Characterization of Precipitated Phases and Carbides’ Coarsening in DH36 Shipbuilding Steel during Tempering Process

Abstract

The effect of tempering temperature on precipitates in DH36 steel was studied by using OM, SEM and TEM analysis. The results showed that DH36 steel after tempering was composed of ferrite, pearlite and precipitated phases. The grain sizes of ferrite have not increased obviously with increasing of tempering temperature. There were two types of carbides: (Nb,Ti)C and cementite, the sizes of which were more than 80 nm and less than 80 nm, respectively. The precipitation temperature of them were 1370 K and 977 K after calculation. As the tempering temperature increased, the average sizes of precipitated phases increased, while volume fractions and precipitation strengthening decreased. Cementite coarsened much faster than (Ti,Nb)C, which made strength and toughness of steel decrease sharply in 973 K. Meanwhile, moderate amount of precipitates made the toughness reach summit at 873 K. To sum up, the steel after 10800 s at 773 K tempering has the best mechanical properties and finest dispersed precipitates.

1. Introduction

An increasing number of offshore platforms and shipbuilding plates come into being with the rapid development of marine industries. Damage of offshore structures gets more serious consequence, such as huge economic losses and the pollution of oceans.¹⁾

DH36 is the higher-quality hull plate. It has high strength, good toughness and excellent welding performance. It has been widely used in construction of large-scale marine and ocean engineering structures.²⁾ However, the application of DH36 is often used in thick size, which leads to the low strength and bad toughness. The requirements of DH36 formulated by classification society were as follows, yield strength is more than 355 MPa, 253 K longitudinal impact energy is more than 34 J, transverse impact energy is more than 24 J.³⁾ Recently, in order to improve the mechanical properties of shipbuilding plate, controlled rolling and cooling combined with microalloying technology is deployed to refine the ferrite grains and control precipitation of second phases.³⁾

The precipitates in low carbon microalloyed steel have been reported a lot. J. Dong⁴⁾ reported that the pinning behavior of MX-type precipitates on austenite in a kind of high strength low carbon microalloyed steel. It was found that Nb-rich precipitates has stronger pinning effect than Ti-rich precipitates. M.S. Mohebbi⁵⁾ reported Nb-rich precipitates nucleate preferentially than Ti-rich precipitates along ferrite grain boundaries during dynamic recrystallization, the average size of which is 12.5 nm. A. Karmakara⁶⁾ studied precipitation behavior of ultra-low carbon Nb-containing steel. The results showed that Nb(C,N) has dispersed distribution and enhances the ferrite structures after deformation. In addition, C.S. Zheng⁷⁾ reported that nano-cementite enhance the ferrite grains and increase the yield strength of eutectoid steel after warm deformation and annealing.

DH36 steel, as a kind of low carbon microalloyed steel, was heat-treated in this article, in order to relieve its residual stress, improve its strengths and toughness in low temperature. Through being tempered at different temperatures, the anisotropy of DH36 steel after rolling was tried to be eliminated, and plasticity could be improved. Meanwhile, characterizations of precipitated phases in DH36 after tempering were investigated, and the calculation of precipitation strengthening was carried out in each steel. The thermodynamic behavior and coarsening kinetics of (Nb,Ti)C and cementite were also studied. Finally, the optimum mechanical properties were figured out.

2. Experimental Procedure

This experiment deployed industry-scaled rolling plate as experiment materials. The process of smelting DH36 shipbuilding plate was: smelting in converter, refining through LF and RH, casting in curved spray continuous casting machine. Then controlled rolling and controlled cooling were used. The finishing rolling temperature was 1333 K, and the final cooling temperature was 983 K. Finally, as-rolled slab of 15 mm thick was available. The chemical compositions of DH36 steel were shown in Table 1.

Table 1 Chemical compositions of DH36 shipbuilding steel (mass%).

The A_c1 and A_c3 can be calculated as following Formulas (1) and (2),⁸⁾   

\begin{align} A_{\text{c}_{1}} &= 723 - 10.7w_{\text{Mn}} - 16.9w_{\text{Ni}} + 29.1w_{\text{Si}} \\ &\quad + 16.9w_{\text{Cr}} + 290w_{\text{As}} + 13.1w_{\text{W}} \end{align} (1)

  

\begin{align} A_{\text{c}_{3}} &= 910 - 203\sqrt{w_{\text{C}}} - 15.2w_{\text{Ni}} + 44.7w_{\text{Si}} \\ &\quad + 104w_{\text{V}} + 31.5w_{\text{Mo}} + 13.1w_{\text{W}} \end{align} (2)

The chemical compositions were taken into Formula (1) and (2), and the A_c1 and A_c3 of DH36 in this experiment has been calculated as 1001 K and 1144 K. The tempering temperature was set to be 773 K, 873 K and 973 K respectively in order to obtain the stable microstructure and improve the mechanical properties of DH36 steel.

Then 4 groups of 20 mm × 200 mm × 200 mm steel sheets were taken respectively, and Fig. 1 shows the size of one sheet. One of the sheets were left as original sample while the other 3 sheets were put into the soaking pit furnace. The heat-treatment processes and sampling number were shown in Fig. 2. The 3 samples were heated in the furnace up to 773 K(1#), 873 K(2#) and 973 K(3#) respectively. They were all kept warming for 10800 s and cooled in the air.

Fig. 1

Sampling positions of DH36 steel for testing.

Fig. 2

Tempering processes of three samples.

Smaller samples were then cut from the 4 plate samples, respectively. The position was shown in Fig. 1. The test for mechanical properties included transverse impact value (Charpy V-notch) tests and tensile strength measurements. The dimensions of the Charpy test samples were 10 × 10 × 55 mm³, and the tensile test samples were 20 × 160 mm² (the central part was 12.5 × 80 mm²). The impact values of the three samples were tested by a ZBC2452-B Pendulum impact testing machine. Tensile strength, yield strength, elongation were examined with a CMT4105 electronic universal testing machine. The tests for each group had repeated 3 times, and the average values were used for analysis and research.

The OM, SEM and TEM sample were also obtained from this plate sample, the dimensions of which were 10 × 10 × 10 mm³, 10 × 10 × 5 mm³ and 10 × 10 × 8 mm³ respectively. The microstructure of the steel was observed with a 9XP-PC optical microscope and field emission scanning electron microscope (JSM-7001F).

The morphology of the carbides in the three samples was examined with a TECNAI-G2 F20 high resolution transmission electron microscope (HR TEM) (manufactured by FEI Company, Hillsboro, OR, USA). The carbon extraction replica method was used to prepare the TEM samples; the specific steps were as follows: first etch the polished metallographic samples in 8% nitric acid alcohol solution, then coat them with a layer of evaporated carbon film, approximately 20–30 nm thick, and finally extract the precipitates by using 10% nitric acid alcohol solution and mount the carbon membrane on a copper grid. The morphology of the precipitates was analyzed by TEM after the samples were dried.

3. Results and Discussions

3.1 Mechanical properties

Following the heat treatment processes of Fig. 2, the mechanical properties of as-rolled and tempering samples from DH36 were listed in Table 2.

Table 2 Mechanical Properties of DH36 steel after tempering and before tempering.

Table 2 showed that as-rolled steel has met the standard set by National Classification Society.⁹⁾ However, it has lower plasticity. Comparing with as-rolled steel, 253 K impact value and plasticity have improved after tempering, especially after 873 K and 773 K. In summary, the low-temperature tempering process is an effective way to improve the strength and toughness, and to establish a new strength/toughness balance of steel.

Table 2 also presented that the plasticity of DH36 maintained 22.5%–23.5%, which was more than the lowest limit of National Classification Society as tempering temperature went up. Although tensile strength and yield strength decreased gradually with increase of tempering temperature, they both met the standard set by National Classification Society.⁹⁾ The 253 K impact value had increased then decreased with tempering temperature went up, but even the lowest value of steel 3# exceeded the lowest limit of standard for 78.1 J. Comparing with the 3 samples, steel 1# had the optimum mechanical properties. The result showed that various heat treatment processes would bring different solid-solid reaction conditions. The evolutions of microstructure and precipitated phases should be considered due to such different mechanical properties of DH36 steel.

3.2 Microstructure of DH36 after tempering

The structure is mainly composed of ferrite and pearlite, as shown in Fig. 3. After analyzing 30 pictures of each sample, the grain sizes were calculated as 12 µm. The grain size has not increased significantly as the tempering temperature went up. Meanwhile the volume fraction of pearlite has not reduced.

Fig. 3

Microstructures of three samples.

The morphologies of ferrite and pearlite in DH36 after holding for 180 min at 773 K were shown in Fig. 4. From Fig. 4, there existed precipitated particles both in grains and along grain boundaries. These particles may generate pinning effect on grain boundary.¹⁰⁾ According to the result of mechanical properties of DH36 in Table 3 and microstructures in Fig. 3 and Fig. 4, it can be concluded that dispersed particles (precipitated phases) may affect mechanical properties of DH36 steel. In the next section, we will characterize the types of these precipitates and consider the effect for the mechanical property changes.

Fig. 4

Microstructure of steel 1# by SEM.

Table 3 Size distributions and strength increments of the three samples.

3.3 Analysis of carbides

3.3.1 Types of carbides

Since there are a lot of research on behavior of nitrides, such as TiN and AlN in low carbon microalloyed steels,^11–13) we mainly focus on carbides and their evolutions during tempering in this experiment.

After observing hundreds of particles with morphologies, SAED patterns and EDS analysis, the carbides in DH36 after tempering were mainly two types, MC and cementite (M₃C).

The morphology of MC-type carbide was shown by TEM in Fig. 5(a) in steel 3#. The size of this carbide was 76.9 nm. Figure 5(b) and Fig. 5(c) showed the SAED and EDS analysis of MC. From Fig. 5(c), we considered this carbide was TiC, along with some Nb. It is reported that both TiC and NbC have FCC structure just like NaCl crystal in Nb–Ti microalloyed low-carbon steels.¹⁴⁾ Actually, MC with Nb, Ti was more common, but its crystal structure was unchanged.¹⁵⁾ Thus, (Ti,Nb)C in this steel ought to be FCC structure. From the direct spot and two specific diffraction spots in Fig. 5(b), the crystalline interplanar spacing was 0.22 nm, and the angle was measured as 90°. Comparing with the PDF card, the Miller indices were defined as (200) and (020). According to the solid-solution products of TiC and NbC¹⁴⁾ as shown in Formula (3) and (4), the precipitation temperatures of TiC and NbC can be calculated as 1292 K and 1370 K respectively in the thermodynamic standard state.   

\begin{equation} \lg([\mathrm{Ti}]\cdot[\mathrm{C}])_{\gamma} = 2.75 - \frac{7000}{T} \end{equation} (3)

  

\begin{equation} \lg([\mathrm{Nb}]\cdot[\mathrm{C}])_{\gamma} = 2.26 - \frac{6770}{T} \end{equation} (4)

Fig. 5

The morphology, SAED, and EDS analysis of (Ti,Nb)C in steel 3#: (a) The morphology of (Ti,Nb)C; (b) SAED analysis; (c) EDS analysis.

(Ti,Nb)C will precipitate at 1292 K∼1370 K theoretically. Thus, there will be large amount of NbC and TiC precipitating at tempering temperatures.

TiC and NbC both belonged to cubic crystal system, and had similar lattice constant. Thus, Ti and Nb realized mutual diffusion during solidification when the steel was cooled from liquid phase to solid phase.¹⁶⁾ (Ti,Nb)C may precipitate from DH36, and can not be decomposed even in γ-Fe and α-Fe. The formation of second phase (Ti,Nb)C can be explained as some niobium dissolved in TiC crystal.¹⁷⁾ Through analyzing large quantities of pictures, most (Ti,Nb)C were square-shaped, belonged to FCC cubic crystal system, and their size was less than 80 nm.

Another type of carbides we detected was cementite (M₃C). The morphology of M₃C was shown in Fig. 6(a). It can be seen that the size of this carbide in steel 3# was 133 × 267 nm². From Fig. 6(c), we concluded that it may be cementite. Generally, there exist some tertiary cementite precipitate after the steel was tempered, and the structure of M₃C is orthorhombic crystal.^10,14,16) From the direct spot and two specific diffraction spots in Fig. 6(b), two values of crystalline interplanar spacing were 0.21 nm and 0.17 nm, and the angle was 72.9°. Comparing with the PDF card, Fe₃C with Miller indices of (211) and (004) has the similar angle, as shown in Fig. 6(b). The lattice constant of Fe₃C was that a = 0.45 nm, b = 0.51 nm, c = 0.67 nm (Fig. 6(b)). From Fig. 6(c), the main content of M₃C was Fe, and some other alloy elements such as Cr, Mn dissolved in it. The crystal structure of Fe₃C became unchanged regardless of dissolution of alloy elements, such as Mn, Cr. This kind of (Fe,Me)₃C is also called cementite.¹⁸⁾

Fig. 6

The morphology, SAED, and EDS analysis of M₃C in steel 3#: (a) The morphology of Fe₃C; (2) SAED analysis; (3) EDS analysis.

The equilibrium phase diagram was calculated by Thermo-Calc software with the TCFE8 database, as shown in Fig. 7. Cementite precipitated at temperature of 977 K from ferrite during cooling process. Meanwhile, austenite gradually decreased with increasing of cementite and ferrite between 900 and 1000 K.

Fig. 7

Phase equilibrium diagram of cementite, austenite and ferrite in DH36.

Thus, cementite precipitated from ferrite at the experimental temperature 773 K, 873 K and 973 K. It can be concluded that the microstructure of DH36 consisted of ferrite, pearlite and precipitated phases (MC and M₃C or cementite) after the steel was tempered and cooled to room temperature. Specifically, the number and morphologies of these precipitates will influence the strength/toughness properties of DH36.

3.3.2 Distributions of precipitated phases and coarsening kinetics of carbides

After different tempering, TEM analysis made by carbon replica was carried out, and Fig. 8 shows the distributions of precipitates in steel 1#, 2# and 3#.

Fig. 8

Precipitate distributions of DH36 after tempering.

From the size distribution of Fig. 8, it can be seen that the average size of precipitates is larger than 10 nm; therefore, the bypass mechanism has the main effect on precipitation strengthening.¹⁹⁾ In order to calculate the contribution of precipitates to yield strength, we employed segment calculation and summation of the results (Table 4). According to the methods of McCall-Boyd²⁰⁾ and the Ashby-Orowan correction model,²¹⁾ the formula for volume fraction and precipitation strengthening of precipitates in DH36 is obtained as Formula (5) and (6). McCall and Boyd²²⁾ analyzed the characteristics of precipitates in ThO₂ by employing carbon extraction replica methods in the 1960s. The McCall-Boyd method is an accurate way to calculate the volume fraction of precipitates with uneven distribution in alloys.   

\begin{equation} f_{\text{i}} = \left(\frac{1.4\pi}{6}\right)\cdot \left(\frac{N_{\text{i}}D_{\text{i}}^{2}}{A}\right) \end{equation} (5)

  

\begin{equation} \sigma_{s} = \left(\sum_{\text{i} = 1}^{n}\sigma_{\text{i}}^{2}\right)^{\frac{1}{2}} {}= \left(\sum_{\text{i}}^{n}\left[\frac{10\mu b}{5.72\pi^{3/2}r_{\text{i}}}f_{\text{i}}^{\frac{1}{2}}\ln\left(\frac{r_{\text{i}}}{b}\right)\right]^{2}\right)^{\frac{1}{2}} \end{equation} (6)

Table 4 Contributions of precipitating strengthening to yield strength of precipitates.

Here, A represents the area of the photos in µm²; N_i represents the amounts of precipitates within a certain range; D_i represents average diameter in nanometers for precipitates within a certain range; r_i represents the average radius in nanometers for precipitates within a certain range; f_i represents the volume fraction, in % of precipitates within a certain range; µ represents a shearing factor (80.26 × 10³ MPa for steel); and b represents the Burgers Vector, with a value of 2.48 × 10⁻⁴ µm.

We have chosen 50 photos of 65 µm² of each sample for statistics in this experiment. The volume fraction of precipitates and the increment of precipitation strengthening of DH36 after tempering were calculated according to Formula (5) and (6). The calculation process was listed in Table 3, and the results of the calculations were shown in Fig. 9 and listed in Table 4.

Fig. 9

Size distribution precipitates after different tempering process.

It can be seen from Fig. 9 that the particles precipitate most frequently in size of 50∼110 nm in three steels. The particle size distribution deviates from a Gaussian distribution. Because it is of little significance of particles larger than 110 nm for calculating precipitation strengthening, we did not take them for statistics. As the tempering temperature went up, there are less particles precipitated in steel, possibly because small particles dissolved and large ones aggregated and grew up during tempering.

From Table 4, as tempering temperature went up, precipitates coarsened from 74 nm to 81 nm, volume fraction of precipitates decreased, and the contribution of precipitates to yield strength also decreased. Thus, steel 1# has the finest precipitation distribution and best precipitation strengthening.

In this experiment, we also found many M₃C particles which sizes were 80∼200 nm. The morphologies and distributions of precipitated phases in steel 1# and 3# by TEM using carbon extraction replica methods were shown as Fig. 10(a) and Fig. 10(b). Some typical particles were chosen by using SAED and EDS analysis to decide their structures. It can be seen that most large particles were M₃C, and smaller ones were (Ti,Nb)C (<80 nm). In addition, the precipitated phases in steel 1# distributed more uniformly than those in steel 3#, and the particle sizes vary widely in steel 3#.

Fig. 10

Morphologies and distributions of precipitated phases in steel 1# and 3# by TEM.

In order to make sure whether MC or M₃C play the key role in strengthening during tempering, Ostwald ripening model was employed to calculate the coarsening rate of MC and M₃C during different tempering temperatures.

By calculating and deducing, the coarsening process of second phases can be expressed as follows,¹¹⁾   

\begin{equation} \overline{r_{t}}^{3} = \overline{r_{0}}^{3} + m^{3}t \end{equation} (7)

$\overline{r_{t}}$ was the average radius of precipitated phase at time t; $\overline{r_{0}}$ was initial radius of precipitated phase; m was the coarsening rate of precipitated phase during Ostwald Ripening Process. m can be expressed as Formula (8),   

\begin{equation} m = \left(\frac{8\sigma V_{\text{P}}^{2}Dx_{0}}{9V_{\text{B}}x_{\text{P}}RT}\right)^{\frac{1}{3}} \end{equation} (8)

In the formula, σ was specific surface energy between alloy compounds and austenite, J/m²; V_p was molar volume of precipitated phases, m³/mol; V_B was molar volume of solid solution element, m³/mol; D was diffusion coefficient of solute element in ferrite, m²/s; x₀ and x_P were equilibrium mole fractions of solute element in ferrite and precipitated phases, respectively. The value of x_P was usually thought to be 1. T was thermodynamic temperature, K; R was gas constant, usually the value of which was 8.314.

Equilibrium mole fractions of solute element in ferrite x₀ can be calculated as follows.

First, for MC-type precipitate, the relationship between [%M] and [%C] can be deduced from formula (3) and (4) as Formula (9) and (10).   

\begin{equation} [\text{% M}] \cdot [\text{% C}] = 10^{A-\frac{B}{T}} \end{equation} (9)

  

\begin{equation} \frac{w_{\text{M}} - [\text{% M}]}{w_{\text{C}} - [\text{% C}]} = \frac{A_{\text{M}}}{A_{\text{C}}} = \omega \end{equation} (10)

w_M and w_C were chemical compositions of alloy element M and carbon in DH36; [%M] and [%C] were the mass fractions dissolved in ferrite; A_M and A_C was the relative atomic mass of alloy element M and carbon. ω was the ideal chemical ratio.

It can be assumed that precipitated phases in DH36 were binary pure substance, and the matching ratio of elements satisfies ideal chemical ratio. Formula (10) can be simplified as follows,   

\begin{equation} \frac{[\text{% M}]}{[\text{% C}]} = \frac{A_{\text{M}}}{A_{\text{C}}} = \omega \end{equation} (11)

Formula (12) was available from eq. (9) and (11).   

\begin{equation} [\text{% M}] = 10^{\frac{A}{2} - \frac{B}{2T}} \cdot \sqrt{\omega} \end{equation} (12)

The relationship between equilibrium mole fraction and mass fraction was as Formula (13).   

\begin{equation} x_{\text{M-0}} = \frac{[\text{% M}]A_{\text{Fe}}}{100A_{\text{M}}} \end{equation} (13)

For M₃C (cementite), x₀ can be calculated as follows, the equilibrium solid solubility of cementite in ferrite was written as Formula (14).¹⁴⁾   

\begin{equation} \lg [\text{%C}]_{\alpha} = 2.38 - \frac{4040}{T} \end{equation} (14)

Formula (14) can be changed as follows,   

\begin{equation} [\text{% C}]_{\alpha} = 10^{2.38 - \frac{4040}{T}} \end{equation} (15)

The relationship between equilibrium mole fraction and mass fraction of carbon was as Formula (16).   

\begin{equation} x_{\text{C-0}} = \frac{[\text{% C}]A_{\text{Fe}}}{100A_{\text{C}}} \end{equation} (16)

Thus, c_M-0 and c_C-0 can be calculated from the above.

Specific interfacial energies of carbides were listed in Table 5.

Table 5 Interfacial Energy between carbides and steel.

The diffusion coefficients of elements in ferrite and their calculated results at 773 K, 873 K and 973 K were listed in Table 6.

Table 6 Diffusion coefficients of elements in austenite and their calculated results.

From Table 6, it can be seen that Ti, Nb, Mn and Cr had the diffusion coefficient of the same order of magnitude. The diffusion coefficient of C was 5∼9 orders of magnitude more than those of Ti, Nb, Mn and Cr. Sadhan Ghosh²³⁾ has verified that coarsening of cementite in Fe–0.6C–2Mn is delayed by dissolution of Mn from the carbide. Thus, it can be concluded that diffusion of alloy elements was becoming the rate controlling process during M₃C coarsening.

Referring to relevant literature,¹⁴⁾ mole fractions of solid-solution elements in ferrite (V_B) and mole fractions of various precipitated phase were listed in Table 7.

Table 7 V_B in austenite and V_P in precipitated phases.

From Table 7, it can be seen that the values of V_B and V_P have little difference.

Finally, specific interfacial energies, equilibrium mass fractions, initial mole fractions and coarsening rates of TiC, NbC and Fe₃C were listed in Table 8, according to Formula (11)∼(16) and data from Table 5∼7.

Table 8 Calculation of coarsening rates of carbides.

From Table 8, it can be seen that coarsening rates of these carbides increased as tempering temperature went up, and Fe₃C coarsened faster than Nb–Ti carbides. It can be concluded that MC play the key role in precipitation strengthening. Coarsening rates of NbC and TiC were comparable. For (Ti,Nb)C compound carbides, the coarsening rate at 973 K basically remained 0.09∼0.14 nm·s$^{\frac{1}{3}}$. And the coarsening rate of Fe₃C was 55.55 nm·s$^{\frac{1}{3}}$. Put this result into Formula (7). After holding for 10800 s, (Ti,Nb)C may grow up almost 4.11∼6.32 nm and cementite may grow up to 2.25 µm. Thus, the diffusion rate of Mn and Cr in ferrite are far lower than that of carbon. The coarsening rate of M₃C will decrease. It can be concluded that alloy elements in DH36 may effectively inhibit the coarsening of M₃C. There existed 200 nm-sized M₃C after holding for 10800 s at 973 K, as shown in Fig. 9(b). Thus it can also be seen that the above analysis fitted well with the TEM observation in this research.

From the analysis above, tensile and yield strength decreased as tempering temperature went up. The reason was that contribution of fine precipitates to the yield strength in 773 K was larger than that in 873 K and 973 K. Meanwhile, toughness increased then decreased with ferrite grain sizes being nearly unchanged when DH36 steel was tempered from 773 K to 973 K. The reason was that there were less precipitates in steel tempered at 873 K than that tempered at 773 K. This will improve the toughness and reach to a summit. As the tempering temperature reached 973 K, fine precipitates dissolved and large precipitates (M₃C) become larger. Therefore, large precipitates deteriorate the toughness of DH36, and make it decrease sharply. The cementite may precipitate from austenite and ferrite, and have not enough energy to form lamella in pearlite. On the contrary, they become the second phases.

At the same time, M₃C (cementite) in Fig. 6(a) was 133 × 267 nm² size, and the coarsening speed of cementite was 29 and 55 nm·s^−1/3 at 873 K and 973 K. Thus, the phenomenon of cementite coarsening was obvious both from calculation and experiment. This will reduce the low temperature toughness of DH36 steel greatly.

4. Conclusion

By testing the mechanical properties, observing microstructure and analyzing precipitated phases of DH36 at different tempering temperatures, the optimum tempering process was obtained. The present studies can be summarized as follows,

1. (1)    The microstructure of DH36 after tempering was composed of ferrite, pearlite and precipitated phases. The grain sizes of ferrite have not increased obviously with increasing of tempering temperature.
2. (2)    The carbides were mainly (Ti,Nb)C and cementite, the sizes of which were less than 80 nm and more than 80 nm respectively. The precipitation temperature of (Ti,Nb)C was below 1370 K, while the precipitation temperature of cementite was below 977 K after calculation. Thus, large quantities of cementite will precipitate during tempering process.
3. (3)    As the tempering temperature increases, the average sizes of precipitated phases increased from 74 nm to 81 nm, while volume fraction of precipitates decreased, and the contribution of precipitates to yield strength also decreased.
4. (4)    The plasticity changed a little from 773 K to 973 K. Because the ferrite grain size grew up slowly, which was pinned by precipitated phases. Meanwhile, tensile and yield strength decreased with the tempering temperature went up. The reason was that precipitated phases coarsened, especially cementite. In addition, moderate distributed and finely dispersed precipitates may improve the toughness of DH36. To sum up, DH36 steel after holding for 10800 s at 773 K has the optimum comprehensive mechanical properties.

Acknowledgments

The authors acknowledge that this research is supported by the natural science foundation of Taiyuan University of Technology, Shanxi, China: Investigation on nanoscale precipitates in DH36 shipbuilding steel and its comprehensive strengthening (2015QN010). This research is also supported by Scientific and Technological Innovation Programs of Higher Education Institutions in Shanxi (201802035).

REFERENCES

• 1)   J.H.  Teng,  D.P.  Wang,  Z.J.  Wang and  F.  Yang: Mater. Des. 118 (2017) 266–278.
• 2)   F.  Yue,  Y.P.  Bao,  H.  Cui and  D.M.  Liu: J. Univ. Sci. Technol. Beijing 29 (2007) 1–5.
• 3)   F.  Hayat and  H.  Uzun: J. Iron Steel Res. Int. 18 (2011) 65–72.
• 4)   J.  Dong,  C.X.  Liu,  Y.C.  Liu,  C.  Li,  Q.Y.  Guo and  H.J.  Li: Fusion Eng. Des. 125 (2017) 415–422.
• 5)   M.S.  Mohebbi,  M.  Rezayat,  M.H.  Parsa,  S.  Nagy and  M.  Nosko: Mater. Sci. Eng. A 723 (2018) 194–203.
• 6)   A.  Karmakar,  S.  Biswas,  S.  Mukherjee,  D.  Chakrabarti and  V.  Kumar: Mater. Sci. Eng. A 690 (2017) 158–169.
• 7)   C.S.  Zheng and  L.F.  Li: Mater. Sci. Eng. A 713 (2018) 35–42.
• 8)   K.W.  Andrews: J. Iron Steel Inst. 203 (1965) 721–727.
• 9)  State Bureau of Quality Technical Supervision: Hull Structural Steel. GB712-2000, (China Standards Press, Beijing, 2000) pp. 7–13.
• 10)   J.  Fu,  G.Q.  Li,  X.P.  Mao and  K.M.  Fang: Metall. Mater. Trans. A 42 (2011) 3797–3812.
• 11)   H.  Ohta and  H.  Suito: ISIJ Int. 47 (2007) 197–206.
• 12)   S.F.  Medina,  M.  Chapa,  P.  Valles,  A.  Quispe and  M.I.  Vega: ISIJ Int. 39 (1999) 930–936.
• 13)   Y.L.  Kang,  H.  Yu,  J.  Fu,  K.  Wang and  Z.  Wang: Mater. Sci. Eng. A 351 (2003) 265–271.
• 14)  Q.L. Yong: The Second Phase of the Steel Materials, (Metallurgical Industry Press, Beijing, 2006) pp. 475–509.
• 15)   J.  Dong,  X.S.  Zhou,  Y.C.  Liu,  C.  Li,  C.X.  Liu and  Q.Y.  Guo: Mater. Sci. Eng. A 683 (2017) 215–226.
• 16)   S.G.  Zheng,  C.  Davis and  M.  Strangwood: Mater. Charact. 95 (2014) 94–104.
• 17)   C.Y.  Chen,  C.C.  Chen and  J.R.  Yang: Mater. Charact. 88 (2014) 69–79.
• 18)  W.X. Song: Metallurgraphy (Second Edition), (Metallurgical Industry Press, Beijing, 2008) pp. 329–337.
• 19)  J.X. Chen: Steelmaking Common Chart Data Manual, (Metallurgical Industry Press, Beijing, 2010) pp. 709–717.
• 20)   G.M.  Carlos,  C.  Capdevila,  F.G.  Caballero and  D.A.C.  Garcia: ISIJ Int. 48 (2008) 1270–1275.
• 21)  Z.Q. Cui and Q.Y. Chun: Metal Science and Heat Treatment, (Machinery Industry Press, Beijing, China, 2007) pp. 177–187.
• 22)  R.J. Seher, H.M. James and G.N. Maniar: Stereology and Quantitative Metallography, ed. by G.E. Pellissier and S.M. Purdy, (ASTM International, West Conshohocken, PA, USA, 1972) pp. 119–137.
• 23)   S.  Ghosh: Scr. Mater. 63 (2010) 273–276.