INTRODUCTION

The integrated casting and rolling facility (ICRF) of JSC Vyksa Steel Works makes coiled stock of 1 to 2.7 mm in thickness from various steels by combining smelting, teeming, direct rolling with rapid cooling and reeling [1–5]. The most popular thickness range of metals used in making electric-welding pipes for oil and gas industries is from 5 to 12.7 mm [6].

A metal with a single chemical formula has traditionally been endowed with various chemical properties by thermal mechanical processing (TMP) in different modes, followed by cooling needed to form different structural states in the metal [7–15]. However, the coiled stock of 5 to 12.7 mm in thickness made using the single technology (chemical composition and TMP modes) at the ICRF can exhibit significant differences in mechanical properties, the most pronounced of which is impact viscosity.

According to the previous studies, the tendency during the dynamic impact bending tests of Sharpy and Mesnager specimens of coiled stock of various thicknesses was as follows: the smaller was the ultimate rolled thickness, the lower were the temperatures at which the metal was exposed to ductile fracture. For example, the coiled stock of 5 to 9 mm in thickness appears to have a cold resistance margin in comparison with the thicker rolled stock of 10 to 12.7 mm. In that case, the grain number under GOST 5639 and the average ferrite grain size were sufficiently close in value.

Based on the foregoing, the goal of the work was to study the influence of structural factor on the cold resistance of the coiled stock of various thicknesses made at the ICRF.

MATERIALS AND METHODS

The material used in the study was the low-carbon microalloyed steel coiled stock with a thickness of 6, 8, 10, and 12 mm made at the ICRF. The steel was smelted in an electric-arc furnace, exposed to ladle refining, teemed by continuous casting to a slab of 105 mm in thickness, heated in a tunnel furnace to the preset temperature and rolled on a continuous hot strip mill 1950 in two stages: first, it was rolled in two stands of the roughing mill; then, it was rolled in six stands of the finishing mill group. After the rolling, the breakdown bars were rapidly cooled and coiled. The coiled stock of various thicknesses is made in various TMP modes with the strain finishing in the lower temperature range of the austenitic region. Whatever the thickness, the stock was characterized by similar temperature parameters of strain in both TMP stages, rapid afterstrain cooling and coiling conditions.

The chemical composition of the metal in % was as follows: ≤0.07 of C; ≤1.5 of Mn; ≤0.05 of Si; ≤0010 of P; ≤0.005 of S; ≤0.008 of N; ≤0.035 of Nb and other nonferrous impurities from the charge (Cr, Ni, Cu).

The mechanical properties of the studied metal exposed to static tension tests and dynamic impact bending tests were estimated on the specimens cut out lengthwise and edgewise relative to the rolling direction. The static tension tests were conducted under GOST 1497 on flat quintuple full-thickness specimens and helped define ultimate tensile resistance σts (N/mm2), yield strength σY (N/mm2), and percentage elongation σ5 (%). The Sharpy and Mesnager specimens were exposed to dynamic impact bending tests under GOST 9454 every 20°C at temperatures from 0 to ‒80°C. Ten specimens were tested at each of the chosen temperatures, which included the reading of impact viscosity values (KCV, KCU, J/cm2) during each test. The respective thicknesses of the impact specimens from the rolled stock of 12, 10, and 8, and 6 mm were 10, 7.5, and 5 mm.

The structural state of the metal was studied by the metallographic method and electron microscopy on polished specimens made lengthwise relative to the rolling direction. The metallographic test of the structure after the chemical etching of the polished specimens in the 4% alcoholic solution of nitrogen acid was conducted using a coded optical microscope Axio Observer D1m with a Thixomet Pro image analysis system.

The electron microscopic examination of the structure was made on unetched polished specimens by the EBSD Technique with the help of an Ultra-55 scanning electron microscope. The maps of crystallographic orientations of crystallites in the Euler space were drawn using the Oxford Instruments HKL Channel 5 system of analyzing electron backscattered diffraction [16, 17]. The orientation maps were drawn lengthwise at a depth of 1/4 rolled thickness at magnifications of 125, 500, and 1500 with a respective step of 500, 200, and 80 nm. The respective areas of the scanned regions were 928 × 696, 232 × 174, and 77 × 58 μm. The larger side of the map is parallel to the rolling direction (RD) and perpendicular to the normal line (NL) to the rolling plane. The orientation maps were automatically modified for reducing the number of nondisplayed and improperly displayed data before the quantitative estimation of the microstructural parameters.

All of the low-angle boundaries (LABs) among the kernel volumes with a misorientation below 2° were excluded from consideration due to the experimental error of the EBSD technique. Crystallite misorientation angle θ of 15° was used as the LAB-HAB differentiation criterion. Thus, the LABs on the EBSD maps were drawn in a misorientation angle range of 2° ≤ θ < 15°, whereas the high-angle boundaries were drawn at θ ≥ 15°. The quantitative evaluation of the LABs and HABs was made by defining grain boundary density ρb in μm–1 by formula (1) for each type of crystallite misorientation:

$${{\rho }_{{\text{b}}}} = \frac{{{{L}_{{\text{b}}}}}}{S},$$
(1)

where Lb is the total length of the grain boundaries, μm; S is the analysis area, μm2.

The substructure of the test steels was evaluated for comparison by drawing Kernel Average Misorientation (KAM) maps on the basis of the EBSD data. These maps characterize the spatial distribution of microstresses and strains in the examined material. The growing density of mislocations and the rising level of microstresses increase the kernel misorientation of the crystalline lattice. Kernel misorientation maps are drawn as described below. The average misorientation relative to the nearest neighboring points is defined for each point on the EBSD map, and this value is attached to the initial pixel and shown in a certain color in the chosen color scheme. Misorientation maps are designed considering angles with less defined values, which is done to avoid the influence of high-angle boundaries. In this work, the angle of 5° was set as the criterion for processing the EBSD maps. Each pixel on the map was assigned with a defined color smoothly changing from red to dark blue; the color characterized the kernel misorientation value. The pixels with a kernel misorientation of 5° were shown in red. The pixels with zero misorientation values were shown in blue.

The crystallographic texture in the examined coiled stock was estimated using the orientation distribution function (ODF) represented as the 3D distribution of Euler angles of orientation density F(φ1, Φ, φ2) by the Bunge method. The ODF was drawn using the Oxford Instruments Channel 5 software suite on the basis of the data of fixed EBSD maps. The crystallographic texture analysis was made by the conventional method for cubic crystals by 2D ODF section at angle φ2 = 45°. This section is the one, where most of the orientations typical of strain textures and recrystallization of VCCL and FCC metals are located (Fig. 1) [18‒20]. In the 2D ODF section, the horizontal legs correspond to certain crystallographic planes {hkl}, whereas the points of these sections correspond to crystallographic directions 〈uvw〉 within these planes.

Fig. 1.
figure 1

Positions of ideal orientations on the 2D sections of the orientation distribution function according to the Bunge method at angle φ2 = 45°.

Full size image

RESULTS AND DISCUSSION

As shown by the microstructural examination of the coiled stock by optical microscopy, the TMP and rapid cooling of the metal of various thicknesses (6, 8, 10, and 12 mm) was attended by the generation of a ferrite-perlitic structure with a perlite content of no more than 3% (Fig. 2). As shown by estimating the ferrite grain size under GOST 5439, all of the examined specimens had similar grain sizes corresponding to number 10–11 at nonhomogeneity numbers from 9 to 14. In the rolled stock with a thickness of 6, 8, and 10 and 12 mm, the respective nonhomogeneity numbers were 10 to 14, 9 to 13, and 8 to 13.

Fig. 2.
figure 2

Microstructure of the tested metal per 1/4 of the rolled thickness of 6 (a), 8 (b), 10 (c), and 12 mm (d).

Full size image

As shown by the static tension test results, the coiled stock with a thickness of 6 and 8 mm has better strength properties and poorer plastic properties than the coiled stock of 10 and 12 mm in thickness (Table 1). The temporary lengthwise and edgewise resistance of the specimens is higher by 10 to 15 mPa on average. The lengthwise and edgewise yield stresses is higher by 30 mPa on average. The percentage elongation is shorter by 2 to 3%.

Table 1.   Mechanical properties of the rolled product
Full size table

The resulting strength properties can point to the fact that the presence of finer ferrite grains in the coiled stock of 6 and 8 mm in thickness can possibly contribute to the grain-boundary hardening of the metal, which helps obtain higher strength levels in the metal of the smaller thickness.

For the results of the dynamic impact bending tests, see Figs. 3 and 4. It is clearly seen that the specimens from the coiled stock of various thicknesses exhibit similar types of impact viscosity relations; however, the values proper are characterized by considerable differences in the impact viscosity levels.

Fig. 3.
figure 3

Dependence of the impact strength (a) and proportion of the viscous component (b) of the Charpy specimens on the test temperature for the rolled product of 6 (1), 8 (2), 10 (3), and 12 mm (4) in thickness.

Full size image
Fig. 4.
figure 4

Dependence of the impact strength (a) and proportion of the viscous component (b) of the Mesnager specimens on the test temperature for the rolled product of 6 (1), 8 (2), 10 (3), and 12 mm (4) in thickness.

Full size image

Ductile-brittle transition temperature T50 of the edgewise Sharpy specimens from the coiled stock with a thickness of 6 and 8 mm is below ‒80°C, whereas respective T50 of the coiled stock with a thickness of 10 and 12 mm is ‒75 and ‒50°C.

Thus, as shown by the dynamic impact bending tests of the Sharpy and Mesnager specimens, the coiled stock of 6 and 8 mm in thickness exhibits the highest cold resistance margin because the temperatures (T50), at which the transition to brittle fracture occurs in this metal, are lower than in the coiled stock with a thickness of 10 and 12 mm. The strongest tendency to brittle fracture is exhibited by the coiled stock with a thickness of 12 mm, which has to do with the features of its structural state, in particular, with the presence of larger ferrite grains.

As shown by examining the microstructure represented as Kikuchi lines, the main structural component of all of the examined specimens is polygonal ferrite (Fig. 5). The coiled stock of 6 and 8 mm in thickness contains thin strips which stretch along the RD and consist from quasi-polygonal ferrite grains and acicular bainite. The coiled stock with a thickness of 10 and 12 mm contains coarse extended sections consisting mainly from bainite of various morphologies and quasi-polygonal ferrite. The maximal respective length and width of those sections in the examined vision fields were about 500 and 50 μm.

Fig. 5.
figure 5

Contrast maps of Kikuchi-paintings per 1/4 of the tested rolled product thickness of 6 (1), 8 (2), 10 (3), and 12 mm (4) in thickness.

Full size image

It is known that the mechanical properties of metal are largely defined by the amount of boundaries [21‒23]. The denser are the boundaries, the smaller is the effective grain size on which the strength, viscosity, and cold resistance of steel depend. Note that LABs inhibit the motion of dislocations and favor strength improvement but do not inhibit the propagation of brittle cracks that is inhibited only by high-angle boundaries.

The grain size of the examined steels was estimated by processing the EBSD maps; it was taken that the boundaries with a minimal misorientation of 15° were high-angle boundaries. For the bar graphs that result from the analysis and show the grain distribution by size, including average (Davg) and average weighted (Dav.wt) size, see Fig. 6.

Fig. 6.
figure 6

Bar graphs of the grain size distribution for the rolled product thickness of 6 (1), 8 (2), 10 (3), and 12 mm (4).

Full size image

For the quantitative characteristics of the rolled stock structure, see Table 2.

Table 2.   Characteristics of the rolled product structure obtained by analyzing the EBSD maps
Full size table

The examined steels have a fairly close average grain size of 6.4 to 8.2 μm. However, the bar graphs of the grain distribution by size clearly show that, with an increase in the rolled thickness, the distribution peak shifts to the region of higher values, the distribution becomes broader and more asymmetrical (Fig. 6). Taking this into consideration, it is more correct to compare average weighted grain diameters and necessary to evaluate the structure for nonhomogeneity.

With the coiled stock thickness rising from 6 to 12 mm, the average weighted grain size rises from 13.3 to 22.8 μm. The longer Dav.wt. in the coiled stock of 8 mm in thickness compared with the coiled stock of 6 mm in thickness is conditioned by the formation of larger ferrite grain, which is clearly seen in the structural comparison in Fig. 5.

In addition to the growing ferrite grain size, the coiled stock of 10 and 12 mm in thickness is characterized by the formation of coarse bainitic areas in its structure, and the length of these areas can reach several hundreds of micrometers. Since the boundaries inside the bainitic colonies are mostly low-angle, the effective bainite grain size is commensurate with the size of the bainite areas proper and reaches about 170 μm in the coiled stock of 10 mm in thickness and about 260 μm in the coiled stock of 12 cm in thickness (Fig. 6).

The structural nonhomogeneity of the metal can be evaluated using parameter Knh defined as the relation of the maximal effective grain size to the size of the grains covering the largest area on the polished specimen. The coiled stock of 10 and 12 mm in thickness with Knh of 13.5 to 20.6 contains large bainite areas, which is why this stock is very different in Knh from the thinner metal with Knh = 5.0–6.2 (Table 2).

For the frequency distributions of crystallite misorientation angles and the results of estimating the grain boundary densities, see Fig. 7. It is easy to see that, with an increase in the rolled thickness, the fraction and density of HABs decrease through the growth of the ferrite grains and generation of coarse bainitic areas in the structure. On the contrary, the amount of LABs is much higher in the specimens with a thickness of 10 and 12 mm than in the stock of 6 and 8 mm in thickness.

Fig. 7.
figure 7

Frequency distributions of the crystallite misorientation angles (a) and the density of grain boundaries (b) in the tested rolled product of 6 (1), 8 (2), 10 (3), and 12 mm (4) in thickness; () the high + low-angle boundaries are shown in dark blue; () the high-angle boundaries are in red; () the low-angle boundaries are in blue.

Full size image

For the orientation maps of the microstructure of the examined stock in the color scheme of a standard stereographic triangle for the RD, see Fig. 8. It is seen that open texture is characteristic only of the coiled stock of 6 mm in thickness: in the orientation map the dominant colors are green and purple corresponding to orientations 〈110〉 || RD and (〈113〉…〈112〉) || RD.

Fig. 8.
figure 8

Orientation microstructural maps of the studied rolled products in the color scheme of a standard stereographic triangle for the rolling direction.

Full size image

It is hard to distinguish dominant orientations on the orientation maps of the other specimens, which points to the generation of feeble crystallographic structure in the coiled stock of 8, 10, and 12 mm in thickness.

CONCLUSIONS

The picture revealed by the tests conducted as part of this study is considered below.

The rolled stock of 6 and 8 mm in thickness has better strength properties than the coiled stock of 10 and 12 mm in thickness (by 10 MPa on average for temporary resistance and by 30 MPa for yield strength).

As shown by the dynamic impact bending tests of Sharpy and Mesnager specimens, the thinner is the stock, the flatter is the impact viscosity—test temperature dependence curve. The rolled stock of 6 and 8 mm in thickness exhibits a better cold resistance characterized by the ductile-brittle transition temperature (T50) (T50 is below ‒80°C). The rolled stock of 12 mm in thickness exhibits the lowest cold resistance (T50 = ‒50°C).

The average ferrite grain size estimated by the grain number under GOST 5639 is similar in the rolled stock of various thicknesses and corresponds to numbers 10–11. In this case, there is a spread in grain sizes. The thinner rolled stock contains smaller grains corresponding to number 14, whereas the thicker rolled stock contains bigger grains corresponding to number 8.

As shown by estimating the grain size by processing the EBSD maps, the rolled stock of 10 and 12 mm in thickness has a lower fraction and density of HABs than the rolled stock of 6 and 8 mm in thickness because the former rolled stock contains coarse bainite sections with mainly LABs.

The orientation maps of the microstructure of the tested rolled stock have shown that the pronounced strain texture is characteristic only of the rolled stock of 6 mm in thickness. The rolled stock of 8, 10, and 12 mm in thickness has a feeble crystallographic texture.

The high brittle fracture resistance of the rolled stock of 6 mm in thickness at subzero temperatures is defined by the high cumulative plastic strain in its austenite at rolling and has to do with the presence of finer ferrite grain separated by HABs as well as by the formed strain texture corresponding to orientation 〈110〉 || RD and (〈113〉…〈112〉) || RD.