Introduction

As today, the use of the duplex stainless steel can represent an efficient alternative to the use of austenitic grade, because of the lower cost due to the lower nickel content and the higher resistance to the stress corrosion cracking. On the other hand, such steel grade often shows an anomalous and poor formability associated to the hot rolling process. Duplex SS products often show an undesiderable necking during the shaping and the plastic drawing. This aspect represent a limiting condition in the manufacture especially when bending and deep drawing are involved. A recent study [1] about the evolution of the microstructure with the manufacturing processes has pointed out that the best formability properties can be obtained after a cold rolling operation, however such procedure implies supplementary costs. Thus, an alternative approach should focus on maximizing the formability properties through a correct control of the hot forming operation. In this study the attention has been focused on the microstructure and the crystallographic modification and evolution featuring duplex stainless steel as a function of the applied rolling parameters, i.e. different asymmetric ratio, different pre-heating temperature and solubilization quenching treatment.

About the asymmetric rolling, a lot of studies (dealing with aluminium cold rolling) show that the application of such technique leads to the formation of a favourable texture for implementing the subsequent plastic deformation processes [2, 3]. However, under certain aspects the absence of symmetry is detrimental and undesired: i.e. strip curvature, geometrical inaccuracies and misalignments, possible decrease of the productivity due to the difficulty of implementation of the reversible working of the slab, etc.. For these reasons, the asymmetric rolling, for many years have been considered as unsuitable. Nonetheless, asymmetric rolling can provide also some advantages: decrease the average rolling pressure, decrease the rolling force and the related torque, improved surface properties of products. Beside, asymmetric rolling involves interesting and favourable consequences on the microstructural features associated to the increase of shear stress strain penetration at the core layers of the machined strip. Several authors have investigated the asymmetric rolling for aluminium alloys [4] and the cold asymmetric rolling of low carbon steel [5, 6], but not information are present about applications of this technique for the rolling of duplex stainless steel grades. Concerning the effects of the reduction levels, several previous works involve the study of both the symmetric and the asymmetric rolling. The former process yields elongation and Lankford ratio properties increasing proportionally with the applied overall reduction. As for the latter technique, the same trend has been produced for lower reduction levels, pointing out a more rapid and effective mechanical improvement with the reduction level.

The aim of this work is to show the benefits induced by asymmetric rolling in the rolled strip and to compare them to those conferred by the conventional symmetrical rolling technique.

Experimental procedure

The asymmetric rolling has been realized imposing different peripheral speeds on the two working rollers. All experiments were carried out using a reversible laboratory mill properly equipped with an individual engine for each rolling cylinder, that allows to control singularly the rotational speed of each cylinder, specifically the upper cylinder was set at a higher peripheral speed. In addition, the cylinders surface was provided properly grinded to work at high temperature without the use of lubricant.

The studied DSS belongs to the grade EN 1.4462 and the experimental parameters include three different pre-heating temperatures and three asymmetric ratios, imposing a single reduction level by several rolling passes. The overall rolling procedure involves no lubrication and a 180° rotation around the rolling direction of the rolled strip after each pass. Moreover, the study involves also the analysis of the influence of solubilization quenching and the SEM, SEM-EBSD investigation dedicated to establish the microstructure modifications. The study attempts to define what are the best operative conditions for the application of symmetric rolling and whether the asymmetric rolling can represent a favourable technological route.

Each of the rolling cylinder, with a diameter of 135 mm, is driven by electrical engine with nominal power of 1.5 kW and is cooled with a water re-circulating system.

The investigation has been performed on coupons of EN 1.4462 duplex stainless steel (DSS 2205) obtained by continuous casting slabs featured by 40 × 150 × 150 mm geometrical size and which chemical composition reported in Table 1.

Table 1 Chemical composition of DSS 1.4462
Full size table

The DSS has been studied in different conditions:

  • hot rolled condition (HR): after heating in a resistance furnace for 2 h, the supplied slabs have been rolled through the laboratory mill. A total reduction level of about 70%, reached in 7 passes, was realized without introduction of any lubricant while the total time undertaken by the rolling procedure from start to end is about 133 s. In literature is already known that the increase of the applied reduction up to 70–90% can play a favourable role on the normal anisotropy, the hardening coefficient and the total elongation. Oppositely, there is no clear information about the temperature effects [7, 8]. The rolling procedure has been performed starting from three temperature levels, respectively 1,200°C, 1,125°C and 1,050°C. The organization of laboratory furnace allows the direct transfer from the furnace door to the rolling strand and the optical digital pyrometers have pointed out a not significant average decrease of 2–3°C. The final rolling temperature has been measured by the same pyrometers system and the cooling rate of the rolled sheet is about 38.8°/min. After rolling, the steel has been cooled down in air to room temperature.

  • asymmetric rolling condition (AR): three different asymmetric ratios were applied imposing different peripheral speeds for each rolling cylinders. The imposed ratios corresponding to 1 (representing the symmetrical condition), 1.2 and 1.4.

  • solubilization annealing treatment (SA): the hot rolled samples have been soaked in furnace for 1 h at 1,050°C and water quenched.

Optical metallography (OM) procedure has been performed to determine morphology and distribution of both structural constituents and to highlight the presence of possible detrimental inter-metallic phases. The chemical etching has been realized through the Beraha reagent (100 ml water, 10 ml HCl and 1 g K2S2O5), by which ferrite appear coloured while austenite remains unaffected. A quantitative analysis of the structural constituents has been performed by means of image analysis software employing the statistic basis provided by ASTM 1245.

The crystallographic textural investigation has been performed by Electron Scan Microscopy (SEM) equipped with Electron Back-Scattered Diffraction (EBSD) probe. Such examination allows to evaluate the distribution of the crystallographic orientation of the phases featured by the analysed samples (austenite and ferrite). The preparation of the coupon surfaces for EBSD analysis includes electrolytic etching with a solution of oxalic acid (10 g in 100 ml water at 6 V for 25 s) and a final polishing for 3 min by a solution of 0.05 μm colloidal silica suspension. The EBSD setup provide an accelerating voltage of 20 keV with a 200x of magnification and pixel size of 25 μm2. Analysis is performed on a section perpendicular to transverse direction (TD) sited in normal-rolling direction plane (ND-RD plane). The area investigated is the core of the rolled sheet and it includes a layers of 10 mm (5 mm up and 5 mm down the middle line of sheet) (Fig. 1). The EBSD analysis has been performed on all the sample completely reduced (because they show the best mechanical properties) and on a sample rolled only 3 passes at 1,200°C in order to evaluate the development and evolution of crystallographic textures.

Fig. 1
figure 1

Frame of reference for the rolled samples

Full size image

In order to determine the mechanical properties of rolled sheet, tensile tests was carried out on completely reduce samples according to the reference indications listed in the ASTM E8 (tensile test), ASTM E646 (determination of hardening index), ASTM E517 (plastic strain ratio coefficient). Specifically the latter norm states:

$$ {\text{r}} = \frac{{{\varepsilon_{\text{w}}}}}{{{\varepsilon_{\text{t}}}}} = \frac{{\ln \left( {{{\text{w}}_0}/{{\text{w}}_{\text{f}}}} \right)}}{{\ln \left( {{{\text{l}}_{\text{f}}}{{\text{w}}_{\text{f}}}/{{\text{l}}_0}{{\text{w}}_0}} \right)}} $$
(1)
$$ {{\text{r}}_{\text{m}}} = \frac{{{{\text{r}}_0} + 2{{\text{r}}_{45}} + {{\text{r}}_{90}}}}{4} $$
(2)

where:

εw :

is the plastic strain measured along the width of the tested specimens

εt :

is the plastic strain measured along the thickness of the tested specimens

w0 :

is the initial width of tensile specimen

wf :

is the final width of tensile specimen

l0 :

is the initial length of tensile specimen

lf :

is the final length of tensile specimen

r:

is the normal anisotropy coefficient (Lankford coefficient)

rm :

is the average anisotropy coefficient (Average Lankford coefficient)

rx :

represent the value of coefficient r taking from x degrees from rolling direction (RD).

All the measured mechanical properties (yield strength, ultimate tensile stress, elongation and Lankford coefficients) underwent the averaging treatment described by Eq. (2).

Results

The microstructural investigation involved in the experimental work consists in the observation of the DSS 2205 alloy under several conditions, namely before and after rolling accounting for the different applied temperatures. The initial microstructure of the DSS 2205 investigated in this work shows a typical rough-hew rolled and solubilized structure (Fig. 2) featuring an average percentage volume fraction of austenite of about 48%.

Fig. 2
figure 2

Microstructure examples for the billet as-received, magnified at 100x (left) and at 200x (right)

Full size image

Following the application of different asymmetric ratios, the work covers all the microstructures resulting from the different applied temperature levels (1,200°C, 1,125°C and 1,050°C), as reported in Figs. 3 and 4.

Fig. 3
figure 3

Micrographies at 100x of hot rolled samples

Full size image
Fig. 4
figure 4

Micrographies at 100x of hot rolled and solubilization annealed samples

Full size image

The observed samples features a morphology sharply pointed out on the RD-ND plane, as expected in material undergone a rolling process. Then, the microstructure is once again modified by the annealing and quenching treatment, which act towards the formation of a similar microstructure for all the samples featured by alternate ferrite-austenite layers. Watching more closely, each phase layer presents the, so-called, “bamboo structure” [7]. In other words, the ferrite grain boundaries are transversal to the layer structure and bridge the gap between the abreast layers, forming a string of grains running along the phase (Fig. 5).

Fig. 5
figure 5

Example of “Bamboo” structure, revealed in samples rolled and quenched

Full size image

From the microstructure images retrieved by optical microscopy the overall volume fraction of austenite island and their shape ratio have been measured by means of image analysis software applying the ASTM E112 and ASTM E1181 standards. The following results are reported in the graph of Figs. 6 and 7.

Fig. 6
figure 6

Average austenite volume fraction

Full size image
Fig. 7
figure 7

Average austenite grain aspect ratio

Full size image

The austenite percentage for the sample investigated appears to decrease with both temperature and asymmetric ratio. Moreover, observing Fig. 8 it clearly appears that the shear deformation related to the asymmetric rolling influences significantly also the ferrite-δ. The main resulting features include, grain size refining and texture modifications. Likely, the grain refining derives from the elevated deformation energy provided by the asymmetric rolling, which is able to promote a fast recrystallization despite the high stacking fault energy of ferrite.

Fig. 8
figure 8

Penetration of shear stress along the thickness as a function of asymmetric ratio

Full size image

As for the austenite percentage, temperature and AR induce a decrease in the aspect ratio of the γ-grains. However, the decrease in aspect ratio seems to be more associated to the passage from symmetric to asymmetric rolling rather than from the AR intensity, as from 1.2 to 1.4 values the reduction is moderated, at best.

Parallel to the optical investigation, the study proceeds with the application of the SEM-EBSD technique, gathering the ODF maps and inverse polar figures related to austenite and ferrite, Figs. 9 and 10 respectively, for all investigated pre-heating temperatures and asymmetric ratios.

Fig. 9
figure 9

Ferrite EBSD images (ODF maps and inverse polar picture) associated to the different asymmetric ratios and pre-heating temperature

Full size image
Fig. 10
figure 10

Austenite EBSD images (ODF maps and inverse polar picture) associated to the different asymmetric ratios and pre-heating temperature

Full size image

Concerning the symmetric rolled samples, the ferrite ODF maps yield for all pre-heating temperatures a similar texture evolution related to the Euler’s angles. The ϕ2 = 0° and ϕ2 = 90° section clearly indicate a weak Cube {011}<001> and a strong Goss {001}<001> component leaning toward the Brass component {011}<211> but developing instead along the α−fibre. At ϕ2 = 20° and ϕ2 = 70° the {130}<001> texture is evident, while at ϕ2 = 45° the ODFs feature a strong Goss component {001}<001> with some traces of γ-fibre strengthening with the temperature. The previous findings are further confirmed by the inverse polar figures featuring the rolling direction (RD) as reference. The inverse polar diagrams convey a <001> direction strongly aligned with RD and a progressive alignment of lattice direction <311>, <211> and <111> along RD (Fig. 9) induced by the increase in thermal levels.

Whereas observing the austenite, still in the case of AR = 1, the ODFs clearly indicate a combination of Cube component {011}<001> and of β-fibre holding the largest share. The orientation distribution remains fairly constant for the different pre-heating temperatures and appears to involve a large number of textural features. Compared with the ferrite ODFs, the more diffused intensity over several textures suggests the occurrence a dynamic recrystallization process involving the austenite and preventing the development of a rolling texture, as further supported by the inverse polar diagrams.

The texture analysis of the asymmetric rolled samples for all the pre-heating temperatures is reported by the ODFs represented in ϕ2 section for the Bunge space. The ferrite phase displays a clear texture evolution increasing the temperature: at 1,050°C the component {4 1 8}<14 4 0> appears rising, while stepping up to 1,125°C is the component {4 1 8}<5 12 4> to prevail, at last once reached the 1,200°C the texture features the development of the Cube {1 0 0}<0 1 0>. Moreover, from the EBSD investigation arises also an intensification with temperature and AR of the Rotated Cube texture {1 0 0}<0 1 1>, usually associated with the recrystallization of the Cube component. A similar trend is reported for the γ-fibre and Goss component, as these texture seems to be boosted, even at the core of the strip, by the asymmetric ratio. To conclude, in ODF representation the ferrite exhibits strong texture intensities. The high stacking fault energy of this phase (typical of bcc elementary cells) promotes the dynamic recovery allowing the retention of the rolling textures [9–12] beside the ones developed by the recrystallization. On the other hand, the austenite exhibits a retention of the Brass component for all thermal levels except 1,200°C, at which the phase promotes also the presence of a Cube component. Furthermore, the presence of Goss and Dillamore components (Fig. 10) clearly suggests a dynamic recrystallization process [9].

Following the rolling process the DSS coupons have been annealed and the EBSD results for the heat-treated samples are reported in Figs. 11 and 12.

Fig. 11
figure 11

Heat-treated ferrite EBSD images (ODF maps and inverse polar picture) associated to the different asymmetric ratios and pre-heating temperature

Full size image
Fig. 12
figure 12

Heat-treated austenite EBSD images (ODF maps and inverse polar picture) associated to the different asymmetric ratios and pre-heating temperature

Full size image

For the asymmetric ratio equal to 1 (symmetric condition), the quantitative image analysis proves that after annealing and quenching the rolled samples starting from the same pre-heating temperature exhibit larger and more rounded austenitic islands. Concerning the results retrieved by means of SEM-EBSD, the ferrite presents the same textures observed for the as-rolled samples (Goss and α-fibre) associated to a strengthening of the γ-fibre, as evident in the ODF maps and indicated by the increase of lattice directions parallel to RD in the inverse polar figures. For the austenite, both ODFs and inverse polar figures indicate an strong retention of the β-fibre and Cube textures. As reported in the figure below, the textures are fairly unaffected by the thermal treatments, as the only significant modification consists in the appearance of Dillamore {11 11 8}<4 4 11> (a first order twin of the Goss orientation). Likely, such retention mechanism is related to the suppressed growth selection [13–17] and to the dynamic recrystallization experienced by the austenitic phase.

The same survey routine has been followed for both AR equal to 1.2 and 1.4. As reported in literature, the annealing textures originate from the preexistent rolling texture. In ferrite steel, but also in the case of austenite-ferrite ones, the rolling textures {0 0 1}<1 0 0> and {1 1 2}<1 0 0> undergo the recovery phenomenon. While the grains featuring {1 1 1}<1 1 0> or {1 1 1}<1 1 2> components and a small share of those presenting the {0 0 1}<1 0 0> orientation tend to recrystallization thanks to the high energy stored during the rolling process, triggering a favorable condition for the nucleation and growth of new grains [12, 18–21].

Recovery tends to conserve the deformation texture, instead recrystallization annihilates the previous textures favoring the rising up of strong {1 1 1}<1 1 2> texture [17, 22]. In this study, the ferrite phase does not show preferential orientations along the α-fiber and thus these components are absent in the samples treated by solution annealing and quenching. On the other hand, a strong {1 1 1}<1 1 2> component appears in association with a γ-fiber weakening. In austenite, Goss and Dillamore components are probably produced by austenite phase twinning.

Finally, together with optical microscopy and EBSD analysis the experimental includes also a mechanical characterization of the alloy, correlating the mechanical properties with the asymmetric ratio and temperature (Fig. 13).

Fig. 13
figure 13

Comparison of mechanical properties vs. pre-heating temperature and asymmetric ratio. Please note that, YS: yield stress (MPa), UTS: ultimate tensile stress (MPa), E%: elongation (%), n: hardening coefficient and rm: av. Lankford coefficient. While the denomination SA stands for “annealed”

Full size image

In the symmetric condition, as a direct consequence of the described microstructural modifications the ultimate tensile and yield strength ratio increase after the rolling process, as confirmed the hardening index n rising. Parallel to this, the increase in rolling temperature clearly affects the plastic strain ratio rm, as decreasing the processing temperature leads the values of rm from about 0.54 (at 1,200°C) to 0.71 (1,050°C). Nevertheless, the formability attitude remains poor due to the high tendency of the rolled strips to thinning. On the other hand, the tensile strength properties do not seem to be significantly affected by the applied rolling temperature, as the values of yield, tensile strength and percentage elongation at fracture are fairly similar after the heat treatment cycle. Concerning instead the effects of the heat treatment, from the experimental tests it has been observed a decrease in the tensile properties, if compared to the as-rolled samples, combined with an favourable enhancement of ductility and Lankford coefficient. On the other hand, while the operative temperature does not appear as a relevant parameter affecting the mechanical properties, the asymmetric ratio (AR) produces noteworthy variations both in the microstructure and in the tensile behaviour. In fact, if compared to the symmetric technique, the speed mismatch of the working rollers alters the morphology so significantly that the alteration are even detectable by the conventional optical microscopy, as the austenite islands are strongly deflected by the non-symmetric plastic flow. The penetration depth of such shear deformation appears to increase as a function of temperature and asymmetric ratio (Fig. 13).

Discussion

Concerning both symmetric (AR = 1) and asymmetric (AR = 1.2 and 1.4) condition, the test-results clearly indicates, in the thermal range investigated, a recovery of the hardened ferrite, from which, thanks to the energy stored in the bulk, austenite may be produced (recrystallization). The occurrence of dynamic recrystallization phenomena has been pointed out by the measure of torque applied by the rolling equipment, showing a decrease of the yield strength (accounted by a decrease in applied torque) usually between the 5th and the 6th pass in the laboratory mill. Increasing the pre-heating temperatures to 1,200°C the softening process appears to occur also between the 6th and the 7th pass, proving a ferrite recovery process favoured by higher thermal levels and associated to a recrystallization interesting both ferrite and austenite. Therefore, at 1,200°C the dynamic recovery and recrystallization clearly develop twice during the applied sequence of reductions. The cause of this phenomenon lies in the different temperature dependence of the stress components for ferrite and austenite. In particular, being the flow stress in austenite virtually temperature independent, the increase in ferrite content, whose stress components are instead temperature dependent, determines a significant material softening [8]. Interestingly, the applied torque values also exhibit a more intense softening associated to the asymmetric rolling, evidence of a larger recrystallization process interesting the overall thickness. Moreover, based upon these observations, it is consistent to assume that the stronger softening effect could entail, beside a dynamic process, also a meta-dynamic recrystallization referred to the strip inter-strand section.

A further evidence of the above-mentioned phenomena consists in the inverse polar figures, which indicate an evident evolution of the orientations due to microstructural transformation involving mostly ferrite rather than austenite [7, 9, 15–17]. Specifically the orientation patterns shown by the two different phases indicate a recrystallization process for austenite and for ferrite (to a lesser extent and only at high temperatures) and a recovery phenomenon involving exclusively ferrite. Such behaviour is consistent with the elevated ferrite stacking fault energy, which requires quite higher temperatures than austenite to activate successfully the recrystallization. Nonetheless, once reached the activation threshold, the ferrite has stored enough energy to induce a noteworthy fast nucleation of new and fine grains.

These assumptions are further confirmed by means of optical microscopy, in fact the observed structure morphology features little austenite islands sited on the elongated and strained ferrite boundaries. However, the same morphology might suggest as well the occurrence of a thermal effect parallel to recrystallization process of the Fe-γ. In other words, the increase in the specimen temperature due to the enthalpy development associated to the deformation appears to be directly responsible of the phases amount, as suggested by the phase diagram. Thus, such observation clearly explains the reported decrease in austenite volume fraction as a function of the pre-heating temperature.

The texture analysis after the solubilization annealing treatment shows a limited ferrite volume fraction interested by recrystallization, even though higher than in the as-rolled samples. The largest part of ferrite undergoes a dynamic recovery that maintains the same texture components developed during the hot rolling process [8]. Whereas, the recrystallization is clearly pointed out by inverse polar figures which show a larger diffusion of crystallographic direction aligned along RD. In brief, while the dynamic recovery conserves the rolling orientations, the recrystallization continuously annihilates the deformation directions hindering the retention of a rolling texture. As expected, the recrystallization seems to manifest more at high thermal levels, as the temperature supplies the energy necessary to induce this morphological evolution. From the ODF maps the products of the recrystallization and recovery are easily recognized. The former process is mainly identified by the γ-fibre, which appears more intensified after the annealing. Likely, such texture intensification is related to the recrystallization of formerly α-fibre components [7, 9, 20]. While the latter is usually associated with the retention of the common rolling textures (Goss, α-fibre and Cube).

The softening observed for the tensile properties following the annealing treatment is certainly related to the recrystallization and recovery of the constituent phases. Specifically, the improvement of the formability properties (rm, n,E%) seems to be always associated to a larger presence of the sharp Goss texture and α-fibre, evidence of a stronger recovery at the expenses of the recrystallization process. The transformation textures of austenite obtained during the solution annealing of this phase are markedly concentrated along the β-fibre and the Cube component. The activation of such crystallographic textures appears to be responsible for the reported increase in the normal anisotropy coefficients. Moreover, for both symmetric (AR = 1) and asymmetric (AR = 1.2 and 1.4) rolling a decrease in pre-heating temperature corresponds to an intensification of those crystallographic textures associated with good anisotropy and resistance against thinning.

Even though the previous findings proved a certain number of similarities among the two techniques, it is crucial to understand that the symmetric and asymmetric rolling differ greatly concerning the texture activation efficiency. According to the experimental the asymmetric route appears to produce a more intense textures development, in particular the Rotated Cube, γ-fibre and Goss component seem to be boosted by the AR, even at the core of the strip. In other words, the asymmetric rolling provides more intense and well-defined crystallographic fibres and is able to achieve a deeper penetration of the shear strain into the rolling strips with respect to the conventional rolling. The second feature comes to be especially crucial for the mechanical properties of the final rolling products, since texture analysis performed by means of SEM-EBSD pointed out a clear correlation between the elongation and Lankford ratios properties and the homogenous distribution of marked and well-defined textures from skin to core of the rolled strips induced by the shear stress on the austenite islands. Interestingly, for asymmetric rolling this specific deformation pattern is even detectable from the macroscopic morphology of the austenite, which is significantly deflected by the applied shear stress. Moreover, it is worth to mention the favourable elongation and Lankford ratios of asymmetrical rolled samples are maintained also after the solution annealing and quenching. Likely, thermal treatment stimulates the suppression of the γ-fibre and the intensification of the Goss and other components typical of the recrystallization process. Thus, solution annealing and quenching yield a good anisotropy increase the thinning resistance, as indicated by the elevated rm values reported by the study and supported by literature [21, 22]. Starting from these arguments, it might be consistent to assume that thanks to the favourable crystallographic textures induced by the deep penetration of the shear strains, the asymmetric route provides mechanical properties that can be matched by the symmetric rolling only at way higher reduction levels.

Conclusions

The performed experiments and rolling trials clearly indicate that the behaviour of the duplex stainless steel and their mechanical properties is strongly influenced by the rolling temperature. Thus, it possible to draw the following conclusions:

  • the solution annealing treatment does not significantly affect the ferrite behaviour, but for an expected softening;

  • the ferrite mainly undergoes a dynamic recovery, as proved by the strong retention of rolling textures, while the austenite seems to be interested by a recrystallization especially after the solution annealing;

  • the improvement of the final formability properties seems to be related to the transformation textures of the austenite, i.e. β-fibre and Cube component, and to the persistence of the strong Goss component in the ferrite. The texture morphology appears to be especially clear after the solution annealing.

Concerning the asymmetric ratio the following conclusions can be listed:

  • the asymmetric rolling appears as a quite interesting route to improve texture and anisotropy microstructure of duplex stainless steel;

  • this technological route induces shear strains even at the core of the rolled strips and the deformation depth increases with the reduction ration and temperature;

  • the annealed and quenched samples show the best mechanical properties and the decrease in temperature seems to enhance the elongation and Lankford properties of the steel, as it hinders the recrystallization of ferrite;

  • the intensification of favourable crystallographic textures and the penetration of shear strain at the core of the strip increase with the speed mismatch of the mill rollers;

  • the asymmetric route induces dynamic recrystallization in the austenite phase and even in a small portion of ferrite, as the shear strains induced at the core imply higher deformation energy development than the conventional rolling technique;

  • the highest asymmetric ratio (AR = 1.4) combined with the lowest operating temperature (starting rolling temperature of 1,050°C) represent the best operative conditions to be applied, as to these conditions correspond the highest Lankford coefficient, overall elongation, hardening coefficient and the lowest yield strength.