Effect of superplastic deformation on precipitation behavior of sigma phase in 3207 duplex stainless steel

Jingkun Li ,Xueping Ren,＊ ,Ynling Zhng ,Hongling Hou ,Xiodn Go,c

a School of Materials Science and Engineering,University of Science and Technology Beijing,Beijing,100083,China

b AVIC Manufacturing Technology Institute,Beijing,100024,China

c Institute for Advanced Materials and Technology,University of Science and Technology Beijing,Beijing,100083,China

Keywords:Duplex stainless steel Superplastic deformation Sigma phase (σ)Precipitation mechanism Hardening effect

ABSTRACT Owing to excellent strength,toughness and corrosion resistance,duplex stainless steels (DSS) are widely used in constructional and petrochemical applications.Sigma phase,which has detrimental impact on the properties,is readily precipitated during hot working of DSS.However,precipitation behavior of sigma phase during superplastic deformation,which is the most significant processing method of DSS,is still unclear.In the current study,the effect of superplastic deformation on the precipitation behavior of sigma phase was investigated in 3207 duplex stainless steel.The result shows that superplastic deformation could prevent sigma phase precipitation generally by increasing mobility of grain boundaries and decreasing misorientation of the sigma phase boundaries,resulting in some sigma phase precipitated on the twin boundaries.Most of the sigma phase precipitated on ferrite-austenite interface with misorientation of 20–25°,while it precipitated in ferrite or austenite with the misorientation of 40°–45°.The orientation relationship between sigma phase and matrix matched well in austenite and on the ferrite/austenite interfaces,while it showed a small misfit in ferrite.The prevention effect of the superplastic deformation on the sigma phase precipitation was beneficial to quasi stable deformation stage,resulting in longer elongation.

1.Introduction

Owing to attractive combination of strength and toughness [1],as well as excellent resistance to corrosion [2],duplex stainless steels are seeing increasing use in constructional and petrochemical applications,such as cooling water pipes,gas tanks and nuclear reactors [3,4].The properties of duplex stainless steels are mainly due to their dual-phase microstructure of body-centered cubic (b.c.c.) ferrite (δ) and face-centered cubic (f.c.c.) austenite (γ) [5].However,the mechanical properties and corrosion resistance of duplex stainless steels are also adversely affected by the formation of intermetallic phases [6].In particular,sigma phase(σ)is readily precipitated,which has detrimental impact on the properties of duplex stainless steels[7].Pezzato et al.[8]showed that pit will nucleate at the primary austenite/secondary austenite interface,thus the presence of secondary phases,e.g.the sigma phase,could reduce the corrosion resistance in DSS.Hosseini et al.[9]emphasized that micro cracks could propagate where the sigma phase precipitated,and selective corrosion occurred in the regions containing fine coral-shaped sigma phase.Therefore,it is important to understand precipitation behavior of sigma phase.

Sigma phase is a hard brittle intermetallic phase,which has higher Cr contents than the matrix[10].Previous studies showed that sigma phase preferentially nucleates on interfaces between ferrite and austenite during hot working[11].The effect of steel chemistry,heat temperature and holding time on sigma phase static precipitation behavior during solution and aging process has been extensively studied[12].Despite significant researches on sigma phase,little is known about dynamic precipitation behavior of sigma phase during hot deformation,which is the aims of this investigation.

Superplasticity is an important property of duplex stainless steels,and more than 1000% elongation could be obtained under certain circumstance[13].Owing to its excellent superplasticity,duplex stainless steel can be hot formed through different forming methods,such as forging,extrusion and diffusion bonding[4].Li et al.[14]investigated the effects of deformation temperature and initial microstructure on superplastic deformation of 2205 duplex stainless steel,and they suggested that the increment of the strain could lead to precipitation of sigma phase during the superplastic deformation progress.However,the precipitation behavior of sigma phase are mainly focus on the static heating process,the effect of superplastic deformation on the precipitation of sigma phase is still unclear.Therefore,the current study aims to investigate the effect of superplastic deformation on precipitation behavior and mechanism of sigma phase,as well as local hardening.

2.Experimental procedures

The as-received material used in the study was a commercially produced 3207 duplex stainless steel with the thickness of 4.0 mm.The chemical compositions were shown in Table 1.After solution-treated at 1250°C for 30 min followed by water quenching,the sheet was then cold rolled into 2.0 mm thickness by 50% reduction.The sheet tensile specimen with the thickness of 2.0 mm and gauge length of 10 mm were machined from the cold rolled sheet.The detailed dimension of the tensile test specimens was shown in Fig.S1a.

Table 1 Chemical compositions of as-received 3207 duplex stainless steel (wt%).

The superplastic deformation was carried out on a SK10-70300 constant temperature tensile testing machine.The temperature control accuracy was ±4°C.Antioxidant was applied on the surface of the specimens before heating.Then the tensile specimen was heated to 950°C and homogenized for 1 min before superplastic deformation.The initial strain rate was 1.5 × 10-3/s,the cross head speed and the temperature kept constant during the deformation process.After fractured,the specimen was water quenched immediately.The length and local width of the specimen were measured,and the elongation was calculated as 944%.Four regions on the specimen were defined in order to investigate the effect of superplastic deformation on the sigma phase precipitation behavior.As shown in Fig.S1b,region 1 was located on the grip of the specimen,without deformation;regions 2 to 4 were located on the gauge of the specimen,with width of 2.56–4.40 mm respectively.

The distribution and crystal orientation of sigma phase were examined by EBSD technique.The EBSD specimens taken from the regions 1 to 4 respectivelywere ground and mechanically polished according to standard procedures and electro-polished for 3–8 min with a solution of chromium anhyd 120 g+orthophosphoric acid (H3PO4) 280 mL +sulphuric acid (H2SO4) 220 mL+deionized water (H2O) 40 mL,at a voltage of 6–8 V and a constant temperature of 80–95°C.The EBSD analyses were carried out with a JEOL JSM-7800F field emission scanning electron microscope (SEM) operated at an acceleration voltage of 20 kV.Oxford Instruments Nordlys Nano EBSD system equipped with Channel 5 software were used for analysis.

The thin foils for TEM observation were machined and thinned by twin jet electropolished in perchloric acid(HClO4)10 mL+ethanol(C2H5OH)90 mL solution at-30°C with applied potential of 50 V,and then examined by FEI Talos F200-field emission TEM operating at 200 kV.

Metallographic specimens of solution-treated samples were ground and polished according to standard procedures and etched by aqua regia.The microstructure of the samples was characterized using an Imager.M2M optical microscope(OM)and a LEO 1450 SEM.

3.Results

3.1.Microstructure of the initial material

The initial microstructure of solution-treated and cold rolled samples obtained by SEM is shown in Fig.S2.The microstructure of solutiontreated sample consisted of ferrite (δ) and austenite (γ).Equiaxed ferrite grains located on coarse austenite matrix uniformly and the fraction of austenite was about 70% measured by Imagetool software(average value of five SEM images).After cold rolling,the grains of both ferrite and austenite were elongated along the rolling direction,and the fraction of the phases did not changed.

3.2.Distribution of sigma phase

Fig.1 shows the distribution and micromorphology of the sigma phase in regions 1–4 after the superplastic deformation.According to the phase maps(Fig.1a,c,1e and 1g),the ferrite was filled with red and the austenite filled with blue.While the high angle boundaries (HAGBs,>10°) was marked as black lines and low angle boundaries(LAGBs,between 2°and 10°)marked as fuchsia lines.The Sigma phase boundaries were marked as green with the misorientation <30°and aqua with the misorientation >30°.The sigma phase formed in different positions is shown in Fig.1.It can be observed that the sigma phase formed not only on the austenite/ferrite interface but also in a single phase (ferrite or austenite)with a spherical or ellipsoid shape.It should be noted that the misorientation of sigma phase in a single phase was much higher than that of on the austenite/ferrite interface in all regions.With the increase of local deformation,the sigma phase had a tendency to coarsen,as shown in Fig.1b,d,1f and 1h.

Although some sigma phase precipitated in ferrite or austenite,most of them precipitated at or near the interface.In the undeformed region 4,the sigma phase mainly precipitated near the austenite/ferrite interface,and it never been observed on twin boundaries,as shown in Fig.1b.However,it is interesting to note that a sigma phase precipitations was observed on a twin boundary in the deformed region 2,as shown in Fig.1d.Previous studies show that the interfaces with low mobility might be preferential nucleation sites of sigma phase [15].Meanwhile,the grain boundary sliding (GBS) during superplastic deformation has a positive effect on improving grain boundaries mobility[16].Therefore,it is reasonable to hypothesize that GBS could prevent the formation of sigma phase on the austenite-ferrite interface.Thus part of sigma phase precipitates have to grow preferencely at twin boundaries with lower mobility rather than some grain boundaries with higher mobility.

In order to further investigate the effect of superplastic deformation on sigma phase precipitation,the phase compositions of regions 1 to 4 are shown in Fig.2.It shows that the fraction of ferrite is 7.73% in undeformed region 1,which was obviously lower than that of 9.47%–10.07%in deformed regions 2–4.While the fraction of sigma phase was 12.85%in undeformed region 1,which is obviously higher than that of 6.86%–8.47% in deformed regions 2–4.Previous study illustrates that sigma phase usually nucleates on austenite-ferrite interfaces and grows into the ferrite as a result of a ferrite →austenite+sigma eutectoid phase transformation [11].Thus the phase transformation behavior was prevented under the effect of superplastic deformation,as shown in Fig.2.

3.3.Misorientation between the sigma phase boundaries

The inverse pole figures (IPF) including orientation information of regions 1–4 are shown in Fig.S3.The grain orientation is described with different colors,while the sigma phase boundaries are marked as black lines.It can be observed that the grains distributed uniform and the orientation was quasi random.However,it clearly showed that the misorientation existed between the sigma phase boundaries.This suggests that the precipitation of sigma phase could lead to misorientation in matrix.

In order to further investigate the misorientation between sigma phase boundaries quantitatively,the misorientation were plotted in Fig.3 for austenite-ferrite interfaces with different precipitation propensities.The average value of the misorientation angle was about 25°in the undeformed region 1.However,it decreased to 21–22.5°in the deformed regions 2–4.Interestingly,it exhibited a similar distribution between average misorientation angles and fraction of sigma phase.Namely,the higher misorientation angles between the sigma phase boundaries were,the larger fraction of sigma phase obtained.It is well known that the GBS and grain rotation during superplastic deformation could weaken the texture intensity,and the grains could also been rotated to a crystal orientation with smaller misfit [17,18].Thus the misorientation angle between sigma phase boundaries might be rotated away from the orientation conducive to the precipitation of sigma phase.

Fig.1.Phase maps (a) (c) (e) (g) and TEM images (b) (d) (f) (h) of regions 1 to 4,respectively.

The misorientation angles of all interfaces exhibited a broad distribution in a range of 0–45°,having peaks at 20–25°and 40–45°,as shown in Fig.3b.Meanwhile,the tendency was similar from regions 1 to 4.The interfaces with the misorientation angles between 40 and 45°revealed a higher tendency to sigma phase precipitation compared to the rest.This misorientation angle range overlaps with the well known Kurdjumov-Sachs (K–S) and Nishiyama-Wasserman (N–W) orientation relationships (ORs) [19].Combined with Fig.1,the sigma phase with K–S and N–W ORs was mainly precipitate in the single phase(ferrite or austenite).While the sigma phase precipitation was promoted in the austenite-ferrite interfaces (with misorientation of 20–25°) that deviate from the K–S and N–W ORs.It is consistent with other reports on duplex stainless steels,where sigma phase mostly precipitated on interfaces far from the K–S and N–W ORs[20].

Fig.2.Phase fractions of regions 1–4.

4.Discussion

4.1.Precipitation mechanisms of sigma phase

As shown in Fig.1,the sigma phase could precipitate in ferrite,austenite and ferrite/austenite interfaces.Furthermore,the precipitation behavior of sigma phase could lead to misorientation in all the three positions.Previous studies on sigma phase precipitation behavior showed that sigma phase precipitation has different orientation relationships with matrix during aging process.When sigma phase precipitate in a single phase,misorientation may near the K–S and N–W ORs;while when it precipitate on ferrite/austenite interfaces,misorientation may far away from the K–S and N–W ORs [19].It can be inferred that sigma phase precipitation behavior may be different in the three positions above.

Thus,in order to investigate precipitation mechanisms of the sigma phase,TEM was used to describe the micromorphology and crystal orientation relationship of the sigma phase.It is well known that the superplastic deformation can be divided into four distinct stages,namely,elastic deformation stage,stable deformation stage,quasi stable deformation stage and localized necking stage [21].Large elongation is obtained in the quasi stable deformation stage [17].Thus the quasi stable deformed region 2 is taken as an example for analysis.

The austenite-ferrite interface with an equiaxed sigma phase precipitation (about 40 nm) was observed by a high resolution transmission electron microscope(HRTEM),as shown in Fig.4a.The result illustrates that the sigma phase had an effect on both ferrite and austenite near the interface.Previous study shows that the sigma phase was enriched in Cr,thus a depletion of Cr could be observed near the interface with sigma phase [5],which may lead to crystal lattice distortion.Therefore,the misorientation between the sigma phase boundaries could be formed,as shown in Fig.1.Williams et al.[22]studied Cr diffusion in ferritic and austenitic steels and found that Cr in ferrite had a diffusion constant D0of 0.15 cm2s-1and activation energy Q of 210 kJ mol-1,whereas Cr in austenite had a diffusion constant D0of 0.27 cm2s-1and activation energy Q of 264 kJ mol-1.The Arrhenius diffusion equation of Cr can be described as DCr=D0e-Q/RT,where R=8.31 J/(mol K),represents the gas constant and T represents the absolute temperature,K.The 3D root-mean-square (RMS) diffusion distance of Cr,=(6DCrt)1/2could be calculated as 74.9 μm in ferrite and 100.2 μm in austenite when the heating time t was calculated as 6360 s.It shows that Cr have a diffusion distance in austenite greater than in ferrite on the experimental condition.So the sigma phase effect zone in austenite(45.7 nm)is wider than that of in ferrite(11.6 nm).

Fig.4b shows the selected area electron diffraction(SAED)pattern of Fig.4a obtained from the Fast Fourier Transform (FFT).The SAED pattern was taken at the interface along [011]σ.The ferrite,austenite and precipitated sigma phase exhibit a cube-cube orientation relationship of (200)δand (020)δ||(311)γ|| (200)σ.It is clear that the orientation relationship between austenite and ferrite is far away from the well known K–S OR.Furthermore,due to the lattice matching relationship,the misortientation between ferrite,austenite and sigma phase is small,as shown in Fig.1.

Fig.5 shows sigma phase in the austenite away from the interfaces.The sigma phase precipitation exhibited equiaxed morphology with diameter about 60 nm,while the interface between sigma phase and austenite was straight,as shown in Fig.5a.Fig.5b shows SAED pattern of white circle area in Fig.5a.The SAED pattern was taken at the interface along [011]σ.The austenite and precipitated sigma phase exhibit orientation relationship ofInterestingly,these orientation relationship is near the K–S and N–W ORs.The K–S and N–W ORs,which differ by only 5.26°,are the most generally used models for sigma phase transformation [11].Thus the peak value of misorientation between 40 and 45°formed,as shown in Fig.4.Fig.5c shows the interface between sigma phase and austenite.The lattice distance of crystal planes near the interface is slightly greater than that inside of the austenite,however,the lattice distortion can hardly been observed.

Fig.6 shows a sigma phase precipitation in ferrite.The precipitation is elliptical with more than 300 nm in the longitudinal axis,and an Orowan loop can be observed around the sigma phase,as shown in Fig.6a.Fig.6b shows SAED pattern of white circle area in Fig.6a,which was taken at the interface along[011]σ.Combined with the HRTEM image of Fig.6b,it is interesting to note that the sigma phase elongated along a direction close to <>σ,which might be attributed to larger precipitates [23].The lattice distortion can be observed along the interface between sigma phase and ferrite.The HRTEM image obtained from magnification of Fig.6c provides further information,as shown in Fig.6d.The crystal planes of ferrite were marked according to crystal lattice distance,while the crystal planes of sigma phase marked according to SAED pattern shown in Fig.6b.It shows that a small misfit of 7.5°existed between()σand (101)δ.It can be assumed that the precipitation of sigma phase in ferrite is more difficult.This inference is satisfy with the sigma phase distribution shown in Fig.1,where less sigma phase can be observed in ferrite.

Fig.3.Misorientation angle characterization:(a) fraction of sigma phase and average misorientation of grains between the intermetallic boundaries;(b) misorientation angle distributions.

Fig.4.HRTEM image of the austenite-ferrite interface with a sigma phase precipitation (a) and related SAED pattern obtained from FFT.

Fig.5.TEM images of a sigma phase observed in the austenite (a) with related SAED pattern (b) and local magnificated HRTEM images (c) (d).

4.2.Hardening effect of sigma phase precipitation

Fig.6.TEM images of a sigma phase observed in the ferrite (a) with related SAED pattern (b) and local magnificated HRTEM images (c) (d).

It is well known that the hardness of materials is generally related to the grain size through a Hall-Petch equation:HV=H0+kHd-1/2,where HVrepresents microhardness of materials,d is the grain size,H0and kHare constants associated with the hardness measurements.In order to investigate strengthening effect of sigma phase precipitation the microhardness of the experimental 3207 duplex stainless steel after solution was tested.The solution was conducted in 1000,1100,1200 and 1300°C for 30 min,in order to obtained equiaxed grains with little dislocations[14].The average grain size of solution-treated samples was measured as 19.3,17.8,7.6 and 7.8 μm respectively,as shown in Fig.S4.Fig.7a shows a plot of HVversus d-1/2for the experimental 3207 duplex stainless steel.The limited data showed the value of H0to be 276.46 HVand kHbe 83.392 HVμm1/2.Thus the local microhardness of the regions 1–4 could be calculated according to Hall-Petch equation.The values were in the range of 309.4–316 HV,which shows small difference.However,the experimental value of regions 1–4 were 372.6,382.3,342.7 and 347.0 HVrespectively,which is higher than the calculated value obviously(Fig.7b).Therefore,it is reasonable to infer that the sigma phase has a strengthening effect on matrix.

Previous studies show that dislocation is an important strengthening mechanism in deformed materials and it may interact with precipitation[24,25].However,it is well known that austenite is a face-centered metal with low stacking fault energy,where dislocation slip is difficult [26].The movement of dislocation is prevented,resulting in a low dislocation density in austenite.Thus only a little number of single dislocation lines can be observed in austenite as shown in Figs.1 and 5.Since the fraction of austenite is more than 80%,the dislocation may have little effect on strengthening of the 3207 duplex stainless steel.

Orowan strengthening,caused by the resistance of closely spaced hard particles to the passing of dislocations,is important in materials with precipitation[27].However,it is not significant if the reinforcing particles arecoarseandthe interparticle spacingislarge[28].Theaveragediameter values of sigma phase (calculated from Fig.1) are 41.3,41.6,49.9 and 397.5 nm for regions 1–4 respectively,which is fine enough.The Orowan strengthening effect can be expressed as equations(1)and(2):

Fig.7.Characterization of microhardness:(a) microhardness HV versus d-1/2;(b) histogram shows theoretical and experimental microhardness.

M is the Taylor factor with a value of 3.1 in steels,λ is average distance of sigma phases measured from Fig.3,G is the shear modulus with a value of 85.5 GPa in duplex stainless steels[29],dpis average particle size of sigma phase,b is the Burgers vector with a value of 0.248 in steels,fVis fraction of sigma phase shown in Fig.2,ν is the Poisson ratio with a value of 0.253 in duplex stainless steels[29].

Thus the increment of yield strength caused by Orowan strengthening is 21.3,12.3,10.3 and 14.4 MPa for regions 1–4,respectively.Since hardness is positively correlated with strength,it can be concluded intuitively that hardening effect of the deformed gauge is lower than the undeformed grip.This phenomenon can also be improved by Fig.2,where the sigma phase fraction of deformed gauge is lower than that of undeformed grip obviously.Therefore the superplastic deformation could prevent the precipitation of sigma phase.This effect could promote the deformation during quasi stable deformation stage,resulting in a longer elongation.

5.Conclusions

The sigma phase precipitation behavior of 3207 duplex stainless steel during superplastic deformation has been studied by using the EBSD technique and TEM.

(1) The superplastic deformation can prevent sigma phase precipitation generally.The grain boundary sliding during superplastic deformation increases the mobility of grain boundaries.It is negative to sigma phase precipitation,resulting in some sigma phase to precipitate on twin boundaries with lower mobility.

(2) Most of the sigma phase precipitates on ferrite-austenite interface with the misorientation of 20–25°,while it also precipitates in ferrite and austenite with misorientation of 40–45°near the K–S and N–W ORs.The ferrite,austenite and precipitated sigma phase exhibit OR of(200)δσand(020)δ||(311)γ||(200) σ on ferrite-austenite interface.The austenite and sigma phase exhibit OR of,which is near the K–S and N–W ORs.While the sigma phase elongated along a directioncloseto<211>σinferrite,and it shows a misfit of7.5°between()σand(101)δ.

(3) After superplastic deformation,the microhardness of the 3207 duplex stainless steel is higher than the theoretical value calculated by Hall-Petch equation.Orowan strengthening is the hardening mechanism.The hardening in undeformed grip is higher than that of the deformed gauge under the effect of sigma phase precipitation.The prevention effect of superplastic deformation on sigma phase precipitation could promote the deformation during quasi stable deformation stage,resulting in a longer elongation.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A.Supplementary data

Supplementary data to this article can be found online at https://do i.org/10.1016/j.pnsc.2020.12.011.