Hot Deformation Behavior Coordination and Processing Maps of 40Cr/Q345B Bimetallic Blank by Centrifugal Casting

Abstract

The compact cast-rolling compound forming of bimetallic ring is an efficient process for manufacturing large bimetallic rings. The difference in hot deformation behavior of the two metals directly affects the coordinated deformation of bimetals during hot rolling. In this paper, hot compression tests of 40Cr/Q345B bimetallic blank produced by centrifugal casting were carried out at temperatures of 950–1200 °C and strain rates of 0.005–5 s⁻¹. Based on the comparisons of flow stress behavior, activation energy, and activation volume, hot deformation behavior coordination of 40Cr/Q345B bimetallic blank was investigated. Processing maps were established to study the optimum processing windows for 40Cr/Q345B bimetallic blank. Under the same deformation conditions, 40Cr shows lower overall flow stress than Q345B, which facilitate the coordinated deformation of the bimetallic blank during rolling, and the flow stress difference between 40Cr and Q345B decreases with the increase of temperature. It was connected with the increase in the second phase particles pinning effect for higher MnS content in Q345B. The common rate-controlling mechanism for 40Cr and Q345B is the thermal breaking of attractive junctions and movement of jogged screw dislocations. The bimetallic blanks exhibit good deformation coordination in the optimum processing window (1030–1100 °C and 0.5–1 s⁻¹ with a strain of 0.6).

1. Introduction

For applications where the inner and outer performance requirements of rings are different, it is difficult for single metal rings to meet the requirements of actual performance. Bimetallic rings can combine the advantages of two metals at the same time, maximize the complementary advantages of materials, improve the performance, service life, and reliability of parts, and effectively save the amount of precious metals to reduce the manufacturing cost of parts. The compact cast-rolling compound forming of bimetallic ring is a new process that combines the two metals together by centrifugal casting and hot rolling [1,2]. Compared with the existing process for producing large bimetallic rings, it is a high-efficiency process with advantages of short process, high quality, low cost, low energy consumption, and less pollution [3,4].

Understanding the hot deformation behavior of metals is crucial for designing hot rolling parameters that are related to the final properties of products. At present, there are many studies on hot deformation behavior. During hot deformation processes, deformation mechanisms including work hardening (WH), dynamic recovery (DRV), and dynamic recrystallization (DRX) have significant effects on the deformation behavior of metals [5,6,7,8,9,10,11,12]. Based on the dislocation density theory, Estrin and Mecking (EM) [13] proposed a physics-based constitutive model describing the WH and DRV of metals [14,15]. Jonas [16,17,18,19] et al. incorporated DRX-induced softening into the constitutive equation by using the Avrami equation, which excellently expresses the hot deformation behavior of metals at various stages. The WH rate, DRV rate, and DRX volume fraction derived during this process can reflect the difficulty of each mechanism, which affects the flow stress. According to kinetics theory, the activation energy of hot deformation describes the energy barrier level that needs to be overcome with respect to certain atomic mechanisms [20,21]. Activation volume can be assumed to be the volume swept by the dislocation slip when thermal activation crosses the barrier [22,23]. Many researchers [24,25,26,27,28,29,30] use the calculated activation energy and activation volume to reflect the difficulty of metal hot deformation, and infer the rate-controlling mechanism. In addition, the establishment and research of the processing map are also very important. Prasad et al. [31] established a processing map based on dynamic material model (DMM), which has been widely used to study the hot formability of various metals and predict the optimal hot forming parameters [32,33,34,35,36]. Therefore, it is of great significance for practical industrial manufacturing. In recent years, in addition to rolled and forged metals, processing maps have been also used to study the hot formability of as-cast metals [37]. Current studies on the hot deformation behavior and the establishment of processing maps focus on a single metal, there are few research reports of bimetallic materials. Li et al. [38,39,40] conducted hot tensile tests on the 2205 duplex stainless/AH36 low-carbon steel bimetal composite (2205/AH36 BC) with a new hot-rolling process, and the hot deformation behavior, microstructure evolution, interfacial characteristics, and fracture mechanical properties at high temperature were systematically studied. Xu et al. [41,42] studied the microstructure and hot deformation behavior characteristics of the high-carbon steel/low-carbon steel bimetal prepared by centrifugal composite casting. Four microstructure regions with different mechanical properties were found. From the perspective of macroscopic deformation, it is believed that the fully austenitized bimetal can be processed like a monolithic material. Based on this assumption, an Arrhenius-type constitutive equation is established. However, the composition and plasticity of the centrifugal casting bimetallic ring blank are not uniform in the whole thickness, and the deformation ability of the inner and outer metal in each direction is different, resulting in the difference in the speed of plastic flow between layers during hot rolling. Ultimately, rolling defects will be caused. Therefore, the hot deformation behavior coordination and hot forming characteristics of bimetallic materials need to be investigated.

In this paper, hot compression tests of 40Cr, Q345B, and bonding layer from 40Cr/Q345B bimetallic blanks produced by centrifugal casting were conducted. The hot deformation behavior and thermal processing map of 40Cr and Q345B were compared. The coordinated deformation behavior of the 40Cr/Q345B bimetallic blank was evaluated. The rate-controlling mechanism was evaluated and the optimum processing parameters were recommended.

2. Materials and Methods

The 40Cr/Q345B bimetallic ring blank was prepared by centrifugal casting, with 40Cr as the outer layer and Q345B as the inner layer. The parameters of centrifugal casting are shown in

Table 1. Process parameters of centrifugal casting to prepare 40Cr/Q345B bimetallic ring blank.

Parameters	Values
Material thickness ratio	1:1
Pouring temperature of outer layer	1560 °C
Pouring temperature of inner layer	1580 °C
Pouring interval	161 s
Mold release temperature	850 °C

. The chemical compositions of 40Cr and Q345B in the bimetallic ring blank are shown in

Table 2. Chemical compositions of 40Cr and Q345B (wt.%).

Materials	C	Cr	Mn	Si	P	S	Ni	Cu	Fe
40Cr	0.445	1.091	0.591	0.301	0.010	0.005	<0.3	≤0.030	Bal.
Q345B	0.168	0.299	1.609	0.461	0.013	0.007	≤0.5	≤0.030	Bal.

. The size of the final bimetallic ring blank was φ360 mm × φ200 mm × 150 mm.

The specimens of 40Cr, Q345B, and bonding layer applied in a compression test were cut from the bimetal ring blank. According to ASTM E209, the specimens were machined into cylinders with a size of φ10 mm × 15 mm. The sampling positions are shown in Figure 1. The graphite and tantalum foils were placed on the flat ends of the specimen to minimize friction between the anvil and the specimen. Isothermal compression tests of specimens were performed on a Gleeble-3500 thermal simulator (DSI Ltd, Poestenkill, NY, USA). The deformation temperatures of 950 °C, 1050 °C, 1100 °C, and 1200 °C, and the strain rates of 0.005 s⁻¹, 0.05 s⁻¹, 0.5 s⁻¹, 1 s⁻¹, and 5 s⁻¹ were selected. The specimens were heated to the target temperature at a heating rate of 5 °C/s and held for 5 min to achieve a stable temperature distribution. The overall reduction is 60% in height. Standard polishing and etching techniques were used to prepare specimens for metallographic examination. Microstructures were observed with the optical microscope. SEM-EDS analysis (SEM, JEOL Ltd, Tokyo, Japan; EDS, Oxford Instruments, Oxford, UK) was carried out on the second phase particles of bimetallic ring blanks. The flow stress curves were converted from the data during hot compression to study the hot deformation behavior and processing maps of 40Cr and Q345B at different deformation conditions.

3. Results

Figure 2 shows the initial microstructure of the bimetallic blank. The microstructure of 40Cr mainly consists of pearlite, and Q345B is composed of ferrite and pearlite. The bonding layer gradually transits from Q345B to 40Cr, with ferrite decreasing and pearlite increasing. The bimetallic blank has formed a good metallurgical bond, and the bonding layer has the characteristics of gradient transition.

Both 40Cr and Q345B contain manganese. At a high temperature, MnS is easily formed due to the strong affinity of Mg and S. Thermodynamic calculations of the phase composition for both 40Cr and Q345B were conducted using the JMatPro software (Version 9.0, Sente Software Ltd., Guildford, UK) at temperatures between 950 and 1200 °C, and the results are shown in Figure 3. It can be concluded that 40Cr and Q345B had a small amount of MnS in addition to transforming into austenite phase between 950 and 1200 °C. The bimetallic specimen was heated to 1050 °C, kept for 5 min, and directly quenched. MnS was found by SEM combined with EDS analysis, as shown in Figure 4. The influence of MnS as a second phase particle on hot deformation behavior and activation energy values will be discussed in the following text.

3.1. Flow Stress Behavior

Figure 5 shows the flow stress curves of 40Cr and Q345B at different deformation conditions. Under the same deformation conditions, the curves of 40Cr and Q345B present the same type and variation trend. The stresses of 40Cr and Q345B decrease with the increase of temperature and the decrease of strain rate. It is worth noting that within the experimental conditions of this study, 40Cr has a lower flow stress than Q345B. Meanwhile, the difference of flow stress between 40Cr and Q345B gradually decreases with the increase of temperature. The difference of flow stress between 40Cr and Q345B directly affect the coordinated deformation of bimetallic blank, and lead to the instability of the forming shape and forming performance. Therefore, it is necessary to analyze the causes of the differences and the variation trend with deformation conditions in order to provide theoretical guidance for formulating reasonable processing parameters.

The deformation mechanisms of WH, DRV, and DRX are reflected in the hot deformation behavior of 40Cr and Q345B, which are shown in Figure 5. Before the critical strain (${\epsilon }_c$), no DRX occurred. At this phase, the flow stress is decided by dislocation density which is the result of interplay between the generation and annihilation of dislocations during WH and DRV, respectively [9,15]. Therefore, according to the dislocation density theory, the EM model is used to describe the flow stress of the WH + DRV stage [13,18,43]:

$${\sigma }_{wh}=\sqrt{{\sigma }_{sat}^2-\left({\sigma }_{sat}^2-{\sigma }_0^2\right)exp\left(-r\epsilon \right)}\left(\epsilon <{\epsilon }_c\right)$$

(1)

${\sigma }_{sat}$ is the saturation stress, and ${\sigma }_0$ is the initial yield stress (when $\epsilon =0$).

The DRV rate $\gamma$ [19] is obtained by sorting and taking the logarithm of Equation (1):

$$\gamma =-\frac{1}{\epsilon }ln\frac{{\sigma }_{sat}^2-{\sigma }^2}{{\sigma }_{sat}^2-{\sigma }_0^2}$$

(2)

The WH parameter D is derived from the derivation method in Reference [17]:

$$D=r\frac{{\sigma }_{sat}^2}{{\left(\alpha \mu b\right)}^2}$$

(3)

when the applied strain increased over to ${\epsilon }_c$, the cumulative dislocation density exceeded the DRX threshold and DRX was generated. The effect of DRX softening on flow stress is affected by DRX percentage. The relationship between DRX volume fraction and flow stress is usually described as [16]:

$$X_D=\frac{{\sigma }_{wh}-\sigma }{{\sigma }_{sat}-{\sigma }_s}\left(\epsilon \ge {\epsilon }_c\right)$$

(4)

where ${\sigma }_{wh}$ is the flow stress predicted by Equation (1), MPa, $\sigma$ is the DRX flow stress, MPa, And ${\sigma }_s$ is DRX actual steady-state stress, MPa.

The WH parameters, DRV rate, and DRX volume fraction of 40Cr and Q345B are obtained through the above calculations. Figure 6 exhibit the D and $\gamma$ values of 40Cr and Q345B obtained by using the flow curves in the WH-DRV phase. The larger D and the smaller $\gamma$ lead to higher stress increase rate. It is obvious that the D of 40Cr and Q345B increase while $\gamma$ decreases with the increase of strain rate. This is mainly attributed to the fact that the dislocation multiplication at high strain rate is more rapid than that at low strain rate, which contributes to greater strain hardening [6]. At the same time, the shortening of deformation time inhibits dislocation annihilation, thereby reducing the $\gamma$. According to Figure 6, it can also be concluded that the D of 40Cr and Q345B decrease and the $\gamma$ increase with the increasing temperature. The reason for these trends is that the thermal activation of metal atoms becomes intense at high temperature, leading to obvious motion of dislocation and vacancy, thus offering greater driving force for dislocation slip and climb [7,8].

In addition, under the same deformation conditions, the D of 40Cr is lower than that of Q345B, while the $\gamma$ is higher than that of Q345B, and the difference gradually decreases with the increasing temperature. This may be due to the solute drag effect and the second phase particles pinning effect. At the studied temperature, the solute drag effect is caused by the difference in the diffusion rate between the solute atoms and the base atoms, which impedes the dislocation motion and accelerates the dislocation multiplication [10]. The atomic radius difference between Mn and Fe is very close to that between Cr and Fe, so the solute drag effect produced by the two elements is close. Carbon, as an interstitial atom, more easily causes lattice distortion. Some reports [24,44] show that the addition of carbon can reduce WH at high temperatures and low strain rates, and increase recovery rate. Meanwhile, MnS, as the second phase particle, has a strong pinning effect on the dislocation motion, which increases the dislocation density by forming more dislocation entanglements and rings [11]. The WH effect is enhanced due to the increase of dislocation density. Moreover, DRV involving rearrangement and annihilation of dislocations is restrained by pinning effect [10]. In the tested temperatures, the MnS content in 40Cr is lower than that in Q345B, which results in high D and low $\gamma$ of 40Cr. The combined strengthening effect of solute and second phase particles makes the stress of 40Cr and Q345B obviously different in WH-DRV stage. As the temperature increases, the thermal activation of metal atoms becomes intense leading to weakened solute strengthening, which reduces the pinning force of dislocation slip and increases DRV [10]. In addition, the MnS contents in 40Cr and Q345B decrease with the increasing temperature (Figure 3), indicating a reduced hindering effect on dislocation motion and a reduced difference in D and $\gamma$ between 40Cr and Q345B.

The applied strain beyond the critical strain, DRX is generated. According to Equation (4), DRX volume fraction $X_D$ of 40Cr and Q345B were calculated. Figure 7 exhibits the curves of $X_D$ with strain at different deformation conditions. The $X_D$ increases with the increasing strain. DRX is easy to occur and the rate is faster at a higher temperature or a lower strain rate. Furthermore, it is clear that the DRX rate of 40Cr is similar to Q345B at the same deformation conditions. DRX can be delayed by solute drag effect as the retard of solute atoms on the moving boundaries [12]. In particular, the higher the carbon content, the slower the DRX. Therefore, the higher carbon content in 40Cr reduces the DRX rate compared to Q345B. At the same time, the second phase particle MnS inhibits dislocation motion and grain boundary migration, which leads to DRX delay by pinning grain boundaries [11]. However, the MnS in Q345B is higher than 40Cr, resulting in a lower DRX rate of Q345B than 40Cr. From the calculated results of the DRX volume fraction, the effects of carbon content and MnS content on DRX are comparable, leading to the DRX kinetics of 40Cr and Q345B being slightly different. In summary, the softening effects of DRX on 40Cr and Q345B are similar, and the stress difference occurs in the WH-DRV stage.

3.2. Coordinated Response to Hot Deformation of Bimetallic Ring Rolling

The 40Cr/Q345B bimetallic ring blank is composed of outer layer, inner layer, and metallurgical bonding layer with gradient transition. The composition, organization, and plasticity of each layer are different, which is similar to a layered structure. The deformation response of the layered structure during the pressing process is shown in Figure 8 [45]. When the individual component layers are very thin and the friction constraint between layers or combined with forced uniform deformation, it is close to the isostrain behavior (Figure 8b) [45]. The thick layer system is approached for isostress behavior as shown in Figure 8c. Each component is subjected to the same stress, and the individual layers act independently [45].

It can be seen from the above analysis that at the test temperature, both 40Cr and Q345B transformed into austenite and contained a small amount of second-phase particles MnS. Although the high-temperature austenite structures of 40Cr and Q345B have similar yield strength, the flow stress of 40Cr is smaller than Q345B due to the difference in the MnS content. This means that the plastic deformation abilities of the outer and inner layers of bimetallic ring blank are different at high temperature. When subjected to the same stress, 40Cr is more easily deformed than Q345B. This is similar to a thick layer system where isostress behavior occurs, which will cause the deformation difference between 40Cr and Q345B. However, through metallurgical bonding between 40Cr and Q345B, the gradient transition of the bonding layer guides the deformation of Q345B to the deformation of 40Cr. The bonding layer forces the materials on both sides to deform uniformly, and plays a role in coordinating the deformation of the outer layer and the inner layer.

Furthermore, the profile shape of the bimetallic ring blank in the rolling process can be represented by Figure 9. According to the ring rolling dynamics, the distribution relationship between the outer wall thickness reduction $\Delta$h₁ and the inner wall thickness reduction $\Delta$h₂ of the bimetallic ring blank is [46]:

$$\Delta h_1/\Delta h_2=\left(\frac{1}{R_1}+\frac{1}{R}\right)/\left(\frac{1}{R_2}-\frac{1}{r}\right)$$

(5)

R is the starting radius of outer contour helix, r is the starting radius of inner contour helix, R₁ is the radius of main roll, R₂ is the radius of mandrel. From the above equation, the deformation of the outer surface is greater than the deformation of the inner surface for a quite long time [47]. The outer 40Cr of the bimetallic ring blank is more easily deformed than the inner Q345B, and the difference in the plastic flow velocity of the outer layer and inner layer is conducive to the rolling process. This facilitates the coordinated deformation of the outer layer and inner layer of the bimetallic ring blank during rolling. It can improve the residual stress and uneven end surface caused by rolling and expansion, which contributes to improve the quality of the formed ring.

3.3. Rate-Controlling Mechanism

Based on kinetic theory, the activation energy Q of hot deformation provides information on the difficulty of atomic rearrangements involved in the rate-controlling mechanism [20,21]. Therefore, activation energy Q is a key parameter that needs to be analyzed, which is of great significance to clearly understand the hot deformation behavior and the corresponding rate-controlling mechanism [21]. The activation energy Q can be obtained during the calculation process for establishing the Arrhenius-type phenomenological constitutive model proposed by Sellars and McTegart [48].

$$\dot{\epsilon }=\{\begin{array}{cc}{A}_1{\sigma }^{n_1}exp\left(-Q/RT\right)& \left(\alpha \sigma <0.8\right)\\ A_{2}exp\left(\beta \sigma \right)exp\left(-Q/RT\right)& \left(\alpha \sigma >1.2\right)\\ A{\left[\sinh \alpha \sigma \right]}^{n}exp\left(-Q/RT\right)& \left(allfor\sigma \right)\end{array}$$

(6)

where $\dot{\epsilon }$ is the strain rate (s⁻¹); σ is peak stress or steady flow stress (MPa); Q is the activation energy of hot deformation (kJ/mol); T is the absolute temperature (K); R is the universal gas constant (8.314 J/mol·K); A (A₁, A₂), β, and α are material constants; and n(n₁) are the stress exponents.

By deriving Equation (6), the material constants n₁, β, and n and the activation energy Q can be calculated as follows:

$$\beta ={\left(\frac{\partial \ln \dot{\epsilon }}{\partial \sigma }\right)}_T$$

(7)

$$n_1={\left(\frac{\partial \ln \dot{\epsilon }}{\partial \ln \sigma }\right)}_T$$

(8)

$$n={\left[\frac{\partial \ln \dot{\epsilon }}{\partial \ln \sinh \left(\alpha \sigma \right)}\right]}_T$$

(9)

$$Q=RnS=R{\left[\frac{\partial \ln \dot{\epsilon }}{\partial \ln \sinh \left(\alpha \sigma \right)}\right]}_T{\left[\frac{\partial \ln \sinh \left(\alpha \sigma \right)}{\partial \left(1/T\right)}\right]}_{\dot{\epsilon }}$$

(10)

According to Equations (7) and (8), the $\beta$ and n₁ are the average slopes of the linear fitting curves of ln$\dot{\epsilon }$ − σ (Figure 10a,b) and ln$\dot{\epsilon }$ − lnσ (Figure 10c,d), respectively, at different temperatures. The value of α can be obtained from α = $\beta$/n₁. According to Equations (9) and (10), the n is the average slope of the linear fitting curves of ln$\dot{\epsilon }$ − ln[sinh(ασ)] (Figure 11a,b) at different temperatures, and S is the average slope of the linear fitting curves of ln[sinh(ασ)] $-$ (1000/T) (Figure 11c,d) at different strain rates. Then, the value of Q can be obtained according to Equation (10). The obtained material constants are shown in

Table 3. Material constants of 40Cr and Q345B in the hyperbolic-sine equation.

Steel	A (s−1)	β	n1	α (MPa−1)	n	Q (kJ/mol)
40Cr	8.73 × 1012	0.094	6.353	0.01480	4.547	366.901
Q345	2.79 × 1013	0.092	6.629	0.01391	4.782	380.307

.

The activation energy of 40Cr (367 kJ/mol) is lower than that of Q345B (380 kJ/mol). In general, if one steel has a greater activation energy than another, its peak stress σ_p is larger when deformed at the same temperatures and strain rates [29]. The results of this study are consistent with that [20]. Notably, the activation energies of 40Cr (367 kJ/mol) and Q345B (380 kJ/mol) are much higher than austenite lattice diffusion energy (270 kJ/mol) and dislocation nuclear diffusion energy (159 kJ/mol) [25]. Some researchers found a similar phenomenon in other low alloy steels [23,28,29]. The reason for this phenomenon may also be that the solute drag effect and the second phase particles pinning effect inhibits dislocation motion. As a result, the material has a very large activation energy. Lattice-diffusion-controlled dislocation climbing or core-diffusion-controlled dislocation climbing is excluded as the rate-controlling mechanism for 40Cr and Q345B due to their high activation energy values. Therefore, the rate-controlling mechanism of 40Cr and Q345B may be dislocation slip or dislocation cross slip.

However, the rate-controlling mechanism cannot be identified only by such a high activation energy values. In this case, since the activation volume also reveals the deformation mechanism [26,27,30], the activation volume can be used to determine the rate-controlling mechanism of 40Cr and Q345B in the range of test deformation conditions. Activation volume can be assumed as the volume swept by the dislocation slip when thermal activation crosses the barrier, and can be described by the formula [22,23]:

$$V=\frac{MKT}{\sigma m}$$

(11)

where M is the Taylor factor, K is the Boltzmann constant, and m is the strain rate sensitivity factor. Figure 12 shows the activation volumes of 40Cr and Q345B within the test conditions. It can be seen that the activation volume values of 40Cr and Q345B are very close, and decrease with the increase of stress, from 1200 to 200 b³. Comparing the experimental results with Reference [26], the cross-slip of dislocation screw dislocations was ruled out (10–10² b³). Finally, the common rate-controlling mechanism for 40Cr and Q345B was considered to be the thermal breaking of attractive junctions and movement of jogged screw dislocations. These are potential barriers related to dislocation interaction, which may be related to inhibition of dislocation motion caused by the solute drag effect and the second phase particle pinning effect in the above analysis.

3.4. Processing Maps

Processing maps can be used to study the hot formability of metals and are of great significance for predicting the optimal processing parameters. In this paper, the hot formability property of 40Cr/Q345B bimetallic blank was studied by establishing the processing maps of 40Cr and Q345B. According to the Dynamic Material Model (DMM), the workpiece under hot deformation can be considered as a nonlinear dissipater of the power. The total power dissipation P consists of two complementary parts, G represents the temperature increase and J represents the dissipation through metallurgical process. The strain rate sensitivity (m) is used to partition the power into G content and J co-content and m can be given by the following equation [31]:

$$\mathrm{m}=\frac{\partial J}{\partial G}=\frac{\dot{\epsilon \partial \sigma }}{\sigma \partial \dot{\epsilon }}=\frac{\partial \left(ln\sigma \right)}{\partial \left(ln\dot{\epsilon }\right)}$$

(12)

The relationships between ln (true stress) and ln (strain rate) at different temperatures and strains can be fitted by using cubic splines.

$$ln\sigma =A+Bln\dot{\epsilon }+C{\left(ln\dot{\epsilon }\right)}^2+D{\left(ln\dot{\epsilon }\right)}^3$$

(13)

The relationships of ln$\sigma$ − ln$\dot{\epsilon }$ for 40Cr and Q345B at strain 0.4 and 0.6, respectively, are shown in Figure 13. Then, according to Equations (12) and (13), the strain rate sensitive factor m is:

$$\mathrm{m}=\frac{\partial \left(ln\sigma \right)}{\partial \left(ln\dot{\epsilon }\right)}=B+2Cln\dot{\epsilon }+3D{\left(ln\dot{\epsilon }\right)}^2$$

(14)

The power dissipation efficiency (η) representing the microstructural evolution in a materials system during hot deformation is defined as [31]:

$$\eta =\frac{J}{J_{max}}=\frac{2m}{m+1}$$

(15)

Under the given strain, the power dissipation map can be obtained by developing a contour map of the power dissipation efficiency against temperature and strain rate. To a certain extent, the higher the η value, the better the hot formability of materials [37]. Nevertheless, plastic instability may result in too high η values, such as wedge crack and pore [49]. Therefore, the flow instability region needs to be determined. The condition for material’s plastic flow instability proposed by Prasad is extensively applied [36]:

$$\xi \left(\dot{\epsilon }\right)=\frac{\partial ln\left(\frac{m}{m+1}\right)}{\partial ln\dot{\epsilon }}+m<0$$

(16)

when $\xi \left(\dot{\epsilon }\right)$ < 0, flow instability is predicted to occur in this deformation region, and the higher value of $\left|\xi \left(\dot{\epsilon }\right)\right|$, the more probability of instability can be obtained. The instability parameters $\xi \left(\dot{\epsilon }\right)$ as a function of temperature and strain rate constitute the instability map. The processing map is obtained by superimposing the instability map on the power dissipation map.

Figure 14 shows the processing maps of 40Cr and Q345B at the strain of 0.4 and 0.6. The counter values represent the power dissipation efficiency at the corresponding conditions, and the shaded regions are the instability regions. The power dissipation efficiencies of 40Cr and Q345B are characterized by the same variation trend with temperature and strain rate at strain of 0.4, as shown in Figure 14a,b. The power dissipation efficiency first increases and then decreases with increasing temperature and strain rate. The maps of 40Cr and Q345B exhibit two domains with higher η, as shown in the blue rectangular regions. Compared with Q345B, the power dissipation efficiency of 40Cr is slightly higher, but Q345B spreads over a wider domain in the temperature. As the strain increases to 0.6 (Figure 14c,d), the power dissipation efficiencies of 40Cr and Q345B decrease slightly. However, there is no obvious change in the power dissipation efficiency, indicating a steady-state deformation at the strain of 0.6.

The instability regions of 40Cr and Q345B appear at high strain rates. This may be due to the dislocations not having enough time to move when deformed at high strain rates, making it easy for the surrounding solute atoms to form a high-density dislocation region. At the same time, it is easy to produce stress concentration at grain boundaries, which leads to cracks [34]. As strain increases (Figure 14c,d), the instability regions of 40Cr and Q345B are reduced, and the decrease of the instability region of 40Cr is more obvious. At the high strain rate, the power dissipation efficiency increases with increasing strain, possibly indicating increased DRX effect. It is well known that DRX is related to the early dislocation annihilation and rearrangement and the formation of new strain-free grains, and the presence of DRX can improve the flow localization which can inhibit the formation of instability characteristics [35].

In addition, in 40Cr and Q345B processing maps, the power dissipation efficiency is higher in domain 2 compared to domain 1, but the contours are very close in temperature. This indicates that in this domain, the material deformation is more sensitive to the temperature, and the processability window in temperature is small, which is difficult to control in industrial production. Therefore, the preferred processing window in industrial mass production is in the domain 2 with a wide range of strain rates and temperatures.

Although the processing maps of 40Cr and Q345B are different, the distribution and variation of the power dissipation efficiency peak regions and the instability regions are basically the same. It indicates that the processing properties of the two steels are similar, which is conducive to the overall coordinated deformation of 40Cr/Q345B bimetallic blank.

It is generally believed that the processing map alone is not sufficient to determine the optimal processing parameters, and confirmation by microstructural characterization is required [33,35,37]. To verify the validity of the established processing map, the compression test of bimetallic specimens (sampling position as shown in Figure 1) were carried out. The microstructure of bimetallic specimens at different deformation conditions is shown in Figure 15. On the left is Q345B, on the right is 40Cr, and in the middle is the bonding layer. The microstructure presents equiaxed and homogeneous characteristics, revealing that the deformation mechanism of stable regions is controlled by DRX. As the temperature increases, the grain boundary migration accelerates, and the grain size tends to coarsen. When the strain rate increases, the growth time of the newly formed DRX grains decreases, leading to the decrease of grain size. According to the microstructure at deformation temperature of 1150 °C and strain rate of 0.05 s⁻¹ shown in Figure 15b, grains are obviously coarsened, which will adversely affect the mechanical properties of metals. Therefore, selecting a low temperature and a high strain rate as much as possible during hot processing can not only refine the grains, but also save energy and improve production efficiency.

The equiaxed and homogeneous grains of 40Cr, Q345B, and the bonding layer are almost the same, and the evolution with the deformation conditions is also consistent. Instabilities such as voids, cracks, or flow localization were not observed at the same time. This indicates that bimetallic blanks exhibit good deformation coordination under these deformation conditions. According to the above analysis, under the same deformation conditions, the strength difference between 40Cr and Q345B caused by the difference in MnS content is small, and 40Cr and Q345B have the same deformation mechanism and microstructure evolution. Furthermore, the properties of the two steels transition through the metallurgical bonding layer. The above characteristics make 40Cr and Q345B have good compatibility in the overall deformation of bimetallic blanks.

Figure 16a,b show the microstructures of bimetallic specimens deformed under instability region, corresponding to bimetallic specimens compressed at 1150 °C, 5 s⁻¹ and at 1050 °C, 5 s⁻¹, respectively. The adiabatic shear band was found in bimetallic specimen at 1150 °C, 5 s⁻¹. Under this deformation condition, 40Cr and Q345B are in the instability region as shown in Figure 14c,b, indicating that the bimetallic blank is unstable. Furthermore, a slip band extending along grain boundaries at high strain rates was observed on the Q345B side of bimetallic specimen compressed at 1050 °C, 5 s⁻¹. Under this deformation condition, only Q345B is in the instability region (Figure 14c,b), indicating that once there is one side of metal in the instability region, the bimetallic blank may have the risk of instability under this deformation condition. This is undesirable for hot deformation of the bimetallic blank. Therefore, it is necessary to control the deformation conditions during hot forming of bimetallic blank within the common optimum processing parameters of 40Cr and Q345B. Based on the analysis above, it is recommended that the optimal processing windows for 40Cr/Q345B bimetallic blank are 1030–1100 °C and 0.5–1 s⁻¹ with a strain of 0.6.

4. Conclusions

The hot deformation behaviors and processing maps of 40Cr and Q345B in the centrifugal casting 40Cr/Q345B bimetallic ring blanks were studied. The following conclusions were obtained:

(1)

Under the same deformation conditions, the flow stress curves of 40Cr and Q345B are consistent. 40Cr shows lower overall flow stress than Q345B due to softening of carbon and pinning of second phase particles (MnS). As the temperature increases, the flow stress difference between 40Cr and Q345B decreases.

(2)

Activation energy and activation volume of 40Cr and Q345B are not significantly different. Therefore, 40Cr and Q345B have the same rate-controlling mechanism, which is believed to be the motion of the creeping screw dislocation and the thermal fracture of the attracting junction.

(3)

In order to ensure that 40Cr and Q345B are in the optimum deformation region during the hot deformation, the optimal processing window for 40Cr/Q345B bimetallic blank at 1030–1100 °C and 0.5–1 s⁻¹ with a strain of 0.6 are recommended. The bimetallic blanks exhibit good deformation coordination in the optimum processing window.

(4)

Easily deformable outer layer than inner layer and gradient transition of bonding layer facilitate the coordinated deformation of the bimetallic ring blank during rolling.