Mechanical properties, formability and corrosion resistance of thermomechanically controlled processed Ti-Nb stabilized IF steel

Materials Science & Engineering A 684 (2017) 22–36

Contents lists available at ScienceDirect

Materials Science & Engineering A journal homepage: www.elsevier.com/locate/msea

Mechanical properties, formability and corrosion resistance of thermomechanically controlled processed Ti-Nb stabilized IF steel

MARK

⁎

Sumit Ghosha, Ajay Kumar Singha, Suhrit Mulaa, , Prasenjit Chandab, Vinay V. Mahashabdeb, T.K. Royb a b

Department of Metallurgical and Materials Engineering, Indian Institute of Technology Rookee, Roorkee 247667, Uttarakhand, India Flat Product Technology Group, TATA Steel, Jamshedpur, Jharkhand 831001, India

A R T I C L E I N F O

A BS T RAC T

Keywords: IF steel Phase controlled rolling Mechanical properties Substructures Transmission electron microscopy Corrosion behaviour

Aim of the present study is to examine the possibilities of achieving a better combination of yield strength (YS), ductility and formability with a better corrosion resistance of a Ti-Nb stabilize IF steel through microstructural refinement by simple phase controlled thermomechanical rolling. The phase controlled thermomechanical rolling could be an industrially reliable method for the production of hot strips, which can substitute cold rolled steel sheets. Accordingly, the phase controlled multi-pass rolling was performed in 3 selected phase regimes (γrecrystallization, γ→α transition & α-regions) on the basis of Ac3/Ar3, Ac1/Ar1 (obtained through Thermo-Calc & Gleeble-3800) and Tnr (from Boratto equation) followed by air cooling. The volume fraction of precipitate(s) correspond to the deformation temperature was estimated using Thermo-Calc Software and the morphology of the precipitates was analyzed by TEM. The strain induced phase transformation of unstable γ occurred during rolling at a high reduction of ~80%, ɛ=1.6 at γ→α transition region. Thus, dynamically recovered stable bimodal equiaxed ferrite structures (fine ferrite ~5 µm embedded with larger size ferrite grains ~32 µm) were obtained after air cooling to room temperature. In case of the rolling at α-region, improvement of the YS ( > 3-fold) is attributed to the formation of ultrafine ferrite grains (1–3 µm) through subgrain structures, strain-induced precipitation of nanosize NbC and/or TiC and micro-shear bands. Very short-annealing (~100 s) at 850 °C followed by forced air cooling was employed in order to simulate continuous annealing process and was found to improve the formability without much affecting its YS. The avoidance of FeTiP phase formation (which deteriorates to form {111} recrystallization texture) and nucleation of ferrite grains within the deformation bands (studied through EBSD study) by a short-annealing treatment are accountable for regaining the formability. The role of strain hardening exponent (n) and plastic-strain-ratio (r) on the deformation characteristics of the thermomechanically treated IF steel were also investigated to correlate the YS and uniform elongation. Furthermore, the rolled (at α-region) + short-annealed samples showed an excellent corrosion resistance due to the formation of dense oxide film on the surface. This is attributed to the dissolution of Fe(Ti+Nb)P precipitates (which are the potential sites for initiation of pits), and formation of fine grains (which facilitate to form dense oxide film on the large surface area).

1. Introduction The major requirements of materials for the automotive applications are deep drawability, high strength, weldability and surface quality [1,2]. It is well-known that the interstitial free (IF) steels are a class of high formable steels, which typically contain very low amounts of interstitial elements (Carbon ≤50 ppm and Nitrogen ≤70 ppm) resulting in excellent deep drawability [1–3]. Production of IF steels requires removal of the interstitial elements through appropriate control of melt chemistry achieved mostly with the

⁎

addition of Ti and/or Nb to remove or minimize N, S and C by forming NbC/TiC, NbN/TiN and NbCS/TiCS or NbCN/TiCN [1–4]. Solid solution strengthening elements, namely, P, Mn and Si are added to the IF steel to increase their strength although Mn and Si additions may result in some loss of deep drawability [4–6]. The benefit of P addition requires tight process controls, since P can precipitate as FeTiP during batch annealing in the temperature range of 873–1123 K (600–850 °C) [7,8]. Formation of the FeTiP precipitate diminishes the P concentration in the matrix leading to a significant loss in strength [9–11]. Moreover, the precipitation of FeTiP is also detrimental for the

Corresponding author. E-mail address: [email protected] (S. Mula).

http://dx.doi.org/10.1016/j.msea.2016.12.034 Received 27 October 2016; Received in revised form 7 December 2016; Accepted 8 December 2016 Available online 09 December 2016 0921-5093/ © 2016 Elsevier B.V. All rights reserved.

Materials Science & Engineering A 684 (2017) 22–36

S. Ghosh et al.

Table 1 Chemical composition (wt%) of the Ti-Nb stabilized IF steel obtained by optical emission spectroscopy analysis. Element

C

Mn

S

P

Si

Al

Nb

Ti

N

Fe

wt%

0.0026

0.14

0.008

0.031

0.007

0.052

0.012

0.042

0.0021

99.7

were conducted to investigate the corrosion resistance of the ultrafine grained IF steel.

formation of the {111} recrystallization texture. The {111} recrystallization texture is advantageous for better deep drawability and hence, the precipitation of FeTiP should be restricted [10–12]. Misra et al. [13] have also reported that the addition of a higher amount P could lead to grain boundary segregation, which would increase the ductileto-brittle transition temperature (DBTT) leading to cold work embrittlement [13,14]. Therefore, proper control of P, suitable annealing temperature and holding time are very important to overcome these problems. Addition of Si on the other hand is not desirable science it deteriorates surface quality, coating adhesion and causes surface oxidation of the IF steel [11]. Yasushi et al. [12] have demonstrated that the high strength can be achieved without Si addition, instead by forming fine Nb(C,N) precipitates, which increase the strength by dispersion hardening in combination with ferrite grain size refinement. Recently, attention has been focused on the improvement of strength of IF steels by microstructural modification through proper control of the hot rolling parameters. Usually, the processing of an IF steel sheet consists of slab reheating, hot rolling, pickling, cold rolling and final annealing. The mechanical properties can be tailored in a wide range by controlling the complex interaction of several metallurgical changes during processing [1–4]. Several investigations have been carried out on microstructure and texture evolution during cold rolling followed by annealing of the IF steels [4,5,10]. The use of hot-rolled steel sheet as a substitute for its cold-rolled counterpart has been considered a cost saving method for the IF steels [15–17], although hot-rolled steel sheet does not have good formability. Some investigations have been carried out on texture control of hot-rolled sheets that propose the finish hot-rolling to occur in the ferrite region. This is beneficial as the subsequent recrystallization annealing is not essential after the finish rolling [18–20]. The physical metallurgy of ferritic hot rolling is significantly different than that of conventional austenitic hot rolling [21–23], and has many advantages such as energy saving, uniformity of temperature control through the thickness etc [17]. Moreover, high intense rolling texture, including normal direction (ND) fibers and rolling direction fibers, that are developed during ferritic rolling of the IF steels, can transform to 〈111〉//ND recrystallization, which is highly beneficial for deep drawability. During high temperature coiling, deformed grains are recrystallized completely and strong 〈111〉//ND recrystallization textures are developed. In order to acquire good drawability, it is necessary to increase the intensity of the 〈111〉//ND recrystallization texture [23]. The present study investigates the effect of hot rolling temperature range (from ferritic to austenite regions) on the microstructure, mechanical properties, formability and corrosion behaviour of ultralow carbon Ti-Nb stabilized IF steel sheets. The path controlled rolling was carried in 3 different phase regions: austenite recrystallization, non-recrystallization and dual phase region. The effect of very shorttime (100 s) annealing at high temperature ~1123 K (850 °C) was also studied with an aim to recover significant amounts of formability without much interference to the yield strength. Forced air cooling was used after the short annealing to avoid formation of the FeTiP precipitates, which suppress the {111} recrystallization textures. Special attention has been given to analyze the effect of rolling temperatures on the formability in light of these two considerations (i.e., FeTiP precipitates and {111} recrystallization textures). An additional piece of this study is to understand the role of strain hardening exponent (n) and plastic strain ratio (r) on deformation characteristics of the thermomechanically treated IF steel, which has not been reported in the literature. Potentiodynamic polarization tests

2. Material and experimental details The Ti-Nb stabilized IF steel used for the present investigation was supplied by TATA Steel, Jamshedpur, India. The chemical composition (wt%) obtained by optical emission spectroscopy analysis (Spectrolab, Germany) is given in Table 1. Dilatometry tests were conducted in the Gleeble-3800 thermomechanical simulator to obtain critical temperatures such as austenite to ferrite start transformation temperature (Ar3) and austenite to ferrite finish transformation temperature during cooling (Ar1). A cylindrical sample (Ф10 mm×80 mm) was heated to a temperature of 1473 K (1200 °C) at a heating rate of 5 °C/s, hold for 1 min and then cooled to room temperature at a slow cooling rate of 1 °C/s. The slope of the curve changes due to the volumetric contraction/expansion of different phases as a result of phase transformation [24]. The values of Ar3 and Ar1 obtained from the cooling curve as shown in Fig. 1 are 974 K (701 °C) and 1133 K (860 °C), respectively; whereas, Ac1 and Ac3 are estimated to be 1188 K (915 °C) and 1198 K (925 °C), respectively. The experimental values Ac1 and Ac3 were correlated with the theoretical values (i.e. ~1183 K (910 °C) and 1203 K (930 °C)) obtained through Thermo-Calc software. Recrystallization stop temperature, Tnr, was calculated using the Boratto equation as follows [25]:

Tnr = 887 + 464C + (6445Nb−644 Nb ) + (732V −230 V ) + 890Ti + 363Al−357Si

(1)

The value of Tnr is found to be ~1243 K (970 °C) for the experimental material. All the elementals inputs (C, Nb, Ti, etc.) to Eq. (1) are in wt%. Based on the calculated critical temperatures (Ar3, Ar1 and Tnr), the controlled rolling was performed using a two high rolling mill at three different phase regimes: (a) pure austenitic at ~1323 K (1050 °C); (b) pure ferrite at ~923 K (650 °C); and (c) austenite -ferrite transition state zone at ~1073 K (800 °C). The rolling schedules are shown schematically in Fig. 3. The specimens used for thermo-mechanical controlled rolling with dimensions of 30 mm×24 mm×10 mm were machined from the cast steel. The samples were first homogenized at

Fig. 1. Dilatometry curve for Ti-Nb stabilized IF steel obtained through Gleeble-3800.

23

Materials Science & Engineering A 684 (2017) 22–36

S. Ghosh et al.

punched from this thin foil using a Gatan disk cutter. Electro-polishing was performed with the FEI twin jet electro-polisher in a solution of 90% methanol+10% perchloric acid at -293 K (−20 °C) operating at 40 V. Vickers hardness measurements were conducted (model: FIEVM50 PC) at an applied load of 10 kg with a dwell time of 30 s. Tensile tests were conducted using the H25 K-S Tinius Olsen with ASTM E8 sub-size samples with a gauge length of 10 mm. A minimum of three samples were analyzed to confirm the reproducibility of the results. The fractured surfaces of the tensile specimens were analyzed with a scanning electron microscope (SEM) (model: ZEISS, 51ADD0048) operating at 15 kV. The corrosion behaviour of the thermomechanically processed samples was evaluated with potentiodynamic polarization tests in a 3.5% NaCl solution at room temperature. The corrosion tests were conducted on well-polished samples with dimensions of 4 cm×4 cm×2 cm. The polarization study was conducted using a Gamry potentiostat (interface 1000). The electrochemical corrosion cell used the specimen as a working electrode, saturated calomel electrode (SCE) as a reference electrode and graphite rod as a counter electrode. The electrodes were placed in a flat cell in such a way that only ~1 cm2 area of the sample was exposed to the test solution. The ratio of solution volume to sample area was kept constant at 300 ml/ cm2. Prior to the experiments, the samples were immersed in the electrolyte for 1 h to establish the open circuit potential. The polarization studies were performed from 300 mV in the cathodic direction to 300 mV in the anodic direction at a scan rate of 0.166 mV/s. The corrosion potential (Ecorr) and the corrosion current density (icorr) were determined from the polarization curves using the Tafel extrapolation method.

Fig. 2. Thermo Calc calculations combined with the SGTE database for calculation of critical temperatures.

3. Results and discussion 3.1. Evaluation of critical temperatures The critical temperatures, Ar3 and Ar1 were determined from the dilatometry curves obtained with the thermomechanical simulator (Gleeble-3800) and are shown in Fig. 1. A slower cooling rate of 1 °C/s instead of the higher heating rate of 5 °C/s was used to get more realistic (close to the equilibrium) values of the Ar3 and Ar1. The curve shows that the dilation in the sample increases until 1193 K (920 °C) during heating; thereafter, there is a decrease in the dilation which suggests contraction (due formation of close packed γ-austenite) in the specimen. As dissolution of the ferrite phase about to complete, the curve dips to a minimum and starts rising again indicating expansion in the single phase γ-austenite. The maximum temperature used in the present work was 1473 K (1200 °C). Cooling of the sample, on the other hand, shows a decrease in the specimen dimension until it reaches 1133 K (860 °C). After that there is an increase in dilation up to 974 K (701 °C), which is followed by a decrease in dilation again due to contraction of the specimen up to room temperature. Thus, Ar3 and Ar1 for the present material were estimated to be 1173 K (860 °C) and 974 K (701 °C), respectively. The critical temperatures, Ac1 (1183 K) and Ac3 (1203 K) were estimated using Thermo-Calc software and the SGTE database analysis as shown in Fig. 2. The experimentally obtained values of Ac1 (1188 K) and Ac3 (1198 K) are in good agreement with the theoretical calculations.

Fig. 3. Schematic design of controlled rolling schedule.

1473 K (1200 °C) for 1 h in order to dissolve the microalloying elements, i.e. Ti and Nb; then the samples were rolled at the desired temperatures. Multiple numbers of passes were given to achieve the total reduction in area (RA) i.e. 50% and 80% RA. After hot rolling, the steel plates were air cooled to room temperature. Complete dissolution of NbC and TiC during the homogenization annealing at 1473 K (1200 °C) was confirmed from the following equations [26]:

log [Nb][C ] = − 6770/ T + 2.26

(2)

log [Ti ][C ] = − 7000/ T + 2.75

(3)

The samples controlled rolled in the ferrite region (both 50% RA and 80% RA) were further annealed at 1123 K (850 °C) for 100 s in order to partially recover and recrystallize the ferrite grains from its heavily deformed structure. Microstructural analysis was performed using optical microscopy (Leica DMI 5000M), electron back scattered diffraction (EBSD) and transmission electron microscopy (TEM). EBSD analysis was performed using FEI-Quanta 200FE-SEM equipped with the TSL data acquisition system. For the EBSD analysis, specimens were mechanically polished on fine cloth using colloidal silica, followed by electropolishing in an electrolyte of 20% perchloric acid in methanol at −40 °C operated at 21 V for 50 s. The EBSD scans were performed on the polished samples with a step size of 0.5 µm, with subsequent analysis conducted through TSL-OIM software. The TEM analysis was carried out using FEI Technai 20 G2S-Twin transmission electron microscope operating at 200 kV. The TEM samples were prepared by mechanical polishing using silicon carbide abrasive papers (800, 1200, 1500 grit size) and thinned down to 100 µm. The 3 mm disk samples were

3.2. Microstructural characterization 3.2.1. Optical microstructures The optical micrograph of the as-cast IF steel sample is shown in Fig. 4a. This cast microstructure is inhomogeneous and predominantly contains small size ferrite grains as the main microconstituent along with several larger size ferrites. Pearlite is not observed in the microstructure due to ultra-low C content (0.0026%) in the steel. The cast billet was homogenized annealed at 1473 K (1200 °C) for 1 h prior 24

Materials Science & Engineering A 684 (2017) 22–36

S. Ghosh et al.

Fig. 4. Optical microstructure of IF steel: (a) as cast sample, (b) homogenized annealed at 1473 K (1200 °C), (c-d) 50% and 80% RA at 1323 K (1050 °C) (γ-recrystallization region), (e– f) 50% and 80% RA at 1073 K (800 °C) (γ→α transition region), (g) 50% RA at 923 K (650 °C) (pure α region), (h) Magnifying image of selected square section of image 4 g, (i) Optical microstructure after 80% RA at 923 K (650 °C), (j) Magnifying image of selected square section of image 4i, (k) Optical microstructure of 50% RA at 923 K (650 °C)+short annealed (l) Optical microstructure of 80% RA at 923 K (650 °C)+short annealed.

the desired temperature before proceeding to the next pass. In each phase regime, the specimens were deformed to 50% and 80% reduction in area (RA) while maintaining a constant desired temperature. In the γ-recrystallization region at ~1323 K (1050 °C), the grain refinement occurred in each pass by continuous dynamic recrystallization from the deformed austenite grains. The hot deformed specimens were air cooled to room temperature after rolling during which the proeutectoid ferrite nucleated along the small recrystallized austenite grain boundaries at around Ar3. The room temperature microstructure of the 50% and 80% deformed specimen consisted of equiaxed ferrites (Fig. 4c and d) with an average grain size of ~37 and 33 µm, respectively. Cuddy et al. [27] also studied the hot deformation behaviour of low carbon (0.12%) Nb microalloyed steel and found a uniform equiaxed ferrite-pearlite microstructure after hot rolling at above the recrystallization temperature followed by air cooling. The 50% RA sample shows relatively coarser and unevenly distributed equiaxed ferrite grains due to the dislocations produced during hot rolling undergoing a recovery process. This is associated with the dislocation movement and absorption at these high temperatures and lead to a low density of dislocations in the microstructure. Thus, a relatively larger size recrystallized grains were developed from the low

to rolling to dissolve carbide particles, if any. The optical micrograph of the homogenized annealed (H-AN) steel is shown in Fig. 4b. The average grain size was determined using line intercept method from 200 such grains, and the grain size was estimated to be 210 and 260 µm, respectively, for the as-cast and H-AN samples. Recrystallized austenite grains were developed from the cast structure and all alloying elements were completely dissolved in the austenite phase during the annealing. Complete dissolution of NbC and TiC during the homogenization annealing was confirmed from the analysis using the Eqs. (2) and (3) (as discussed earlier). Cooling followed by H-AN led to the formation of pro-eutectoid ferrites which were nucleated at the austenite grain boundaries at Ar3 and grew further as the sample cooled down to room temperature. Alloying elements, such as Ti and Nb again precipitated as carbides/nitrides during the cooling process. Based on the recrystallization stop temperature of ~1243 K (970 °C), and critical temperatures of 974 K (701 °C) (Ar1) and 1133 K (860 °C) (Ar3) (obtained from dilatometry analysis), the HAN samples were control rolled with multiple passes at three different temperatures: 1323 K (1050 °C) (γ-recrystallization region), 1073 K (800 °C) (γ→α transition region) and 923 K (650 °C) (α-region). After each pass, the sample was kept in the furnace for few minutes to attain

25

Materials Science & Engineering A 684 (2017) 22–36

S. Ghosh et al.

Fig. 4. (continued)

formation of complex carbides/nitrides that restrict grain growth and also due to the solute drag effect of Nb in solid solution. Complete recrystallization does not take place in the current steel between the rolling passes in the γ→α transition region (just below Ar3) at 1073 K (800 °C), and the deformation strain is retained from one pass to the next pass. After 50% RA, some ferrite grains were elongated along the rolling direction (Fig. 4e) and some are recrystallized (strain induced ferritic transformation just started). As the %RA increases (to 80%), the retained strain increases which increases number of nucleation sites for γ→α transformation. Therefore, a fine ferritic (FF) grain structure is likely to form, which can also be observed in Fig. 4f. The microstructure consists of FF (av. grain size of ~5 µm) embedded with larger size ferrite grains (av. size ~32 µm, estimated from 300 grains). At 800 °C just before the deformation, the microstructure consists of mainly smaller size nucleated ferrites (just starting to nucleate from γ) and retained austenite. As γ-austenite is a low stacking fault energy material, a strong dislocation substructure is produced inside the remaining austenite grains during rolling [33]. Ferrite grains are nucleated within these dislocation substructures (due to high energy present) and FF grains are formed via strain-induced transformation of the austenite. Thus, a higher deformation stress tends to (i) increase γ→α transformation and (ii) start strain-induced phase transformation of ferrite during deformation. As a result, control rolling of the IF steel in this region followed by normal air cooling can produce fine equiaxed

density available nucleation sites. On the other hand, the microstructure of 80% RA sample shows relatively finer equiaxed and uniformly distributed ferrite grains due to the larger amount of accumulated dislocation density available from the higher deformation that leads to more nucleation sites in the deformed austenite grains. A finer sized equiaxed austenite grains are developed through dynamic recrystallization in this case [28]. The austenite grains are expected to resist coarsening during the intermittent heating period due to the presence of fine carbide and/or nitride precipitates [29]. As example, Arribas et al. [30] found that the presence of TiN particles inhibited austenite grain growth during interpass delay times and maintained fine recrystallized austenite grains. Hu et al. [31] demonstrated that inhibiting grain growth may be attributed to the solute drag effect promoted by Nb in solid solution. Gong et al. [32] examined the dissolution kinetics of NbC and (Ti,Nb)C in HSLA steels during holding hot rolled steel at 1473 K (1200 °C). They reported that rate of austenite grain coarsening was less for the Nb–Ti steel than that for the Nb steel, because of the high temperature stability of (Ti,Nb)C. Furthermore, the excessive grain growth of proeutectoid ferrites was also expected to be inhibited during the cooling process due to the presence of these complex carbides. TEM analysis showed the presence of uniformly distributed TiC/NbC precipitates in the current steel (discussed later in section 3.1.3 TEM analysis section). Thus, the fine recrystallized microstructure obtained is due to the 26

Materials Science & Engineering A 684 (2017) 22–36

S. Ghosh et al.

the range of 4–10 µm (avg. size ~4.2 µm) along with some (~15%) submicron level grains of < 1 µm. Zhang et al. [37] studied the effect of short-time annealing at high temperature (at 850 °C) after 80% cold deformation of Nb-Ti IF steel. They found a partially recrystallized fine irregular and elongated grains after annealing for short duration ~90– 120 s at 850 °C; while a fully recrystallized equiaxed grain structure mostly consisting of 4–5 µm size was obtained when annealed for 150 s at the same temperature. After annealing for 180 s at the same temperature, bimodal grain sizes with mostly in the range of 6–7 µm were reported to form. Nowadays, in modern steel industries, continuous-annealing process is used in order to produce good product uniformity and better surface cleanliness than the conventional batch annealing process for the production of thin sheets [37].

ferrite grains (~5 µm) distributed in the larger size ferrite (32 µm). The major grain refinement was obtained when the thermomechanical treatment was given in the pure ferritic region at 923 K (650 °C) (below Ar1). In case of 50% rolling, most of the ferrite grains were elongated along the rolling direction and deformation bands can be observed in the specimen (Fig. 4g) with a few recrystallized grains within the deformed bands. The microstructure contained two kinds of ferrite grains: approximately 80% of the grains with a large aspect ratio (elongated) and 20% of the grains (subgrains) with an average size of ~1–3 µm (magnified view can be seen in Fig. 4h). In case of the 80% deformation, a large amount of subgrains along with the formation of microshear bands can be observed (Fig. 4i–j). Repeated deformation of the coarse-grained ferrite leads to a highly strain hardened condition associated with a high dislocation density. Subsequently, a large number of dislocation cells or subgrains are developed both at the ferrite boundaries and within the deformed grains. These subgrains finally recover and recrystallize to generate new equiaxed strain-free grains [20–22]. It is well known that lowering the deformation temperature can promote deformation induced ferrite transformation (DIFT) [22] and consequently decrease the ferrite grain size. The microshear bands developed during deformation create new dislocations, increase accumulation of dislocations and reorganization between them [23]. It also increases the grain boundary misorientation, transformation of low angle boundaries to high angles grain boundaries and finally to the formation of ultra-fine ferrite grains [23]. Sekban et al. studied the effect of friction stir processing (FSP) on the microstructural evolutions of Ti-stabilize IF steel and reported to form a fine grain size distribution in the range of 3–7 µm at the central part of the FSPed region. The formation of such fine grains was ascertained to the deformation- and temperature-induced grain subdivision mechanism [34]. The development of a bimodal grain structure was also achieved in a low C (0.15%) Nb-Ti stabilize microalloyed steel during deformation in the dual phase region ~1073 K (800 °C) [35]. The bimodal grain structures are mainly developed due to the formation of microshear bands followed by dynamic recrystallization of austenite and DIFT [35]. In addition, the presence of a high density of dislocations encourages nucleation of Nb/Ti-carbonitride precipitates [4,5], which has been confirmed by TEM analysis (discussed later). The carbonitride precipitates play an important role in the microstructural refinement [4,5] by resisting the grain growth. After deformation in the ferritic region, some specimens were short time annealed followed by forced air cooling to improve formability of the steel (discussed later). It is well-known that the recrystallization and growth of ferrite grains are highly influenced by the annealing time; indeed, grain growth occurred very rapidly at high temperatures [28]. As a result, optimum annealing time of 100 s at 1123 K (850 ℃) was chosen in the present study to avoid deterioration of mechanical properties. The samples were forced air-cooled to room temperature following the annealing. For 50% RA + short-time-aged samples, a bimodal grain size distribution consisting of larger (~25 µm) and finer (6–7 µm) grains were observed (Fig. 4k). It is found that few ferrite grains were partially recrystallized in this case, and only a few ferrite grains coarsened during the 100 s annealing used. In case of 80% deformation + short time annealing, most of the grains were seen to be extremely fine along with a large number of recrystallized grains, which are observed in between the shear bands (Fig. 4l). The ultrafine grains could not be resolved by OM and the grain size (~1 µm) was analyzed by EBSD and TEM analysis (discussed later). Saray et al. [36] investigated post deformation annealing experiments after multi-pass deformation through equal-channel angular extrusion (ECAE) of a Tistabilized IF steel. Low temperature (500–650 °C) annealing experiments were carried out for various duration from 12 to 60 min with an aim to obtain partially recrystallized bi-modal microstructure. And the desired partially recrystallized microstructure was obtained when the annealing experiment was done at 600 °C for 12 min. The resultant microstructure was found to consist of bimodal recrystallized grains in

3.2.2. EBSD analysis Electron back scattered diffraction (EBSD) analysis was conducted for the α-region rolled (50% and 80% RA)+short annealed (S-AN) specimens after forced air cooling; since these samples showed significantly refined microstructures that could not be resolved by OM. The inverse pole figure map with superimposed grain boundary map of both specimens depict a severely deformed bimodal ferrite grain structures, as shown in Fig. 5a and c, respectively. Fine grain ferrites embedded within the larger size ferrites regions are clearly evident from the EBSD images (Fig. 5a and c). The different grain colours indicate the orientation of each grain as shown in the inset unit triangle. The magnified view (in Fig. 5b) of the 50% deformed+short annealed specimen showed partially finer grains (~5 µm) along with larger sized grains (~32 µm). Fig. 5c and its magnified view in Fig. 5d and e show EBSD micrographs of the 80% RA+S-AN specimen showing morphological distribution of the bimodal grains. It reveals that the ultrafine grains (~30%) with a size of 1–3 µm are distributed in larger sized ferrites (70%) of ~26 µm size. A large number of ultrafine ferrite grains can clearly be observed in the magnified EBSD images in Fig. 5d and e. The EBSD micrographs show the new ferrite grains developed along the grain boundaries of the pre-existing larger ferrite grains. The average grain size distribution and misorientation angle for both the samples are shown in Fig. 5f and g, respectively, which showed that the majority of the grains had mostly low angle grain boundaries (Fig. 5g). The analysis shows that the misorientation angle of the grain boundaries (Fig. 5g) is comparatively higher for the 80% rolled sample as compared to that of the 50% deformed specimen. On the other hand, a large amount of high angle boundaries in the 80% deformed sample can be attributed to the presence of comparatively finer sized grains due to a higher degree of recrystallization in the 80% rolled sample in the α-ferrite region [33,37]. 3.2.3. TEM analysis The volume fractions of probable precipitates formed at different deformation temperatures were estimated with the SGTE database of the Thermo-Calc Software and are shown in Fig. 6. The amount and type of the precipitates formed changed with the processing temperature. For example, the amount of (Ti+Nb)C/N phase formation decreased as the processing temperature increased beyond 1113 K (840 °C); whereas, the probability of formation of Ti4C2S2 and TiS phases increased at temperatures above 1123 K (850 °C). The detailed microstructural features of the IF steel processed at different temperatures are shown in TEM micrographs in Fig. 7–11. The formation of strain-induced dynamic recrystallized ferrites along with elongated ferrite grains can clearly be observed from the bright field TEM image (in Fig. 7) of the sample deformed in the transition (γ→α) region. Figs. 8a and 9a show the bright field TEM images of 50% and 80% rolled (at 923 K (650 °C)) specimens, respectively, followed by air cooling. Ultrafine ferrite grains with a high density of dislocations were formed (Fig. 8a) in sample rolled in the 80% ferritic region. Fig. 8b shows the TEM image of a sample rolled to 80% in the pure ferritic 27

Materials Science & Engineering A 684 (2017) 22–36

S. Ghosh et al.

Fig. 5. (a) EBSD image of ferritic region controlled rolled (50%RA) + short annealed specimen, (b) magnifying EBSD image of selected square section of (a), (c) EBSD image of ferritic region controlled rolled (80% RA) + short annealed specimen, (d–e) magnifying EBSD image of selected square section of (c), (f) average grain size distribution, (g) grain boundary misorientation profile.

28

Materials Science & Engineering A 684 (2017) 22–36

S. Ghosh et al.

The SAED spot pattern (appearing as a regular square shape) along the [001] zone axis (Fig. 9c) also confirms the formation of TiC, while the background spot pattern indicates the presence of a relatively large grained ferrite matrix. The TiC particles are distributed evenly in the ferrite matrix as shown in Fig. 9b and average could be around 20 nm. On the other hand, it is difficult to distinguish the presence of TiS and Ti4C2S2 particles and their morphology in the steel. The formation of recovered equiaxed grains (~1 µm) from the dislocation substructures can be ascertained from Fig. 9d and were formed due to recrystallization at the optimum annealing conditions. Fig. 10a shows the SAED pattern corresponding to Fe(Ti+Nb)P particles obtained from TEM analysis of 80% RA ferritic region rolled sample. The dark field TEM image in Fig. 10b was generated from the diffraction spot encircled in Fig. 10a and shows the presence of ~30– 50 nm size Fe(Ti+Nb)P particles distributed within the matrix. Pampa et. al [7,8] reported that Fe(Ti+Nb)P particles affect the sheet formability in two stages: First, Fe(Ti+Nb)P particles precipitate during warm rolling and decrease the (111) texture, while formation of the Fe(Ti+Nb)P phase allows less Ti available for TiC precipitation during recrystallization annealing. Thus, a higher amount of resultant interstitial carbon in the matrix leads to the formation of a weaker {111} texture, which is associated with lower formability. Secondly, Fe(Ti +Nb)P dissolved at a temperature above 1073 K (800 °C) promotes the formation of {111} texture again. The TEM analysis confirmed the regions where microshear bands formed and localization of precipitates occurred in the IF steel during deformation in the pure ferritic region (Fig. 11a). During annealing, the dark region in Fig. 11a, which comprises of dislocations, shear bands and precipitates, tends to become uniformly piled up as shown in Fig. 11b. The nucleation for recrystallization can start easily in the piled up region, which is essentially a high angle grain boundary area [7,8].

Fig. 6. Volume fraction of the second-phase particles (e.g. precipitate) evaluated by SGTE database of the Thermo-Calc study.

3.3. Mechanical properties 3.3.1. Hardness and tensile properties The mechanical properties of the controlled rolled specimens were estimated by Vickers hardness measurements and tensile tests; and the results are compared with those of the as-cast and the homogenized annealed (H-AN) specimens. Fig. 12 shows the Vickers hardness of the samples processed at different conditions. The average hardness value of the H-AN specimen was only 70 HV and increased rapidly to a high range of hardness (120–150 HV) after 80% RA due to strain hardening and grain size refinement during controlled rolling at different critical temperatures. The hardness increased to 115 and 121HV (1.6 and 1.7 fold) for the 50% and 80% RA samples, respectively, when rolling was conducted at ~1323 K (1050 °C). The hardness values increased further when rolling was conducted at lower temperatures i.e. at just below Ar3 (~1073 K) and in the pure ferritic region (~923 K). The maximum hardness was measured to be 151HV for the sample 80% rolled at 923 K (650 °C). The increased hardness is ~2.2 times higher than that of the H-AN specimen. This increase in hardness is primarily due to the continuous refinement of ferrite grains at lower rolling temperature. Tensile tests were conducted to understand the total plasticity of the materials processed in the various conditions. The engineering stress-strain curves of the as-cast, H-AN and controlled rolled specimens are compared and shown in Fig. 13. The summarized mechanical properties for all the samples are shown in Table 2. The yield strength of the as-cast sample was 165 MPa with a ductility of ~34%, while the annealed sample showed a slightly lower YS of 141 MPa with a corresponding elongation of ~46%. Decreasing the YS with corresponding increase in the elongation of the H-AN specimen correlated well with the formation of relatively homogenized and larger size grains (~260 µm; Fig. 4b) than that of the as-cast structure (210 µm; Fig. 4a). The YS (Fig. 13) of the specimen rolled for 50% and 80% RA in the γ-recrystallized region increased to 302 and 323 MPa, respectively. The YS further increased to 351 and

Fig. 7. TEM image of transition region (γ → α) control rolled (50% RA) sample.

region and short annealed for 100 s. Fine equiaxed grains of ~500 nm were formed following recrystallization in this optimum annealing condition. It is known that many of the deformed elongated grains and substructure boundaries recover into subgrain boundaries [20–22]. This leads to reordering of dislocations together with some loss of dislocations as the recovery proceeds and recrystallization occurs to form more subgrains [20–22]. The presence of fcc niobium carbide (NbC) along with its selected area electron diffraction (SAED) pattern is presented in Fig. 8c. Apart from the spot pattern (appearing as regular hexagon) along the NbC [111] zone axis, the SAED pattern also reveals a ring pattern due to the presence of fine ferrite grains. Distribution and size of NbC precipitates can be ascertained from the TEM image shown in Fig. 8d. The image was generated (putting the aperture over the spot) from the diffracted spot as encircled in Fig. 8c. It can be noticed (from Fig. 8d) that the ultrafine NbC precipitates were ~10 nm (Fig. 8d) and were distributed uniformly in the fine ferrite matrix. On the other hand, the specimen 50% rolled in the pure ferrite region showed relatively larger ferrites (Fig. 9a) as compared to that of the 80% rolled sample. The presence of a dislocation substructure can also be observed from Fig. 9a; while the highly magnified view of the same shows the formation of strain induced TiC precipitates (Fig. 9b).

29

Materials Science & Engineering A 684 (2017) 22–36

S. Ghosh et al.

Fig. 8. (a) TEM bright field image of fine grain with high dislocation structure of ferritic rolled (80% RA) specimen at 923 K (650 °C), (b) TEM bright field image of fine recovered equiaxed grain after short annealing, (c) SAD spot pattern of NbC from [111] zone axis and ring pattern from the background fine ferrite matrix, (d) TEM dark field image, recorded from the diffracted beam from the spot as encircled in (c).

improving the YS and the larger size ferrite grains controlling the ductility. The formation of dual size ferrites and a good combination of YS and ductility were also observed by Godha et al. [22] in C-Mn steel (0.13C, 1.32Mn, 0.0032N), when rolling was done in the intercritical regime at 1003 K (730 °C). The enhancement of the YS and UTS of the γ-region controlled rolled samples is mainly due to the formation of smaller size equiaxed ferrites which formed from the fine recrystallized austenite grains. It is well-known that the higher grain boundary area of the recrystallized γ provides a large number of nucleation sites for the formation of fine ferrite grains during cooling [28,29]. Hence, the room temperature microstructure consists mainly of small sized equiaxed ferrite grains in the γ-region rolled sample. The YS of the γ→α transition region control rolled sample is found to be greater than that of the γ-regime controlled rolled samples mainly due to an additional refinement of ferrite grains (Fig. 4e–f). The evolutions of the fine ferrite grains are mainly due to an increase straining in the γ→α transformation and strain-induced phase transformation of the ferrite during deformation in this region. The strain-induced phase transformation of ferrite is also reported to improve mechanical properties in a TRIP steel, which was controlled rolled to just below the Ac3 temperature [19]. Intercritical region controlled rolled followed by quenching and partitioning of an HSLA steel showed an ultrahigh YS (1700 MPa) with significant ductility (15%) due to the development of a fine grained ferritic structure by DIFT [33]. Overall, it can be concluded that the YS of the sample rolled in the pure α-region was found to increase significantly with only a slight reduction in the ductility. The bit reduction in the ductility may

379 MPa, respectively, for the 50% and 80% RA specimens, when rolled in the austenite-ferrite (γ→α) transition region. The specimen deformed in pure ferritic region exhibited a maximum improvement in the YS and the corresponding values were estimated to be 381 and 407 MPa, respectively, for the 50% and 80% RA samples. It is worthwhile to mention that the higher YS of 407 MPa is ~ 3 times higher than that of the H-AN specimen (141 MPa). The ferritic rolled+short annealed sample followed by forced air cooling showed a slightly reduced YS of 381 and 407 MPa, respectively, for 50% and 80% RA samples, as compared to that of the ferritic rolled sample. It is observed (Fig. 13) that the YS and UTS of the specimens rolled in the 3 phase regions improved with only a slight reduction in the ductility compared with that of the H-AN sample (46%). For example, the tensile ductility of the recrystallized 80% rolled specimen decreased to 33%, with a corresponding average grain size of ~35 µm (Fig. 4d). The specimen rolled for 80% RA in the γ→α region showed a ductility of 30% and contained ferrite grains of ~5 µm size embedded within larger ferrite grains of ~32 µm (Fig. 4f). The specimen rolled for 80% in the pure α-region exhibited a tensile ductility of 27%. An improved ductility of 30% was obtained for the 80% ferritic rolled specimen after a short annealing (s-AN) for 100 s. Earlier it was mentioned that the microstructure of the specimen deformed at the pure α-region contained 30% ultrafine ferrite grains (~1–3 µm) with 70% larger sized ferrites (~25–27 µm). Consequently, it can be observed that the YS of the ferritic (α) rolled specimen is noticeably improved without much reduction in the ductility mostly due to the development of a bimodal grain structure; with the fine ferrite grains 30

Materials Science & Engineering A 684 (2017) 22–36

S. Ghosh et al.

Fig. 9. (a) TEM bright field image of fine grain with high dislocation structure of ferritic rolled (50% RA) specimen at 923 K (650 °C), (b) Enlarge view of the square section of (a) along with TEM dark field image indicates strain induce precipitates of TiC, (c) TEM bright field image of fine recovered equiaxed grain after short annealing, (d) SAD spot pattern of TiC from [001] zone axis along with sport pattern from the background fine ferrite matrix.

ing sample.

affect its formability. Very short-annealing (100 s) at 1123 K (850 °C) followed by forced air cooling was found to recover the formability (discussed in the next section) significantly without much reduction in the YS. Finally, it can be concluded that the enhancement of mechanical properties of thermo-mechanically controlled specimens is attractive and well-corroborated with the microstructure of the correspond-

3.3.2. Formability and strain hardening behaviour The strain hardening exponent (n) indicates the capability of the metal to undergo plastic deformation prior to necking/fracture and the formability of sheet metals is strongly influenced by the value of n and

−−

Fig. 10. (a) SAD spot pattern of Fe(Ti+Nb)P from [022 ] zone axis along with ring pattern from the background of the fine ferrite matrix of 80%RA ferritic rolled sample, (b) TEM dark field image, recorded from the diffracted beam from the spot as encircled in (a).

31

Materials Science & Engineering A 684 (2017) 22–36

S. Ghosh et al.

Fig. 11. (a) micro-shear bands formation, (b) The piling up of the shear bands, dislocations and the precipitates.

then decreased gradually as the processing temperature decreased. It decreased to 0.122 after 80% deformation in the ferritic region indicating a deterioration of strain hardening ability (the elongation also decreased to 8.4%), while the YS correspondingly increased to 421 MPa. It can be observed that the value of n is found to be higher (0.175) for the specimen rolled in the γ-recrystallization region as compared to that of the specimens deformed in the two other regions (i.e. γ→α transition region (0.12) and pure α region (0.091)). Usually the materials having low YS and high uniform elongation show high work-hardening capacity [38]. Dynamic recrystallization in the γrecrystallization region allows formation of new strain-free γ grains, which subsequently transform to fine strain-free ferrites during cooling. On the other hand, deformation of ferrites at 923 K leads to high rate of strain hardening followed by formation of fine ferrite grains from the dislocation substructures. Therefore, the strain hardening ability of the specimen deformed in the γ-recrystallization region is higher, while the formability of the specimen deformed in the ferrite region is poor. The extent of work hardening is generally less in the controlled rolled steel [38,39]. This can be attributed to an increase in the strain rate sensitivity with strain, especially for the materials with high strain rate sensitivity like IF steel [38]. However, IF steel generally possesses reasonably high n values (greater than 0.2) in the annealed condition leading to a high formability. A short-annealing of 100 s at a higher temperature of 1123 K followed by forced air cooling could be very effective and increased the n value of the ferritic rolled sample from 0.122 to 0.168. It can be observed from Table 2 that the total elongation of the thermomechanically treated samples lies within a narrow range between 33 to 27%, while the YS increased significantly

Fig. 12. Vickers hardness of the controlled rolled specimens compared with homogenized annealed and cast conditions.

plastic strain ratio (r).Table 2 summarizes the values of n and r (obtained from the true stress-true strain curves) along with the mechanical (tensile) properties of the thermomechanically treated IF steel sheets. It can be seen from Fig. 14 that the values of YS increase with a corresponding decrease in uniform elongation and strain hardening exponent (n). The strain hardening exponent (n) of the as-cast sample was estimated to be 0.207, which increased to 0.254 corresponding to a uniform elongation of 30% after the H-AN treatment. The value of n

Fig. 13. Engineering stress–strain curves of 50% RA and 80% RA controlled rolled specimens.

32

Materials Science & Engineering A 684 (2017) 22–36

S. Ghosh et al.

Table 2 The Mechanical properties of the homogenized annealed and controlled rolled IF steel specimens in different critical zones. UTS (MPa)

El (%)

YS (MPa)

Temp. (°C)

50 R

80 R

50 R

80 R

50 R

80 R

50 R

80 R

50 R

80 R

1050 800 650 650+S-AN H-AN As-cast

302 ± 5 351 ± 4 401 ± 6 381 ± 3 141 ± 3 165 ± 4

323 ± 5 379 ± 5 421 ± 5 407 ± 6

349 ± 3 403 ± 7 447 ± 8 420 ± 5 256 ± 5 281 ± 3

378 ± 3 425 ± 3 479 ± 3 450 ± 4

35 ± 2 32 ± 1 29 ± 1 31.5 ± 1 46 ± 1.7 34 ± 1.5

33 ± 2 30 ± 2 27 ± 2 30 ± 1

0.175 0.143 0.132 0.148 0.254 0.207

0.163 0.135 0.122 0.139

1.77 1.56 1.41 1.68 2.11 1.91

1.64 1.45 1.32 1.57

and lies in the range of 323–421 MPa. Sekban et al. [34] obtained the maximum YS of 381 MPa with a considerable decline of strain hardening exponent (n=0.12) after FSP of Ti-stabilize IF steel. Compared to the reported value of n (=0.12), the strain hardening exponent, n is found to be higher for all the rolled specimens in the present study. The minimum value of n (=0.122) was obtained for the ferritic region controlled 80% rolled specimen, which showed significantly higher YS of 421 MPa. The low strain hardening coefficient of the FSPed sample was explained to the decreasing contribution of dislocation interaction on the strength during the plastic deformation [34]. The formability of sheet metal materials is highly influenced by the − anisotropy of the material (r ), which can be estimated from the plastic − strain ratio (r). The values of r was evaluated using the specimens prepared according to the standard of ASTM E517 specification [38]. The r-value, width and gauge length of the specimens were accurately measured at an elongation of 20% (before the maximum load is reached). The final width and gauge length (r-value tests are conducted only in uniform plastic deformation range) were measured and the plastic strain ratio (r) was calculated from following equation:

r = εw / εt = εw /−(εw + ε1) = ln(wf / w0 )/ln(l0w0 / l f wf )

(r0 + 2r45 + r90 ) 4

(r )

(4) 3.4. Fractography analysis

where, w0, l0 are initial width and length, wf, lf are final width and length, ɛw= true width strain, ɛt = true thickness strain, ɛl = true length strain. The tensile samples were prepared in three particular directions with reference to the rolling direction (i.e. parallel, 45° and transverse to the rolling direction) to evaluate r0, r45 and r90. The normal − anisotropy (r ) was calculated by using the standard formula [40].

r =

(n)

N content. Also controlled thermomechanical processing+short annealing develops favourable {111} crystallographic texture in these sheets [37]. While steel sheets with low YS and large uniform elongation demonstrated a high formability [38], it was found that controlled rolling in different phase regions affected the formability of IF steel. It can be seen (Table 2) that as the deformation temperature decreases, − the value of r gradually decreases with a corresponding decrease in the strain hardening exponent (n). Especially, the sample rolled for 80% in the ferrite region showed a significant increase in YS (421 MPa) but the − r decreased to 1.32. This could be due to the formation of Fe(Ti+Nb)P phase which is not favourable to the formation of {111} recrystallization texture during rolling in the ferritic region. However, since the Fe(Ti+Nb)P phase can be dissolved above a temperature of 1073 K (800 °C) [37], heating the specimens above this temperature for a short period of time can restore the formability by promoting the formation of the {111} recrystallization texture. It is well-known that the {111} − crystallographic texture has a direct relation with r [10,41]. The 80% ferritic rolled specimens after the short-annealing (at 1073 K for 100 s) − showed an increase r value of 1.57 without a significant decrease in the YS (407 MPa). This is due to the formation of ultrafine (1–3 µm) grains (Fig. 6c) which helps to retain the high YS while simultaneously promoting the {111} crystallographic textures due to the dissolution of the Fe(Ti+Nb)P precipitates. Gao et al. [41] also reported an improvement of formability (without affecting YS) in 17% Cr ferritic stainless steel due to the formation of in-grain shear bands and {111} recrystallized textures when hot rolling was carried out at a low finishing temperature (973 K). Saray et al. [37] also observed a limited formability, inadequate strain hardenability and low uniform elongation in the UFG IF steel processed by ECAE. They reported that the low temperature (600–650 °C) post-ECAE annealing treatment showed a strong effect on improvement of the formability of IF steel. This effect became more pronounced with increasing the annealing temperature and/or time. For example, post-ECAE annealing at 600–650 °C was found to be effective to improve the formability when the annealing was done for 60 min. But, subsequently the YS decreased to a very low range (201–320 MPa) due to the coarsening of the recrystallized grains. Therefore, the post deformation annealing above 1073 K (800 °C) for very short duration (e.g. 100 s) is much effective to improve the formability of IF steel without deterioration of the YS.

Fig. 14. Variation of YS and uniform elongation (UE) with strain hardening exponent.

−

−

Rolling

The fractured surface of the selected tensile specimens was examined in a scanning electron microscope to analyze the failure mode and the fractographs are shown in Fig. 15a-d. The presence of dimples in all specimens indicates that the fracture occurred in a ductile manner. Relatively larger size and deeper dimples can be observed (in Fig. 15b) on the fractured surface of the 50% ferritic rolled sample compare to the 80% rolled sample (Fig. 15a). In case of the 80% rolled sample, the fractured surface consists of a large number of fine dimple marks, which is due to a decrease in tensile elongation (Fig. 4i,j) as a result of grain size refinement and strain hardening associated with the plastic deformation [28,38]. The fractured beha-

(5)

where, the subscript indicates the orientation of the specimen axis with − respect to the rolling direction. High value of normal anisotropy (r ) greater than 2.0 indicates an excellent drawability of sheet metals. A − high r value is expected for the IF steel sheets due to its very low C and 33

Materials Science & Engineering A 684 (2017) 22–36

S. Ghosh et al.

Fig. 15. SEM fractographs of tensile test specimen of controlled rolled at ferritic region: (a) 50% rolled, (b) 80% rolled, (c) 50% rolled + S-AN (d) 80% rolled+S-AN respectively.

annihilation of metastable pits [42] occur. Xia et al. [43] reported a similar behaviour for coarse grained plain C-steel that undergone pitting corrosion in an aerated chloride solution. It can be noticed that the rolled samples showed stabilization in OCP earlier than that of the H-AN sample, which indicates early passivation in the treated samples. The H-AN sample exhibited a more negative OCP of 895 mVSCE when compared with that of the other 3 specimens (692 mVSCE for 50% ferritic rolled, 684 mVSCE for 80% ferritic rolled and 667 mVSCE for rolled+short annealed). The ferritic rolled+short annealed sample showed about 228 mV more positive OCP than that of the H-AN indicating an improved resistance to corrosion.

viour correlates well with the tensile elongation of the corresponding specimen. After the short-annealing, both the 50% and 80% rolled samples showed relatively more deeper dimple marks along the uniformly distributed fine dimples (Fig. 15c and d). 3.5. Corrosion behaviour 3.5.1. Open circuit potential (OCP) The change in open circuit potential (OCP) as a function of time in an aerated solution of 3.5 wt% NaCl for the differently treated samples is shown in Fig. 16a. The figure shows that the frequency of oscillations in OCP is more in the initial stage. These oscillations are related to the activation and repassivation processes. In this stage, initiation and

Fig. 16. (a) Variation in open circuit potential (OCP) as a function of time and (b) Potentiodynamic polarization scans for the selected thermo-mechanical treated samples in 3.5 wt% NaCl solution.

34

Materials Science & Engineering A 684 (2017) 22–36

S. Ghosh et al.

rolled and H-AN samples. Krishna et al. [47] also observed that the precipitates act as a cathode for galvanic corrosion and dislocations offer potential sites for initiation of pits, which are reduced during the short annealing in the present study. In the controlled rolled structures, the passive potential is higher as compared to that of the H-AN sample due to the reduction in the grain size thereby increasing crystalline defects, such as, grain boundary area etc. These defects believe to decrease in net potential difference between matrix and insoluble dispersoids because of reduction of the corrosive ions per galvanic cell. Therefore, the corrosion resistance of the thermomechanically treated IF steel was found to improve and agreed with the mostly reported works demonstrating that decreasing the grain size would increase the corrosion resistance [47–51].

3.5.2. Potentiodynamic polarization The potentiodynamic polarization test curves of the H-AN, 50% and 80% ferritic rolled and 80% ferritic rolled+short annealed samples are shown in Fig. 16b. The corresponding Ecorr and icorr values were acquired from the corresponding polarization curve using Tafel extrapolation method [44]. The effect of grain size on corrosion behaviour (exposed to 3.5 wt% NaCl medium) of the IF steel can clearly be identified from corresponding polarization curve (Fig. 16b). The corrosion rate (CR) was measured using the following formula:

CR(inmilperyear) = 0.13 × icorr(μAcm −2) × [eq.wt. of steel/density of steel] (6) It is observed that the corrosion rate of the annealed sample (avg. grain size=260 µm) is considerably higher (16.25 mpy) as compared to that of the fine grained (1–3 µm) counterpart (11.67 mpy). Recently, it is reported that the ultrafine grained structure promotes the passivity easily and exhibits higher corrosion resistance [45]. In the present study, the controlled rolled specimens showed relatively more noble corrosion potential (−0.753 to −0.61 VSCE) as compare to that of the HAN sample (−0.849 VSCE). The oxidation occurs at the region where number of crystalline defects like grain boundary and dislocations are available [46]. The controlled thermomechanical treatment produced fine/ultrafine grains. This led to increase in the grain boundary area and promotes easy formation of oxide films on the surface of fine grained material [47,48]. It is also reported that the corrosion resistance of IF steel processed by ECAP and accumulative roll bonding (ARB) improved as compared to that of the as-cast IF steel [48]. It is believed that the development of high angle grain boundaries and high dislocation density upon heavy deformation enables the formation of dense oxide surface films [47,48]. In the present study, the localized corrosion rate decreased in the controlled rolled specimens, because of formation of dense oxide film on the large grain boundary area. Also, the large amount of residual stresses believed to facilitate the growth of dense oxide films [47,48]. Ferritic rolled samples resulted in a cathodic shift in the Ecorr from −0.753 to −0.792 VSCE and decreased in the icorr from 21.90 to 19.60 µA/cm2 corresponding to that of the H-AN coarse grained sample (Ecorr=-0.849 VSCE, icorr=35.20 µA/cm2). The Ecorr and icorr values of the ferritic rolled+short annealed sample, respectively, are −0 .61VSCE and 17.40 µA/cm2, which confirmed an improvement in the corrosion resistance further. The short-annealing of the rolled sample led to an anodic shift in the Ecorr from −0.792 to -0.61VSCE and reduced in Icorr from 19.60 µA/cm2 to 17.40 µA/cm2. The anodic shift in Ecorr for the rolled+short annealed sample, when compared with that of the rolled sample, is primarily due to the partial dissolution (e.g., Fe(Ti +Nb)P) of strengthening precipitates upon short annealing. Also, the fine grained structure has more number of grain boundaries, which in turn reduces the chloride concentration per grain boundary area resulting in lesser current density [47,48]. The obtained result in the present investigation is in agreement with that reported by Krishna et al. [49] and Singh et al. [47]. They reported the similar phenomena in case of corrosion behaviour of an ultrafine grained Al–4Zn–2Mg and Al 2024 alloys processed by cryorolling [49] and multiaxial cryoforging [47], respectively. The intergranular corrosion of IF steel samples is mainly influenced by the net electrochemical potential difference between the matrix and grain boundary and size and shape of the precipitate phases [50,51]. Fig. 16b shows that the ferritic rolled+short annealed sample exhibited more corrosion resistant compared to that of the other rolled samples. During high temperature annealing, some precipitates (e.g. Fe(Ti+Nb) P) might be dissolved into the matrix along with the recovery of dislocations (e.g., formation subgrains); and therefore, net electrochemical potential difference between the grain boundary and matrix is reduced. Hence, ferritic rolled+short annealed sample exhibited a relatively positive corrosion potential compared to that of the ferritic

4. Conclusions In the present study, Nb-Ti stabilized IF steel was processed by controlled rolling in 3 different phase regimes with an aim to improve its mechanical properties by grain size refinement. A good combination of YS and formability as well as better corrosion resistance have been achieved for the ferritic rolled and ferritic rolled+short annealed samples. The following findings are highlighted as the important outcomes of the study: a) Control rolling in the γ-region (1323 K) leads to grain size refinement by rapid dynamic recrystallization of the deformed austenite, which is followed by the γ/α transformation to form fine equiaxed grains (avg. size of ~33–37 µm) during cooling. On the other hand, rolling in a temperature region of ~1073 K (800 °C) (where no recrystallization takes place), the α/γ transformation starts from the deformed austenite grains producing fine ferrite grains (~5 µm) upon cooling. Finally, ultrafine ferrite grains formed via DIFT along with the microshear bands in the ferritic region rolled sample (923 K). b) The 80% rolled sample in the ferritic region (923 K) shows the highest improvement of YS (427 MPa) without a significant decrease in total elongation (27%) compared to that of the other two 80% rolled samples deformed in transition and γ regions. The improvement in the YS simultaneously retaining a high ductility is accomplished with a of bimodal grain structures obtained by controlled rolling in this region. The ultrafine ferrites (average size ~ 1–3 µm) provides the high YS, while the comparatively larger sized ferrite grains (average size ~ 25–27 µm) promote ductility. Precipitation hardening (formation of fine strain induced precipitates (NbC+TiC) also plays an important role to enhance the YS. c) The improved strength of the ferritic region controlled rolled steel is associated with a decrease in the uniform elongation and thus resulted a lower formability. A short annealing of 100 s at a higher temperature of 1123 K followed by forced air cooling was found to recover the formability significantly without much change in the YS. The formation of {111} recrystallization textures (require for high formability) could be recovered by dissolution of the Fe(Ti+Nb)P phase at high temperature followed by forced air cooling. d) The passive potential is higher for the controlled rolled structure as compared to that of the homogenized annealed (H-AN) sample due to the reduction in grain size thereby increasing grain boundary area. These defects decreased the net potential difference between matrix and insoluble dispersoids. The decrease in the grain size also decreased the corrosive ions per galvanic cell. Therefore, the corrosion resistance of the thermomechanically treated IF steel was found to improve. Finally, it can be concluded that ferritic rolling followed by a short annealing seems to be a promising production strategy for thin hot rolled strips with desirable deep drawing properties along with the improved mechanical properties. 35

Materials Science & Engineering A 684 (2017) 22–36

S. Ghosh et al.

[25] F.Boratto, R.Barbosa, S.Yue, J.J.Jonas, Effect of chemical composition on the critical temperature of microalloyed steels, in: Proceedings of the International Conference THERMEC-pp. 383–390, 1988. [26] A. Ghosh, S. Das, S. Chatterjee, B. Mishra, P.R. Rao, Influence of thermomechanical processing and different post-cooling techniques on structure and properties of an ultra low carbon Cu bearing HSLA forging, Mater. Sci. Eng. A 348 (2003) 299–308. [27] L.J. Cuddy, Grain refinement of Nb steels by control of recrystallization during hot rolling, Metall. Mater. Trans. A 15 (1984) 87–98. [28] V. Raghaban, Materials Science and Engineering, 5th ed, PHI Learning, 2004, pp. 225–400. [29] J.H. Yang, Q.Y. Liu, D.B. Sun, X.Y. Li, Recrystallization behavior of deformed austenite in high strength microalloyed pipeline steel, J. Iron Steel Res. Int. 16 (2009) 75–80. [30] M. Arribas, B. Lopez, J.M.R. Ibabe, Additional grain refinement in recrystallization controlled rolling of Ti-microalloyed steels processed by near-net-shape casting technology, Mater. Sci. Eng. A 485 (2008) 383–394. [31] J. Hu, L.X. Du, H. Xie, X.H. Gao, R.D.K. Misra, Microstructure and mechanical properties of TMCP heavy plate microalloyed steel, Mater. Sci. Eng. A 607 (2014) 122–131. [32] P. Gong, E.J. Palmiere, W.M. Rainforth, Dissolution and precipitation behaviour in steels microalloyed with niobium during thermomechanical processing, Acta Mater. 97 (2015) 392–403. [33] H. Liu, H. Sun, B. Liu, D. Li, F. Sun, X. Jin, An ultrahigh strength steel with ultrafine-grained microstructure produced through intercritical deformation and partitioning process, Mater. Des. 83 (2015) 760–767. [34] D.M. Sekban, O. Saray, S.M. Aktarer, G. Purcek, Z.Y. Ma, Microstructure, mechanical properties and formability of friction stir processed interstitial-free steel, Mater. Sci. Eng. A 642 (2015) 57–64. [35] S. Ghosh, S. Mula, Thermomechanical processing of low carbon Nb–Ti stabilized microalloyed steel: microstructure and mechanical properties, Mater. Sci. Eng. A 646 (2015) 218–233. [36] O. Saray, G. Purcek, I. Karaman, H.J. Maier, Improvement of formability of ultrafine-grained materials by post-SPD annealing, Mater. Sci. Eng. A 619 (2014) 119–128. [37] H. Zhang, L.F. Qiao, The development of annealing time for super fine grain and high strength IF steel, Adv. Mater. Res. 631–632 (2013) 613–616. [38] G.E. Dieter, Mechanical Metallurgy, 2nd ed., Mc Graw-Hill Book Co., New York, 1976, pp. 325–337. [39] R. Song, D. Ponge, D. Raabe, Improvement of the work hardening rate of ultrafine grained steels through second phase particles, Scr. Mater. 52 (2005) 1075–1080. [40] A.K. Gupta, D.R. Kumar, Formability of galvanized interstitial-free steel sheets, J. Mater. Process. Technol. 172 (2006) 225–237. [41] F. Gao, Z. Liu, H. Liu, G. Wang, Texture evolution and formability under different hot rolling conditions in ultra purified 17%Cr ferritic stainless steels, Mater. Charact. 75 (2012) 93–100. [42] E.L.S.M. Sherif, A Comparative study on the electrochemical corrosion behavior of iron and X-65 steel in 4.0 wt % sodium chloride solution after different exposure intervals, Molecules 19 (2014) 9962–9974. [43] Z. Xia, K.C. Chou, Z.S. Smialowska, Pitting corrosion of mild steel in CO2containing NaCl brine, Corrosion 45 (1989) 636–642. [44] M.G. Fontena, Corrosion EngTata McGraw-Hill Edition, 3rd ed (2005), 2005, pp. 265–301. [45] K.D. Ralston, N. Birbilis, Effect of grain size on corrosion: a review, Corrosion 66 (2010) 075005–075013. [46] R.W. Revie, H.H. Uhlig, Corrosion and Corrosion Control, 4th ed, John Wiley and Sons Inc, 2008, pp. 106–109. [47] A.K. Singh, S. Ghosh, S. Mula, Simultaneous improvement of strength, ductility and corrosion resistance of Al2024 alloy processed by cryoforging followed by ageing, Mater. Sci. Eng. A 651 (2016) 774–785. [48] B. Hadzima, M. Janecek, Y. Estrin, H.S. Kim, Microstructure and corrosion properties of ultrafine-grained interstitial free steel, Mater. Sci. Eng. A 462 (2007) 243–247. [49] K.G. Krishna, K. Sivaprasad, T.S.N.S. Narayanan, K.C.H. Kumar, Localized corrosion of an ultrafine grained Al–4Zn–2Mg alloy produced by cryorolling, Corros. Sci. 60 (2012) 82–89. [50] K. dehghani, M. Hosseini, A. Nekahi, Comparing the corrosion behavior of nanograined and coarse-grained interstitial free steels, Int. J. Mater. Res. 104 (2013) 999–1006. [51] G.P. Singh, A.P. Moon, S. Sengupta, G. Deo, S. Sangal, K. Mondal, Corrosion behavior of IF steel in various media and its comparison with mild steel, J. Mater. Eng. Perform. 24 (2015) 1961–1974.

Acknowledgment The authors highly acknowledge IIC, Indian Institute of Technology Roorkee for providing the instrumental facilities to carry out the research work. References [1] R. Mendoza, J. Huante, M. Alanis, C. Gonzalez-Rivera, J.A. Juarez-Islas, Processing of ultra-low carbon steels with mechanical properties adequate for automotive applications in the as-annealed condition, Mater. Sci. Eng. A 276 (2000) 203–209. [2] K. Tsunoyama, Metallurgy of Ultra-low-C interstitial-free sheet steel for automobile applications, Phys. Status Solidi A 167 (1998) 427–433. [3] H. Takechi, Metallurgical aspects on interstitial free sheet steel from industrial view points, ISIJ Int. 34 (1994) 1–8. [4] S. Hoile, Processing and properties of mild interstitial free steels, Mater. Sci. Technol. 16 (2000) 1079–1093. [5] R.K. Ray, P. Ghosh, An overview on precipitation and texture formation in IF and IFHS steels during processing, Mater. Manufac. Process. 25 (2010) 195–201. [6] A. Uenishi, C. Teodosiu, Solid solution softening at high strain rates in Si-and/or Mn-added interstitial free steels, Acta Mater. 51 (2003) 4437–4446. [7] P. Ghosh, B. Bhattacharya, R.K. Ray, Comparative study of precipitation behavior and texture formation in cold rolled-batch annealed and cold rolled-continuous annealed interstitial free high strength steels, Scr. Mater. 56 (2007) 657–660. [8] P. Ghosh, R.K. Ray, B. Bhattacharya, S. Bhargava, Precipitation and texture formation in two cold rolled and batch annealed interstitial-free high strength steels, Scr. Mater. 55 (2006) 271–274. [9] X. Song, Z. Yuan, J. Jia, D. Wang, P. Li, Z. Deng, Effect of phosphorus grain boundaries segregation and precipitations on mechanical properties for Ti-IF steel after recrystallization annealing, J. Mater. Sci. Technol. 26 (2010) 793–797. [10] R.K. Ray, J.J. Jonas, R.E. Hook, Cold rolling and annealing textures in low carbon and extra low carbon steels, Int. Mater. Rev. 39 (1994) 129–172. [11] J.S. Rege, M. Hua, C.I. Garcia, A.J. Deardo, The segregation behavior of phosphorus in Ti and Ti1Nb stabilized interstitial-free steels, ISIJ Int. 40 (2000) 191–199. [12] T. Yasushi, U. Toshiaki, N. Yasunobu, A new type of high strength steel for exposed panels—high-strength steel with excellent formability, superior surface precision after press forming, and uniform surface appearance, JFE GIHO No. 4 (2004) 15–21. [13] R.D.K. Misra, Temperature-time dependence of grain boundary segregation of phosphorus in interstitial-free steels, J. Mater. Sci. Lett. 21 (2002) 1275–1279. [14] X.M. Chen, S.H. Song, L.Q. Weng, S.J. Liu, K. Wang, Relation of ductile-to-brittle transition temperature to phosphorus grain boundary segregation for a Tistabilized interstitial free steel, Mater. Sci. Eng. A 528 (2011) 8299–8304. [15] R.D.K. Misra, G.C. Weatherly, J.E. Hartmann, A.J. Boucek, Ultrahigh strength hot rolled microalloyed steels: microstructural aspects of development, Mater. Sci. Technol. 17 (2001) 1119–1129. [16] A. Ghosh, B. Mishra, S. Chatterjee, Development of low carbon microalloyed ultra high strength steels, Intl. Symp. microalloying for new steel processes and applications, Mater. Sci. Forum 500–501 (2005) 551–558. [17] S. Ghosh, A.K. Singh, S. Mula, Effect of critical temperatures on microstructures and mechanical properties of Nb–Ti stabilized IF steel processed by multiaxial forging, Mater. Des. 100 (2016) 47–57. [18] A. Basuki, E. Aernoudt, Influence of rolling of TRIP steel in the intercritical region on the stability of retained austenite, J. Mater. Process. Technol. 89–90 (1999) 37–43. [19] A. Basuki, E. Aernoudt, Effect of deformation in the intercritical area on the grain refinement of retained austenite of 0.4C trip steel, Scr. Mater. 40 (1999) 1003–1008. [20] S.C. Hong, K.S. Lee, Influence of deformation induced ferrite transformation on grain refinement of dual phase steel, Mater. Sci. Eng. A 323 (2002) 148–159. [21] C.J. Barrett, B. Wilshire, The production of ferritically hot rolled interstitial-free steel on a modern hot strip mill, J. Mater. Process. Technol. 122 (2002) 56–62. [22] S. Godha, K. Watanabe, Y. Hashimoto, Effects of the intercritical rolling on structure and properties of low carbon steel, ISIJ Trans. 21 (1981) 6–15. [23] A.D. Paepe, J.C. Herman, V. Leroy, Deep drawable ultra low carbon Ti-IF steels hot rolled in the ferrite region, Steel Res. 68 (1997) 479–486. [24] C.G. Andres, F.G. Caballero, C. Capdevila, L.F. Alvarez, Application of dilatometric analysis to the study of solid–solid phase transformations in steels, Mater. Charact. 48 (2002) 101–111.

36