- Email: [email protected]
Hot compression deformation characteristics and microstructural evolution of a Co–Cr–Mo–C alloy: Effect of precipitate and martensitic transformation
Author’s Accepted Manuscript Hot compression deformation characteristics and microstructural evolution of a co-cr-mo-C alloy: Effect of precipitate and martensitic transformation F.Z. Hassani, M. Ketabchi, G.R. Ebrahimi, S. Bruschi www.elsevier.com/locate/msea
PII: DOI: Reference:
S0921-5093(16)30091-0 http://dx.doi.org/10.1016/j.msea.2016.01.084 MSA33271
To appear in: Materials Science & Engineering A Received date: 20 October 2015 Revised date: 22 January 2016 Accepted date: 26 January 2016 Cite this article as: F.Z. Hassani, M. Ketabchi, G.R. Ebrahimi and S. Bruschi, Hot compression deformation characteristics and microstructural evolution of a co-cr-mo-C alloy: Effect of precipitate and martensitic transformation, Materials Science & Engineering A, http://dx.doi.org/10.1016/j.msea.2016.01.084 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Hot compression deformation characteristics and microstructural evolution of a Co-Cr-Mo-C alloy: Effect of precipitate and martensitic transformation F.Z. Hassaniª, M. Ketabchi*ª, G.R. Ebrahimib, S. Bruschic ªMining and Metallurgical Engineering Department, Amirkabir University of Technology (Tehran Polytechnic), Tehran, Iran b
Department of Materials and Polymer Engineering, Hakim Sabzevari University, Sabzevar,
9617976487, Iran c
Department of Industrial Engineering, University of Padua, Via Venezia 1, 35131 Padua, Italy
*Corresponding author: [email protected] Tel: +989132005610; Fax: +982166405846 Postal address: Amirkabir University of Technology (Tehran Polytechnic), Hafez Street, Enghelab Avenue, Tehran, Iran.
Abstract The hot compression behavior of a Co-Cr-Mo-C alloy was studied in the temperature range of 1100-1200 ˚C and the strain rate of 0.001-1 s-1. The influence of processing parameters including temperature and strain rate on the microstructure of the alloy was investigated. The flow curves of the hot compressed samples at the investigated deformation temperatures and strain rates exhibited single peak stress. The effect of the strain rate and temperature of deformation on the 1
flow stress was studied using a hyperbolic sine-type equation. The calculated activation energy was about 744.5984 kJ/mol. The evidenced microstructural characteristics combined with the shape of the flow curves indicate that there is a competition between dynamic recrystallization (DRX) and precipitation phenomena, especially at low strain rates, i.e. 0.001 s-1. The formation of precipitates at the grain boundaries prevent the mobility of boundaries and also cause the partial inhibition of the dynamic recrystallization. Whereas, long time for the precipitates coarsening causes the predomination of dynamic recrystallization. Thus, the serrated boundaries in microstructure along with oscillation in the flow curves take place due to this competition. Keywords: Cobalt-base alloy; Hot deformation; Recrystallization; Precipitates, Martensitic transformation. 1. Introduction Co-Cr-Mo alloys are usually used to produce artificial joints thanks to their excellent biocompatibility, wear and mechanical properties [1-4]. These alloys are commonly used in the as-cast condition, where they exhibit some main defects, such as shrinkages, coarse grain size and chemical heterogeneity [5]. Therefore, their mechanical properties in the as-cast condition are extremely low due to the distribution of these defects. As an example of possible consequences, low mechanical properties can cause the early failure of the implant device [1]. The duplex structure of Co-Cr-Mo alloys consists of a metastable γ (fcc) phase and a ε (hcp) phase. The metastable γ phase at room temperature could experience a martensitic transformation γ (FCC)→ε (HCP), which is induced by quenching from the stable γ (FCC) phase (above 1100˚C) [6]. The martensitic transformation, however, results in poor workability in case of cold deformation [7]. Therefore, it is necessary to eliminate the alloys casting defects and
2
enhance their mechanical properties for practical applications by applying proper thermomechanical treatments [8]. The improvement of the mechanical properties can be obtained by applying plastic deformation combined with a subsequent recrystallization heat treatment. Whatever the procedure applied to improve the alloys mechanical properties, a clear understanding of the microstructural changes occurring during thermo-mechanical processing is mandatory [8]. Dynamic recrystallization (DRX) as a softening mechanism plays a significant role in thermo-mechanical processing [9], besides having an important influence on the microstructural evolution [10, 11]. The process parameters, namely temperature, strain rate and strain, have strong effect on the degree of DRX during the microstructural evolution [9]. Moreover, the competition between recrystallization and precipitation is the common mechanism in the materials which are predisposed to precipitation of second phase [10]. So, due to presence of strong carbide formers, such as the Cr and Mo, these alloys present a reasonable potential for precipitation to occur. In biomedical Co-Cr-Mo alloys the effect of hot compression parameters on the grain refinement with low percent of carbon (0.02%) was reported [12, 13]. However, further research has not been conducted yet to characterize the effect of precipitates and hot compression parameters on the microstructural evolution of Co-Cr-Mo-C alloys characterized by a high percent of carbon (0.35%), which causes indeed the increase of the stability of the γ phase and also the formation of carbides [14]. Therefore, the purpose of this study is to disclose the microstructural and mechanical properties of Co-Cr-Mo-C alloy at different strain rates and temperatures and also determining the competition between DRX and precipitation of second phase. Thus, in this study, the effects of thermo-mechanical parameters, namely strain, strain rate, and temperature, on both the mechanical and microstructural characteristics of a Co-Cr-Mo-C alloy characterized 3
by a high carbon content is evaluated. Moreover, the competition between DRX and second phases precipitation is discussed as a function of the temperature change in the hot compression tests. Finally, the constitutive relationship to predict the flow stress behavior is assessed and the related coefficients identified. 2. Materials and methods 2.1 Material The composition of the Co-Cr-Mo-C alloy object of the present study is given in Table. 1. Fig. 1a shows the microstructure of the alloy in the as-received condition. The XRD pattern indicates a single γ phase in the microstructure (Fig. 1b), proving that the martensitic transformation did not occur. The average grain size of the γ phase was approximately 52 μm. 2.2 Hot compression tests Cylindrical compression specimens (with a diameter of 7.5 mm and a height of 13.4) were prepared for the hot compression tests. All the tests were performed on a Zwick/Roell 250
adopting a final true strain of 0.8 at temperatures ranging from 1100°C to 1200 ̊C, at different strain rate of 0.001, 0.01, 0.1 and 1 s-1. Before the compression, the specimens were heated to 1200 ̊C at a rate of 5 ̊C s-1 and held at this temperature for 300 s. Then they were cooled to the
deformation temperature with the cooling rate of 1; C/s. After soaking for 300 s at the testing temperature, continuous hot compression tests were carried out, keeping the deformation temperature constant. At the end of the deformation, the specimens were immediately quenched in water to retain the high temperature deformed microstructure. 2.3 Characterization methods
4
The deformed specimens were sectioned in the parallel direction with respect to the compression axis, and polished using standard metallography techniques. The polished surface was then etched at room temperature by immersion in a 5:1 solution of HCl:H2O2. In order to specify the carbide phases the polished samples were also stain etched by immersion in a staining solution containing one part of 20 vol % potassium permanganate in distilled water and one part of 8.0 vol % sodium hydroxide in distilled water [15]. The microstructural observations were performed by using a Field Emission Scanning Electron Microscope (FESEM). And the secondary electron detector (Everhart-Thornley: ETD detector) at different magnification was used to detect the pictures. The microscope were also equipped with Energy Dispersive X-ray (EDX) spectroscopy to provide more accurate characterization. For the EDX analysis, the working distance of sample was set-up between 9.5 to 10 mm. The size of the grains was determined by using image tool software. The identification of the alloy phases was performed using the X-Ray Diffraction (XRD) technique (a Cu Kα radiation λ=1.54184 A;
in a Philips
machine) and 2ɵ scans between 40˚ and 60˚ with step size of 0.02 2ɵ S-1. The samples for XRD measurements were prepared by conventional mechanical grinding and polishing. The volume fraction of the athermal ε-martensite was determined according to the method of Sage and Gullaud by measuring the integrated intensities of the principal peaks of the hcp phase [16-18] (see equation 1). ℎ"#($%%) =
'(101*1)+,'(101*1)+,- + 1.5'(200)3,,
(1)
Results and discussion 3.1 Flow characteristic
5
Typical true stress–strain curves obtained from hot compression tests of Co-Cr-Mo-C alloy at various deformation temperatures and strain rates are shown in Fig. 2. As expected, the flow stress increases with strain rate and decreases with temperature. Three stages of behavior may be observed in the flow curve, namely work hardening, flow softening and steady state [19]. The work hardening is the predominant phenomenon at the initial stages of deformation: the rapid multiplication of dislocations causes indeed the flow stress to increase [9, 20, 21]. When the strain reaches to critical value (εc), the dynamic restoration, becomes the predominant mechanism and starts partially offsetting the work hardening effect. At higher strains when there is a balance between the DRX and precipitation, a steady-state regime is obtained. All the flow curves exhibit one peak, with the peak strain and stress decreasing at increasing temperature and decreasing strain rate. Fig. 3 shows the variation of the peak stress as a function of the testing temperature at different strain rates. Higher generation rate of dislocations are due by higher strain rate, caused by the promotion of nucleation sites for new nuclei rather than the growth of previously formed grains. Moreover, the density of dislocations increases with rising strain rate [22] and the mobility of High-Angle Grain Boundaries (HAGBs) decreases with increasing strain rate. However, hot deformation is a thermally activated process [20], therefore, the migration of grain boundaries and movement of dislocation is accelerated at increasing temperature. This means that the work hardening rate is lower at higher temperatures, and the stress increases in the first stage at a lesser rate than at lower temperatures. Therefore, the work hardening rate and strengthening effect due to the dislocation interaction gradually decrease at increasing temperature and decreasing strain rate.
6
Some oscillations can be observed at the lowest strain rate, i.e. 0.001 s −1. This is due to the competition between DRX and precipitation [22, 23]. Carbides formation may be caused by the pinning effect on the grain boundaries that prevents the DRX to occur [22]. When the curved grain boundary tries to release from carbides, the flow stress increases, before reaching a plateau. In fact the precipitates growth when they reach their critical radius or dissolve when they reach a suitable temperature causing the stress to decrease. Thus, these plateaus can be attributed to the stress relaxation due to the continuous growth or dissolution of the precipitates. 3.2 Microstructural observations Fig. 4 shows the microstructure of the specimens subjected to the hot compression at 1100 ̊C,
1150 ̊C and 1200 C ̊ at a strain rate of 0.1 s-1.
As expected, the microstructures changes with the testing temperature. In fact, strong carbide formers present in the alloy composition, such as the Cr and Mo, cause the competition between precipitation and DRX. A larger amount of precipitate phase can be observed in the compressed specimens than in the as-received material, proving that the degree of deformation play a role in the competition between DRX and precipitation. The precipitation causes the pinning effect on the HAGBs [10, 22]. Moreover, precipitation and the mobility of boundaries are diffusioncontrolled. Thus, the frequency of competition between mobile boundaries and precipitates increases at increasing deformation temperature [24]. SEM and EDS micro-chemical analyses were carried out to confirm the presence of carbides. Fig. 5 shows the EDS micro-chemical maps of the specimen deformed to 55% (ε=0.8) at 1100 ̊C and strain rate of 0.01 s-1. The white particles in the SEM image are distributed inside the grains and also along the grain boundaries. By analyzing the distribution of some elements in composition mapping, they were Cr, Mo and
7
C enriched and were Co depleted. It can be seen the segregation of Cr, Mo and C at the grain boundaries and inside the grains where the precipitates exactly were existed. It means that a portion of carbon precipitated at grain boundaries and inside of the grains instead of resolving into the matrix. For more clarification the chemical composition of the precipitates and matrix are also reported in Table 2. The EDS analysis of the precipitates shows the segregation of Cr and C: precipitates are found to be enriched in Cr and C when compared with the Co-matrix. In the results reported by L.E. Ramırez et al [25] the main secondary phases in this alloy are identified as M(M: Cr or Mo element)23C6 carbides. Other studies [26, 27] identified M6C carbide as minor constituents. Shin Go Minet et al [28] reported that the precipitates detected in Co-Cr-Mo-C alloys depended on the carbon content as well as the temperature of the heat treatment. Carbon causes the stabilization of M23C6-type carbides in Co-Cr-Mo-C alloys [29]. Grain boundaries and also stacking faults provide appropriate sites for the precipitation of M23C6 carbides [30, 31], and M23C6 carbides are primarily observed in the Co-Cr-Mo-C alloys heat treated from 1175°C to 1225 C ̊ [32].
Fig. 4a shows the presence of precipitates at the grain boundaries and inside the grains caused by the retardation of the DRX occurrence. At higher temperature (Fig. 4b) the pinning effect of the precipitates on the mobile boundaries is decreased due to the dissolution by the diffusion mechanism and also increment of their size. Therefore, at this temperature, DRX will occur more easily. As can be seen in Fig. 4, the percentage of carbides was reduced at increasing temperature. The amount of grain boundary carbides and also fine blocky carbides inside the
grains decreased in the samples hot compressed at1200 ̊C compared to the ones deformed at 1100 ̊C. For the Co-Cr-Mo alloy with a high percent of carbon, 1200°C is the starting
temperature for M23C6 carbides dissolution [33, 34]. However, the dissolution of M23C6 depends 8
on the temperature as well as on the holding time. Thus, at 1200 ̊C for 600 sec in this study, a partial carbide dissolution happened. Moreover, the carbides shape changed from an elongated to a globular shape at increasing temperature. This confirms the study of Taylor and Waterhous
[33] who reported that the temperature of 1150 ̊C is the temperature at which the carbides morphology changes. EDS analysis of carbides and matrix at 1100°C and 1200°C are shown in Fig. 6 pointed out by red circle. According to this analysis, the carbon atomic percent of the grain boundary carbides
and the ones inside the grains at 1100 ̊C (Fig. 6a and b respectively) is lower than at 1200 ̊C (Fig.
6c) and also the carbon percent in the matrix at 1100 ̊C (Fig. 6d) is higher than at 1200 ̊C (Fig. 6e). Thus, it can be concluded that most of the carbides at 1100 ̊C are M23C6-type and most of them at 1200 ̊C are M6C-type. This is in accordance with the results of Weeton and Signorelli
[35] who worked on the Co-Cr-Mo-C alloy with high percent of carbon (0.29) that pointed out that M23C6 transform to M6C during heat treatment. For more clarification the stain etching was also performed. Based on the stain etching at 1100˚C the volume percent of M23C6 carbides (brown particles) was more than M6C carbides (red, green or blue particles).Moreover according to the time-temperature-precipitate diagram pointed out by Taylor and Waterhouse [34],1200 ˚C is the dissolution temperature of M23C6 carbides. Wimmer et al [31] pointed out that the M23C6 carbides precipitate at the grain boundaries and inside the grains but, the M6C ones exist only inside the grains. The results here reported also confirm the presence of M23C6 carbides at the grain boundaries especially at lower temperatures. However, the amount of M23C6 carbides at the grain boundaries is less at 1200 rather than at 1100°C. The M23C6 carbides tend to transform to M6C carbides at temperatures in the range between 1165°C and 1230 ̊C [36]. Thus, it seems that the most of the carbides (the grain 9
boundary and also fine blocky ones) at 1100 ̊C are M23C6 phase which transformed at the higher temperature to M6C according to the reaction of M23C6+M→ M6C+γ. Therefore, most of the carbides at 1200°C are M6C carbides that exist inside the grains. This is due to the above mentioned transformation and also because of the promotion of formation of M6C by increasing the thermal treatment temperature [36]. Moreover, as previously mentioned, the temperature of 1200°C is the appropriate one for the dissolution of M23C6 carbides. And a lot of grain boundary M23C6 carbides were dissolved at 1200 ̊C. Again, the presented results are in good agreement
with literature works, as the one by Clemow and Daniell [36] who stated that the M6C carbides are a stable phase in the Co-Cr-Mo-C alloys. As illustrated in Fig. 4a, new grains smaller than the starting ones are formed at the grain boundaries, with the necklace DRX being in progress. The original grains of the specimens subjected to the hot compression at 1150 ̊C are replaced by new grains, which are equiaxed ones
with sharp boundaries is characteristic of a completed DRX process [10]. It is clearly observed that the average grain size compared to the starting material is significantly decreased after hot
compression. At lower temperatures (i.e. 1100 ̊C) as can be seen in SEM image of Fig. 5 pointed out by a yellow circle, the refined grain boundaries are trapped by the carbides. At increasing temperature (Fig. 4c), coarsening of the new grains occurs: this is due to the dissolution of some precipitates as previously mentioned. Thus, their effect on the mobility of grains was decreased. It means that the recrystallized grain size increased at increasing the deformation temperature. The final grain sizes of the specimens hot compressed at strain rate of 0.1 and temperatures of 1150 ̊C and 1200 ̊C were about 9.5 and 24.4 μm, respectively, both smaller that the initial grain size (52 μm).
10
The effect of the strain rate on the microstructure of the specimens deformed at 1200 ̊C is shown in Fig. 7: all the specimens exhibit fully recrystallized γ-phase microstructures with equiaxed grains, except the one deformed at strain rate of 0.001 s-1 (Fig. 7a). It can be seen from the XRD patterns of the specimens deformed at different strain rates in Fig. 7 that the diffraction peaks (101*1) and (101*0) of hcp ε-phase appear at 2ϴ values of 46.65˚ and
41˚, respectively. The intensity of the (101*1) peaks at 0.001 s-1 is higher than that ones at the other strain rates. The values of the martensite volume fractions (fHCP) at different strain rates in 1200˚C are reported in Table 3 calculated according to the Sage and Guillaud [6, 16, 17, 37] method. This is athermal martensite produced after quenching as a consequence of the rapid cooling from the annealing temperature (T˃1100˚C) [38]. The developement of the athermal
martensite produced after quenching is dependent on the grain size [17, 39], increasing at increasing grain size [17, 37, 39, 40]. However, the transformation is limited and the volume fraction of martensite does not proceed beyond of 0.4-0.5 [6]. The grain size of hot deformed Co-Cr-Mo-C alloy at temperature of 1200 ˚C and different strain rates are given in Table 3. As can be seen, higher volume fractions of athermal hcp martensite at the lowest strain rate applied in this study is due to the larger grain size of alloy compared to the ones obtained at higher strain rates. The γ→ε transformation leads to greater work hardening [41] due to less active slip systems compared to the γ phase, which makes difficult the plastic deformation [42]. The alloy low Stacking Fault Energy (SFE) due to the ε martensite transformation plays a crucial role in enhancing the work hardening. As it is known, a low SFE causes difficulties in the stress relaxation by cross slip. Therefore, a higher work hardening of Co-Cr-Mo-C alloy at strain rate
11
of 0.001 s-1 rather than at the other strain rates, due to the formation of ε-martensite, leads to higher rate of flow hardening. Therefore, the rate of hardening compensates the rate of softening, a little amount of flow softening is produced and a steady state flow appears more quickly (Fig. 2c). The serrated boundaries are observed at the lowest strain rate, indicated by arrows in Fig. 7a. At intermediate strain rates, i.e. 0.1 s-1 (Fig. 7c), the amount of serrated boundaries decreases remarkably. These changes in the grain boundaries are due to the competition between DRX and precipitation. At the highest strain rate (Fig. 7d), the time for precipitation is too short and the driving force for DRX is higher. Therefore, DRX is the dominant mechanism [24]. Whereas, at the lowest strain rate there is enough time for the precipitation occurrence. But the long time for coarsening of new precipitates also causes lower pinning effect at the grain boundaries in some regions. Therefore, the serrated boundaries appear, which is the consequence of the continuous competition between DRX and precipitation at different regions of the grain boundaries. In fact, long time for the precipitates coarsening before starting the second cycle of precipitation causes the carbides release from the boundaries. Moreover, there is enough time for the beginning of subsequent cycles of fine precipitation. These fine carbides, which are precipitated, cause the binding of the mobile boundaries. These processes are repeated more times at lower strain rates at different regions of the grain boundaries, resulting in the serrated grain boundaries. When the strain rate is increased, the DRX grains become finer, as a consequence of the shorter time at higher strain rate needed for growing up the grains. 3.3 Constitutive equation of the flow stress
12
The material flow stress was described through constitutive equations. The base equation used in this study was proposed by Zener and Hollomon [43], who introduced the so-called ZenerHollomon parameter as a function of the strain rate and temperature according to equation (2). Z = ε̇ exp 8 <=F(σ) 9
(2)
:;
Where Q is an activation energy (KJ/mol), ε̇ is strain rate (s-1), (R=8.31 J/K mol) is the molar gas constant, T is the absolute temperature (K), σ is the flow stress (MPa). In hot working processes, several constitutive equations have commonly been applied as bellow [44].
F(σ)= A΄σn΄
ασ<0.8
(3)
F(σ)= A΄΄exp(βσ)
ασ>1.2
(4)
F(σ)= A( sinh(ασ)n)
For all σ
(5)
In which A, A΄, A΄΄, α, β, n and n΄ are the material constants, α=β/n. by substituted Eq (3), and Eq (4) and Eq (5) in the Eq (2), and taking the natural logarithm the following equations will be achieved. Lnε̇ + Lnε̇ +
Q = nA΄ + n΄ ln ?RT Q = lnA΄΄ + β?RT
Lnε̇ + :; = lnA + nln(sinh(α?D )) 9
13
ασ<0.8
(6)
ασ>1.2
(7)
For all σ
(8)
For obtaining the values of n΄, β and n the following equations can be used, which are the slope of lines of equations (6-8). É = G
HIJK̇
HIJMN
P=G E=G
HIJK̇ HMN
OT
OT
HIJK̇
(9) (10) OT
(11)
HIJ(SUJV(WXY ))
The graphs of lnε̇ vs lnσP and lnε̇ vs σP at different temperatures are shown in Fig. 8 a and Fig. 8b, respectively. The following mean values of n΄ and β are obtained: 5.7591 MPa-1 and 0.0525 MPa-1 respectively. Then α=β/n΄ is identified as 0.0091 MPa-1. Fig. 8c shows the lnε̇ vs ln(sinh(α?D )) plot. From the slope of Fig. 8c and substituting the value of α, the value of n is
obtained as 4.1970 MPa-1s-1
The Q parameter can be identified from the equilibrium (12-14) as follows: ^_E?D [ = \E΄ ] a 1 ^ 8`< ḃ [ = \P ]
(12)
^?D a 1 ^ 8`< ḃ
(13)
^ ln(sinh(c?D )) [ = \E ] a 1 ^ 8`< ḃ
(14)
The lnσP vs (1/T), σP vs (1/T) and ln(sinh(ασP)) vs (1/T) plots are shown in Fig. 9a, Fig. 9b and Fig. 9c, respectively. The obtained average values of Q are: 707.0674, 744.5984 and 728.3985 (kJ mol-1) from the Eqs. (12-14). By considering the correlation coefficient (R2), from the Fig. 14
9a-c, it can be considered that the Fig. 9b or Eq. 13 shows a better fitting with the experimental data. Therefore, it can be stated that the activation energy of the studied Co-Cr-Mo-C alloy is about 744.5984 (kJ mol-1). The Zener–Hollomon parameter for the investigated Co-Cr-Mo-C alloy can be estimated by Eq. 2 and is represented as follows: 744.5984 d = ḟ exp g q = 3.9199 − 13"#$.%$&' \`
(15)
744.5984 ) = *̇ exp , 0 = 1.5924 − 4 exp(0.0525 "# ) -/ 744.5984 ) = *̇ exp , 0 = 0.0388[sinh(0.0093"# )]<.'&%> -/
(16) (17)
The lnZ vs lnσP, lnZ vs σP and lnZ vs ln(sinh(ασP)) were drown. The correlation coefficients (R2) were 0.9269, 1 and 1, respectively. By considering the R2, it can be stated that the Eqs. 16 and 17 show the best agreement with the experimental data. Moreover, as previously was mentioned, Eq. 13 which was mainly extracted from Eq. 4 was better fitted with experimental data. Therefore, the hot deformation equation of Co-Cr-Mo-C alloy over the temperature range of 1100-1200 ̊C can be written as follows. −744.5984 *̇ = 1.5924 − 4 exp , 0 exp(0.0525 "# ) -/
(18)
4. Conclusions The stress-strain curves of the Co-Cr-Mo-C alloy investigated in this study exhibite the typical DRX behavior with a single peak stress at different temperatures and strain rates. §
Cr and Mo as strong carbide formers cause precipitation during the hot deformation. Thus, there was a competition between the dynamic recrystallization and precipitation. 15
The precipitates cause the pinning effect at the high angle grain boundaries. This phenomenon, in turn, causes the retardation or inhibition of the dynamic recrystallization. §
In the range of low temperatures and strain rates the deformation is accompanied to a competition between precipitation and dynamic recrystallization. This produces serrated boundaries in the microstructure and also some kind of oscillations in the stress-strain curves.
§
In addition to precipitation, dynamic recrystallization is affected by temperature. Completed dynamic recrystallization occurs at 1150°C. The temperature increase causes the increase of the DRX grain size.
§
The γ→ε transformation produced at strain rate of 0.001 s-1 leads to higher rate of flow hardening due to less active slip systems compared to the γ phase. Therefore, the rate of hardening compensates the rate of softening, a little amount of flow softening is produced and a steady state flow appears more quickly than the other strain rates
§
The following constitutive equation expresses the hot working characteristic of the investigated alloy:
*̇ = 1.5924 − 4 exp @
−744.5984
AB
C exp(0.0525 "# )
16
References
[1] J. Giacchi, O. Fornaro, H. Palacio, Materials Characterization, 68 (2012) 49-57. [2] K. Yamanaka, M. Mori, A. Chiba, Journal of the Mechanical Behavior of Biomedical Materials, 29 (2014) 417-426. [3] Z. Guo, X. Pang, Y. Yan, K. Gao, A.A. Volinsky, T.-Y. Zhang, Applied Surface Science, (2015). [4] K. Yamanaka, M. Mori, A. Chiba, Materials & Design, 57 (2014) 421-425. [5] J. Giacchi, C. Morando, O. Fornaro, H. Palacio, Materials Characterization, 62 (2011) 53-61. [6] R. Turrubiates-Estrada, A. Salinas-Rodriguez, H. Lopez, Journal of Materials Science, 46 (2011) 254-262. [7] M. Mori, K. Yamanaka, S. Sato, K. Wagatsuma, A. Chiba, Metallurgical and Materials Transactions A, 43 (2012) 3108-3119. [8] J. Immarigeon, K. Rajan, W. Wallace, Metallurgical Transactions A, 15 (1984) 339-345. [9] Y. Han, G. Liu, D. Zou, R. Liu, G. Qiao, Materials Science and Engineering: A, 565 (2013) 342-350. [10] G.R. Ebrahimi, A. Momeni, M. Jahazi, P. Bocher, Metallurgical and Materials Transactions A, 45 (2014) 2219-2231. [11] A. Marandi, R. Zarei-Hanzaki, A. Zarei-Hanzaki, H. Abedi, Materials Science and Engineering: A, 607 (2014) 397-408. [12] K. Yamanaka, M. Mori, A. Chiba, Metallurgical and Materials Transactions A, 43 (2012) 4875-4887. [13] K. Yamanaka, M. Mori, S. Kurosu, H. Matsumoto, A. Chiba, Metallurgical and Materials Transactions A, 40 (2009) 1980-1994. [14] S.-H. Lee, N. Nomura, A. Chiba, Materials transactions, 49 (2008) 260-264. [15] R. Varano, J. Bobyn, J. Medley, S. Yue, Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine, 220 (2006) 145-159. [16] A. Mani, H. Lopez, Materials Science and Engineering: A, 528 (2011) 3037-3043. [17] P. Huang, H. Lopez, Materials Letters, 39 (1999) 249-253. [18] C. Balagna, S. Spriano, M. Faga, Materials Science and Engineering: C, 32 (2012) 18681877. [19] S. Abbasi, A. Shokuhfar, Materials letters, 61 (2007) 2523-2526. [20] Y. Han, H. Wu, W. Zhang, D. Zou, G. Liu, G. Qiao, Materials & Design, 69 (2015) 230240. [21] Y. Wang, W. Shao, L. Zhen, L. Yang, X. Zhang, Materials Science and Engineering: A, 497 (2008) 479-486. [22] A. Momeni, K. Dehghani, H. Keshmiri, G. Ebrahimi, Materials Science and Engineering: A, 527 (2010) 1605-1611. [23] H. Sun, Y. Sun, R. Zhang, M. Wang, R. Tang, Z. Zhou, Materials & Design, 64 (2014) 374380. [24] A. Momeni, K. Dehghani, Metallurgical and Materials Transactions A, 42 (2011) 19251932. [25] L. Ramırez, M. Castro, M. Méndez, J. Lacaze, M. Herrera, G. Lesoult, Scripta materialia, 47 (2002) 811-816. [26] T. Kilner, R. Pilliar, G. Weatherly, C. Allibert, Journal of biomedical materials research, 16 (1982) 63-79. [27] S. Mineta, S. Namba, T. Yoneda, K. Ueda, T. Narushima, Metallurgical and Materials Transactions A, 43 (2012) 3351-3358. 17
[28] S. Mineta, S. Namba, T. Yoneda, K. Ueda, T. Narushima, Metallurgical and Materials Transactions A, 41 (2010) 2129-2138. [29] K. Yamanaka, M. Mori, A. Chiba, Materials Science and Engineering: C, 40 (2014) 127134. [30] S.-H. Lee, E. Takahashi, N. Nomura, A. Chiba, Materials transactions, 47 (2006) 287-290. [31] M. Wimmer, J. Loos, R. Nassutt, M. Heitkemper, A. Fischer, Wear, 250 (2001) 129-139. [32] S. Mineta, S. Namba, T. Yoneda, K. Ueda, T. Narushima, Metallurgical and Materials Transactions A, 42 (2011) 1941-1949. [33] R. Taylor, R. Waterhouse, Journal of materials science, 18 (1983) 3265-3280. [34] R. Taylor, R. Waterhouse, Journal of materials science, 21 (1986) 1990-1996. [35] J. Weeton, R. Signorelli, Trans. ASM, 47 (1955) 815-852. [36] A. Clemow, B. Daniell, Journal of biomedical materials research, 13 (1979) 265-279. [37] A. de J. Saldı́var Garcı́a, A. Manı́ Medrano, A. Salinas Rodrı́guez, Scripta Materialia, 40 (1999) 717-722. [38] A. Saldivar, PhD Thesis, CINVESTAV-Saltillo, Mexico, 1996. [39] P. Huang, A. Salinas-Rodriguez, H. Lopez, Materials science and technology, 15 (1999) 1324-1330. [40] H. Lopez, A. Saldivar-Garcia, Metallurgical and Materials Transactions A, 39 (2008) 8-18. [41] X. Wu, N. Tao, Y. Hong, J. Lu, K. Lu, Scripta materialia, 52 (2005) 547-551. [42] A. Saldívar-García, H. López, Journal of Biomedical Materials Research Part A, 74 (2005) 269-274. [43] A. Rollett, F. Humphreys, G.S. Rohrer, M. Hatherly, Recrystallization and related annealing phenomena, Elsevier, 2004. [44] Y. Lin, M.-S. Chen, J. Zhong, Journal of materials processing technology, 205 (2008) 308315.
18
Figure caption Fig. 1 Microstructure of the alloy in the as-received condition: (a) SEM micrograph and (b) Xray diffraction (XRD) profiles. Fig. 2 True stress-true strain curves at different strain rates and (a) 1100˚C, (b) 1150 ˚C and (c) 1200 ˚C. Fig. 3 Variation of the peak stress as a function of the temperature at different strain rates. Fig. 4 SEM micrographs of the CO-Cr-Mo-C alloy hot deformed at 0.1s-1 and temperatures of (a) 1100 ̊C, (b) 1150 ̊C and (c) 1200 ̊C Fig. 5 SEM images and chemical composition mapping of C, Cr, Mo, and Co elements in the hot compressed sample at 1100 ̊C and strain rate of 0.01 s-1 Fig. 6 SEM micrographs and EDX chemical analyses of the Co-Cr-Mo-C alloy hot compressed at a strain rate of 0.1 s-1: (a) grain boundary carbide at 1100 C ̊ , (b) carbide inside the grain at 1100 ̊C, (c) carbide inside the grain at 1200 ̊C, (d) matrix at 1100 ̊C and (e) matrix at 1200 ̊C. Fig. 7 SEM micrographs of the CO-Cr-Mo-C alloy hot deformed at 1200 ̊C and strain rates of (a) 0.001 s-1, (b) 0.01 s-1 (c) 0.1 s-1 and (d) 1 s-1 Fig. 8 The variation of the peak stress with the strain rate at different temperatures: (a) lnε̇ vs lnσP , (b) lnε̇ vs σP, (c) lnε̇ vs ln(sinh(α"# )). Fig. 9 Slopes of Eqs. (12-14) (a) lnσP vs (1/T), (b) σP vs (1/T) and (c) ln(sinh(ασP)) vs (1/T)
19
Table caption Table 1 Chemical composition of the Co-Cr-Mo-C alloy object of the present study (wt%) Table 2 Chemical composition of the matrix and precipitates in the Co-Cr-Mo-C alloy hot compressed at 1100 ̊C at 0.01 s-1 (EDS analysis). Table 3 Grain size and athermal martensite volume fractions of the CO-Cr-Mo-C alloy hot deformed at 1200˚C at different strain rates.
Highlights The hot deformation behavior of Co-28Cr-6Mo-0.3C alloy was investigated. Strong carbide formers such as Cr and Mo caused to precipitation during hot deformation. · The microstructure at the low strain rates showed the serrated boundaries due to the competition between precipitation and DRX. · The activation energy for hot deformation of CO-28Cr-6Mo-0.3C alloy was calculated to be 744.5984 kJ/mol. Element Cr Mo C Si Mn Co wt% 28.5 6 0.33 <1 <1 Balance · ·
(Atomic %) Matrix Precipitate
Co 59.48 14.06
Strain rate (s-1)
Grain size (μm)
0.001 0.01 0.1 1
38.5 29 24.4 17.6
Cr 27.12 56.43
Mo 4.45 4
C 10.94 25.51
Athermal martensite volume fractions (fHCP) 0.15 0.07 0.02 0.04
20
Fig. 1a
Fig. 1b
Fig. 2a
Fig. 2b
Fig. 2c
Fig. 3
Fig. 4a
Fig. 4b
Fig. 4c
Fig. 5
Fig. 6a
Fig. 6b
Fig. 6c
Fig. 6d
Fig. 6e
Fig. 7a
Fig. 7b
Fig. 7c
Fig. 7d
Fig. 8a
Fig. 8b
Fig. 8c
Fig. 9a
Fig. 9b
Fig. 9c
PII: DOI: Reference:
S0921-5093(16)30091-0 http://dx.doi.org/10.1016/j.msea.2016.01.084 MSA33271
To appear in: Materials Science & Engineering A Received date: 20 October 2015 Revised date: 22 January 2016 Accepted date: 26 January 2016 Cite this article as: F.Z. Hassani, M. Ketabchi, G.R. Ebrahimi and S. Bruschi, Hot compression deformation characteristics and microstructural evolution of a co-cr-mo-C alloy: Effect of precipitate and martensitic transformation, Materials Science & Engineering A, http://dx.doi.org/10.1016/j.msea.2016.01.084 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Hot compression deformation characteristics and microstructural evolution of a Co-Cr-Mo-C alloy: Effect of precipitate and martensitic transformation F.Z. Hassaniª, M. Ketabchi*ª, G.R. Ebrahimib, S. Bruschic ªMining and Metallurgical Engineering Department, Amirkabir University of Technology (Tehran Polytechnic), Tehran, Iran b
Department of Materials and Polymer Engineering, Hakim Sabzevari University, Sabzevar,
9617976487, Iran c
Department of Industrial Engineering, University of Padua, Via Venezia 1, 35131 Padua, Italy
*Corresponding author: [email protected] Tel: +989132005610; Fax: +982166405846 Postal address: Amirkabir University of Technology (Tehran Polytechnic), Hafez Street, Enghelab Avenue, Tehran, Iran.
Abstract The hot compression behavior of a Co-Cr-Mo-C alloy was studied in the temperature range of 1100-1200 ˚C and the strain rate of 0.001-1 s-1. The influence of processing parameters including temperature and strain rate on the microstructure of the alloy was investigated. The flow curves of the hot compressed samples at the investigated deformation temperatures and strain rates exhibited single peak stress. The effect of the strain rate and temperature of deformation on the 1
flow stress was studied using a hyperbolic sine-type equation. The calculated activation energy was about 744.5984 kJ/mol. The evidenced microstructural characteristics combined with the shape of the flow curves indicate that there is a competition between dynamic recrystallization (DRX) and precipitation phenomena, especially at low strain rates, i.e. 0.001 s-1. The formation of precipitates at the grain boundaries prevent the mobility of boundaries and also cause the partial inhibition of the dynamic recrystallization. Whereas, long time for the precipitates coarsening causes the predomination of dynamic recrystallization. Thus, the serrated boundaries in microstructure along with oscillation in the flow curves take place due to this competition. Keywords: Cobalt-base alloy; Hot deformation; Recrystallization; Precipitates, Martensitic transformation. 1. Introduction Co-Cr-Mo alloys are usually used to produce artificial joints thanks to their excellent biocompatibility, wear and mechanical properties [1-4]. These alloys are commonly used in the as-cast condition, where they exhibit some main defects, such as shrinkages, coarse grain size and chemical heterogeneity [5]. Therefore, their mechanical properties in the as-cast condition are extremely low due to the distribution of these defects. As an example of possible consequences, low mechanical properties can cause the early failure of the implant device [1]. The duplex structure of Co-Cr-Mo alloys consists of a metastable γ (fcc) phase and a ε (hcp) phase. The metastable γ phase at room temperature could experience a martensitic transformation γ (FCC)→ε (HCP), which is induced by quenching from the stable γ (FCC) phase (above 1100˚C) [6]. The martensitic transformation, however, results in poor workability in case of cold deformation [7]. Therefore, it is necessary to eliminate the alloys casting defects and
2
enhance their mechanical properties for practical applications by applying proper thermomechanical treatments [8]. The improvement of the mechanical properties can be obtained by applying plastic deformation combined with a subsequent recrystallization heat treatment. Whatever the procedure applied to improve the alloys mechanical properties, a clear understanding of the microstructural changes occurring during thermo-mechanical processing is mandatory [8]. Dynamic recrystallization (DRX) as a softening mechanism plays a significant role in thermo-mechanical processing [9], besides having an important influence on the microstructural evolution [10, 11]. The process parameters, namely temperature, strain rate and strain, have strong effect on the degree of DRX during the microstructural evolution [9]. Moreover, the competition between recrystallization and precipitation is the common mechanism in the materials which are predisposed to precipitation of second phase [10]. So, due to presence of strong carbide formers, such as the Cr and Mo, these alloys present a reasonable potential for precipitation to occur. In biomedical Co-Cr-Mo alloys the effect of hot compression parameters on the grain refinement with low percent of carbon (0.02%) was reported [12, 13]. However, further research has not been conducted yet to characterize the effect of precipitates and hot compression parameters on the microstructural evolution of Co-Cr-Mo-C alloys characterized by a high percent of carbon (0.35%), which causes indeed the increase of the stability of the γ phase and also the formation of carbides [14]. Therefore, the purpose of this study is to disclose the microstructural and mechanical properties of Co-Cr-Mo-C alloy at different strain rates and temperatures and also determining the competition between DRX and precipitation of second phase. Thus, in this study, the effects of thermo-mechanical parameters, namely strain, strain rate, and temperature, on both the mechanical and microstructural characteristics of a Co-Cr-Mo-C alloy characterized 3
by a high carbon content is evaluated. Moreover, the competition between DRX and second phases precipitation is discussed as a function of the temperature change in the hot compression tests. Finally, the constitutive relationship to predict the flow stress behavior is assessed and the related coefficients identified. 2. Materials and methods 2.1 Material The composition of the Co-Cr-Mo-C alloy object of the present study is given in Table. 1. Fig. 1a shows the microstructure of the alloy in the as-received condition. The XRD pattern indicates a single γ phase in the microstructure (Fig. 1b), proving that the martensitic transformation did not occur. The average grain size of the γ phase was approximately 52 μm. 2.2 Hot compression tests Cylindrical compression specimens (with a diameter of 7.5 mm and a height of 13.4) were prepared for the hot compression tests. All the tests were performed on a Zwick/Roell 250
adopting a final true strain of 0.8 at temperatures ranging from 1100°C to 1200 ̊C, at different strain rate of 0.001, 0.01, 0.1 and 1 s-1. Before the compression, the specimens were heated to 1200 ̊C at a rate of 5 ̊C s-1 and held at this temperature for 300 s. Then they were cooled to the
deformation temperature with the cooling rate of 1; C/s. After soaking for 300 s at the testing temperature, continuous hot compression tests were carried out, keeping the deformation temperature constant. At the end of the deformation, the specimens were immediately quenched in water to retain the high temperature deformed microstructure. 2.3 Characterization methods
4
The deformed specimens were sectioned in the parallel direction with respect to the compression axis, and polished using standard metallography techniques. The polished surface was then etched at room temperature by immersion in a 5:1 solution of HCl:H2O2. In order to specify the carbide phases the polished samples were also stain etched by immersion in a staining solution containing one part of 20 vol % potassium permanganate in distilled water and one part of 8.0 vol % sodium hydroxide in distilled water [15]. The microstructural observations were performed by using a Field Emission Scanning Electron Microscope (FESEM). And the secondary electron detector (Everhart-Thornley: ETD detector) at different magnification was used to detect the pictures. The microscope were also equipped with Energy Dispersive X-ray (EDX) spectroscopy to provide more accurate characterization. For the EDX analysis, the working distance of sample was set-up between 9.5 to 10 mm. The size of the grains was determined by using image tool software. The identification of the alloy phases was performed using the X-Ray Diffraction (XRD) technique (a Cu Kα radiation λ=1.54184 A;
in a Philips
machine) and 2ɵ scans between 40˚ and 60˚ with step size of 0.02 2ɵ S-1. The samples for XRD measurements were prepared by conventional mechanical grinding and polishing. The volume fraction of the athermal ε-martensite was determined according to the method of Sage and Gullaud by measuring the integrated intensities of the principal peaks of the hcp phase [16-18] (see equation 1). ℎ"#($%%) =
'(101*1)+,'(101*1)+,- + 1.5'(200)3,,
(1)
Results and discussion 3.1 Flow characteristic
5
Typical true stress–strain curves obtained from hot compression tests of Co-Cr-Mo-C alloy at various deformation temperatures and strain rates are shown in Fig. 2. As expected, the flow stress increases with strain rate and decreases with temperature. Three stages of behavior may be observed in the flow curve, namely work hardening, flow softening and steady state [19]. The work hardening is the predominant phenomenon at the initial stages of deformation: the rapid multiplication of dislocations causes indeed the flow stress to increase [9, 20, 21]. When the strain reaches to critical value (εc), the dynamic restoration, becomes the predominant mechanism and starts partially offsetting the work hardening effect. At higher strains when there is a balance between the DRX and precipitation, a steady-state regime is obtained. All the flow curves exhibit one peak, with the peak strain and stress decreasing at increasing temperature and decreasing strain rate. Fig. 3 shows the variation of the peak stress as a function of the testing temperature at different strain rates. Higher generation rate of dislocations are due by higher strain rate, caused by the promotion of nucleation sites for new nuclei rather than the growth of previously formed grains. Moreover, the density of dislocations increases with rising strain rate [22] and the mobility of High-Angle Grain Boundaries (HAGBs) decreases with increasing strain rate. However, hot deformation is a thermally activated process [20], therefore, the migration of grain boundaries and movement of dislocation is accelerated at increasing temperature. This means that the work hardening rate is lower at higher temperatures, and the stress increases in the first stage at a lesser rate than at lower temperatures. Therefore, the work hardening rate and strengthening effect due to the dislocation interaction gradually decrease at increasing temperature and decreasing strain rate.
6
Some oscillations can be observed at the lowest strain rate, i.e. 0.001 s −1. This is due to the competition between DRX and precipitation [22, 23]. Carbides formation may be caused by the pinning effect on the grain boundaries that prevents the DRX to occur [22]. When the curved grain boundary tries to release from carbides, the flow stress increases, before reaching a plateau. In fact the precipitates growth when they reach their critical radius or dissolve when they reach a suitable temperature causing the stress to decrease. Thus, these plateaus can be attributed to the stress relaxation due to the continuous growth or dissolution of the precipitates. 3.2 Microstructural observations Fig. 4 shows the microstructure of the specimens subjected to the hot compression at 1100 ̊C,
1150 ̊C and 1200 C ̊ at a strain rate of 0.1 s-1.
As expected, the microstructures changes with the testing temperature. In fact, strong carbide formers present in the alloy composition, such as the Cr and Mo, cause the competition between precipitation and DRX. A larger amount of precipitate phase can be observed in the compressed specimens than in the as-received material, proving that the degree of deformation play a role in the competition between DRX and precipitation. The precipitation causes the pinning effect on the HAGBs [10, 22]. Moreover, precipitation and the mobility of boundaries are diffusioncontrolled. Thus, the frequency of competition between mobile boundaries and precipitates increases at increasing deformation temperature [24]. SEM and EDS micro-chemical analyses were carried out to confirm the presence of carbides. Fig. 5 shows the EDS micro-chemical maps of the specimen deformed to 55% (ε=0.8) at 1100 ̊C and strain rate of 0.01 s-1. The white particles in the SEM image are distributed inside the grains and also along the grain boundaries. By analyzing the distribution of some elements in composition mapping, they were Cr, Mo and
7
C enriched and were Co depleted. It can be seen the segregation of Cr, Mo and C at the grain boundaries and inside the grains where the precipitates exactly were existed. It means that a portion of carbon precipitated at grain boundaries and inside of the grains instead of resolving into the matrix. For more clarification the chemical composition of the precipitates and matrix are also reported in Table 2. The EDS analysis of the precipitates shows the segregation of Cr and C: precipitates are found to be enriched in Cr and C when compared with the Co-matrix. In the results reported by L.E. Ramırez et al [25] the main secondary phases in this alloy are identified as M(M: Cr or Mo element)23C6 carbides. Other studies [26, 27] identified M6C carbide as minor constituents. Shin Go Minet et al [28] reported that the precipitates detected in Co-Cr-Mo-C alloys depended on the carbon content as well as the temperature of the heat treatment. Carbon causes the stabilization of M23C6-type carbides in Co-Cr-Mo-C alloys [29]. Grain boundaries and also stacking faults provide appropriate sites for the precipitation of M23C6 carbides [30, 31], and M23C6 carbides are primarily observed in the Co-Cr-Mo-C alloys heat treated from 1175°C to 1225 C ̊ [32].
Fig. 4a shows the presence of precipitates at the grain boundaries and inside the grains caused by the retardation of the DRX occurrence. At higher temperature (Fig. 4b) the pinning effect of the precipitates on the mobile boundaries is decreased due to the dissolution by the diffusion mechanism and also increment of their size. Therefore, at this temperature, DRX will occur more easily. As can be seen in Fig. 4, the percentage of carbides was reduced at increasing temperature. The amount of grain boundary carbides and also fine blocky carbides inside the
grains decreased in the samples hot compressed at1200 ̊C compared to the ones deformed at 1100 ̊C. For the Co-Cr-Mo alloy with a high percent of carbon, 1200°C is the starting
temperature for M23C6 carbides dissolution [33, 34]. However, the dissolution of M23C6 depends 8
on the temperature as well as on the holding time. Thus, at 1200 ̊C for 600 sec in this study, a partial carbide dissolution happened. Moreover, the carbides shape changed from an elongated to a globular shape at increasing temperature. This confirms the study of Taylor and Waterhous
[33] who reported that the temperature of 1150 ̊C is the temperature at which the carbides morphology changes. EDS analysis of carbides and matrix at 1100°C and 1200°C are shown in Fig. 6 pointed out by red circle. According to this analysis, the carbon atomic percent of the grain boundary carbides
and the ones inside the grains at 1100 ̊C (Fig. 6a and b respectively) is lower than at 1200 ̊C (Fig.
6c) and also the carbon percent in the matrix at 1100 ̊C (Fig. 6d) is higher than at 1200 ̊C (Fig. 6e). Thus, it can be concluded that most of the carbides at 1100 ̊C are M23C6-type and most of them at 1200 ̊C are M6C-type. This is in accordance with the results of Weeton and Signorelli
[35] who worked on the Co-Cr-Mo-C alloy with high percent of carbon (0.29) that pointed out that M23C6 transform to M6C during heat treatment. For more clarification the stain etching was also performed. Based on the stain etching at 1100˚C the volume percent of M23C6 carbides (brown particles) was more than M6C carbides (red, green or blue particles).Moreover according to the time-temperature-precipitate diagram pointed out by Taylor and Waterhouse [34],1200 ˚C is the dissolution temperature of M23C6 carbides. Wimmer et al [31] pointed out that the M23C6 carbides precipitate at the grain boundaries and inside the grains but, the M6C ones exist only inside the grains. The results here reported also confirm the presence of M23C6 carbides at the grain boundaries especially at lower temperatures. However, the amount of M23C6 carbides at the grain boundaries is less at 1200 rather than at 1100°C. The M23C6 carbides tend to transform to M6C carbides at temperatures in the range between 1165°C and 1230 ̊C [36]. Thus, it seems that the most of the carbides (the grain 9
boundary and also fine blocky ones) at 1100 ̊C are M23C6 phase which transformed at the higher temperature to M6C according to the reaction of M23C6+M→ M6C+γ. Therefore, most of the carbides at 1200°C are M6C carbides that exist inside the grains. This is due to the above mentioned transformation and also because of the promotion of formation of M6C by increasing the thermal treatment temperature [36]. Moreover, as previously mentioned, the temperature of 1200°C is the appropriate one for the dissolution of M23C6 carbides. And a lot of grain boundary M23C6 carbides were dissolved at 1200 ̊C. Again, the presented results are in good agreement
with literature works, as the one by Clemow and Daniell [36] who stated that the M6C carbides are a stable phase in the Co-Cr-Mo-C alloys. As illustrated in Fig. 4a, new grains smaller than the starting ones are formed at the grain boundaries, with the necklace DRX being in progress. The original grains of the specimens subjected to the hot compression at 1150 ̊C are replaced by new grains, which are equiaxed ones
with sharp boundaries is characteristic of a completed DRX process [10]. It is clearly observed that the average grain size compared to the starting material is significantly decreased after hot
compression. At lower temperatures (i.e. 1100 ̊C) as can be seen in SEM image of Fig. 5 pointed out by a yellow circle, the refined grain boundaries are trapped by the carbides. At increasing temperature (Fig. 4c), coarsening of the new grains occurs: this is due to the dissolution of some precipitates as previously mentioned. Thus, their effect on the mobility of grains was decreased. It means that the recrystallized grain size increased at increasing the deformation temperature. The final grain sizes of the specimens hot compressed at strain rate of 0.1 and temperatures of 1150 ̊C and 1200 ̊C were about 9.5 and 24.4 μm, respectively, both smaller that the initial grain size (52 μm).
10
The effect of the strain rate on the microstructure of the specimens deformed at 1200 ̊C is shown in Fig. 7: all the specimens exhibit fully recrystallized γ-phase microstructures with equiaxed grains, except the one deformed at strain rate of 0.001 s-1 (Fig. 7a). It can be seen from the XRD patterns of the specimens deformed at different strain rates in Fig. 7 that the diffraction peaks (101*1) and (101*0) of hcp ε-phase appear at 2ϴ values of 46.65˚ and
41˚, respectively. The intensity of the (101*1) peaks at 0.001 s-1 is higher than that ones at the other strain rates. The values of the martensite volume fractions (fHCP) at different strain rates in 1200˚C are reported in Table 3 calculated according to the Sage and Guillaud [6, 16, 17, 37] method. This is athermal martensite produced after quenching as a consequence of the rapid cooling from the annealing temperature (T˃1100˚C) [38]. The developement of the athermal
martensite produced after quenching is dependent on the grain size [17, 39], increasing at increasing grain size [17, 37, 39, 40]. However, the transformation is limited and the volume fraction of martensite does not proceed beyond of 0.4-0.5 [6]. The grain size of hot deformed Co-Cr-Mo-C alloy at temperature of 1200 ˚C and different strain rates are given in Table 3. As can be seen, higher volume fractions of athermal hcp martensite at the lowest strain rate applied in this study is due to the larger grain size of alloy compared to the ones obtained at higher strain rates. The γ→ε transformation leads to greater work hardening [41] due to less active slip systems compared to the γ phase, which makes difficult the plastic deformation [42]. The alloy low Stacking Fault Energy (SFE) due to the ε martensite transformation plays a crucial role in enhancing the work hardening. As it is known, a low SFE causes difficulties in the stress relaxation by cross slip. Therefore, a higher work hardening of Co-Cr-Mo-C alloy at strain rate
11
of 0.001 s-1 rather than at the other strain rates, due to the formation of ε-martensite, leads to higher rate of flow hardening. Therefore, the rate of hardening compensates the rate of softening, a little amount of flow softening is produced and a steady state flow appears more quickly (Fig. 2c). The serrated boundaries are observed at the lowest strain rate, indicated by arrows in Fig. 7a. At intermediate strain rates, i.e. 0.1 s-1 (Fig. 7c), the amount of serrated boundaries decreases remarkably. These changes in the grain boundaries are due to the competition between DRX and precipitation. At the highest strain rate (Fig. 7d), the time for precipitation is too short and the driving force for DRX is higher. Therefore, DRX is the dominant mechanism [24]. Whereas, at the lowest strain rate there is enough time for the precipitation occurrence. But the long time for coarsening of new precipitates also causes lower pinning effect at the grain boundaries in some regions. Therefore, the serrated boundaries appear, which is the consequence of the continuous competition between DRX and precipitation at different regions of the grain boundaries. In fact, long time for the precipitates coarsening before starting the second cycle of precipitation causes the carbides release from the boundaries. Moreover, there is enough time for the beginning of subsequent cycles of fine precipitation. These fine carbides, which are precipitated, cause the binding of the mobile boundaries. These processes are repeated more times at lower strain rates at different regions of the grain boundaries, resulting in the serrated grain boundaries. When the strain rate is increased, the DRX grains become finer, as a consequence of the shorter time at higher strain rate needed for growing up the grains. 3.3 Constitutive equation of the flow stress
12
The material flow stress was described through constitutive equations. The base equation used in this study was proposed by Zener and Hollomon [43], who introduced the so-called ZenerHollomon parameter as a function of the strain rate and temperature according to equation (2). Z = ε̇ exp 8 <=F(σ) 9
(2)
:;
Where Q is an activation energy (KJ/mol), ε̇ is strain rate (s-1), (R=8.31 J/K mol) is the molar gas constant, T is the absolute temperature (K), σ is the flow stress (MPa). In hot working processes, several constitutive equations have commonly been applied as bellow [44].
F(σ)= A΄σn΄
ασ<0.8
(3)
F(σ)= A΄΄exp(βσ)
ασ>1.2
(4)
F(σ)= A( sinh(ασ)n)
For all σ
(5)
In which A, A΄, A΄΄, α, β, n and n΄ are the material constants, α=β/n. by substituted Eq (3), and Eq (4) and Eq (5) in the Eq (2), and taking the natural logarithm the following equations will be achieved. Lnε̇ + Lnε̇ +
Q = nA΄ + n΄ ln ?RT Q = lnA΄΄ + β?RT
Lnε̇ + :; = lnA + nln(sinh(α?D )) 9
13
ασ<0.8
(6)
ασ>1.2
(7)
For all σ
(8)
For obtaining the values of n΄, β and n the following equations can be used, which are the slope of lines of equations (6-8). É = G
HIJK̇
HIJMN
P=G E=G
HIJK̇ HMN
OT
OT
HIJK̇
(9) (10) OT
(11)
HIJ(SUJV(WXY ))
The graphs of lnε̇ vs lnσP and lnε̇ vs σP at different temperatures are shown in Fig. 8 a and Fig. 8b, respectively. The following mean values of n΄ and β are obtained: 5.7591 MPa-1 and 0.0525 MPa-1 respectively. Then α=β/n΄ is identified as 0.0091 MPa-1. Fig. 8c shows the lnε̇ vs ln(sinh(α?D )) plot. From the slope of Fig. 8c and substituting the value of α, the value of n is
obtained as 4.1970 MPa-1s-1
The Q parameter can be identified from the equilibrium (12-14) as follows: ^_E?D [ = \E΄ ] a 1 ^ 8`< ḃ [ = \P ]
(12)
^?D a 1 ^ 8`< ḃ
(13)
^ ln(sinh(c?D )) [ = \E ] a 1 ^ 8`< ḃ
(14)
The lnσP vs (1/T), σP vs (1/T) and ln(sinh(ασP)) vs (1/T) plots are shown in Fig. 9a, Fig. 9b and Fig. 9c, respectively. The obtained average values of Q are: 707.0674, 744.5984 and 728.3985 (kJ mol-1) from the Eqs. (12-14). By considering the correlation coefficient (R2), from the Fig. 14
9a-c, it can be considered that the Fig. 9b or Eq. 13 shows a better fitting with the experimental data. Therefore, it can be stated that the activation energy of the studied Co-Cr-Mo-C alloy is about 744.5984 (kJ mol-1). The Zener–Hollomon parameter for the investigated Co-Cr-Mo-C alloy can be estimated by Eq. 2 and is represented as follows: 744.5984 d = ḟ exp g q = 3.9199 − 13"#$.%$&' \`
(15)
744.5984 ) = *̇ exp , 0 = 1.5924 − 4 exp(0.0525 "# ) -/ 744.5984 ) = *̇ exp , 0 = 0.0388[sinh(0.0093"# )]<.'&%> -/
(16) (17)
The lnZ vs lnσP, lnZ vs σP and lnZ vs ln(sinh(ασP)) were drown. The correlation coefficients (R2) were 0.9269, 1 and 1, respectively. By considering the R2, it can be stated that the Eqs. 16 and 17 show the best agreement with the experimental data. Moreover, as previously was mentioned, Eq. 13 which was mainly extracted from Eq. 4 was better fitted with experimental data. Therefore, the hot deformation equation of Co-Cr-Mo-C alloy over the temperature range of 1100-1200 ̊C can be written as follows. −744.5984 *̇ = 1.5924 − 4 exp , 0 exp(0.0525 "# ) -/
(18)
4. Conclusions The stress-strain curves of the Co-Cr-Mo-C alloy investigated in this study exhibite the typical DRX behavior with a single peak stress at different temperatures and strain rates. §
Cr and Mo as strong carbide formers cause precipitation during the hot deformation. Thus, there was a competition between the dynamic recrystallization and precipitation. 15
The precipitates cause the pinning effect at the high angle grain boundaries. This phenomenon, in turn, causes the retardation or inhibition of the dynamic recrystallization. §
In the range of low temperatures and strain rates the deformation is accompanied to a competition between precipitation and dynamic recrystallization. This produces serrated boundaries in the microstructure and also some kind of oscillations in the stress-strain curves.
§
In addition to precipitation, dynamic recrystallization is affected by temperature. Completed dynamic recrystallization occurs at 1150°C. The temperature increase causes the increase of the DRX grain size.
§
The γ→ε transformation produced at strain rate of 0.001 s-1 leads to higher rate of flow hardening due to less active slip systems compared to the γ phase. Therefore, the rate of hardening compensates the rate of softening, a little amount of flow softening is produced and a steady state flow appears more quickly than the other strain rates
§
The following constitutive equation expresses the hot working characteristic of the investigated alloy:
*̇ = 1.5924 − 4 exp @
−744.5984
AB
C exp(0.0525 "# )
16
References
[1] J. Giacchi, O. Fornaro, H. Palacio, Materials Characterization, 68 (2012) 49-57. [2] K. Yamanaka, M. Mori, A. Chiba, Journal of the Mechanical Behavior of Biomedical Materials, 29 (2014) 417-426. [3] Z. Guo, X. Pang, Y. Yan, K. Gao, A.A. Volinsky, T.-Y. Zhang, Applied Surface Science, (2015). [4] K. Yamanaka, M. Mori, A. Chiba, Materials & Design, 57 (2014) 421-425. [5] J. Giacchi, C. Morando, O. Fornaro, H. Palacio, Materials Characterization, 62 (2011) 53-61. [6] R. Turrubiates-Estrada, A. Salinas-Rodriguez, H. Lopez, Journal of Materials Science, 46 (2011) 254-262. [7] M. Mori, K. Yamanaka, S. Sato, K. Wagatsuma, A. Chiba, Metallurgical and Materials Transactions A, 43 (2012) 3108-3119. [8] J. Immarigeon, K. Rajan, W. Wallace, Metallurgical Transactions A, 15 (1984) 339-345. [9] Y. Han, G. Liu, D. Zou, R. Liu, G. Qiao, Materials Science and Engineering: A, 565 (2013) 342-350. [10] G.R. Ebrahimi, A. Momeni, M. Jahazi, P. Bocher, Metallurgical and Materials Transactions A, 45 (2014) 2219-2231. [11] A. Marandi, R. Zarei-Hanzaki, A. Zarei-Hanzaki, H. Abedi, Materials Science and Engineering: A, 607 (2014) 397-408. [12] K. Yamanaka, M. Mori, A. Chiba, Metallurgical and Materials Transactions A, 43 (2012) 4875-4887. [13] K. Yamanaka, M. Mori, S. Kurosu, H. Matsumoto, A. Chiba, Metallurgical and Materials Transactions A, 40 (2009) 1980-1994. [14] S.-H. Lee, N. Nomura, A. Chiba, Materials transactions, 49 (2008) 260-264. [15] R. Varano, J. Bobyn, J. Medley, S. Yue, Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine, 220 (2006) 145-159. [16] A. Mani, H. Lopez, Materials Science and Engineering: A, 528 (2011) 3037-3043. [17] P. Huang, H. Lopez, Materials Letters, 39 (1999) 249-253. [18] C. Balagna, S. Spriano, M. Faga, Materials Science and Engineering: C, 32 (2012) 18681877. [19] S. Abbasi, A. Shokuhfar, Materials letters, 61 (2007) 2523-2526. [20] Y. Han, H. Wu, W. Zhang, D. Zou, G. Liu, G. Qiao, Materials & Design, 69 (2015) 230240. [21] Y. Wang, W. Shao, L. Zhen, L. Yang, X. Zhang, Materials Science and Engineering: A, 497 (2008) 479-486. [22] A. Momeni, K. Dehghani, H. Keshmiri, G. Ebrahimi, Materials Science and Engineering: A, 527 (2010) 1605-1611. [23] H. Sun, Y. Sun, R. Zhang, M. Wang, R. Tang, Z. Zhou, Materials & Design, 64 (2014) 374380. [24] A. Momeni, K. Dehghani, Metallurgical and Materials Transactions A, 42 (2011) 19251932. [25] L. Ramırez, M. Castro, M. Méndez, J. Lacaze, M. Herrera, G. Lesoult, Scripta materialia, 47 (2002) 811-816. [26] T. Kilner, R. Pilliar, G. Weatherly, C. Allibert, Journal of biomedical materials research, 16 (1982) 63-79. [27] S. Mineta, S. Namba, T. Yoneda, K. Ueda, T. Narushima, Metallurgical and Materials Transactions A, 43 (2012) 3351-3358. 17
[28] S. Mineta, S. Namba, T. Yoneda, K. Ueda, T. Narushima, Metallurgical and Materials Transactions A, 41 (2010) 2129-2138. [29] K. Yamanaka, M. Mori, A. Chiba, Materials Science and Engineering: C, 40 (2014) 127134. [30] S.-H. Lee, E. Takahashi, N. Nomura, A. Chiba, Materials transactions, 47 (2006) 287-290. [31] M. Wimmer, J. Loos, R. Nassutt, M. Heitkemper, A. Fischer, Wear, 250 (2001) 129-139. [32] S. Mineta, S. Namba, T. Yoneda, K. Ueda, T. Narushima, Metallurgical and Materials Transactions A, 42 (2011) 1941-1949. [33] R. Taylor, R. Waterhouse, Journal of materials science, 18 (1983) 3265-3280. [34] R. Taylor, R. Waterhouse, Journal of materials science, 21 (1986) 1990-1996. [35] J. Weeton, R. Signorelli, Trans. ASM, 47 (1955) 815-852. [36] A. Clemow, B. Daniell, Journal of biomedical materials research, 13 (1979) 265-279. [37] A. de J. Saldı́var Garcı́a, A. Manı́ Medrano, A. Salinas Rodrı́guez, Scripta Materialia, 40 (1999) 717-722. [38] A. Saldivar, PhD Thesis, CINVESTAV-Saltillo, Mexico, 1996. [39] P. Huang, A. Salinas-Rodriguez, H. Lopez, Materials science and technology, 15 (1999) 1324-1330. [40] H. Lopez, A. Saldivar-Garcia, Metallurgical and Materials Transactions A, 39 (2008) 8-18. [41] X. Wu, N. Tao, Y. Hong, J. Lu, K. Lu, Scripta materialia, 52 (2005) 547-551. [42] A. Saldívar-García, H. López, Journal of Biomedical Materials Research Part A, 74 (2005) 269-274. [43] A. Rollett, F. Humphreys, G.S. Rohrer, M. Hatherly, Recrystallization and related annealing phenomena, Elsevier, 2004. [44] Y. Lin, M.-S. Chen, J. Zhong, Journal of materials processing technology, 205 (2008) 308315.
18
Figure caption Fig. 1 Microstructure of the alloy in the as-received condition: (a) SEM micrograph and (b) Xray diffraction (XRD) profiles. Fig. 2 True stress-true strain curves at different strain rates and (a) 1100˚C, (b) 1150 ˚C and (c) 1200 ˚C. Fig. 3 Variation of the peak stress as a function of the temperature at different strain rates. Fig. 4 SEM micrographs of the CO-Cr-Mo-C alloy hot deformed at 0.1s-1 and temperatures of (a) 1100 ̊C, (b) 1150 ̊C and (c) 1200 ̊C Fig. 5 SEM images and chemical composition mapping of C, Cr, Mo, and Co elements in the hot compressed sample at 1100 ̊C and strain rate of 0.01 s-1 Fig. 6 SEM micrographs and EDX chemical analyses of the Co-Cr-Mo-C alloy hot compressed at a strain rate of 0.1 s-1: (a) grain boundary carbide at 1100 C ̊ , (b) carbide inside the grain at 1100 ̊C, (c) carbide inside the grain at 1200 ̊C, (d) matrix at 1100 ̊C and (e) matrix at 1200 ̊C. Fig. 7 SEM micrographs of the CO-Cr-Mo-C alloy hot deformed at 1200 ̊C and strain rates of (a) 0.001 s-1, (b) 0.01 s-1 (c) 0.1 s-1 and (d) 1 s-1 Fig. 8 The variation of the peak stress with the strain rate at different temperatures: (a) lnε̇ vs lnσP , (b) lnε̇ vs σP, (c) lnε̇ vs ln(sinh(α"# )). Fig. 9 Slopes of Eqs. (12-14) (a) lnσP vs (1/T), (b) σP vs (1/T) and (c) ln(sinh(ασP)) vs (1/T)
19
Table caption Table 1 Chemical composition of the Co-Cr-Mo-C alloy object of the present study (wt%) Table 2 Chemical composition of the matrix and precipitates in the Co-Cr-Mo-C alloy hot compressed at 1100 ̊C at 0.01 s-1 (EDS analysis). Table 3 Grain size and athermal martensite volume fractions of the CO-Cr-Mo-C alloy hot deformed at 1200˚C at different strain rates.
Highlights The hot deformation behavior of Co-28Cr-6Mo-0.3C alloy was investigated. Strong carbide formers such as Cr and Mo caused to precipitation during hot deformation. · The microstructure at the low strain rates showed the serrated boundaries due to the competition between precipitation and DRX. · The activation energy for hot deformation of CO-28Cr-6Mo-0.3C alloy was calculated to be 744.5984 kJ/mol. Element Cr Mo C Si Mn Co wt% 28.5 6 0.33 <1 <1 Balance · ·
(Atomic %) Matrix Precipitate
Co 59.48 14.06
Strain rate (s-1)
Grain size (μm)
0.001 0.01 0.1 1
38.5 29 24.4 17.6
Cr 27.12 56.43
Mo 4.45 4
C 10.94 25.51
Athermal martensite volume fractions (fHCP) 0.15 0.07 0.02 0.04
20
Fig. 1a
Fig. 1b
Fig. 2a
Fig. 2b
Fig. 2c
Fig. 3
Fig. 4a
Fig. 4b
Fig. 4c
Fig. 5
Fig. 6a
Fig. 6b
Fig. 6c
Fig. 6d
Fig. 6e
Fig. 7a
Fig. 7b
Fig. 7c
Fig. 7d
Fig. 8a
Fig. 8b
Fig. 8c
Fig. 9a
Fig. 9b
Fig. 9c