Effect of Initial Microstructures on Hot Deformation Behavior ... - MDPI

metals Article

Effect of Initial Microstructures on Hot Deformation Behavior and Workability of Ti2AlNb-Based Alloy Sibing Wang, Wenchen Xu *, Yingying Zong, Xunmao Zhong and Debin Shan * School of Materials Science and Engineering & National Key Laboratory for Precision Hot Processing of Metals, Harbin Institute of Technology, Harbin 150001, China; [email protected] (S.W.); [email protected] (Y.Z.); [email protected] (X.Z.) * Correspondence: [email protected] (W.X.); [email protected] (D.S.); Tel.: +86-451-8641-8732 (W.X.)





Received: 12 May 2018; Accepted: 22 May 2018; Published: 25 May 2018

Abstract: In order to study the effect of initial microstructures on the hot deformation behavior and workability of Ti2 AlNb alloy, the isothermal compression experiments of as-rolled and solutiontreated Ti–19Al–23Nb–0.5Mo alloys were conducted in the temperature range of 900–1100 ◦ C and strain rate range of 0.001–10 s−1 . Subsequently, the processing maps of different state materials were established based on dynamic material model (DMM) and Prasad’s instability criterion. The suitable regions for hot working were determined in the processing maps, which was verified through high-temperature tensile test. The results show that although the solution-treatment could be used to improve the ductility of as-rolled Ti2 AlNb alloy at room temperature, the as-rolled microstructure exhibited better hot workability at high temperature compared to the solution-treated microstructure. Keywords: Ti2 AlNb alloy; microstructure; solid solution; workability; processing map

1. Introduction Ti2 AlNb alloys are attractive candidate materials for use in aviation and automobile industries because of its excellent properties, such as high specific strength, good oxidation resistance, and sufficient creep resistance at elevated temperatures [1,2]. Due to the introduction of amounts of lamellar O phase (orthorhombic structure with CmCm symmetry based on Ti2 AlNb) with superior creep resistance, Ti2 AlNb-based alloys can be used in the temperature range of 600–750 ◦ C [3,4]. Besides, the hexagonal close-packed (hcp) α2 phase (DO19 structure based on Ti3 Al), and body-centered cubic (bcc) phase β (disordered structure) or B2 phase (ordered structure) also exist in the Ti2 AlNb alloys. The α2 phase is a brittle phase with high hardness and B2 phase is a plastic phase, which can increase the strength and improve the ductility of Ti2 AlNb alloys, respectively. In the case of high temperature, the B2/β phase undergoes an ordered-disorder transition [5]. The phase transition temperature depends strongly on the alloy composition. L.A. Bendersky and K. Muraleedharan et al. [6,7] found that when the content of Nb element in the alloy exceeds 20%, the transition temperature of B2/β phase increases sharply. Actually, the mechanical properties of Ti2 AlNb alloy are quite sensitive to their microstructure morphologies [8,9]. Xue Chen et al. [10] quantitatively investigated the coarsening behavior of the lamellar O phase and its effect on the tensile properties of Ti–22Al–25Nb alloy. O.G. Khadzhieva et al. [11] analyzed the phase and microstructure transformation occurring upon the ageing of a titanium alloy based on the O-phase. S. Emura et al. [12] successfully controlled the prior B2 grain size of a Ti–22Al–27Nb alloy to be in the range of 8–49 µm by means of the pinning effect of spherical a2 particles against grain growth. Due to the inherent brittleness of intermetallic compounds, the Ti2 AlNb alloys possess high flow stress and low ductility at room temperature, so they need to be processed at elevated temperature. However, the microstructure and properties of Ti2 AlNb alloys exhibit a strong dependence on the Metals 2018, 8, 382; doi:10.3390/met8060382

www.mdpi.com/journal/metals

Metals 2018, 8, 382

2 of 16

processing parameters, such as deformation temperature, strain and strain rate, thus it is important to control and optimize the hot working parameters during hot deformation of Ti2 AlNb alloys. Moreover, different initial microstructures usually show different hot-workability in Ti2 AlNb alloys. Nevertheless, there is little study on the dependence of initial microstructures on the hot deformation behavior and workability of Ti2 AlNb alloys, which restricts their thermal mechanical processing and wide application until now. In this study, the hot deformation behaviors of a Ti2 AlNb-based alloy with the chemical composition of Ti–19Al–23Nb–0.5Mo (at. %), as-rolled and solution-treated, were studied, respectively. The evolution of microstructure and mechanical properties during thermal deformation were analyzed. Meanwhile, the hot processing maps of the Ti2 AlNb alloy with different initial microstructures were established. Furthermore, the difference of the deformation behaviors and thermal processing performance between two initial microstructures were discussed. Thereafter, the suitable hot-working conditions were determined, which could provide effective guidance for hot processing of the Ti2 AlNb alloys. 2. Materials and Methods The alloy sheet used in the present work was provided by Baoji Titanium Industry Co, Ltd. (Baoji, China), 480 mm in length, 240 mm in width, and 11.7 mm in thickness. The chemical composition of the as-rolled sheet was measured to be Ti–19Al–23Nb–0.5Mo in atomic percent through inductively coupled plasma mass spectrometer (ICP) (Optima 5300 DV, PerkinElmer, Waltham, MA, USA), as listed in Table 1. In order to determine the solution temperature of O phase, the solid solution was performed at 890–1090 ◦ C with interval of 20 ◦ C for 3 h, followed by water cooling to about 20 ◦ C. The mechanical properties of the Ti2 AlNb alloy under as-rolled and solution-treated conditions were analyzed through the microhardness test and tensile experiment. The microhardness of as-rolled and solution-treated alloys was measured on a HVS-1000A microhardness tester provided by Laizhou Huayin Testing Instrument Co., Ltd. (Laizhou, China), with a Vickers indenter under a test load of 1000 g and a dwell time of 15 s according to the ASTM standard E384-99. Table 1. Measured chemical composition of Ti2 AlNb alloy. Elements

Ti

Al

Mo

Nb

Fe

wt. % at. %

46.31 48.52

10.27 19.0

0.95 0.50

42.42 22.91

0.08 0.07

The cylindrical specimens for isothermal compression were spark machined from the as-rolled and solution-treated sheet, 7.8 mm in diameter and 11.7 mm in height. The isothermal compression test was conducted in the temperature range from 900 ◦ C to 1100 ◦ C with 50 ◦ C intervals and the strain rate range from 0.001 s−1 to 10 s−1 on a Gleeble-1500D thermal simulator (DSI Company, Sao Paulo, MN, USA). Before the onset of hot compression, the specimens were heated to deformation temperature at a heating rate of 10 ◦ C/s and held for 5 min. Subsequently, the specimens were compressed to a true strain of 0.7 in low vacuum atmosphere of 1 × 10−3 Torr and then quenched into water of about 20 ◦ C. In order to reduce the friction, graphite paper was used as lubricant between the crosshead and specimen during hot compression. After hot compression, the specimens were sectioned parallel to the compression axis for microstructural observation using an Olympus-PMG3 optical microscope (OM) (Olympus Corporation, Tokyo, Japan). The electron backscattered diffraction (EBSD) measurement was carried out on a Quanta 200FEG scanning electron microscope (FEI Company, Hillsborough, OR, USA), and the data with the confidence (CI) value higher than 0.5 were analyzed using the software of TSL-OIM (Analysis 6.1, EDAX Company, Mahwah, NJ, USA). The TEM analysis was conducted on a Talos F200x transmission electron microscope (FEI Company, Hillsborough, OR, USA) operated at 200 KV. The grain size of as-rolled and solution-treated specimens was measured by the linear intercept method using optical micrographs in three individual fields. The specimens for EBSD analysis were

Metals 2018, 8, 382

Metals 2018, 8, x FOR PEER REVIEW

3 of 16

3 of 16

microscope (FEI Company, Hillsborough, OR, USA) operated at 200 KV. The grain size of as-rolled

first mechanically groundspecimens and polished, and thenbyelectropolished in a solution of perchloric acid, and solution-treated was measured the linear intercept method using optical methylmicrographs alcohol andinbutanol with the ratioThe of specimens 6:60:34 for s atanalysis the temperature of about −20 ◦ C. three individual fields. for120 EBSD were first mechanically groundspecimens and polished, thentest electropolished in a solution of perchloric methyl alcohol and The Ti2 AlNb forand TEM were mechanically milled to aboutacid, 70 µm, followed by twin-jet butanol with the ratio of 6:60:34 for 120 s at the temperature of about −20 °C. The Ti2AlNb electropolishing using a solution of perchloric acid, methyl alcohol, and butanol with the ratio of specimens for TEM test were mechanically milled to about 70 μm, followed by twin-jet 6:60:34 at −20 ◦ C and 25 V. electropolishing using a solution of perchloric acid, methyl alcohol, and butanol with the ratio of 6:60:34 at −20 °C and 25 V.

3. Results and Discussion

3. Results and Discussion

3.1. Microstructure and Properties of As-Rolled (AR) and Solution-Treated (ST) Ti2 AlNb Alloy 3.1. Microstructure and Properties of As-Rolled (AR) and Solution-Treated (ST) Ti2AlNb Alloy

Figure 1 shows the OM, SEM, and TEM microstructures of as-rolled Ti2 AlNb alloy, respectively. 1 shows the OM, SEM, and TEM microstructures of as-rolled Ti2AlNb alloy, respectively. The white αFigure 2 phase particles were dispersedly distributed in the matrix (see Figure 1a). The microstructure The white α2 phase particles were dispersedly distributed in the matrix (see Figure 1a). The of the as-rolled sheet was mainly consisted of lamellar B2/O phases, equiaxed α2 (dark), B2 (light), microstructure of the as-rolled sheet was mainly consisted of lamellar B2/O phases, equiaxed α2 and rim-O phases (gray), as depicted in Figure 1b. Figure 1d–f were the electron diffraction patterns of (dark), B2 (light), and rim-O phases (gray), as depicted in Figure 1b. Figure 1d–f were the electron B2, α2 and O phases, respectively, Figure 1c. diffraction patterns of B2, α2 andselected O phases,from respectively, selected from Figure1c.

Figure 1. Microstructures of as-rolled Ti-19Al-23Nb-0.5Mo alloy: (a) OM (optical microscope) image,

Figure 1. Microstructures of as-rolled Ti-19Al-23Nb-0.5Mo alloy: (a) OM (optical microscope) image, (b) SEM (scanning electron microscope) micrograph of α2, B2 and O phases, (c) TEM (transmission (b) SEMelectron (scanning electron microscope) of α(d–f), andfractions O phases, (transmission 2 , B2area B2 and O phases, of α2(c) , O,TEM B2 phases in microscope) micrograph of α2, micrograph electronFigure microscope) micrograph of α2 , B2 and O phases, (d–f), area fractions of α2 , O, B2 phases in 1c. Figure 1c. Figure 2 presents the SEM microstructures of the Ti2AlNb alloy at different solution temperatures (890–1090 °C) for 40min. As shown in Figure 2a,b, when the solution temperature was Figure 2 presents the SEM microstructures of the Ti2 AlNb alloy at different solution temperatures below 930 °C, the dark α2 phase was surrounded by light gray rim-O phase due to the occurrence of ◦ C) for 40 min. As shown in Figure 2a,b, when the solution temperature was below 930 ◦ C, (890–1090 the peritectoid transformation at the interface of α2 phase and B2 phase [13]. Also, the lamellar O the darkphase α2 phase was surrounded by light gray rim-O due to the occurrence of the was visible in the B2 phase matrix. When the phase solution temperature reached 950 °C, peritectoid the microstructure consisted large B2 grains with grain of about 2–5 μm and reducedwas α2 visible transformation at themainly interface of α2ofphase and B2 phase [13].size Also, the lamellar O phase phase distributed the the B2 grain boundaries, but the lamellar phase disappeared (see Figure mainly ◦ C, in the B2 phase matrix.along When solution temperature reachedO950 the microstructure 2c). It indicates that the transformation temperature of O phase was around 950 °C in this Ti2AlNb consisted of large B2 grains with grain size of about 2–5 µm and reduced α2 phase distributed along alloy. With further increasing of solution temperature, the B2 phase continued to coarsen while the the B2 αgrain boundaries, but the lamellar O phase disappeared (see Figure 2c). It indicates that 2 phase reduced gradually (see Figure 2d–f). When the solution temperature increased to 1050 °C

the transformation temperature of O phase was around 950 ◦ C in this Ti2 AlNb alloy. With further increasing of solution temperature, the B2 phase continued to coarsen while the α2 phase reduced gradually (see Figure 2d–f). When the solution temperature increased to 1050 ◦ C or higher, only B2 grains existed in the matrix, suggesting that the α2 phase transition temperature was less than 1050 ◦ C (see Figure 2f,g). The grain size of B2 phase grew rapidly from 173 µm to 273 µm when the temperature increased from 1050 ◦ C to 1090 ◦ C, which should be ascribed to the lack of the pinning effect of the secondary α2 phase distributed along B2 grain boundaries.

Metals 2018, 8, x FOR PEER REVIEW

4 of 16

or higher, only B2 grains existed in the matrix, suggesting that the α2 phase transition temperature was less than 1050 °C (see Figure 2f,g). The grain size of B2 phase grew rapidly from 173 μm to 273 Metals μm 2018,when 8, 382the temperature increased from 1050 °C to 1090 °C, which should be ascribed to the lack4 of 16 of the pinning effect of the secondary α2 phase distributed along B2 grain boundaries.

Figure 2. SEM microstructures of solution-treated Ti2AlNb alloy: (a) 890 °C; (b) 930 °C; (c) 950 °C; (d)

Figure 2. SEM microstructures of solution-treated Ti2 AlNb alloy: (a) 890 ◦ C; (b) 930 ◦ C; (c) 950 ◦ C; 970 °C; (e) 990 °C; (f) 1030 °C; (g) 1050 °C; and (h) 1090 °C. (d) 970 ◦ C; (e) 990 ◦ C; (f) 1030 ◦ C; (g) 1050 ◦ C; and (h) 1090 ◦ C.

In order to improve the ductility of the Ti2AlNb alloy, the solid solution was conducted at 980 °C for 3 h to the eliminate O of phase coarsening B2 phase. The SEM and TEM ◦ In order to improve ductility the Tiwithout 2 AlNb alloy, the solid solution was conducted at 980 C microstructures of solution-treated sheet are shown in Figure 3. Through comparing Figures 1b and of for 3 h to eliminate O phase without coarsening B2 phase. The SEM and TEM microstructures

solution-treated sheet are shown in Figure 3. Through comparing Figures 1b and 3a, it can be seen that the O phase disappeared completely after solid solution. Meanwhile, the α2 phase grains exhibited a spheroidization tendency and the dislocation density reduced clearly, as shown Figures 1c and 3b.

Metals 2018, 8, x FOR PEER REVIEW Metals 2018, 8, x FOR PEER REVIEW

5 of 16 5 of 16

3a, it can be seen that the O phase disappeared completely after solid solution. Meanwhile, the α2 3a, it can be seen that the O phase disappeared completely after solid solution. Meanwhile, the α2 phase grains exhibited a spheroidization tendency and the dislocation density reduced clearly, Metals 2018, 8, 382 5 ofas 16 phase grains exhibited a spheroidization tendency and the dislocation density reduced clearly, as shown Figures 1c and 3b. shown Figures 1c and 3b.

Figure solution-treated Ti–19Al–23Nb–0.5Mo Ti–19Al–23Nb–0.5Mo alloy alloy (980 (980 ◦°C/3 Figure 3. 3. Microstructure Microstructure of of the the solution-treated C/3 h): h): (a) (a) SEM SEM Figure 3. Microstructure of the solution-treated Ti–19Al–23Nb–0.5Mo alloy (980 °C/3 h): (a) SEM 2 phase and B2 phase. microstructure, (b) TEM micrograph of α microstructure, (b) TEM micrograph of α phase and B2 phase. microstructure, (b) TEM micrograph of α22 phase and B2 phase.

Figure 4 shows the mechanical properties of the Ti2AlNb alloy under as-rolled and the mechanical properties of the Ti2of AlNb under as-rolled and solution-treated Figure 4 4shows shows the mechanical properties thealloy Ti2AlNb alloy under as-rolled and solution-treated conditions. Compared to the as-rolled alloy, the tensile strength of solution-treated conditions. Compared to theCompared as-rolled alloy, tensilealloy, strength solution-treated alloy decreased solution-treated conditions. to thethe as-rolled the of tensile strength of solution-treated alloy decreased in RD (rolling direction) and TD (transverse direction) direction from 1110 MPa and in RDdecreased (rolling direction) and TD (transverse direction) direction fromdirection 1110 MPa and1110 1157MPa MPa to alloy in RD (rolling direction) and TD (transverse direction) from and 1157 MPa to 1020 MPa and 1058 MPa, respectively. The elongation in RD and TD direction 1020 MPa The elongation in RD andelongation TD direction 5.7% and 1157 MPaand to 1058 1020MPa, MParespectively. and 1058 MPa, respectively. The in increased RD and from TD direction increased from 5.7% and 2.4% to 9.5% and 4.1%, respectively. After solid solution in the α2+B2/β 2.4% to 9.5% and5.7% 4.1%,and respectively. Afterand solid4.1%, solution in the α2 +B2/β two-phase region, the α average increased from 2.4% to 9.5% respectively. After solid solution in the 2+B2/β two-phase region, the average hardness of solution-treated samples was 326 (±9) HV (MPa), which hardness ofregion, solution-treated samples was of 326 (±9) HV (MPa),samples which was lower thatwhich of the two-phase the average hardness solution-treated wasmuch 326 (±9) HVthan (MPa), was much lower than that of the as-rolled Ti2AlNb samples 408 (±12) HV (MPa). The change of as-rolled Ti2lower AlNb samples 12)as-rolled HV (MPa). The change of mechanical properties be partly was much than that408 of (± the Ti2AlNb samples 408 (±12) HV (MPa).should The change of mechanical properties should be partly attributed to the weakening of the secondary-phase attributed toproperties the weakening of the effect asofthethe lamellar /lenticular mechanical should be secondary-phase partly attributedstrengthening to the weakening secondary-phase strengthening effect as the lamellar /lenticular O phase dissolved into the matrix of B2 phase O phase dissolved of /lenticular B2 phase completely after solidinto solution. In addition, static strengthening effectinto as the matrix lamellar O phase dissolved the matrix of B2 phase completely after solid solution. In addition, static recrystallization of β (B2) phase occurring in solid recrystallization of β (B2) phaseInoccurring solidrecrystallization solution could of eliminate the work hardening by completely after solid solution. addition,in static β (B2) phase occurring in solid solution could eliminate the work hardening by decreasing dislocation density in as-rolled sheet. decreasing dislocation density in as-rolled sheet. solution could eliminate the work hardening by decreasing dislocation density in as-rolled sheet.

Figure 4. Tensile strength and elongation in two directions (RD and TD) of AR (as-rolled) and ST Figure 4. two directions (RD and TD)TD) of AR (as-rolled) andand ST ST 4. Tensile Tensilestrength strengthand andelongation elongationinin two directions (RD and of AR (as-rolled) (solution-treated) alloy. (solution-treated) alloy. alloy.

3.2. Flow Stress Behavior of As-Rolled (AR) and Solution-Treated (ST) Alloy 3.2. (ST) Alloy Alloy 3.2. Flow Flow Stress Stress Behavior Behavior of of As-Rolled As-Rolled (AR) (AR) and and Solution-Treated Solution-Treated (ST) Figure 5 shows the true stress-strain curves for as-rolled and solution-treated samples Figure 5shows shows stress-strain for as-rolled and solution-treated samples Figure 5at thethe truetrue stress-strain curvescurves for with as-rolled and solution-treated compressed compressed the temperatures of 900–1100 °C interval 50 °C. For bothsamples the as-rolled and ◦ ◦ compressed at the temperatures of 900–1100 °C with interval 50 °C. For both the as-rolled and at the temperatures of 900–1100 C with intervalwith 50 C. For bothtemperature the as-rolledand anddecreasing solution-treated solution-treated alloys, the flow stress decreased increasing strain solution-treated alloys, the flow stress decreased with increasing temperature and decreasing strain alloys,indicating the flow stress with increasing temperature and decreasing strain rate, indicating that rate, that decreased the flow stress was sensitive to deformation temperature and strain rate. rate, indicating that the flow stress was sensitive to deformation temperature and strain rate. the flow stress was sensitive to deformation temperature and strain rate. Actually, the flow stress-strain Actually, the flow stress-strain curves exhibited two typical deformation characteristics in the hot Actually, the flow stress-strain curves exhibited two typical deformation characteristics in the hot curves exhibited two typical deformation characteristics in the hot compression experiment. When hot compression experiment. When hot compression was conducted at 950–1050 °C with the strain rate ◦ − 1 compression experiment. When hot compression was conducted at 950–1050 °C with thetemperature strain rate −1 −1 compression was conducted at 950–1050 C with the strain rate lower than 0.01 s or the lower than 0.01 s or the temperature higher than 1100 °C with the strain rate lower than 1 s , the −1 lower the the temperature than 11100 strain rate lowerexhibited than 1 s−1 , the higherthan than 0.01 1100s◦ Corwith strain ratehigher lower than s−1 , °C thewith true the stress-strain curves single peak stress at the initial deformation stage, followed by the decrease of flow stress to a relatively

MetalsMetals 2018, 2018, 8, 3828, x FOR PEER REVIEW

6 of 166 of 16

true stress-strain curves exhibited single peak stress at the initial deformation stage, followed by the

steady value. of It represents deformation characteristic caused by sufficient dynamic decrease flow stress typical to a relatively steadysoftening value. It represents typical deformation softening characteristic(DRX). causedUnder by sufficient dynamic recrystallization (DRX). Under thedecreased other deformation recrystallization the other deformation conditions, the flow stress continuously conditions, flowdue stress decreased continuously beyond the peak value due to incomplete beyond the peakthe value to incomplete dynamic recrystallization or flow instability [14,15]. ◦ dynamic recrystallization or flow instability [14,15]. Under the temperature range of 900–1050 C, the as-rolled alloy showed higher peak flow stress the temperature range of 900–1050 °C, the as-rolled alloy showed higher peak flow comparedUnder with the solution-treated alloy, as shown in Figure 6. However, the difference of peak stress stress compared with the solution-treated alloy, as shown in Figure 6. However, the difference of reduced with the increasing of deformation temperature. The possible reason is that the reduction of peak stress reduced with the increasing of deformation temperature. The possible reason is that the dispersively finedistributed lamellar Ofine phase and α phaseand accelerated the decreasing of flow stress reductiondistributed of dispersively lamellar O2phase α2 phase accelerated the decreasing of as-rolled the increase of deformation led to the reduction of peak of flow alloy stress with of as-rolled alloy with the increase temperature, of deformationwhich temperature, which led to the stressreduction differenceofbetween as-rolled and solution-treated alloy. When the temperature to 1100 peak stress difference between as-rolled and solution-treated alloy.increased When the ◦ C, both the secondary O phase and α phase disappeared in the B2 matrix, thus the solution-treated temperature increased to 1100 °C, both the secondary O phase and α 2 phase disappeared in the B2 2 thus the solution-treated alloystress even relative exhibitedtoa the higher peak flow relative the 5e alloymatrix, even exhibited a higher peak flow as-rolled alloy,stress as shown in to Figures as-rolled alloy, as shown in Figures 5e andslip 6. Itwas is well that grain boundary slip was the and 6. It is well known that grain boundary theknown main deformation mode of metals at high main deformation of metals at high In this study, the grain alloy size ofunder temperature [16,17]. In mode this study, the grain sizetemperature of as-rolled[16,17]. and solution-treated Ti2 AlNb as-rolled and solution-treated Ti2AlNb alloy under 1100 °C was 88 μm and 95 μm, respectively. The 1100 ◦ C was 88 µm and 95 µm, respectively. The larger grain size of solution-treated Ti2 AlNb alloy larger grain size of solution-treated Ti2AlNb alloy should result in less internal grain boundaries, should result in less internal grain boundaries, which was not conducive to plastic deformation and which was not conducive to plastic deformation and thus led to higher flow stress of thus led to higher flow stress of solution-treated alloy compared to as-rolled alloy. solution-treated alloy compared to as-rolled alloy.

Figure 5. Typical stress-strain curves at different temperatures and strain rates of as-rolled (AR) and solution-treated (ST) Ti2 AlNb alloy, (a) 900 ◦ C, (b) 950 ◦ C, (c) 1000 ◦ C, (d) 1050 ◦ C, and (e) 1100 ◦ C.

Metals 2018, 8, x FOR PEER REVIEW

7 of 16

Typical stress-strain curves at different temperatures and strain rates of as-rolled (AR) and7 of 16 MetalsFigure 2018, 8,5. 382 solution-treated (ST) Ti2AlNb alloy, (a) 900 °C, (b) 950 °C, (c) 1000 °C, (d) 1050 °C, and (e) 1100 °C.

Figure 6. The peak stress and strain characterization at different temperatures and strain rates of Figure 6. The peak stress and strain characterization at different temperatures and strain rates of as-rolled and solution-treated alloy: (a) difference of peak stress ( Δ = σ pr − σ ps , Δ is the difference of as-rolled and solution-treated alloy: (a) difference of peak stress (∆ = σpr − σps , ∆ is the difference of peak stress of as-rolled alloy, ispeak the stress peak stress of solution-treated peak stress, σσprpris is peak stress, thethe peak stress of as-rolled alloy, andand σps isσ the of solution-treated alloy), ps (b) Relative difference of peakofstress = (σpr δ is the difference of peak , δrelative is the relative difference alloy), (b) Relative difference peak(δstress ( δ−= σ(σpspr)/σ − σpsps× ) /100%, σ ps × 100% stress, σpr is the peak stress of as-cast alloy, and σps is the peak stress of solution treated alloy). of peak stress, σ pr is the peak stress of as-cast alloy, and σ ps is the peak stress of solution treated alloy). Analysis 3.3. Kinetic

Theoretically, 3.3. Kinetic Analysisplastic deformation is controlled by the thermally activated process. At all the stress levels, the relationship between flow stress, temperature and strain rate can be expressed by hyperbolic Theoretically, plastic deformation is controlled by the thermally activated process. At all the sine function [18]: stress levels, the relationship between flow stress, temperature . Q actand strain rate can be expressed by ε = A[sinh(α · σp )]n exp(− ). (1) hyperbolic sine function [18]: RT It also can be given by the Zener-Hollomon parameter as follows [19]: Q (1) ε = A[sinh(α ⋅ σ p )]n exp( − act ) . RT . Q act Z = ε exp( ) = A[sinh(α · σp )]n , (2) RT parameter as follows [19] : It also can be given by the Zener-Hollomon .

∂ ln ε where Z is the Zener-Hollomon parameter; A,Qαactand n [sinh( are materials ∂σ , Z = ε exp( )=A α ⋅ σ p )]n , constants, and α = β/n1 , β = (2) . . ∂ ln ε RT n1 = ; ε is the strain rate; T is the deformation temperature; R is the gas constant; Qact is the ∂ ln σ

activation and σp is the peak stress. According to nEquation (1), theconstants, activation and energy where Z isenergy; the Zener-Hollomon parameter; A, and are materials αQ = act β /can n1 , be obtained by taking partial differential equation of Equation (1), which can be expressed as [20]: ∂ ln ε ∂ ln ε β= , n1 = ; ε is the strain rate; T is thedeformation temperature; R is the gas constant;    . ∂σ ∂ ln σ ∂ ln[sinh(ασp )] ∂ ln ε Q act and = R σ is the peak stress. According to Equation . (3) Qact is the activation energy; (1), the activation . p∂ ln[sinh( ασp )] ∂(1/T )] ε T energy Qact can be obtained by taking partial differential equation of Equation (1), which can be . The αasvalues could be calculated through linear fitting of the relationships of σp − ln ε and expressed [20]: . ln σp − ln ε, as shown in Figure 7a,b. Subsequently, the deformation activation energy could be . )]  ln[sinh(   (∂ασ ∂ ln ε  ln[sinh(ασ)] − 1/T, as shown p obtained through linear fitting )] − lnασ ε and Qactof=the R curves of ln[sinh (3)    . ln[sinh( )] (1 )] ασ T ∂ ∂ in Figure 7c–e. Besides, the stress constant n p could ε according to the curves of T  be calculated ln[sinh(ασp )] − ln Z shown in Figure 7f. In addition, the correlation coefficients of as-rolled and and The α values could calculated through of linear the relationships of σ p − ln ε were ◦C solution-treated alloys forbethe linear regression lnZ fitting and ln[of sinh (ασ p )] during 900–1000 as shown in Figurewhich 7a,b. reached Subsequently, the 0.9944, deformation activation energy could ◦be ln σ p − and ln ε ,0.98511, 0.9878 respectively, 0.9934 and respectively, during 1000–1100 C, indicating the reliability of kinetic analysis of the Ti AlNb alloy by hyperbolic sine law. 2 obtained through linear fitting of the curves of ln[sinh( ασ )] − ln ε and ln[sinh(ασ )] − 1 / T , as shown

in Figure 7c–e. Besides, the stress constant n could be calculated according to the curves of ln[sinh(ασ p )] − ln Z shown in Figure 7f. In addition, the correlation coefficients of as-rolled and solution-treated alloys for the linear regression of lnZ and ln[sinh(ασ p )] during 900–1000 °C were 0.9878 and 0.98511, respectively, which reached 0.9934 and 0.9944, respectively, during 1000–1100 °C, indicating the reliability of kinetic analysis of the Ti2AlNb alloy by hyperbolic sine law.

Metals 2018, 8, 382 Metals 2018, 8, x FOR PEER REVIEW

8 of 16 8 of 16

Figure 7. The flow behaviors for as-cast and solution-treated Ti–19Al–23Nb–0.5Mo alloy: (a) . Figure 7. The flow behaviors for as-cast and solution-treated Ti–19Al–23Nb–0.5Mo alloy: (a) σp − ln ε; σ − ln ε ; .(b) ln σ − ln ε ; (c) and (d). ln[sinh(ασ )] − ln ε ; (e) ln[sinh(ασ )] − 1 / T ; and (f) (b) lnp σp − ln ε; (c) andp(d) ln[sinh(ασ )] − ln ε; (e) ln[sinh(ασ )] − 1/T; and (f) ln Z − ln[sinh(ασp )]. ln Z − ln[sinh(ασ p )] .

Table 2 lists the corresponding values of various parameters in the constitutive equation of Table 2 lists the corresponding values of various parameters in the constitutive equation of as-rolled and solution-treated alloys. It can be seen that after solid solution, the activation energy as-rolled and solution-treated alloys. It can be seen that after solid solution, the activation energy decreased from 727 kJ/mol and 594 kJ/mol to 510 kJ/mol and 470 kJ/mol, respectively, in the decreased from 727 kJ/mol and 594 kJ/mol to 510 kJ/mol and 470 kJ/mol, respectively, in the temperature range of 900–1000 ◦ C and 1000–1100 ◦ C. Clearly, both the activation energies of as-rolled temperature range of 900–1000 °C and 1000–1100 °C. Clearly, both the activation energies of and solution-treated Ti2 AlNb alloys during 900–1000 ◦ C are higher than 1000–1100 ◦ C in this study. as-rolled and solution-treated Ti2AlNb alloys during 900–1000 °C are higher than 1000–1100 °C in Also, [21]Xiong calculated thecalculated apparentthe activation energies of forged ingots were to be 789 kJ/mol thisXiong study.Ma Also, Ma[21] apparent activation energies of forged ingots were to in be the789 α2 kJ/mol + B2/βin+the O phase region and 436 kJ/mol in the α + B2/β region, respectively. This should 2 α2 + B2/β + O phase region and 436 kJ/mol in the α2 + B2/β region, respectively. ◦ C and gradual disappearance of be This partly caused the dissolution of lamellar O phase above O 1000 should beby partly caused by the dissolution of lamellar phase above 1000 °C and gradual ◦ α2 disappearance phase until 1050 In addition, both activation energies activation of as-rolled alloy inofthe of α2C.phase until 1050 °C.the Indeformation addition, both the deformation energies ◦ ◦ temperature range of 900–1000 C and 1000–1100 C °C were than °C that of solution-treated as-rolled alloy in the temperature range of 900–1000 andhigher 1000–1100 were higher than that alloy. of The possible reason is that dissolution lamellar O phase with excellent creepwith resistance may solution-treated alloy. Thethe possible reason of is that the dissolution of lamellar O phase excellent cause theresistance decreasing ofcause deformation activation energy. creep may the decreasing of deformation activation energy.

Metals 2018, 8, 382

9 of 16

Table 2. Parameter of as-rolled and solution-treated alloy. State

α

Stress Constant n

Qact (kJ/mol)

As-rolled (900–1000 ◦ C) As-rolled (1000–1100 ◦ C) Solution-treated (900–1000 ◦ C) Solution-treated (1000–1100 ◦ C)

4.1 × 10–3 9.4 × 10− 3 4.6 × 10− 3 9.2 × 10− 3

4.0 2.9 3.8 3.3

727 (±31) 594 (±29) 510 (±25) 470 (±24)

On the above basis, the dependence of the peak stress on the deformation temperature and strain rate of as-rolled Ti2 AlNb alloy during 900–1000 ◦ C and 1000–1100 ◦ C, respectively, could be expressed as: ε = 3.5 × 1019 [sinh(4.1 × 10−3 σp )]

.

4.0

.

2.9

ε = 3.7 × 1020 [sinh(9.4 × 10−3 σp )] ◦C

· exp(−

727 ), RT

(4)

· exp(−

594 ). RT

(5)

The constitutive equation of solution-treated Ti2 AlNb alloy in the temperature range of 900–1000 and 1000–1100 ◦ C, respectively, could be written as: ε = 6.3 × 1018 [sinh(4.6 × 10−3 σp )]

.

3.8

.

3.3

ε = 1.6 × 1015 [sinh(9.2 × 10−3 σp )]

· exp(−

510 ), RT

(6)

· exp(−

470 ). RT

(7)

Usually, deformation mechanism could be represented by stress exponent in the constitutive relationship of materials. For metals, the deformation process was controlled by dislocation slide or dislocation climb when the stress exponent reached three or more [22]. Also, Boehlert et al. [9,23,24] pointed out that when the stress exponent was in the in the range of 3.5–7.2, hot deformation mechanism was dominated by dislocation slide or climb. Therefore, dislocation slide and climb may control plastic deformation during 900–1100 ◦ C for both as-rolled and solution-treated Ti2 AlNb alloy in the present study. 3.4. Hot Processing Map of As-Rolled and Solution-Treated Alloy 3.4.1. The Theory for Processing Map Based on dynamic material modeling (DMM), the processing map was established by Prasad et al., which is widely used to study the deformation behavior of metal materials [25–28]. Further, Mahmudi et al. [29,30] utilized the theory of a processing map to predict shear deformation, which also exhibited good applicability. According to the DMM, the workpiece is considered to be a power dissipation and the total input power (P) consists of two complementary parts, wherein the G content represents the power dissipation for plastic deformation and J co-content represents the power dissipation through microstructure evolution, such as dynamic recovery, dynamic recrystallization, and phase transformation. The relationship can be described as: P = σε = G + J =

Z ε. 0

.

σdε +

Z σ . 0

εdσ.

(8)

For given deformation temperature and strain, the flow stress can be expressed as: .m

σ = Kε , .

(9)

where K is the material constant; σ is the flow stress; ε is the strain rate; m is the strain rate sensitivity, which is thought to be independent of strain rate and can be described as follows:

Metals 2018, 8, 382

10 of 16

m=

∂(ln σ ) dJ = . . dG ∂(ln ε)

(10)

The J co-content can be expressed as: .

J = σ εm/(m + 1).

(11) .

For the ideal linear dissipation body, m = 1 and J co-content reaches the maximum: Jmax = 12 σ ε. The efficiency of non-linear dissipator can be defined as: η=

J 2m . = Jmax ( m + 1)

(12)

The instability map is developed based on the extremum principle of irreversible thermodynamics applied for the large plastic flow body. The instability criterion can be defined as follows: .

ξ (ε) =

∂ ln( mm+1 ) .

∂(ln ε)

+ m < 0.

(13)

.
The variation of the instability parameter ξ ε with temperature and strain rate constitutes the instability map, from which the instability regions can be obtained. For the stable flow:

2m/η > 1.

(14)

The superposition of the power dissipation map over the instability map gives a processing map, in which different domains can be correlated with different microstructural mechanisms. 3.4.2. Hot Processing Map Figure 8 shows the processing maps of as-rolled and solution-treated Ti2 AlNb alloy based on Prasad’s criterion at the true strain of 0.7, respectively. It can be seen that the processing maps of as-cast and solution-treated alloy based on Prasad’s criterion exhibited similar peak power dissipation regions at 900–1100 ◦ C and 0.001–0.1 s−1 , as shown in Figure 8a,b. Both the instability regions in as-rolled alloy and solution-treated alloy were located under the temperatures of 900–1100 ◦ C and strain rates of 0.3–10 s−1 , as shown in Figure 8c,d. However, there was an evident difference existing between the stability regions of the two processing maps. The maximum peak dissipation efficiency of as-rolled alloy (1075–1100 ◦ C and 0.001–0.25 s−1 ) and solution-treated alloy (1075–1100 ◦ C and 0.01–0.25 s−1 ) was 0.6 and 0.5, respectively. It indicates that more deformation work was dissipated through microstructure evolution such as dynamic recovery and dynamic recrystallization, during hot compression of as-rolled alloy. Moreover, the region III for as-rolled alloy was wider than that for the solution-treated alloy, which suggests the processing area for as-rolled alloy should be larger. 3.4.3. Instability Region The processing map exhibited a large domain of flow instability at the strain rates higher than 1.0 which was probably caused by the occurrence of adiabatic shear bands or localized plastic flow [25,31,32]. Figure 9a,d show the macrostructures of hot compressed specimens of as-rolled and solution-treated alloy at 900 ◦ C and 10 s−1 , respectively. Figure 9b,f show the macrostructures of as-rolled and solution-treated alloy at 1100 ◦ C and 10 s−1 , respectively. It can be seen from Figure 9d that adiabatic shear bands were present at an angle of 45◦ to the compressive axis, which was often observed in titanium alloys at low temperatures and high strain rates. In general, shear bands were easily observed due to the adiabatic heating induced by lower thermal conductivity of titanium based alloy relative to other metal materials. The deformation heat generated during hot deformation with higher strain rate could not be released in time, resulting in localized temperature rising. s−1 ,

Metals 2018, 8, 382

11 of 16

Metals 2018, 8, x FOR PEER REVIEW of 16 Prasad et al. [25] also suggested that titanium alloys exhibit flow instability due to the11formation Metals 2018, 8, x FOR PEER REVIEW 11 of 16 adiabatic shear bands at high strain rates. Clearly, it can be observed that flow localization occurred that titanium alloys exhibit flow instability due to the formation adiabatic shear bands at high strain in Figure 9a,b,d,e, which indicates that Ti–19Al–23Nb–0.5Mo wasbands veryindicates that titanium alloys exhibit flowthat instability due to theoccurred formation shear atsensitive high strain rates. Clearly, it can be observed flowthe localization in adiabatic Figurealloy 9a,b,d,e, which thatto the rates. Clearly, it can be observed that flow localization occurred in Figure 9a,b,d,e, which indicates that deformation temperature and strain rate. Since flow instability should be avoided during hot working, the Ti–19Al–23Nb–0.5Mo alloy was very sensitive to the deformation temperature and strain rate. Since the Ti–19Al–23Nb–0.5Mo alloy was very sensitive to the deformation temperature and strain rate. Since it is flow necessary to control deformation temperatures and strain rates during hot working instability should be avoided during hot working, it is necessary to control deformationof the flow instability be during avoidedhotduring hot working, is necessary to control deformation temperatures andshould strain rates working of the presentitalloy. present alloy.

temperatures and strain rates during hot working of the present alloy.

Figure 8. Processing maps of as-rolled solution-treated 2AlNb alloy: (a) power dissipation Figure 8. Processing maps of as-rolled andand solution-treated Ti2TiAlNb alloy: (a) power dissipation map Figure Processing of as-rolled and solution-treated Ti2AlNb alloy; alloy: ((ca)) processing power dissipation dissipation map of solution-treated of map of 8. as-rolled alloy;maps (b) power of as-rolled alloy; (b) power dissipation map of solution-treated alloy; (c) processing mapmap of as-rolled dissipation map of solution-treated alloy; (c) processing map of map of as-rolled (b) powermap of solution-treated alloy. as-rolled alloy; andalloy; (d) processing alloy;as-rolled and (d) alloy; processing of solution-treated alloy. processing map of solution-treated alloy. and (d)map

Figure 9. Microstructures occurring in the instability domain of Ti–19Al–23Nb–0.5Mo alloy, ε = 0.7

9. Microstructures occurring°C-10 thesinstability domain of Ti–19Al–23Nb–0.5Mo alloy, 9b, ε = (0.7 −1, (c) microstructure a) 9.AR-900 °C-10 s−1, (boccurring ) AR-1100 the specimen in Figure (Figure Figure Microstructures ininthe instability domain of of Ti–19Al–23Nb–0.5Mo alloy, εd)= 0.7 −1, (b) AR-1100 −1°C-10 s−1, (c) microstructure of the specimen in Figure 9b, (d) a) AR-900 °C-10 s (ST-900 −1 ◦ C-10 1 , (c) microstructure °C-10 s ◦, C-10 and (f)s− microstructure of the specimen in Figure 9e. in Figure 9b, s ,s(−e1) ,ST-1100 (a) AR-900 °C-10 (b) AR-1100 of the specimen ST-900 °C-10 s−1, (e) ST-1100 °C-10 s−1, and (f) microstructure of the specimen in Figure 9e. (d) ST-900 ◦ C-10 s−1 , (e) ST-1100 ◦ C-10 s−1 , and (f) microstructure of the specimen in Figure 9e.

Metals 2018, 8, 382 Metals 2018, 8, x FOR PEER REVIEW

12 of 16 12 of 16

3.4.4. Stability Region 3.4.4. Stability Region

Figures 10 and 11 show the EBSD images in three regions, i.e., region I, II, and III, of as-rolled Figures 10 andalloy, 11 show the EBSD images in three regions, region I, II, denoted and III, ofB2, as-rolled and solution-treated respectively. The red, green, and i.e., yellow color α2 (Ti3 Al) and solution-treated alloy, respectively. The red, green, and yellow color denoted B2, (Ti3Al) and O (Ti2 AlNb), respectively, in Figures 10d–f and 11d–f. The mean grain size of as-rolled and and O (Ti2AlNb), respectively, in Figures 10d–f and 11d–f. The mean grain size of as-rolled and solution-treated alloy under 900 ◦ C and 0.001s−1−1 was quite small (~1 µm) (see Figure 10a,g and solution-treated alloy under 900 °C and 0.001s was quite small (~1 μm) (see Figures 10a,g and Figure 11a,g). Obviously, hot compression contributed to the microstructure refinement of as-rolled 11a,g). Obviously, hot compression contributed to the microstructure refinement of as-rolled and and solution-treated alloys,which which should beneficial improvement of the comprehensive solution-treated alloys, should be be beneficial for for thethe improvement of the comprehensive mechanical properties of Ti AlNb alloy workpieces. mechanical properties of2Ti2AlNb alloy workpieces. As shown in Figures 10b 10b andand 11b,11b, both thethe microstructures ofof as-rolled As shown in Figures both microstructures as-rolledand andsolution-treated solution-treated alloy alloy strip-like exhibited morphology, strip-like morphology, and amounts of DRXcould grains found the exhibited and amounts of DRX grains becould foundbealong thealong strip-shaped −1 and boundaries 0.1 s−1. However, is an evident difference in hot grainstrip-shaped boundariesgrain under 1000 ◦ Cunder and 1000 0.1 s°C . However, there isthere an evident difference in the the hot compressed microstructures of as-rolled and solution-treated alloys. Large quantities of compressed microstructures of as-rolled and solution-treated alloys. Large quantities of dispersively dispersively distributed O phase remained in the as-rolled alloy (see Figure 10e), while nearly no O distributed O phase remained in the as-rolled alloy (see Figure 10e), while nearly no O phase could be phase could be found in the solution treated alloy (see Figure 11e). Since small O phase remaining found in the solution treated alloy (see Figure 11e). Since small O phase remaining in as-rolled alloy in as-rolled alloy limited the growth of B2-phase grain, the mean grain size of B2 phase (49 μm) in limited the growth of B2-phase grain,than the mean size of B2 phase alloy as-rolled alloy was much smaller that ingrain solution-treated alloy (49 (75 µm) μm),inasas-rolled illustrated in was muchFigures smaller than that in solution-treated alloy (75 µm), as illustrated in Figures 10h and 11h. 10h and 11h.

Figure 10. Electron back scattering diffraction (EBSD) images of the specimens deformed at different

Figure 10. Electron back scattering diffraction (EBSD) images of the specimens deformed at-1different s−1 (a); − 1000 °C-0.1 s ◦(b), deformation conditions of the as-rolled Ti2AlNb alloy: Grain size at 900 °C-0.001 ◦ C-0.001 1 (a); deformation conditions of the as-rolled Ti AlNb alloy: Grain size at 900 s 1000 C-0.1 s-1 2 and 1100 °C-0.1 s−1 (c); Phase composition of Figure 10a (d); Figure 10b (e); and Figure 10c (f); ◦ − 1 (b), and 1100 C-0.1 s (c); Phase composition of Figure 10a (d); Figure 10b (e); and Figure 10c (f); Misorientation and grain size fraction of Figure 10a (g); Figure 10b (h); and Figure 10c (i). Wherein α2, B2 Misorientation sizeinfraction of Figure 10acolor, (g); Figure 10b (h); and Figure 10c (i). Wherein α2, and O phaseand wasgrain displayed green, red, and yellow respectively. B2 and O phase was displayed in green, red, and yellow color, respectively.

When deformed under 1100 ◦ C and 0.1 s−1 , the recrystallized grains grew obviously (see Figures 10c and 11c) and α2 phase could not be found (see Figures 10f and 11f) both in as-rolled and

MetalsMetals 2018, 2018, 8, 3828, x FOR PEER REVIEW

13 of 16 13 of 16

When deformed under 1100 °C and 0.1 s−1, the recrystallized grains grew obviously (see Figure 10c and Figure 11c) and phase small could not be found Figures 10f andcould 11f) both in as-rolled and 10f, solution-treated alloys. In addition, amounts of (see residual O phase be found in Figure solution-treated alloys. In addition, small amounts of residual O phase could be found in Figure 10f, which disappeared completely in Figure 11f. Similarly, the pinning effect of O phase limited the which disappeared completely in Figure 11f. Similarly, the pinning effect of O phase limited the coarsening of B2 grains, which led to smaller grain size (88 µm) in as-rolled alloy than solution-treated coarsening of B2 grains, which led to smaller grain size (88 μm) in as-rolled alloy than alloysolution-treated (95 µm), as illustrated Figures 10i andin11i. alloy (95 in μm), as illustrated Figures 10i and 11i.

Figure 11. Electron back scattering diffraction (EBSD) images of the specimens deformed at different

Figure 11. Electron back scattering diffraction (EBSD) images of the specimens deformed at−1different deformation conditions of the solution-treated Ti2AlNb alloy: Grain size at 900 °C-0.001 s (a); deformation conditions of the solution-treated Ti2 AlNb alloy: Grain size at 900 ◦ C-0.001 s−1 (a); of Figure 11a (d); Figure 11b (e); and 1000 °C-0.1 s−1(b); and 1100 °C-0.1s−1 (c); Phase composition ◦ C-0.1 s−1 (b); and 1100 ◦ C-0.1s−1 (c); Phase composition of Figure 11a (d); Figure 11b (e); 1000 Figure 11c (f); Misorientation and grain size fraction of Figure 11a (g), Figure 11b; (h) and Figure 11c and Figure 11c (f); size fraction Figure 11a (g),color, Figure 11b; (h) and Figure 11c α2,Misorientation B2 and O phaseand wasgrain displayed in green,ofred, and yellow respectively. (i). Wherein (i). Wherein α2, B2 and O phase was displayed in green, red, and yellow color, respectively. 3.4.5. Comparative Analysis of Hot Workability

3.4.5. Comparative Analysis of Hot Workability Figure 12 shows the true stress-true strain curves of the two Ti2AlNb alloys in the tensile test at the temperatures °C,stress-true 1000 °C and 1100curves °C, and the two strainTirates of 0.00 s and 0.1 s , Figure 12 shows of the900 true strain of the 2 AlNb alloys in the tensile test respectively. It can be seen that the ductility of the Ti 2 AlNb alloys in as-rolled state was less than at the temperatures of 900 ◦ C, 1000 ◦ C and 1100 ◦ C, and the strain rates of 0.00 s−1 and 0.1 s−1 , 82.2% (corresponding to true strain of 0.6) under the temperature of 900 °C and strain rate of respectively. It can be seen that the ductility of the Ti AlNb alloys in as-rolled state was less than 0.001s−1, which limited large plastic deformation of 2the Ti2AlNb alloys, such as bulk forming −1 , 82.2%processes (corresponding true strain of 0.6) etc.). underNevertheless, the temperature 900 ◦ C processes, and strainsuch rate of (i.e., hottoforging, extrusion, sheetofforming as 0.001s hot which limited bending, large plastic deformation of the alloys, as bulk forming 2 AlNbin stamping, and flow forming, could beTi selected regionsuch I for the Ti2AlNb alloys. processes When the (i.e., hot forging, extrusion, etc.). Nevertheless, sheet forming processes, such as hot stamping, bending, temperature increased to 1000 °C, the ductility of as-rolled alloy reached 125% (corresponding to −1 true forming, strain of could 0.8) atbe higher strain rate of I0.1 . Therefore, region When II should be proper forincreased the and flow selected in region forsthe Ti2 AlNb alloys. the temperature ◦ C, the Ti2AlNb alloy to undertake large plastic deformation, as forging andtoextrusion. Boehlert to 1000 ductility of as-rolled alloy reached 125%such (corresponding true strain of 0.8)etatal. higher and rolling of three O + BCC Ti2AlNb alloysalloy in the straininvestigated rate of 0.1 the s−1 .forging Therefore, regionprocess II should be proper for the Ti2 AlNb to temperature undertake large plastic deformation, such as forging and extrusion. Boehlert et al. investigated the forging and rolling process of three O + BCC Ti2 AlNb alloys in the temperature range of 932–1000 ◦ C, which indicates that the sub-transus processing of the Ti2 AlNb alloys was feasible for obtaining uniformly refined microstructure without large prior BCC grains [33]. When the temperature reached 1100 ◦ C, there was −1

−1

Metals 2018, 8, 382

Metals 2018, 8, x FOR PEER REVIEW

14 of 16

14 of 16

range of 932–1000 °C, which indicates that the sub-transus processing of the Ti2AlNb alloys was

a decrease in the elongation of as-rolled Ti2 AlNb alloys, which should be attributed to the coarsening feasible for obtaining uniformly refined microstructure without large prior BCC grains [33]. When of B2 grain at higher temperature. Therefore, region III should be inappropriate for hot processing of the temperature reached 1100 °C, there was a decrease in the elongation of as-rolled Ti2AlNb alloys, the present as-rolled Ti2 AlNb to alloy. which should be attributed the coarsening of B2 grain at higher temperature. Therefore, region III Noticeably, although the solution-treatment could be as-rolled used to improve the ductility of the as-rolled should be inappropriate for hot processing of the present Ti2AlNb alloy. alloy at room temperature, the solution-treated alloy exhibited inferior compared to Noticeably, although the solution-treatment could be used to improveelongation the ductility of the ◦ ◦ as-rolled at room temperature, solution-treated alloy solid exhibited inferior elongation as-rolled alloyalloy in the temperature rangethe of 900–1100 C. After solution at 980 C for 3 h, compared to as-rolled alloyalloy in thewas temperature range of 900–1100 °C. After solid solution 980 °C (see the microstructure of Ti2 AlNb coarsened evidently compared to initial as-rolledatmaterial for 13 and h, the Ti2AlNb alloy was evidently compared to initial as-rolled alloy Figures 3),microstructure which causedoflower elongation incoarsened the solution-treated alloy relative to as-rolled material (see Figures 1 and 3), which caused lower elongation in the solution-treated alloy relative during the hot tensile test. Besides, the pinning effect of remaining O phase and α2 phase at 1000 ◦ C to as-rolled alloy during the hot tensile test. Besides, the pinning effect of remaining O phase and α2 ◦ and 1100 C, respectively, restricted the excessive growth of B2 grains in as-rolled Ti2 AlNb alloy (see phase at 1000 °C and 1100 °C, respectively, restricted the excessive growth of B2 grains in as-rolled Figures 10 and 11), which contributed to better ductility in as-rolled alloy compared to solution-treated Ti2AlNb alloy (see Figures 10 and 11), which contributed to better ductility in as-rolled alloy alloy.compared Clearly, the microstructure show better workability relative to the solution-treated to as-rolled solution-treated alloy. Clearly, the hot as-rolled microstructure show better hot microstructure for the present Ti AlNb alloy. 2 workability relative to the solution-treated microstructure for the present Ti2AlNb alloy.

Figure 12. Tensile stress-strain curves at different temperatures and strain rates of as-rolled (AR)

Figure 12. Tensile stress-strain curves at different temperatures and strain rates of as-rolled (AR) and and solution-treated (ST) Ti2AlNb alloy. solution-treated (ST) Ti2 AlNb alloy.

4. Conclusions

4. Conclusions

The hot deformation behaviors of Ti–19Al–23Nb–0.5Mo alloy with as-rolled and The hot deformation behaviorswere of Ti–19Al–23Nb–0.5Mo alloy with as-rolledexperiment and solution-treated solution-treated microstructures investigated using isothermal compression in the temperature were rangeinvestigated of 900–1100 °C andisothermal strain rate range from 0.001–10 s−1, based on temperature which the hotrange microstructures using compression experiment in the ◦ C maps processing were established hot workability wason evaluated for the 2AlNb alloy with were of 900–1100 and strain rate range and fromthe 0.001–10 s−1 , based which the hotTi processing maps different microstructures. The main conclusions can be drawn as follows: established and the hot workability was evaluated for the Ti2 AlNb alloy with different microstructures. (1) The as-rolled Ti2AlNb alloy mainly consisted of lamellar B2/O phases, equiaxed α2, B2, and The main conclusions can be drawn as follows: rim-O phases in the as-rolled state. After solution treated at 950 °C and 1050 °C, the lamellar O (1) The as-rolled Ti2 AlNb alloy mainly consisted of lamellar B2/O phases, equiaxed α2 , B2, phase and equiaxed α 2 phase disappeared successively. The microstructure of as-rolled Ti2AlNb ◦ C and 1050 ◦ C, the lamellar O and rim-O phases in the After solution treated 950 after alloy transformed intoas-rolled coarsenedstate. B2 grain and spheroidized α2 at phase solid solution at 980 °C phaseforand α2 phase successively. The microstructure ofalloy. as-rolled Ti2 AlNb alloy 3 h,equiaxed which increased thedisappeared ductility and decreased the strength of the Ti2AlNb transformed into coarsened B2 grain and α2 as-rolled phase after solid solution at peak 980 ◦flow C for 3 h, (2) Under the temperature range ofspheroidized 900–1050 °C, the alloy showed higher stress compared the solution-treated With theof increase deformation which increased the with ductility and decreasedalloy. the strength the Ti2of AlNb alloy. temperature, the ◦ C, the reduction lamellar O phaserange and equiaxed α2 phase accelerated decreasing flow stress thestress (2) Underof the temperature of 900–1050 as-rolledthe alloy showedof higher peak in flow as-rolled alloy thus minimized theWith peakthe stress difference between the two Ti2AlNb compared with the and solution-treated alloy. increase of deformation temperature, thealloys. reduction When the temperature increased 1100 accelerated °C, the solution-treated alloyofeven higher peak alloy of lamellar O phase and equiaxed α2to phase the decreasing flowexhibited stress inathe as-rolled flow stress relative to the as-rolled alloy probably because larger grain size in the solution-treated and thus minimized the peak stress difference between the two Ti2 AlNb alloys. When the temperature alloy could result in less internal grain boundaries and thus hinder grain boundary slip more increased to 1100 ◦ C, the solution-treated alloy even exhibited a higher peak flow stress relative to strongly at high temperature.

the as-rolled alloy probably because larger grain size in the solution-treated alloy could result in less internal grain boundaries and thus hinder grain boundary slip more strongly at high temperature. (3) Four constitutive equations were formulated to describe the flow stress of as-rolled and solution-treated alloy, respectively, as a function of deformation temperature and strain rate.

Metals 2018, 8, 382

15 of 16

The deformation activation energy were calculated to be 727 and 594 kJ/mol during 900–1000 ◦ C and 1000–1100 ◦ C, respectively, for as-rolled alloy. Correspondingly, the activation energy was reduced to 510 and 470 kJ/mol, respectively, for solution-treated alloy. (4) Two processing maps based on the Prasad’s instability criterion were developed for as-rolled and solution-treated alloy, respectively. The sheet forming process could be conducted in region I (near 900 ◦ C/0.001 s−1 ) due to the occurrence of dynamic recrystallization, and large plastic deformation, such as forging and extrusion, could be carried out in region II (near 1000 ◦ C/0.1 s−1 ) due to the high elongation of Ti2 AlNb alloy. When the strain rate increased higher than 1.0 s−1 , the material exhibited flow instability associated with the adiabatic shear bands and flow localization, which should be avoided during hot processing. (5) Although the solution-treatment could be used to improve the ductility of the as-rolled alloy at room temperature, the solution-treated alloy exhibited inferior elongation compared to as-rolled alloy in the temperature range of 900–1100 ◦ C. The as-rolled microstructure show better hot workability at high temperature relative to the solution-treated microstructure for the Ti2 AlNb alloy. Author Contributions: W.X. conceived and designed the experiments; X.Z. and S.W. performed the experiments; W.X. and S.W. analyzed the data and wrote the paper; D.S. and Z.Y. provided guidance and all sorts of support during the work. Acknowledgments: This work is supported by the National Natural Science Foundation of China (No. 51775137). Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3. 4.

5. 6.

7. 8.

9. 10.

11. 12.

Dey, S.R.; Roy, S.; Suwas, S.; Fundenberger, J.J.; Ray, R.K. Annealing response of the intermetallic alloy Ti–22Al–25Nb. Intermetallics 2010, 18, 1122–1131. [CrossRef] Gogia, A.K.; Nandy, T.K.; Banerjee, D.; Carisey, T.; Strudel, J.L.; Franchet, J.M. Microstructure and mechanical properties of orthorhombic alloys in the Ti-Al-Nb system. Intermetallics 1998, 6, 741–748. [CrossRef] Jia, J.; Zhang, K.; Liu, L.; Wu, F. Hot deformation behavior and processing map of a powder metallurgy Ti–22Al–25Nb alloy. J. Alloys Compd. 2014, 600, 215–221. [CrossRef] Xue, C.; Zeng, W.; Xu, B.; Liang, X.; Zhang, J.; Li, S. B2 grain growth and particle pinning effect of Ti-22Al-25Nb orthorhombic intermetallic alloy during heating process. Intermetallics 2012, 29, 41–47. [CrossRef] Shen, J.; Feng, A. Recent advances on microstructural controlling and hot forming of Ti2 AlNb-based alloys. Acta Metall. Sin. 2013, 49, 1286–1294. [CrossRef] Bendersky, L.A.; Roytburd, A.; Boettinger, W.J. Phase transformations in the (Ti, Al)3 Nb section of the Ti-Al-Nb system—I. Microstructural predictions based on a subgroup relation between phases. Acta Metall. Mater. 1994, 42, 2323–2335. [CrossRef] Muraleedharan, K.; Nandy, T.K.; Banerjee, D.; Lele, S. Transformations in a Ti-24Al-15Nb alloy: Part II. A composition invariant βo → O transformation. Metall. Trans. A 1992, 23, 417–431. [CrossRef] Wang, W.; Zeng, W.; Xue, C.; Liang, X.; Zhang, J. Microstructure control and mechanical properties from isothermal forging and heat treatment of Ti–22Al–25Nb (at.%) orthorhombic alloy. Intermetallics 2015, 56, 79–86. [CrossRef] Boehlert, C.J.; Miracle, D.B. Part II. The creep behavior of Ti-Al-Nb O + Bcc orthorhombic alloys. Metall. Mater. Trans. A 1999, 30, 2349–2367. [CrossRef] Chen, X.; Weidong, Z.; Wei, W.; Xiaobo, L.; Jianwei, Z. Coarsening behavior of lamellar orthorhombic phase and its effect on tensile properties for the Ti–22Al–25Nb alloy. Mater. Sci. Eng. A 2014, 611, 320–325. [CrossRef] Khadzhieva, O.G.; Illarionov, A.G.; Popov, A.A. Effect of aging on structure and properties of quenched alloy based on orthorhombic titanium aluminide Ti2 AlNb. Phys. Met. Metall. 2014, 115, 12–20. [CrossRef] Emura, S.; Araoka, A.; Hagiwara, M. B2 grain size refinement and its effect on room temperature tensile properties of a Ti–22Al–27Nb orthorhombic intermetallic alloy. Scr. Mater. 2003, 48, 629–634. [CrossRef]

Metals 2018, 8, 382

13.

14. 15. 16.

17. 18. 19. 20.

21.

22. 23. 24.

25. 26. 27. 28. 29. 30. 31. 32. 33.

16 of 16

Wei, W.; Weidong, Z.; Chen, X.; Xiaobo, L.; Jianwei, Z. Designed bimodal size lamellar o microstructures in Ti2 AlNb based alloy: Microstructural evolution, tensile and creep properties. Mater. Sci. Eng. A 2014, 618, 288–294. [CrossRef] Xia, X.; Chen, Q.; Zhang, K.; Zhao, Z.; Ma, M.; Li, X.; Li, Y. Hot deformation behavior and processing map of coarse-grained Mg–Gd–Y–Nd–Zr alloy. Mater. Sci. Eng. A 2013, 587, 283–290. [CrossRef] Xia, X.; Chen, Q.; Huang, S.; Lin, J.; Hu, C.; Zhao, Z. Hot deformation behavior of extruded Mg–Gd–Y–Nd–Zr alloy. J. Alloys Compd. 2015, 644, 308–316. [CrossRef] Quast, J.P.; Boehlert, C.J. Comparison of the microstructure, tensile, and creep behavior for Ti-24Al-17Nb-0.66Mo (atomic percent) and Ti-24Al-17Nb-2.3Mo (atomic percent) alloys. Metall. Mater. Trans. A 2007, 38, 529–536. [CrossRef] Boehlert, C.J.; Majumdar, B.S.; Seetharaman, V.; Miracle, D.B. Part I. The microstructural evolution in Ti-Al-Nb O+Bcc orthorhombic alloys. Metall. Mater. Trans. A 1999, 30, 2305–2323. [CrossRef] Jin, Z.; Cui, Z. Investigation on dynamic recrystallization using a modified cellular automaton. Comput. Mater. Sci. 2012, 63, 249–255. [CrossRef] Xu, W.; Jin, X.; Shan, D.; Chai, B. Study on the effect of solution treatment on hot deformation behavior and workability of Mg-7Gd-5Y-0.6Zn-0.8Zr magnesium alloy. J. Alloys Compd. 2017, 720, 309–323. [CrossRef] Zhang, Y.; Jiang, S.; Liang, Y.; Hu, L. Simulation of dynamic recrystallization of NiTi shape memory alloy during hot compression deformation based on cellular automaton. Comput. Mater. Sci. 2013, 71, 124–134. [CrossRef] Ma, X.; Zeng, W.; Xu, B.; Sun, Y.; Xue, C.; Han, Y. Characterization of the hot deformation behavior of a Ti–22Al–25Nb alloy using processing maps based on the murty criterion. Intermetallics 2012, 20, 1–7. [CrossRef] Sherby, O.D.; Taleff, E.M. Influence of grain size, solute atoms and second-phase particles on creep behavior of polycrystalline solids. Mater. Sci. Eng. A 2002, 322, 89–99. [CrossRef] Appel, F.; Christoph, U.; Oehring, M. Creep deformation in two-phase titanium aluminide alloys. Mater. Sci. Eng. A 2002, 329, 780–787. [CrossRef] Tang, F.; Nakazawa, S.; Hagiwara, M. The effect of quaternary additions on the microstructures and mechanical properties of orthorhombic Ti2 AlNb-based alloys. Mater. Sci. Eng. A 2002, 329, 492–498. [CrossRef] Prasad, Y.V.R.K.; Seshacharyulu, T. Processing maps for hot working of titanium alloys. Mater. Sci. Eng. A 1998, 243, 82–88. [CrossRef] Srinivasan, N.; Prasad, Y.V.R.K.; Rama Rao, P. Hot deformation behaviour of Mg–3Al alloy—A study using processing map. Mater. Sci. Eng. A 2008, 476, 146–156. [CrossRef] Sivakesavam, O.; Prasad, Y.V.R.K. Hot deformation behaviour of as-cast Mg–2Zn–1Mn alloy in compression: A study with processing map. Mater. Sci. Eng. A 2003, 362, 118–124. [CrossRef] Rao, K.P.; Prasad, Y.V.R.K.; Suresh, K. Hot working behavior and processing map of a γ-tial alloy synthesized by powder metallurgy. Mater. Des. 2011, 32, 4874–4881. [CrossRef] Sarebanzadeh, M.; Mahmudi, R.; Roumina, R. Constitutive analysis and processing map of an extruded Mg–3Gd–1Zn alloy under hot shear deformation. Mater. Sci. Eng. A 2015, 637, 155–161. [CrossRef] Karami, M.; Mahmudi, R. Hot shear deformation constitutive analysis and processing map of extruded Mg–12Li–1Zn Bcc alloy. Mater. Des. 2014, 53, 534–539. [CrossRef] Li, A.B.; Huang, L.J.; Meng, Q.Y.; Geng, L.; Cui, X.P. Hot working of Ti–6Al–3Mo–2Zr–0.3Si alloy with lamellar α+β starting structure using processing map. Mater. Des. 2009, 30, 1625–1631. [CrossRef] Poletti, C.; Degischer, H.P.; Kremmer, S.; Marketz, W. Processing maps of Ti662 unreinforced and reinforced with TiC particles according to dynamic models. Mater. Sci. Eng. A 2008, 486, 127–137. [CrossRef] Boehlert, C.J. The effects of forging and rolling on microstructure in O + Bcc Ti-Al-Nb alloys. Mater. Sci. Eng. A 2000, 279, 118–129. [CrossRef] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).