Constitutive Characteristics, Microstructure, and Texture Evolution of As-Cast 42CrMo Alloy in Nonisothermal Multipass Compression

The nonisothermal multipass deformation behavior of as-cast 42CrMo alloy was studied with declining temperature, constant pass strain, varying strain rate, and interval time. The stresses are used to develop the constitutive model. As the ﬁnishing temperature increases from 990 ° C to 1070 ° C, the stress decreases gradually and the softening eﬀect increases, which results in a large grain size and inhomogeneous microstructure. The low angle grain boundaries transform into high angle grain boundaries through absorbing dislocations. The noticeable stress softening in a high strain rate is attributed to the thermal softening, dynamic recovery, and dynamic recrystallization. The thermal softening is no longer considered to be the main interpass softening mechanism at a low strain rate. The interval time has a negligible eﬀect on the stress, but the signiﬁcant changes in grain size and texture component are caused by the interpass softening. The average grain size is approximately 40 μ m, and the distorted grain boundaries and small ﬁne grains are found in the interval times of 0.5–5 s, implying the dynamic recovery and grain growth. The near {001} < 110 > and {110} < 112 > orientation exerts an important inﬂuence on the grain reﬁnement.


Introduction
42CrMo alloy is widely used in the production of ring components such as bearing and flange due to the high strength, excellent fracture toughness, and abrasive resistance [1,2]. e 42CrMo ring parts are manufactured by the radial-axial hot ring rolling (RAHRR). e RAHRR is a complex process with the characteristics of nonisothermal, multipass, and accumulative deformation and so forth. e wall thickness of ring blank is gradually reduced with a progressively narrowing gap between the driven roll and mandrel. e height is controlled by feeding the upper axial roll towards the lower axial roll [3]. e continuous expansion in diameter and shape in cross section is produced.
Extensive studies on the multipass deformation behavior based on multiroll continuous rolling and multidirection forging have been conducted using compression test [4][5][6][7][8][9]. e isothermal and nonisothermal multipass compression tests of 7150 aluminum alloy were carried out, and the deformation behavior, constitutive relations, and microstructure evolutions were analyzed in detail [4]. e microstructure evolution of 2519 aluminum alloy was studied using isothermal multipass compression and verified that the recrystallization fraction increased with increasing pass numbers [5]. e stress decreased with increasing finishing temperature during 7-pass compression [6], and the subgrains transformed into the grains with straight grain boundaries. e effect of grain size on mechanical properties of CrMo thick plate was predicted by multipass deformation [7]. e evolution of grain size of F40 ship plate steel was also studied by multipass deformation and indicated that the degree of grain refinement increased with increasing pass numbers [8].
e mechanisms of grain refinement in Cu-Cr-Zr alloy at different multiaxial forging temperatures were revealed, and then the grain orientation was determined [9]. However, there are few studies focusing on the multipass deformation in RAHRR, especially the rolling based on as-cast ring blank. e as-cast ring blank is hotrolled directly to the required geometrical dimension, and both the thermoplasticity and mechanical properties are improved in RAHRR. e process has some advantages including saving material and energy, reducing times of heating, and low emission [10]. e dynamic and static recrystallizations occur in radial and axial roll gaps, while the metadynamic recrystallization plays a dominant role outside deformation areas in RAHRR. e metadynamic recrystallization behavior of 42CrMo alloy was studied to reveal the response of grain refinement to deformation parameters [11]. e effects of temperature and interval time on the microstructure of as-cast 42CrMo alloy were also clarified using interrupted isothermal compression [12]. In addition, for industry multipass processing, the nonisothermal process is always undergone with gradually declining temperature [4,13]. e interaction between hardening and softening effects of as-cast 42CrMo alloy in RAHRR presents the complicated microstructure and texture evolution. e microstructure and texture are affected by the discontinuous characteristics in deformation intervals during multipass rolling [14][15][16]. e grain refinement and size distribution are also closely related to the texture evolution. e softening behavior under interpass deformation results in some distinct differences when compared with that under single-pass continuous compression.
us, it is critical to clarify the mechanisms of stress softening and grain refinement in discontinuous multipass compression first.
In this study, the nonisothermal multipass compression of as-cast 42CrMo ally was performed on a Gleeble-3500 thermal simulator, and the effects of finishing temperature, strain rate, and interval time on the stress, microstructure, and texture were studied. e corresponding constitutive model was established. e softening behavior and microstructure evolution mechanism were clarified in detail by electron backscatter diffraction (EBSD) and transmission electron microscopy (TEM) techniques. e results will provide new insights for the microstructure control and parameter optimization in RAHRR.

Material and Experimental Procedure
A commercial 42CrMo alloy with the main chemical composition (wt.%) of 0.44 C, 0.28Si, 0.72Mn, 1.13Cr, 0.22Mo, 0.012P, 0.007S, 0.15Ni, 0.12Cu, and (bal.) Fe was used in this study. e as-cast ring blank with the dimensions of 240 mm in outer diameter, 120 mm in inner diameter, and 45 mm in height was prepared by sand-casting process [17,18]. In the sand casting, a pouring temperature of 1530-1540°C and a pouring rate of 16.8-18.5 kg/s were selected. e average grain size is approximately 87 μm, as shown in Figure 1. Cylindrical specimens with a diameter of 10 mm and a height of 15 mm were machined from the same radial positions of the ring blank to guarantee the uniformity of microstructure [19]. e nonisothermal multipass (5 passes, continuous cooling) compression tests were carried out on a Gleeble-3500 thermal simulator at the initial temperature of 1150°C. Figure 2 shows the detailed experimental processes of hot compression tests. Before compression, all specimens were heated to 1150°C at a heating rate of 10°C/s and soaked for 2 min to eliminate thermal gradients. e finishing temperatures of 1070°C, 1030°C, and 990°C and the corresponding temperature decrements of 20°C, 30°C, and 40°C in interpasses were applied, respectively. e true strain of 0.183 in each pass and interval times (tp) of 0.5 s, 1 s, and 5 s are used. e strain rates were 0.1 s − 1 and 1 s − 1 . e specimen numbers and detailed deformation parameters are listed in Table 1.
e nonisothermal multipass compression tests are similar to the methods of Qin et al. 2016 [19], but the strain rate and interval time are selected as the variables to study their effect on the flow behavior in this study. e graphite foils with a thickness of 0.5 mm were positioned between the specimen and anvils to reduce the contact friction [20]. A thermocouple was welded to the center of each sample to allow the temperature to be tracked and the load-stroke data to be collected with accuracy [21]. en, the specimen was compressed at the selected parameters. To retain the deformed microstructure, all the specimens were immediately water-quenched to ambient temperature.
True stress-strain curves were recorded automatically using the standard equations from the load-stroke data 100 μm Figure 1: e initial microstructure of as-cast 42CrMo ring blank [18].  Advances in Materials Science and Engineering measured in these compression tests [22]. e specimens were sectioned parallel to the longitudinal compression axis. e section was ground, polished, and etched in a solution of saturated picric acid. e microstructures were observed by VHX-600E optical microscopy. e average grain sizes were measured using the line intercept method. e texture was determined by EBSD system equipped at a FEI Nano 430 field-emission scanning electron microscopy (SEM). e EBSD specimens were electropolished in a solution of 5% perchloric acid and 95% ethyl alcohol at 30 V and 20°C. in foils for TEM were prepared by mechanical polishing to a thickness of about 60 μm followed by double-jet electropolishing in a solution of 10% perchloric acid and 90% acetic acid at 10 V and − 25°C. TEM observations were performed on a FEI Tecnai G2 F20 transmission electron microscopy operating at 200 kV.

Multipass Flow Behavior.
According to the results of isothermal compression [10,12,23,24], nonisothermal multipass compression tests were carried out to simulate industry multipass processing more precisely. Figures 3(a)-3(c) show the true stress-strain curves and temperature curves of as-cast 42CrMo alloy at different finishing temperatures. e stress increases with increasing strain rate, while it decreases with increasing finishing temperature at a given initial temperature. ey imply that the stress is sensitive to the strain rate and finishing temperature during nonisothermal multipass deformation. Rising stress curves and decreasing pass temperature across the cumulative strain with increasing passes number are all presented in Figure 3. e rise in stress is not only the declining temperature effect but also the more prevalent retained deformed regions [25]. In addition, at the same finishing temperature and strain rate, the stresses in each pass do not show significant change with the interval time. However, under diminishing temperature conditions with the constant pass true strain of 0.183, the interpass softening is enhanced at the interval time of 5 s. is result is dissimilar to that under the isothermal multipass deformation; that is, longer interval time causes less softening [21]. e flow softening during nonisothermal multipass deformation is the net result of microstructure softening by recrystallization and hardening by accumulative strain; therein, the flow hardening is due to the decrease in temperature. At the high strain rate of 1 s − 1 , the decreasing stress occurs after peak stress in each pass, which might be strongly depending on the thermal softening [26]. e variations of peak stresses with finishing temperature at interval time of 1 s are compared under the same pass number, as shown in Figure 3(d). It can also be found that, for a constant strain rate, the peak stress increases gradually with increasing passes number. e difference in peak stress of different finishing temperature is inconspicuous at the smaller passes number (≤2), but it increases with increasing passes number. For example, under the strain rate of 0.1 s − 1 and finishing temperatures of 990°C and 1070°C, the difference value of peak stress increases from 0.8 MPa at 1 pass to 22 MPa at 5 passes. It is mainly due to the fact that the static recrystallization (SRX) becomes weakened during the interpass. e peak stress in each pass rises compared with the one under the true stain of 0.183. e mixed microstructures and new grains in all passes are similar to those in the isothermal conditions. But the shapes of the nonisothermal stress curves considerably differ from those under isothermal stress curves [26].

Constitutive
Characteristics. Nonisothermal multipass stresses have been used to derive the constitutive models, which have been further applied to multistage simulations of a declining temperature rolling schedule [4,[27][28][29]. e effect of temperature and strain rate on the stress is described by the Zener-Hollomon parameter (Z) according to the following equations [30,31]: where σ is the stress (MPa); _ ε is the strain rate (s − 1 ); Q is the activation energy (kJ/mol); T is the absolute temperature (K); R is the universal gas constant (8.314 J/mol·K); A, A 1 , and A 2 are the material constants (s − 1 ); β and α are also the material constants (MPa − 1 ); n and n 1 are the stress exponents, and α is calculated as α � β/n 1 . Equations (1)-(3) are suitable for a low stress level (ασ < 0.8), a high stress level (ασ > 1.2), and a wide range of stress level, respectively. en, β and n 1 are taken as the average values of the slopes of the ln_ ε versus σ plots and ln _ ε versus ln σ plots at a series of temperatures, respectively. According to our previous results [24], α � 0.00785 MPa − 1 is adopted in this study. e relationship between stress and strain rate is given by substituting equation (1) into (3): e natural logarithm and partial differential of both sides of equation (4) are performed, respectively, to derive the values of n and Q:   According to the stress curves during nonisothermal multipass compression, n and Q are derived by equations (5) and (6). n is the average slope of ln _ ε Figures 4(a) and 4(b). e values of n and Q are calculated as 8.1718 and 441.198 KJ/mol, respectively. In addition, taking the natural logarithm of both sides of equation (3), A is expressed as follows:

versus ln[sin h(ασ)] plots and S is the average slope of ln[sin h(ασ)] versus (1/T) plots, as shown in
ln Z � ln A + n ln sin h(ασ).

Microstructural Characterization and Texture Evolution.
Except for the difference in the stress curves, the microstructure is the main characteristic of the nonisothermal deformation over the isothermal process [32]. Figure 5 shows the microstructures of as-cast 42CrMo alloy at different nonisothermal multipass deformation conditions. At the high finishing temperature of 1070°C and interval time of 0.5 s, the large average grain sizes are found regardless of strain rate. In these conditions, the grains approximately 73 μm in diameter at the strain rate of 0.1 s − 1 are slightly larger compared with those at the strain rate of 1 s − 1 , as presented in Figures 5(a)  Advances in Materials Science and Engineering 5 e DRX occurs at the interval time of 0.5 s, which results in the finer and homogeneous distribution grains with the average size being approximately 40 μm. In addition, the DRV and grain growth are dominated at the interval times of 1 s and 5 s. e distorted grain boundaries and small fine grains are found in Figures 5(d) and 5(e). is may be related to the larger grains growing with high activity at the expense of suppressing the smaller DRX fine grains. It also can be seen from Figure 5(f ) that the microstructure is similar to that in Figure 5(c). e low finishing temperature of 990°C leads to the more fine and homogeneous grain approximately 35 μm in diameter when compared with that under the finishing temperatures of 1030°C and 1070°C.
In the nonisothermal multipass compression, the grain refinement and texture evolution are affected by the finishing temperature, interval time, and strain rate. EBSD data was analyzed with postprocessing TSL-OIM software. e different textures extracted from φ 2 � 45°sections of orientation distribution functions (ODFs) were discussed. Figure 6 shows the texture evolution of as-cast 42CrMo alloy under different nonisothermal multipass compression conditions. At the finishing temperature of 1070°C and interval time of 0.5 s, the texture is mainly composed of {113} e results are consistent with the fine and homogeneous distribution in grain size, as presented in Figure 5(f ).
Underlying the mechanical behavior is the microstructure evolution, which depends on the mechanisms and controls the final mechanical properties [27]. In RAHRR, the accumulated deformation is almost the result of dislocation motion, leading to considerable interpass softening in grain boundaries sliding and strain hardening in the elongated grains. e refined grains with a high dislocation density are found in Figures 5(f ) and 7(a). e microstructure is characterized by an unstable state with high dislocation density and serious lattice distortion. As the finishing temperature increases, the stored energy is consumed by the softening effect during the interpass. e nucleation rate of recrystallization decreases because of decreasing energy and driving force. However, the grain size increases gradually with increasing finishing temperature. It can be seen from Figures 7(a) and 7(b) that the low angle grain boundaries (LAGBs) transform into high angle grain boundaries (HAGBs) through absorbing dislocations, resulting in the decrease in dislocation density. It can be concluded that rising temperature increases the DRV and recrystallization rate and thereby decreases the density of dislocations and hence reduces the driving force for recrystallization [25,33]. However, it is evident that its opposing effect of accelerating the rearrangements was involved in nucleation and migration of grain boundaries [25,[34][35][36]. ese results are closely related to the stress softening at the high strain rate of 1 s − 1 , as shown in Figure 3. e grains grow through the migration of dislocations and HAGBs. e dominant softening mechanisms are attributed to the combined effect of thermal softening, DRV, and DRX. As the strain rate decreases to 0.1 s − 1 , the dislocation density is further reduced and the fraction of HAGBs is increased, as presented in Figure 7(c). However, the thermal softening is no longer  Advances in Materials Science and Engineering 7 considered to be the main interpass softening mechanism at a low strain rate of 0.1 s − 1 . Moreover, previous multistage testing shows that as the interval is decreased, softening declines so that the stress becomes almost coincident with the single-stage continuous deformation for isothermal tests and for declining temperature tests [25,37]. is arises in the stress within the interval time of 0.5 s preventing any static restoration [38].

Conclusions
e nonisothermal multipass deformation behavior of ascast 42CrMo alloy was studied and the corresponding constitutive model was derived. e softening behavior and microstructure evolution mechanism were clarified in detail.
e main conclusions are drawn as follows: (1) As the finishing temperature increases from 990°C to 1070°C, the stress decreases gradually and the softening effect increases, which results in a large grain size and inhomogeneous microstructure. e LAGBs transform into HAGBs through absorbing dislocations.
(2) e stress softening in a high strain rate is attributed to the thermal softening, DRV, and DRX. e thermal softening is no longer considered to be the main interpass softening mechanism at a low strain rate. e near {001}<110> and {110}<112> orientation exerts an important influence on the grain refinement.
(3) e interval time has a negligible effect on the stress, but the significant changes in grain size and texture component are caused by the interpass softening. e average grain size is approximately 40 μm and both the distorted grain boundaries and small fine grains are found in the interval times of 0.5-5 s, implying the DRV and grain growth.

Data Availability
e experimental data used to support the findings of this study are included within the article.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.