Managing Structure and Properties of Pipe Steel Alloyed with Chromium in the Process of Cooling After Thermomechanical Rolling

A study was conducted concerning the kinetics of phase transformations and structure of grade K60 (X70) pipe steels alloyed with small amounts of chromium (K60Cr) and molybdenum (K60Mo). It was shown that K60Mo steel differs by a higher structural content of M/A components. The laboratory tests and microstructure studies of the obtained samples by color etching, TEM, and XRD methods have shown that by using the improved cooling conditions after thermomechanical rolling of K60Cr steel, it becomes possible to obtain samples having different matrix microstructure (polygonal, quasi-polygonal, or bainitic ferrite) and various secondary phases (M/A or cementite), which yields different mechanical properties ranging from 467 to 633 MPa (yield point – σ_y ); 681 to 804 MPa (ultimate strength – σ_t ); 22 to 31.8% (percent elongation – δ_e); 8.8 to 17.4% (proportional elongation – δ_p); and 0.68 to 0.88 (σ_y /σ_t ratio). It was established that the presence of M/A component instead of cementite in the structure contributes to an increase in strength and decrease in σ_y /σ_t ratio without a considerable decline in cold resistance. At JSC “Vyksa Metallurgical Plant” (VMZ), K60Cr steel, which is more economical compared to K60Mo, and newly developed technology were used to produce commercial batches of rolled products and large diameter pipes, which demonstrated the specified σ_y /σ_t ratio and higher impact toughness and cold resistance values during a drop weight tear test (DWTT).

The formation of the modern grade К56-К65 (Х65-Х80) pipe steel structure (mainly, ferrite-bainite) is achieved by using an accelerated cooling of the rolled stock after thermomechanical (controlled) rolling. In this case, the steel composition includes elements shifting the perlitic transformation region toward the lower cooling rates [1]. A lagging austenite breakdown in the perlitic region is associated with the low diffusion rate of alloying elements in austenite, inhibition of carbon diffusion depending on the alloying element content, and reduction in γ → α transformation rate, which is determined by the interaction between the atoms of alloying element and iron atoms in solid solution [2, 3]. The most pronounced conversion inhibition effect in the perlitic region is observed during alloying with molybdenum and chromium [1]. To ensure a favorable combination of strength, plasticity, and toughness, it is important to be able to form a meaningful fraction of martensite/austenite (MA) component [4, 5]. In general, alloying with small (0.1 to 0.3%) portions of molybdenum and chromium positively affects the steel structure, since it helps expanding the “technological window” of cooling rates, which enables a uniform ferrite-bainite structure in large cross-section rolled products. In this case, the use of chromium as an alloying element is more economically feasible, since it is not considered a critical element and has lower cost compared to molybdenum.

Two commercially produced steels were chosen as a material for studying the phase transformations: molybdenum-alloyed (K60Mo) and chromium-alloyed (K60Cr1, K60Cr2) steel (Table 1).

Table 1. Chemical Composition of Studied Steels

A study of the kinetics of austenite transformation in steels during continuous cooling was conducted using an automated deformation dilatometer DIL 805 A/D. Cylindrical samples measuring 5 mm in diameter and 10 mm in length selected from commercial rolled stock were used. The dilatometric samples were heated in a furnace under the protective atmosphere to 1250°C, soaked for 30 minutes and quenched in water. Next, a thermomechanical treatment (TMT) mode was simulated by heating the samples to 1100°C (5 min), performing two rounds of deformation at 12% compression ratio per each round (preliminary deformation stage), followed by additional three rounds of 12% – deformation. The total deformation ratio was 64%. Prior to the start of cooling at the rates of 0.5 to 80°C/sec, deformed samples at 810°C were cooled for 15 seconds to 800°C to simulate the conditions of moving the metal sheet to an accelerated cooling unit.

The sample microstructure was studied using a metallographic inverted microscope AxioObserverD1m at 1000x magnification. The studied surface, prepared according to a standard procedure for producing metallographic specimens, was subjected to etching in 4% alcohol solution of nitric acid to reveal the ultimate structure. The number of structural components and size of ferrite grains were evaluated based on the microphotographs using a Thixomet image analyzer according to GOST 5639. The Vickers hardness under 1 kg load was measured using an automatic microhardness tester KV 30. The microstructure was studied according to a combined procedure developed for steels with M/A structural sites [6]. The structural studies were conducted by optical microscopy with color etching and TEM technique using JEM 200CX transmission electron microscope at an accelerating voltage of 120 kV. The fraction of residual austenite was determined by X-ray structural analysis using a Geigerflex XRD-meter with a diffracted beam monochromator and Philips fine-focus tube. The quantitative parameters of the microstructure were evaluated using the ImageExpertPro 3 software.

Experiments were conducted on a laboratory rolling mill DUO-300 using workpieces measuring 80 × 90 × 150 mm cut out of the commercial K60Cr2 steel slab (see Table 1). The heating temperature prior to rolling was 1180°C. The rolling was performed using a two-stage mode with fixed reductions in thickness to 15 mm.

The deformation was completed in the lower part of the γ -region (790 ± 10°C). The accelerated cooling parameters were varied widely: the cooling was initiated from the γ - (770 to 780°C) and (γ + α )- (705 to 715°C) regions; the cooling rate varied from 5 to 90°C/sec, and the accelerated cooling completion temperature varied from 650 to 90°C. A single-stage (continuous) and two-stage accelerated cooling (with pauses at different temperatures) was used.

The mechanical tensile tests were performed at room temperature using cylindrical and flat samples according to GOST 1497, and the impact bending tests were performed using the V-notch samples (KSV) according to GOST 9454.

The plotted thermokinetic transformation diagrams (TTD) of deformed austenite under continuous cooling for steels K60Mo and K60Cr1 are shown in Fig. 1.

Fig. 1.

CCT diagram of deformed austenite transformation in K60Cr1 steels (solid line) and K60Mo steel (dashed line).

For both steels, the appearance of the diagrams, as well as the ratio of the structural components formed as a result of cooling at 2°C/sec, are practically identical. Both steels are characterized by a low austenite resistance to ferrite transformation (the ferrite region does not close even at a cooling rate of 80°C/sec), and a wide range of cooling rates resulting in bainite formation in the steel structure. As the sample cooling rate increases, the hardness values tend to go up, which is associated with an increased structural dispersion and higher volumetric content of bainite. In general, the hardness of the K60Mo steel samples is slightly higher than that of the K60Cr1 samples (Fig. 2).

Fig. 2.

Effect of the cooling rate on the hardness of K60Cr1 (1) and K60Mo (2) steels.

The degree of supercooling has a different effect on the growth of critical point (Ar₃) for molybdenum- and chromium-containing steel. For example, at the cooling rates of up to 15°C/sec, the onset temperature of the γ → α transformation in K60Mo steel is higher compared to K60Cr1 steel. However, starting with 15°C/sec, the critical point (Ar₃) for the molybdenum-containing steel appears to be lower than for the chromium-containing steel.

An almost identical structural dispersion of both alloyed steels was observed. At a cooling rate of 1°C/sec, a ferrite grain mostly corresponding to numbers 10–11 is formed in the structure of these steels. As the cooling rate increases, the ferrite morphology in the structure of both steels changes from polygonal to quasi-polygonal and needle-like (bainitic), while the dispersion of the structure increases as well as. According to the previous studies [7], the M/A component content of the molybdenum-containing steel appears to be slightly higher compared to chromium-containing steel in the entire range of cooling rates, which is likely associated with different effects of molybdenum and chromium on carbon diffusion. As the cooling rate increases, the dispersion of M/A sites becomes higher for both steels.

The results of testing the mechanical properties of К60Cr2 steel strips rolled on the laboratory rolling mill showed that under the same rolling conditions, the variation in the post-formation cooling parameters enables a wide change in the combination of properties, such as: yield point – from 467 to 633 N/mm²; ultimate strength – from 681 to 804 N/mm²; percent elongation – from 22 to 31.8%; proportional elongation – from 8.8 to 17.4%; yield point – to – ultimate strength ratio – from 0.68 to 0.88; and cold-shortness threshold (T₅₀) – from –50°С to – 130°С.

The variation of mechanical properties within wide range is caused by the significant changes in both the microstructure of the ferrite matrix, and high-carbon secondary phases of the studied steel depending on the post-deformation cooling conditions [8]. The ferrite matrix microstructure of the experimental rolled stock samples consisted of different types of ferrite: polygonal (PF), quasi-polygonal (QPF), lath and granular bainitic ferrite (LBF, GBF). Different types of secondary high-carbon phases and structures were obtained, such as lath martensite (LM); different forms of the M/A component with various rations of residual austenite (A) and twinned martensite (TM), as well as A and TM in pure form. In a number of samples, precipitates and interlayers of cementite (C) were observed. The M/A component ratio determined by the image analysis method using color etching constituted 0.2 to 7.1%, while the residual austenite ratio determined by XRD analysis was 0 to 6%. The main types of microstructural components obtained in the experimental samples are shown in Figs. 3 and 4.

Fig. 3.

Major varieties of ferrite matrix microstructure in K60Cr2 steel (TEM): (a) polygonal ferrite; (b) quasi-polygonal ferrite; (c) bainitic ferrite, lath-like crystals (light-field view, × 15,000).

Fig. 4.

Major types of secondary structural components in K60Cr2 steel (TEM) – dark-field view: (a) cementite precipitated along the boundaries and inside the ferrite blocks, × 30,000; (b) twinned martensite, × 30,000, in martensite reflex; (c) austenite in bainitic ferrite medium, × 15,000, in austenite reflex; (d) austenite in ferrite – TM + A “isles,” × 15,000, in austenite reflex; (e) TM + A “isle”, × 15,000, in martensite reflex; (f) lath martensite (LM + A) with residual austenite between laths, × 15,000, in austenite reflex.

The study of the sample microstructure after continuous cooling from the γ -region showed that in case of accelerated cooling (AC) below the onset temperature of bainitic transformation, an increase in the cooling rate leads to a gradual variation in the ferrite matrix microstructure. At the cooling rates from 6 to 12°C/sec, the microstructure reveals the formation of rather large grains of polygonal and quasi-polygonal ferrite, while at the cooling rates of 20°C/sec and higher, a lath bainitic ferrite microstructure is formed. At the intermediate cooling rates, the matrix microstructure is mixed. The higher the cooling rate, the higher the dispersion of the ferrite microstructure.

The cooling from the (γ + α )-region is characterized by the presence of polygonal ferrite formed during cooling in air prior to the beginning of accelerated cooling (AC). In this case, the specifics of the cooling rate effect on the ferrite matrix of the structures formed during the AC process are similar to those during cooling from the γ-region. Due to a higher carbon content in austenite after the partial ferritic transformation, the matrix may contain lath martensite in addition to lath bainitic ferrite after the end of accelerated cooling from the (γ + α )-region (cooling rates ≥ 20°C/sec).

The formation of high-carbon structural components is affected by both the morphology of the ferrite matrix, within which they precipitate, and the cooling rate along with the Ac end temperature. In the matrix formed by polygonal and quasi-polygonal ferrite, the size of the high-carbon component sites, located along the boundaries and at the triple boundary intersections, is larger, and their shape is closer to equiaxial. In the lath BF- or LM-matrix, the high-carbon phase precipitates in the form of small sites or thin interlayers between the ferrite laths.

Provided the cooling rate is sufficient for inhibiting the diffusion processes in the carbon-rich austenite sites, from which the secondary phase is formed, a decrease in the end of cooling temperature with result in the following sequence of changes in the high-carbon secondary phases and structural components: lamellar perlite; degenerated perlite; high-carbon bainite or cementite; and M/A component. In this case, the fraction of residual austenite after cooling will be determined by the concentration of carbon in M/A component sites. The higher the carbon content, the higher the fraction of stable austenite in the M/A structure. The examples of microstructures obtained under continuous cooling conditions are shown in Fig. 5.

Fig. 5.

Microstructures of K60Cr steel samples obtained during continuous cooling to 200–450°C (LePera etchant): (a) cooling from γ -region (v_c = 6.5°C/sec), QPF + M/A microstructure; (b) cooling from γ -region (v_c = 19°C/sec), BF + QPF + M/A microstructure; (c) cooling from γ -region (v_c = 42°C/sec), BF + M/A microstructure; (d) cooling from (γ + α)-region (v_c = 40°C/sec), PF + LM microstructure.

The use of a two-stage accelerated cooling changes somewhat the mechanisms of structure formation, since after the first AC stage, the bainite transformation region undergoes the process of redistribution of carbon from the matrix to untransformed austenite. The matrix microstructure is relaxed and does not change with further cooling. At the second cooling stage, the high-carbon structures are formed depending on the cooling rate (T_c) and morphology of the ferrite matrix. The difference when using a two-stage cooling from the γ - and (γ + α)-regions, same as in case of continuous cooling, is that in case of cooling from the (γ + α)-region, polygonal ferrite is formed first, while the above-described processes occur during breakdown of austenite, which remains untransformed prior to the beginning of accelerated cooling. The examples of microstructures obtained during two-stage cooling are shown in Fig. 6.

Fig. 6.

Microstructures of K60Cr2 steel samples obtained during two-stage cooling (LePera etchant): (a) cooling from γ -region (v_c = 7°C/sec), QPF + M/A microstructure; (b) cooling from γ -region (v_c = 26°C/sec), BF + M/A microstructure; (c) cooling from γ -region (v_c = 7°C/sec), QPF + BF + M/A microstructure; (d) cooling from (γ + α)-region (v_c = 27°C/sec), PF + QPF + M/A microstructure.

The comparison between the types of microstructures obtained under comparable cooling conditions (see Fig. 5(a), (b), (d) and Fig. 6(a), (b), (d)) has shown that during a two-stage cooling, the MA component content in the microstructure increases compared to continuous cooling. However, by varying the process parameters, it becomes possible to obtain similar structures during both continuous and two-stage cooling.

For the structure of the samples cooled from the γ-region at sufficiently high cooling rates, it is typical to see ferrite matrix, predominantly consisting of bainitic ferrite of the lath morphology (BF). The matrix of the structure of the samples treated at the low rates of cooling from the γ -region consists of less dispersed quasipolygonal ferrite. When cooling from the (γ + α )-region, ferrite is mostly polygonal, and the fraction of bainitic or quasi-polygonal ferrite constitutes less than 30%. When using a conventional thermomechanical treatment technology, the high-carbon phase composition contains mostly cementite, while in case of the specially developed technology, structural cementite is replaced with M/A component. Depending on the cooling parameters (rate, break temperature, number of stages, etc.), the dispersion of bainitic and quasi-polygonal ferrite, as well as the dispersion and composition of high-carbon phases change, resulting in different levels of strength properties, plasticity, toughness and cold resistance.

We analyzed the effect of the type of matrix microstructure and secondary phases on the level of mechanical properties and cold resistance of sheets obtained during laboratory rolling on the DUO 300 rolling mill using both conventional and new technologies. To perform the analysis, the sheets were divided into two groups: the first group had a matrix mainly consisting of BF forming during intensive accelerated cooling, and the second group had a matrix mostly consisting from PF formed prior to the beginning of accelerated cooling, of QPF formed during low-intensity accelerated cooling. The subgroups included sheets mainly having martensite and residual austenite (M/A component) as high-carbon phases or containing cementite structures. Figures 7 and 8 show the relationship between mechanical properties and cold resistance of the experimental sheets depending on the types of structural components of the matrix and secondary phases.

Fig. 7.

Effect of the microstructure type on mechanical properties of K60Cr2 steel: (a) σ_y – σ_t relationship; (b) σ_y – δ₅ relationship; (c) δ_p – σ_y /σ_t relationship.

Fig. 8.

Effect of the microstructure type on impact toughness and cold resistance of K60Cr steel samples with different yield point: (a) σ_t – KCV⁻²⁰ relationship; (b) σ_t – cold-shortness threshold (T₅₀) relationship.

As can be seen from Figs. 7 and 8, the samples with different microstructures have different combinations of mechanical properties and cold resistance affected by both the type of the main matrix component (PF, QPF, or BF), and the type of the secondary phases (C or M/A). At the same level of ultimate strength, the BF-matrix samples have a higher yield point than the PF- and QPF-matrix samples (see Fig. 7(a)). The domination of the secondary phase containing cementite in the BF-matrix samples is characterized by even higher yield point (σ_y ) at a comparable values of ultimate strength (σ_t ) (i.e., higher σ_y /σ_t ratio), than in case of the M/A-type secondary phases. For PF- and QPF-matrix samples, the presence of the cementite-containing structural component results in a decrease of both σ_y and σ_t values at a constant σ_y /σ_t ratio. Despite of the high σ_y values, the plasticity of the BF-matrix samples is almost the same as that of the samples with (PF, QPF) + M/A structure (see Fig. 7(b)). The samples with (PF, QPF) + C structure have the highest percent elongation (> 30%), however, their strength characteristics are the lowest (see Fig. 7(a)). The analysis of the relationship between δ_p and σ_y /σ_t (see Fig. 7(c)) has shown that the highest proportional elongation at the lowest σ_y /σ_t ratio is observed in case of the lower-strength sheets having (PF, QPF) + C structure. The combination of (PF, QPF) matrix and M/A secondary phase provides a lower σ_y /σ_t ratio compared to the BF-matrix sheets at a comparable δ_p value. The most unfavorable correlation between δ_p and σ_y /σ_t ratio is observed in cased of the sheets with BF + C structure.

In terms of providing high impact toughness, the higher values were expectedly demonstrated by the BFmatrix samples, which have also provided a higher yield point (see Fig. 8(a)): the KSV values at – 20°C were in excess of 300 J/cm². The PF- and QPF-matrix samples had, on average, the lower KSV⁻²⁰ values (205 to 368 J/cm²). The effect of the type of secondary phases (C or M/A) on impact toughness has not been established. The effect of σ_y on transition temperature T₅₀ is close to linear. The samples with higher yield point have higher cold resistance. As the σ_y value decreases and BF-matrix is replaced with QPF- and PF-matrix, T₅₀ increases. No significant effect of the type of secondary phases on cold resistance was revealed, although the samples with the cementite-containing secondary phase and PF-matrix have a slightly more favorable σ_y /T₅₀ ratio compared to those with the same matrix and M/A component.

Thus, the analysis has shown that in case of the sheets produced according to the conventional technology and sheets containing M/A component in the structure, the type of matrix determines the relationship between the strength properties, toughness, and cold resistance. The matrix composed of dispersed bainitic ferrite with high dislocation density and smaller effective grain size causes a higher level of yield point, impact energy, and cold resistance. The presence of M/A component in the bainitic ferrite microstructure instead of cementite reduces the yield point due to a change of the shape of the stress-strain diagram as a result of increased density of dislocations around martensitic sites [9]. The presence of martensite in the structure also increases the ultimate strength value, while the presence of residual austenite in the structure (< 6%) may contribute to an increase in plastic properties due to TRIP-effect [10].

If a polygonal and quasi-polygonal ferrite-containing matrix is obtained instead of BF-matrix, the effective grain size increases, which, according to well-established behavior [11], results in both the lower yield point and impact energy, as well as higher T₅₀ value of brittle fracturing. On the other hand, the presence of low-dislocation density ferrite increases the plastic properties. During formation of M/A component instead of the cementite-contining secondary phases, the PF- and QPF-matrix steels also demonstrated an increase in ultimate strength and, accordingly, a decrease in σ_y /σ_t ratio.

The performed analysis showed that the use of various thermomechanical treatment options in case of K60Cr2 steel can provide a wide range of mechanical properties of the rolled products made of this steel. Depending on the type of microstructure, different combinations of strength properties, cold resistance and impact toughness can be obtained, which allows satisfying various customer requirements with the same chemical composition of steel.

By using chromium-alloyed steel and based on the results obtained at JSC “Vyksa Metallurgical Plant” (VMZ), a new product has been developed and introduced to the market in mass volumes: a thick-sheet rolled stock and electric-welded pipes with the diameter of 1420 mm with wall thickness of more than 20 mm. By using a new technology of producing microstructure with M/A component, it became possible to significantly improve the combination of mechanical properties. The average rolled stock property values were as follows: ultimate strength – 650 N/mm²; σ_y /σ_t ratio – 0.77; percent elongation – 24.3%; and impact toughness (KSV⁻²⁰) – 401 J/cm². Compared to the previously manufactured standard product, with the ultimate strength higher by 35 N/mm², the new product has the σ_y /σ_t ratio lower by 0.10; percent elongation – higher than by 1% (abs.); and impact toughness (KSV⁻²⁰) – higher by 50 J/cm². At the same time, the cold resistance during the drop weight tear test (DWTT) is guaranteed to – 40°С, inclusively, which could not be provided by the sheets produced according to the standard technology. The results of testing of several hundreds of sheet samples produced according to the new technology have shown a better strength-to-plasticity ratio (Fig. 9).

Fig. 9.

Strength-to-plasticity ratio for two commercial batches of rolled products: 1 – standard product samples; 2 – new product samples made of К60Cr steel.

Conclusions

1. The comparison of the kinetics of phase transformations and microstructure during continuous cooling of chromium- and molybdenum-alloyed pipe steels has shown that the resulting structure is almost identical in terms of dispersion and hardness, and the difference is mainly due to the presence of M/A component sites, the content of which is higher in molybdenum-containing steel.

2. The experiments carried out using the laboratory rolling mill to roll and cool the workpieces made of chromium-alloyed steel, while varying the cooling parameters within a wide range and studying the microstructure of the obtained samples, have shown that depending on the cooling modes following thermomechanical rolling, a different matrix structure is formed, including PF, QPF, BF, or the combinations thereof, with different secondary phases (M/A or C). The cooling process parameters determine the type and morphology of both the ferrite matrix and secondary high-carbon phases, which allows obtaining the optimal structure in the studied steel. The content of M/A component in the structure can reach 7.3%, and the content of residual austenite – 6%.

3. The test results showed that by varying the cooling parameters under a constant rolling conditions, it became possible to widely change the combination of properties of studied chromium-containing steel: yield point – from 467 to 633 N/mm²; ultimate strength – 681 to 804 N/mm²; percent elongation – from 22 to 31.8%; proportional elongation – from 8.8 to 17.4%; yield point – to – ultimate strength ratio – from 0.68 to 0.88; and cold-shortness threshold (T₅₀) – from –50 to –130°С.

4. The analysis of the effect of microstructure type on the properties showed that the presence of bainitic ferrite matrix in the microstructure of K60Cr steel increases the yield point, impact toughness and cold resistance, while the matrix containing polygonal and quasi-polygonal ferrite improves the plasticity and σ_y /σ_t ratio, but reduces the impact toughness and cold resistance. The presence of M/A component instead of cementite in the structure contributes to an increase in strength and decrease in the σ_y /σ_t ratio. No significant negative effect of M/A component in the concentrations up to 5–7% on the impact toughness and cold resistance of K60Cr steel was revealed.

5. At JSC “Vyksa Metallurgical Plant” (VMZ), the new technology and lightly chromium-alloyed steel were used to produce the commercial batches of rolled products and pipes with the diameter of 1420 mm and wall thickness of more than 20 mm. By using the new technology, the combination of mechanical properties was significantly improved. The average rolled stock property values were as follows: ultimate strength – 650 N/mm²; σ_y /σ_t ratio – 0.77, percent elongation – 24.3%; and impact toughness (KSV⁻²⁰) – 401 J/cm².