Contents lists available at ScienceDirect

Intermetallics

journal homepage: www.elsevier.com/locate/intermet

Effects of heat treatment on the microstructure and properties of cold-forgedCoNiFe medium entropy alloy

X.L. Ana, H. Zhaoa, T. Daia, H.G. Yua, Z.H. Huanga, C. Guoa, Paul K. Chub, C.L. Chua,∗

a School of Materials Science and Engineering and Jiangsu Key Laboratory for Advanced Metallic Materials, Southeast University, Nanjing, 211189, ChinabDepartment of Physics and Department of Materials Science and Engineering, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong, China

A R T I C L E I N F O

Keywords:Two-dimensional (2D) forgingMEAsCoNiFeAnnealingWork-hardening microstructure and properties

A B S T R A C T

The effects of heat treatment on the microstructure and properties of cold-forged two-dimensional (2D) CoNiFemedium entropy alloy (MEA) are determined. Compared to the as-cast specimen with a columnar crystalstructure, the cold-forged CoNiFe MEA has a heavily fragmented microstructure with deformation twins. As theannealing temperature is increased, the grain size becomes larger markedly. Annealing at 900 °C yields a fullyrecrystallized microstructure with a large population of annealing twins and the new orientation exhibits a first-order twin relationship (60°< 111> rotation) during recrystallization. Moreover, the CoNiFe HEA annealed at900 °C possesses excellent ductility (ε=50%) and work-hardening ability (σUTS-σY=246MPa, σUTS/σY= 0.5),which depend on the annealing twins, dislocations, as well as micro-shear bands in the grains. Analysis of thefracture surface indicates that the main failure mechanism is ductile. Meanwhile, no phase separation occurs asthe temperature is raised from 0 °C to 1000 °C as shown by the expansion rate versus temperature relationship,indicating that the materials have good stability at high temperature. The thermal expansion coefficient (CTE) ofthe sample annealed at 1100 °C for 1 h is 12.1× 10−6 K−1 which is less than that of traditional metals(14.4–16×10−6 K−1).

1. Introduction

Development of new structural materials is spurred by technologicaladvancement and high entropy alloys (HEAs) which consist of multipleelements have aroused much interest [1]. HEAs generally have five ormore principal elements in equal or near-equal atomic concentrationsand the concentration of each element is between 3% and 35%. Thefour effects of high entropy, sluggish diffusion, severe lattice distortion,and cocktail effects have been shown to play important roles in theproperties of HEAs [2–7]. HEAs have many attractive properties such asthe excellent oxidation resistance [8,9], thermal stability [10], ele-vated-temperature strength [11,12], fracture toughness [13–16], cor-rosion resistance [17], and surconductivity [5]. However, most HEAshave drawbacks in practice because of the low ductility and high brit-tleness especially at room temperature but medium entropy alloys(MEAs) composed of two to four principal elements with mixing en-tropy between 1R and 1.5R may deliver better performance [18–20].

The properties of CoNiFe MEAs have been studied [21]. Rathi et al.[22] prepared nano-crystalline equiatomic NiCoFe alloy powder bymechanical alloying and the materials possessed superior magneticproperties at ambient temperature. Tsau et al. [23] studied the

microstructure and polarization behavior of CoNiFe and CoNiFeCr al-loys which had the FCC granular structure and the corrosion resistanceof both alloys in 1MH2SO4 was better than that of 304 stainless steel.However, the mechanical properties of CoNiFe MHAs at room tem-perature have been found to be not good enough [24]. Cold working byfor example, forging, followed by recrystallization is one of the simplethermal mechanical processes (TMP) to control the strength and duc-tility balance. In fact, two-dimensional (2D) forging is a simple andindustrially viable method that is an effective slight plastic deformationtechnique in contrast to severe plastic deformation (SPD) such as high-pressure torsion [25].

In this work, CoNiFe MEA samples are prepared by two-dimensional(2D) forging followed by heat treatment. Our objective is to study theeffects of annealing on the microstructure and properties of cold-forgedCoNiFe MEA to improve the mechanical properties, especially theductility at room temperature. The relationship between annealingtwins, dislocations as well as micro-shear bands in the grains and me-chanical behavior is determined.

https://doi.org/10.1016/j.intermet.2019.106477Received 7 January 2019; Received in revised form 19 April 2019; Accepted 19 April 2019

∗ Corresponding author.E-mail address: clchu@seu.edu.cn (C.L. Chu).

Intermetallics 110 (2019) 106477

0966-9795/ © 2019 Published by Elsevier Ltd.

T

http://www.sciencedirect.com/science/journal/09669795https://www.elsevier.com/locate/intermethttps://doi.org/10.1016/j.intermet.2019.106477https://doi.org/10.1016/j.intermet.2019.106477mailto:clchu@seu.edu.cnhttps://doi.org/10.1016/j.intermet.2019.106477http://crossmark.crossref.org/dialog/?doi=10.1016/j.intermet.2019.106477&domain=pdf

2. Materials and methods

2.1. Experimental materials and the annealing process

The CoNiFe MEA samples were produced by vacuum arc meltingunder argon from 99.9% pure Co, Ni, and Fe metals. Before melting, thechamber was evacuated to a pressure of 1× 10−5 Torr and back-filledwith argon to 500 Torr. High-purity Ti was used as a getterer prior tomelting to mitigate contamination from residual oxygen and nitrogen.The ingots were remelted under electromagnetic stirring at least fivetimes to enhance the uniformity of the chemical composition and thencooled naturally in the furnace.

Before forging, the samples with dimensions of 30mm(length)× 20mm (width)× 15mm (height) were cut from the ingots.Then 2D forged CoNiFe MEAs samples were prepared on an industrialair pneumatic hammer machine at a load of 100 kg and hammer fre-quency of 180 times per minute. The schematic of the 2D forgingprocedure and heat treatment is shown in Fig. 1. The final forgedsamples had dimensions of 50mm (length)× 40mm (width)× 4.5 mm(height). The amount of strain (forging ratio) was derived by the re-duced height or increased cross section increasing of the forged speci-mens as follows:

= =φ SS

HHf

0

1

1

0 (1)

where S0, H0 and S1, H1 are the cross sections and heights of the initialand final samples, respectively. The forging ratio φf was 30% accordingto the above formula.

2.2. Microstructural characterization

The metallographic sections were prepared by standard mechanicalpolishing procedures. An Olympus metallurgical microscope was usedto examine the grain morphology of the samples and the grain size wasmeasured using the Nano-measurement software. The microstructureand composition of the alloys were determined by field-emissionscanning electron microscopy (EF-SEM, FEI Sirion), energy-dispersiveX-ray spectrometry (EDS) (Sirion), and transmission electron micro-scopy (TEM, FEI Tecnai G2 at 200 kV). Electron backscattering dif-fraction (EBSD) was performed on the SEM using an EDAX Hikari superEBSD camera in conjunction with the acquisition software EDAX TSLDC7 and analysis software a TSL OIM. Prior to TEM examination, thespecimens were mechanically ground to a thickness below 100 μm withSiC paper, punched into 3mm diameter disks, and then thinned by ionmilling to electron transparency. The phases were determined by X-raydiffraction (XRD, D8-Discover, Germany and Bruker-AXS) using Cu Kαradiation (λ=1.54183Å) in the 2θ range of 20–90° at a scanningspeed of 8°/min, 40 kV, and 20mA.

2.3. Determination of mechanical properties

The uniaxial tensile specimens with a gauge length of 8mm werecut from the alloy by electrical discharge machining and the tensiletests were performed in triplicate on a CMT5105 universal electronictensile testing machine at a strain rate of 1× 10−3 s−1 at room tem-perature. The yield strength (YS), ultimate tensile strength (UTS), andelongation were determined from the tensile stress-strain curves.Cylindrical samples 2mm in diameter and 10mm tall were cut from thealloy om an electrical discharge machine and the thermal expansiontests were performed by thermo-mechanical analysis (TMA) at aheating rate of 5 °C s−1 from 0 °C to 1000 °C.

3. Results

Fig. 2 illustrates the microstructural evolution of the as-cast CoNiFeMEA and after different annealing processes. The transformation fromcolumnar crystals in the as-cast CoNiFe MEA to a heavily fragmentedmicrostructure with deformation twins (DT) is presented in Fig. 5a–eand the grain size increases with annealing temperature. The meangrain sizes of the specimens annealed at 900 °C, 1000 °C, and 1100 °Care 42 μm, 74 μm, and 90 μm, respectively, as shown in Fig. 2f. Similarto the one annealed at 900 °C, the samples annealed at higher

Fig. 1. Schematic of (a) 2D forging procedure and (b) Heat treatment.

Fig. 2. OM grain morphology of CoNiFe MEAs after different treatment processes: (a) As-cast; (b) Cold forging; (c–d) Cold forging and then annealing at 900 °C,1000 °C, and 1100 °C for 1 h, respectively; (f) Grain sizes of cold-forged samples annealed at different temperature.

X.L. An, et al. Intermetallics 110 (2019) 106477

2

temperature show many annealing twins (AT). A similar temperaturetrend during grain growth was also observed by Sathiaraj et al. [26]from CoNiCr MEAs. Athiaraj et al. found that annealing at 700 °Cyielded a fully recrystallized ultrafine grain microstructure with a largefraction of annealing twins and extensive grain coarsening occurredafter annealing at 1100 °C. Chuang et al. [27] revealed a face-centeredcubic (fcc) crystal structure and low stacking fault energy (SFE) are keyto the production of alloys with a large density of annealing twins.Sathiaraj et al. [28] showed that the microstructure and texture ofternary alloys are different from those of quaternary and quinary alloysbecause CoNiFe MEAs are not low SFE alloys. The quinary and qua-ternary alloys with lower SFE can form more TBs during deformationand annealing produces greater microstructural fragmentation of re-finement and texture transition. Pure metal or copper-type texture isobserved from high to medium SFE alloys during the deformation state,whereas low SFE alloys exhibit a strong brass or alloy-type texture [29].

The EBSD maps (inverse pole figure, IPE) of the microstructuresannealed at 800 °C, 900 °C, and 1000 °C for 1 h are shown in Fig. 3 (a1),(b1), and (c1). The misorientation angle distributions curves of thevarious microstructures (Fig. 3 (a2), (b2), and (c2)) show that themisorientation angle distributions have a strong preference for 60°,indicating a large number of annealing twins (ATs) in the grains asshown by SEM (Fig. 3a, b, c). The 60°< 111> and38.9°< 110> angle-axis misorientations corresponding to the Σ3 andΣ9 boundaries result in strong and weak peaks, respectively, implyingthat a lot of TBs exist in the forged-CoNiFe MEAs. The fraction of Σ3(about 35%) in the sample treated at 900 °C is larger than those treatedat 1000 °C (about 2%) and 1100 °C (about 6%). According to Sathiarajet al. [29,30], annealing twins resulting in a large fraction of annealingtwin boundaries (Σ3) are observed from CoCrFeMnNi HEAs with low

SFE and Ni with medium SFE. The twin fraction of ternary CoNiFeMEAs is lower than that of the quinary HEAS but higher than that ofpure Ni and it is consistent with the difference in the SFE values: Fe-CoNiCrMn(γSFE= 20mJm−2) < FeCoNi(γSFE= 70mJm−2) < Ni(γSFE= 130mJm−2) [31,32].

The grain boundary characteristic distributions (GBCDs) of theforged FeCoNi MEAs annealed at different temperature show that thefraction of low-angle grain boundaries (LAGBs) such as the Σ1 bound-aries (misorientation across them in the range of 2–15°) is quite largeand increases with annealing temperature. The evolution of LAGBs is ingood agreement with the general tendency of the proportion of LAGBsto increase and high energy grain boundaries to decrease during graingrowth [33]. The reason is that the Σ1 boundaries have relatively lowenergies. Because grain growth reduces the total interfacial area, moregrain boundaries are eliminated rather than created. Dillon et al. [34]showed that higher-energy grain boundaries were preferentiallyeliminated from the network during grain growth thus producing alarger relative population of low energy grain boundaries. Besides, themotion and impact of the incoherent Σ3 formed during annealing leadto the formation of Σ1 which improves the fraction of LAGBs.

Fig. 4 shows the XRD diffraction patterns of the forged-CoNiFeMEAs annealed at different temperature for 1 h. All the diffractionpeaks can be indexed assuming a single FCC phase with a latticeparameter a= 0.362, indicating that no phase transformation occursduring cold forging and annealing except the texture.

3.1. Mechanical properties of CoNiFe MEAs

Fig. 5a shows the engineering stress–strain curves of the forged-CoNiFe MEAs after annealing at different temperature. Compared to the

Fig. 3. SEM images and IPF map of the forged-CoNiFe MEAs annealed for 1 h at (a) 900 °C, (b) 1000 °C, and (c) 1100 °C (“1” showing the IPF map and “2” showingthe misorientation angle distributions curves).

X.L. An, et al. Intermetallics 110 (2019) 106477

3

untreated one, the annealed samples have much improved ductility.The fracture surface is analyzed to determine the fatigue characteristicssuch as crack initiation sites, crack propagation, and final fracture. Theunannealed sample shows brittle fracture characteristics according toFig. 5d, but the annealed samples exhibit ductile failure as shown inFig. 5(e, f, g). As the annealing temperature is increased from 800 °C to1100 °C, the ultimate tensile strength σUTS decreases from 540MPa to430MPa and tensile ductility σy decreases from 50% to 40%. The yieldstrength values are 276, 217, and 201MPa after annealing at 900,1000, and 1100 °C, respectively. The sample annealed at 900 °C shows

superior ductility in comparison with similar HEAs reported in the lit-erature [24].

According to the true stress-strain curves, the strain-hardening rates(dσ/dε) versus true strain curves of the forged-CoNiFe MEAs annealedat different temperature are presented in Fig. 5b. The strain-hardeningrates of the annealed alloys exhibit similar trends that can be separatedinto two stages. In stage I, the strain-hardening rates decrease linearlyand in stage II, the strain-hardening rates diminish with increasingstrain. Heterogeneous deformation caused by grain size and sub-structure is generally believed to be the reason and the alloy annealedat 900 °C shows excellent work-hardening ability (σUTS - σY= 246MPa,σUTS/σY= 0.5). The higher work-hardening rate can be ascribed to twinboundaries which can pin the dislocations and promote dislocationaccumulation for strain hardening [35]. The outstanding work-hard-ening ability is crucial to the reliability in engineering applicationsbecause its ensure a large safety margin against fracture.

4. Discussion

4.1. Relationship between mechanical behavior and microstructure

According to the tensile test, the forged CoNiFe MEA treated at900 °C possesses excellent mechanical properties. Fig. 6 shows the TEMmicrographs and corresponding SAED patterns of the forged-CoNiFeMEA treated at 900 °C. The rings in the SAED pattern (Fig. 5a) suggest a[011] zone axis for the FCC structure consistent with XRD (Fig. 4) andbased on the interlunar spacing in the SAED pattern, the lattice para-meter is calculated to be 3.60 A. The TEM image and correspondingSAED patterns of the annealing twins are shown in Fig. 6c. Annealingtwins are common in cold-working face-centered cubic metals and af-fect the mechanical properties and corrosion behavior [36]. Manymodels have been proposed to explain the formation of annealing twins

Fig. 4. XRD diffraction patterns of the forged-CoNiFe MEAs annealed at dif-ferent temperature for 1 h.

Fig. 5. Mechanical properties of the forged-CoNiFe MEAsannealed at different temperature for 1 h: (a) Engineeringstress-strain curves; (b) True stress-strain and strainhardening rates; (c) Transverse section and longitudinalsection of the fracture surface of the forged-CoNiFe MEAannealed at 1000 °C after the tensile test; (d–g) Fracturesurfaces of the forged-CoNiFe MEAs (Untreated and an-nealed at 900 °C, 1000 °C, and 1100 °C).

X.L. An, et al. Intermetallics 110 (2019) 106477

4

[37,38] and the growth accident model has been validated experi-mentally [35,39,40]. This model asserts that an annealing twinboundary forms during migration of grain boundaries because of astacking error. With regard to the mechanism of annealing twinning inFCC materials, it is generally accepted that twinning is initiated by pre-existing dislocations that dissociate into partial dislocations. After thefirst glide of partial dislocations requiring the maximum stress, thesecond glide of partial dislocations takes place on the successive {111}plane. Previous research [41] has indicated that the migration distanceand velocity of the grain boundary migration are two key factors pro-moting the generation of annealing twins in the growth accident model.Meanwhile, the twin density has a strong tie with both the grainboundary migration distance and velocity. Gleiter [42] designed aformula which could accurately determine the annealing twin densityand Mahajan [43] suggested that twinning was only influenced by thetemperature due to the effects on grain size. Jin et al. [41] presentedevidence that the annealing twin density increased during re-crystallization but decreased during grain growth and the observation isconsistent with ours.

Annealing twins have a large effect on the motion of dislocations. Ahigh density of dislocations is observed from the areas adjacent to theTBs (Fig. 6d) indicative of strong interactions between the dislocationsand TBs [44]. Lu et al. [45] reported that strengthening and workhardening of alloy were attributed to high-density TBs interfering withor obstructing dislocation propagation consistent with our results(Fig. 5b). Lu et al. investigated the influences of twin thickness, grainsize, as well as strain rate on the work-hardening behavior and foundthat decreasing twin thickness or increasing grain size and increasingstrain rate or decreasing deformation temperature might enhance work-hardening ability and subsequent mechanical performance. With regardto the forged-CoNiFe MEA treated at 900 °C for 1 h, because the twinvolume fractions are relatively large, their contributions to the totaltensile strain are larger. Moreover, the new twin boundaries that formduring straining (“dynamic Hall-petch”) play an important role inproviding steady strain hardening which delays the onset of neck and

enhances uniform elongation before fracture [46].Excellent work-hardening strain is observed from the forged-CoNiFe

MEAs due to dislocation accumulation near the TBs as observed duringthe experiments conducted at room temperature. The BF TEM images inFig. 6d to show the morphology and state of dislocations in theboundaries. A large density of extended dislocations accumulates andpiles up at the boundary forming dislocation wall. The long and straightdislocations and high density of dislocation walls suggest strong inter-actions in the multiple slip system at room temperature [47]. The si-multaneous movement, intersection, and entangling of dislocations onthe two active {111} planes lead produce volume structural defectswhich retard dislocation movement and increase the strength and strainhardening capacity of alloy [48,49].

The shear band also plays a critical role in plastic deformation.Previous studies [50] show that the shear bands occur on planes or-iented roughly at 45° to the principal stress. Cui et al. [51] reported thatthe work-hardening capability was linked to the shear bands and analloy with a large shear band density and critical shear offset exhibitbetter plasticity in general. The sheer band in the forged-CoNiFe MEAstreated at 900 °C examined by TEM is shown in Fig. 5e.

4.2. Fracture mechanism

The fracture surfaces of the samples treated at 900 °C for 1 h un-dergo tensile tests and the results are presented in Fig. 7. Fracture startsapproximately in the geometrical middle of the tensile specimens andends at the specimen edge. Based on the crack propagation and fracturecharacteristics, the fracture surface can be divided into three regions(Fig. 7a). The crack origin is shown in region “I” and many microvoidsappear in region “I” (Fig. 6(b1)). The crack originates from these mi-crovoids and propagate from the center to outside. Cracks are usuallyinitiated at defects on the surface or at the corner of the regions whichfavor crack nucleation [52]. The fracture surface in region “II” (Fig. 7c)exhibits typical ductile fracture with dimples. Many uniform ovaldimples can be observed from Fig. 7(c3). Since the obvious neck be-havior occurs during the tensile process, region “II” is the final

Fig. 6. Microstructure characterization of the forged-CoNiFe MEAs treated at900 °C for 60min: (a) TEM image showing an FCC matrix with the insetshowing the corresponding SAED pattern of the FCC [011] zone axis; (b) AT andcorresponding SAED pattern along the ZM= [011]M and ZT= [0īī] zone axes;(c) BF TEM showing the dislocations and grain boundary; (d) BF TEM showingthe shear band.

Fig. 7. Fracture surfaces of the forged-CoNiFe MEAs treated at 900 °C for 1 h:(a) Tensile test of the fracture surface with apparent shear edges divided intoregions: I, II, and III; (b) Fracture surface at high magnification (b1) and lowmagnification (b2) for “region I”; (c) Fracture surface at high magnification (c1,c2) and low magnification (c2) for “region II”; (d) Fracture surface at highmagnification (d1) and low magnification (d2) for “region III”.

X.L. An, et al. Intermetallics 110 (2019) 106477

5

propagation area thus indicating a ductile fracture mechanism. Thecharacteristics of sheer tearing are shown in region “III” and the frac-ture and major stress have a 45° angle. Our results suggest that an-nealing twins and dislocation slips are the main factors for the excep-tional ductility achieved from the forged-CoNiFe MEAs treated at900 °C for 1 h.

4.3. Thermomechanical behavior

Fig. 8 shows the thermal expansion rate of the forged-CoNiFe MEAsannealed at different temperature. The thermal expansion rate in-creases almost linearly with temperature. The CTE is a temperature-dependent property and for moat metals and alloys, gradually increaseswith temperature. However, the coefficient shows discontinuities asphase changes occur. For example, a sharp contraction occurs as heatediron-based materials go through the ferrite to austenite transition. It isbecause the crystal changes from body-centered cubic to the morecompact face-centered cubic structure [53]. There are no peaks ordiscontinuities in the thermal expansion rate-temperature curve (Fig. 8)suggesting absence of phase transformation with increasing tempera-ture. Therefore, it can be inferred indirectly that the forged-CoNiFeMEAs annealed at different temperature are stable at high temperature.

The thermal expansion coefficient (CTE) of the forged-CoNiFe MEAsannealed for 1 h at 900 °C, 1000 °C, and 1100 °C are 14.5×10−6 K−1,14.0×10−6 K−1, 12.1×10−6 K−1, respectively, which decreaseslightly compared with the untreated one (15×10−6 K−1). The CTE ofthe sample annealed at 1000 °C is relatively small compared to tradi-tional stainless steel which generally has CTE between 14.4 and16×10−6 K−1.

5. Conclusion

Effective slight plastic deformation - two dimensional (2D) forgingand subsequent annealing at different temperature are conducted toinvestigate the microstructure and properties of the CoNiFe mediumentropy alloy (MEA). The as-cast and forged CoNiFe MEA samples havea face-centered cubic (FCC) crystal structure. As the annealing tem-perature is raised from 800 °C to 1100 °C, the ultimate tensile strengthσUTS decreases from 540MPa to 430MPa and tensile ductility σy de-creases from 50% to 40%. The mechanical properties depend on thegrain size and annealing twins affect the dislocation motion. The Σ3annealing TBs corresponding to misorientation distribution at60°< 111> are observed from the annealed CoNiFe MEAs. For the

CoNiFe MEA sample annealed at 900 °C for 1 h, the efficiency of grainrefinement by recrystallization is enhanced and the annealing twins,dislocations, as well as micro-shear bands inside the grains are re-sponsible for the outstanding ductility and work hardening ability. TheCTE of the forged-CoNiFe MEA sample annealed at 1000 °C is relativelylow compared to traditional stainless steel that generally has CTE be-tween 14.4 and 16× 10−6 K−1.

Acknowledgements

The research was jointly supported by the National Natural ScienceFoundation of China (Grant Nos. 51771054 and 31570961), State KeyProgram of National Natural Science Foundation of China (Grant No.51631003), National Key Research and Development Program of China(Grant No. 2016YFC1102402), National Key R&D Program of China(Grant No. 2018YFB1106100), and Hong Kong Research Grants Council(RGC) General Research Funds (GRF) No. CityU f11205617.

References

[1] J.W. Yeh, S.K. Chen, S.J. Lin, et al., Nanostructured High-entropy alloys withmultiple principal elements Novel Alloy design concepts and outcomes, Adv. Eng.Mater. 6 (5) (2004) 299.

[2] D. Miracle, J. Miller, O. Senkov, et al., Exploration and development of high entropyalloys for structural applications, Entropy 16 (1) (2014) 494–525.

[3] B.E. Macdonald, Z. Fu, B. Zheng, et al., Recent progress in high entropy alloy re-search, JOM (J. Occup. Med.) 69 (10) (2017) 2024–2031.

[4] Fabrication of the high-entropy alloys and recent research trends: a review, KoreanJ. Met. Mater. 55 (10) (2017).

[5] D.B. Miracle, O.N. Senkov, A critical review of high entropy alloys and relatedconcepts, Acta Mater. 122 (2017) 448–511.

[6] J.-W. Yeh, Alloy design strategies and future trends in high-entropy alloys, JOM (J.Occup. Med.) 65 (12) (2013) 1759–1771.

[7] F. Zhang, C. Zhang, S.L. Chen, et al., An understanding of high entropy alloys fromphase diagram calculations, Calphad 45 (2014) 1–10.

[8] H.M. Daoud, A.M. Manzoni, R. Völkl, et al., Oxidation behavior ofAl8Co17Cr17Cu8Fe17Ni33, Al23Co15Cr23Cu8Fe15Ni15, andAl17Co17Cr17Cu17Fe17Ni17Compositionally complex alloys (High-Entropy al-loys) at elevated temperatures in air, Adv. Eng. Mater. 17 (8) (2015) 1134–1141.

[9] C.M. Rost, E. Sachet, T. Borman, et al., Entropy-stabilized oxides, Nat. Commun. 6(2015) 8485.

[10] B. Schuh, F. Mendez-Martin, B. Völker, et al., Mechanical properties, microstructureand thermal stability of a nanocrystalline CoCrFeMnNi high-entropy alloy aftersevere plastic deformation, Acta Mater. 96 (2015) 258–268.

[11] Z. Wu, H. Bei, G.M. Pharr, et al., Temperature dependence of the mechanicalproperties of equiatomic solid solution alloys with face-centered cubic crystalstructures, Acta Mater. 81 (2014) 428–441.

[12] Y. L, Y. Dong, A promising new class of high-temperature alloys: eutectic high-entropy alloys, Sci. Rep. (2014), https://doi.org/10.1038/srep06200.

[13] Z. Li, K.G. Pradeep, Y. Deng, et al., Metastable high-entropy dual-phase alloysovercome the strength-ductility trade-off, Nature 534 (7606) (2016) 227–230.

[14] Z. Zhang, M.M. Mao, J. Wang, et al., Nanoscale origins of the damage tolerance ofthe high-entropy alloy CrMnFeCoNi, Nat. Commun. 6 (2015) 10143.

[15] Z. Zhang, H. Sheng, Z. Wang, et al., Dislocation mechanisms and 3D twin archi-tectures generate exceptional strength-ductility-toughness combination in CrCoNimedium-entropy alloy, Nat. Commun. 8 (2017) 14390.

[16] C. Minghao, T. i. Hung, W. o. Ren, et al., Microstructure and wear behavior ofAlxCo1.5CrFeNi1.5Tiy high-entropy alloys, Acta Mater. 59 (16) (2011) 6308–6317.

[17] Y. Shi, B. Yang, P. Liaw, Corrosion-resistant high-entropy alloys: a review, Metals 7(2) (2017) 43.

[18] D.Z. Yang Zhou, Xi Jin, Lu Zhang, Xingyu Du, Bangsheng Li, Design of Non-equiatomic Medium Entropy Alloy. Scientific Reports, (2018).

[19] B. Gludovatz, A. Hohenwarter, K.V. Thurston, et al., Exceptional damage-toleranceof a medium-entropy alloy CrCoNi at cryogenic temperatures, Nat. Commun. 7(2016) 10602.

[20] A. Gali, E.P. George, Tensile properties of high- and medium-entropy alloys,Intermetallics 39 (2013) 74–78.

[21] S. Samanifar, M. Almasi Kashi, A. Ramazani, et al., Reversal modes in FeCoNi na-nowire arrays: correlation between magnetostatic interactions and nanowireslength, J. Magn. Magn. Mater. 378 (2015) 73–83.

[22] A. Rathi, V.M. Meka, T.V. Jayaraman, Synthesis of nanocrystalline equiatomicnickel-cobalt-iron alloy powders by mechanical alloying and their structural andmagnetic characterization, J. Magn. Magn. Mater. 469 (2019) 467–482.

[23] C.H. Tsau, S.X. Lin, C.H. Fang, Microstructures and corrosion behaviors of FeCoNiand CrFeCoNi equimolar alloys, Mater. Chem. Phys. (186) (2016) 534–540.

[24] P. Li, A. Wang, C.T. Liu, Composition dependence of structure, physical and me-chanical properties of FeCoNi(MnAl) x high entropy alloys, Intermetallics 87 (2017)21–26.

[25] S. Praveen, J.W. Bae, P. Asghari-Rad, et al., Ultra-high tensile strength nanocrys-talline CoCrNi equi-atomic medium entropy alloy processed by high-pressure

Fig. 8. Thermal expansion rates of the forged-CoNiFe MEAs annealed at dif-ferent temperature.

X.L. An, et al. Intermetallics 110 (2019) 106477

6

http://refhub.elsevier.com/S0966-9795(19)30025-1/sref1http://refhub.elsevier.com/S0966-9795(19)30025-1/sref1http://refhub.elsevier.com/S0966-9795(19)30025-1/sref1http://refhub.elsevier.com/S0966-9795(19)30025-1/sref2http://refhub.elsevier.com/S0966-9795(19)30025-1/sref2http://refhub.elsevier.com/S0966-9795(19)30025-1/sref3http://refhub.elsevier.com/S0966-9795(19)30025-1/sref3http://refhub.elsevier.com/S0966-9795(19)30025-1/sref4http://refhub.elsevier.com/S0966-9795(19)30025-1/sref4http://refhub.elsevier.com/S0966-9795(19)30025-1/sref5http://refhub.elsevier.com/S0966-9795(19)30025-1/sref5http://refhub.elsevier.com/S0966-9795(19)30025-1/sref6http://refhub.elsevier.com/S0966-9795(19)30025-1/sref6http://refhub.elsevier.com/S0966-9795(19)30025-1/sref7http://refhub.elsevier.com/S0966-9795(19)30025-1/sref7http://refhub.elsevier.com/S0966-9795(19)30025-1/sref8http://refhub.elsevier.com/S0966-9795(19)30025-1/sref8http://refhub.elsevier.com/S0966-9795(19)30025-1/sref8http://refhub.elsevier.com/S0966-9795(19)30025-1/sref8http://refhub.elsevier.com/S0966-9795(19)30025-1/sref9http://refhub.elsevier.com/S0966-9795(19)30025-1/sref9http://refhub.elsevier.com/S0966-9795(19)30025-1/sref10http://refhub.elsevier.com/S0966-9795(19)30025-1/sref10http://refhub.elsevier.com/S0966-9795(19)30025-1/sref10http://refhub.elsevier.com/S0966-9795(19)30025-1/sref11http://refhub.elsevier.com/S0966-9795(19)30025-1/sref11http://refhub.elsevier.com/S0966-9795(19)30025-1/sref11https://doi.org/10.1038/srep06200http://refhub.elsevier.com/S0966-9795(19)30025-1/sref13http://refhub.elsevier.com/S0966-9795(19)30025-1/sref13http://refhub.elsevier.com/S0966-9795(19)30025-1/sref14http://refhub.elsevier.com/S0966-9795(19)30025-1/sref14http://refhub.elsevier.com/S0966-9795(19)30025-1/sref15http://refhub.elsevier.com/S0966-9795(19)30025-1/sref15http://refhub.elsevier.com/S0966-9795(19)30025-1/sref15http://refhub.elsevier.com/S0966-9795(19)30025-1/sref16http://refhub.elsevier.com/S0966-9795(19)30025-1/sref16http://refhub.elsevier.com/S0966-9795(19)30025-1/sref17http://refhub.elsevier.com/S0966-9795(19)30025-1/sref17http://refhub.elsevier.com/S0966-9795(19)30025-1/sref18http://refhub.elsevier.com/S0966-9795(19)30025-1/sref18http://refhub.elsevier.com/S0966-9795(19)30025-1/sref19http://refhub.elsevier.com/S0966-9795(19)30025-1/sref19http://refhub.elsevier.com/S0966-9795(19)30025-1/sref19http://refhub.elsevier.com/S0966-9795(19)30025-1/sref20http://refhub.elsevier.com/S0966-9795(19)30025-1/sref20http://refhub.elsevier.com/S0966-9795(19)30025-1/sref21http://refhub.elsevier.com/S0966-9795(19)30025-1/sref21http://refhub.elsevier.com/S0966-9795(19)30025-1/sref21http://refhub.elsevier.com/S0966-9795(19)30025-1/sref22http://refhub.elsevier.com/S0966-9795(19)30025-1/sref22http://refhub.elsevier.com/S0966-9795(19)30025-1/sref22http://refhub.elsevier.com/S0966-9795(19)30025-1/sref23http://refhub.elsevier.com/S0966-9795(19)30025-1/sref23http://refhub.elsevier.com/S0966-9795(19)30025-1/sref24http://refhub.elsevier.com/S0966-9795(19)30025-1/sref24http://refhub.elsevier.com/S0966-9795(19)30025-1/sref24http://refhub.elsevier.com/S0966-9795(19)30025-1/sref25http://refhub.elsevier.com/S0966-9795(19)30025-1/sref25

torsion, Mater. Sci. Eng., A 735 (2018) 394–397.[26] W.S.G. Dan Sathiaraj, Aurimas Pukenas, Rolf Schaarschuch, R. Jose Immanuel,

S.K. Panigrahi, J. Arout Chelvaned, S.S. Satheesh Kumard, Effect of annealing onthe microstructure and texture of cold rolled medium-entropy alloy, Intermetallics101 (2018) 87–98.

[27] T.-H. Chuang, H.-C. Wang, C.-H. Tsai, et al., Thermal stability of grain structure andmaterial properties in an annealing-twinned Ag–8Au–3Pd alloy wire, Scripta Mater.67 (6) (2012) 605–608.

[28] G.D. Sathiaraj, M.Z. Ahmed, P.P. Bhattacharjee, Microstructure and texture ofheavily cold-rolled and annealed fcc equiatomic medium to high entropy alloys, J.Alloy. Comp. 664 (2016) 109–119.

[29] P. P Bhattacharijee, G. D Sathiaraj, M. Zaid, et al., Microstructure and textureevolution during annealing of equiatomic CoCrFeMnNi high-entropy alloy, J. Alloy.Comp. 587 (7) (2014) 544–552.

[30] P.P. Bhattacharjee, S.K. Sinha, A. Upadhyaya, Effect of sintering temperature ongrain boundary character distribution in pure nickel, Scripta Mater. 56 (1) (2007)13–16.

[31] A.J. Zaddach, C. Niu, C.C. Koch, et al., Mechanical properties and stacking faultenergies of NiFeCrCoMn high-entropy alloy, JOM (J. Occup. Med.) 65 (12) (2013)1780–1789.

[32] S. Huang, W. Li, S. Lu, et al., Temperature dependent stacking fault energy ofFeCrCoNiMn high entropy alloy, Scripta Mater. 108 (2015) 44–47.

[33] P.J. Hhumphreys, A network mod for recovery and recrystallisation, Scripta Mater.27 (1992) 1557–1562.

[34] S.J. Dillon, G.S. Rohrer, Mechanism for the development of anisotropic grainboundary character distributions during normal grain growth, Acta Mater. 57 (1)(2009) 1–7.

[35] Q. Li, J.R. Cahoon, N.L. Richards, On the calculation of annealing twin density,Scripta Mater. 55 (12) (2006) 1155–1158.

[36] B.B. Rath, M.A. Imam, C.S. Pande, Nucleation and growth of twin interfaces in fccmetals and alloys, Mater. Phys. Mech. 1 (2000) 61–66.

[37] Y. Jin, B. Lin, M. Bernacki, et al., Annealing twin development during re-crystallization and grain growth in pure nickel, Mater. Sci. Eng., A 597 (8) (2014)295–303.

[38] A.J.S. Mukul Kumar, Wayne E. King, Microstructural evolution during grainboundary engineering of low to medium stacking fault energy fcc materials, ActaMater. 50 (2002) 2599–2612.

[39] K.H. Song, Y.B. Chun, S.K. Hwang, Direct observation of annealing twin formation

in a Pb-base alloy, Mater. Sci. Eng., A 454 (16) (2007) 629–636.[40] Q. Li, J.R. Cahoon, N.L. Richards, Effects of thermo-mechanical processing para-

meters on the special boundary configurations of commercially pure nickel, Mater.Sci. Eng., A 527 (1) (2009) 263–271.

[41] Y. Jin, B. Lin, M. Bernacki, et al., Annealing twin development during re-crystallization and grain growth in pure nickel, Mater. Sci. Eng., A 597 (2014)295–303.

[42] H. Gleiter, The formation of annealing twins, Acta Metall. 17 (12) (1969)1421–1428.

[43] C.S.P.S. Mahhjan, M.A. Imam, Formation of annealing twin in fcc crystal, ActaMater. 45 (4) (1997) 2633–2638.

[44] Y. Ma, F. Yuan, M. Yang, et al., Dynamic shear deformation of a CrCoNi medium-entropy alloy with heterogeneous grain structures, Acta Mater. 148 (2018)407–418.

[45] L. Lu, Z.S. You, K. Lu, Work hardening of polycrystalline Cu with nanoscale twins,Scripta Mater. 66 (11) (2012) 837–842.

[46] G. Laplanche, A. Kostka, O.M. Horst, et al., Microstructure evolution and criticalstress for twinning in the CrMnFeCoNi high-entropy alloy, Acta Mater. 118 (2016)152–163.

[47] Y. Lu, X. Gao, L. Jiang, et al., Directly cast bulk eutectic and near-eutectic highentropy alloys with balanced strength and ductility in a wide temperature range,Acta Mater. 124 (2017) 143–150.

[48] M.M.M. Z Zhang, J. wang, B. Gludovatz, Nanoscale origins of the damage toleranceof the high-entropy alloy CrMnFeCoNi, Nat. Commun. 6 (2015) 10143.

[49] I. Moravcik, J. Cizek, Z. Kovacova, et al., Mechanical and microstructural char-acterization of powder metallurgy CoCrNi medium entropy alloy, Mater. Sci. Eng.,A 701 (2017) 370–380.

[50] C.C. Hays, P. Kim, W.L. Johnson, Shear, Band Pattern Formation in Bulk MetallicGlasses Containing In-Situ Formed Ductile Phase Dendrite Dispersions, (2000).

[51] J.W. Cui, R.T. Qu, F.F. Wu, et al., Shear band evolution during large plastic de-formation of brittle and ductile metallic glasses, Phil. Mag. Lett. 90 (8) (2010)573–579.

[52] M.A. Hemphill, T. Yuan, G.Y. Wang, et al., Fatigue behavior of Al0.5CoCrCuFeNihigh entropy alloys, Acta Mater. 60 (16) (2012) 5723–5734.

[53] J.D. James, J.A. Spittle, S.G.R. Brown, et al., A review of measurement techniquesfor the thermal expansion coefficient of metals and alloys at elevated temperatures,Meas. Sci. Technol. 12 (12) (2001) R1.

X.L. An, et al. Intermetallics 110 (2019) 106477

7

http://refhub.elsevier.com/S0966-9795(19)30025-1/sref25http://refhub.elsevier.com/S0966-9795(19)30025-1/sref26http://refhub.elsevier.com/S0966-9795(19)30025-1/sref26http://refhub.elsevier.com/S0966-9795(19)30025-1/sref26http://refhub.elsevier.com/S0966-9795(19)30025-1/sref26http://refhub.elsevier.com/S0966-9795(19)30025-1/sref27http://refhub.elsevier.com/S0966-9795(19)30025-1/sref27http://refhub.elsevier.com/S0966-9795(19)30025-1/sref27http://refhub.elsevier.com/S0966-9795(19)30025-1/sref28http://refhub.elsevier.com/S0966-9795(19)30025-1/sref28http://refhub.elsevier.com/S0966-9795(19)30025-1/sref28http://refhub.elsevier.com/S0966-9795(19)30025-1/sref29http://refhub.elsevier.com/S0966-9795(19)30025-1/sref29http://refhub.elsevier.com/S0966-9795(19)30025-1/sref29http://refhub.elsevier.com/S0966-9795(19)30025-1/sref30http://refhub.elsevier.com/S0966-9795(19)30025-1/sref30http://refhub.elsevier.com/S0966-9795(19)30025-1/sref30http://refhub.elsevier.com/S0966-9795(19)30025-1/sref31http://refhub.elsevier.com/S0966-9795(19)30025-1/sref31http://refhub.elsevier.com/S0966-9795(19)30025-1/sref31http://refhub.elsevier.com/S0966-9795(19)30025-1/sref32http://refhub.elsevier.com/S0966-9795(19)30025-1/sref32http://refhub.elsevier.com/S0966-9795(19)30025-1/sref33http://refhub.elsevier.com/S0966-9795(19)30025-1/sref33http://refhub.elsevier.com/S0966-9795(19)30025-1/sref34http://refhub.elsevier.com/S0966-9795(19)30025-1/sref34http://refhub.elsevier.com/S0966-9795(19)30025-1/sref34http://refhub.elsevier.com/S0966-9795(19)30025-1/sref35http://refhub.elsevier.com/S0966-9795(19)30025-1/sref35http://refhub.elsevier.com/S0966-9795(19)30025-1/sref36http://refhub.elsevier.com/S0966-9795(19)30025-1/sref36http://refhub.elsevier.com/S0966-9795(19)30025-1/sref37http://refhub.elsevier.com/S0966-9795(19)30025-1/sref37http://refhub.elsevier.com/S0966-9795(19)30025-1/sref37http://refhub.elsevier.com/S0966-9795(19)30025-1/sref38http://refhub.elsevier.com/S0966-9795(19)30025-1/sref38http://refhub.elsevier.com/S0966-9795(19)30025-1/sref38http://refhub.elsevier.com/S0966-9795(19)30025-1/sref39http://refhub.elsevier.com/S0966-9795(19)30025-1/sref39http://refhub.elsevier.com/S0966-9795(19)30025-1/sref40http://refhub.elsevier.com/S0966-9795(19)30025-1/sref40http://refhub.elsevier.com/S0966-9795(19)30025-1/sref40http://refhub.elsevier.com/S0966-9795(19)30025-1/sref41http://refhub.elsevier.com/S0966-9795(19)30025-1/sref41http://refhub.elsevier.com/S0966-9795(19)30025-1/sref41http://refhub.elsevier.com/S0966-9795(19)30025-1/sref42http://refhub.elsevier.com/S0966-9795(19)30025-1/sref42http://refhub.elsevier.com/S0966-9795(19)30025-1/sref43http://refhub.elsevier.com/S0966-9795(19)30025-1/sref43http://refhub.elsevier.com/S0966-9795(19)30025-1/sref44http://refhub.elsevier.com/S0966-9795(19)30025-1/sref44http://refhub.elsevier.com/S0966-9795(19)30025-1/sref44http://refhub.elsevier.com/S0966-9795(19)30025-1/sref45http://refhub.elsevier.com/S0966-9795(19)30025-1/sref45http://refhub.elsevier.com/S0966-9795(19)30025-1/sref46http://refhub.elsevier.com/S0966-9795(19)30025-1/sref46http://refhub.elsevier.com/S0966-9795(19)30025-1/sref46http://refhub.elsevier.com/S0966-9795(19)30025-1/sref47http://refhub.elsevier.com/S0966-9795(19)30025-1/sref47http://refhub.elsevier.com/S0966-9795(19)30025-1/sref47http://refhub.elsevier.com/S0966-9795(19)30025-1/sref48http://refhub.elsevier.com/S0966-9795(19)30025-1/sref48http://refhub.elsevier.com/S0966-9795(19)30025-1/sref49http://refhub.elsevier.com/S0966-9795(19)30025-1/sref49http://refhub.elsevier.com/S0966-9795(19)30025-1/sref49http://refhub.elsevier.com/S0966-9795(19)30025-1/sref50http://refhub.elsevier.com/S0966-9795(19)30025-1/sref50http://refhub.elsevier.com/S0966-9795(19)30025-1/sref51http://refhub.elsevier.com/S0966-9795(19)30025-1/sref51http://refhub.elsevier.com/S0966-9795(19)30025-1/sref51http://refhub.elsevier.com/S0966-9795(19)30025-1/sref52http://refhub.elsevier.com/S0966-9795(19)30025-1/sref52http://refhub.elsevier.com/S0966-9795(19)30025-1/sref53http://refhub.elsevier.com/S0966-9795(19)30025-1/sref53http://refhub.elsevier.com/S0966-9795(19)30025-1/sref53

Effects of heat treatment on the microstructure and properties of cold-forged CoNiFe medium entropy alloyIntroductionMaterials and methodsExperimental materials and the annealing processMicrostructural characterizationDetermination of mechanical properties

ResultsMechanical properties of CoNiFe MEAs

DiscussionRelationship between mechanical behavior and microstructureFracture mechanismThermomechanical behavior

ConclusionAcknowledgementsReferences