under this low strain rate deformation condition. This is consistent with the hot working range of the tested steel recommended above.