Materials Characterization 96 (2014) 84–92

Contents lists available at ScienceDirect

Materials Characterization

j ourna l homepage: www.e lsev ie r .com/ locate /matchar

Experimental and theoretical investigation on the compressive behaviorof aluminum borate whisker reinforced 2024Al composites

Cheng Yang, Yingying Zong ⁎, Zhenzhu Zheng, Debin Shan ⁎National Key Laboratory for Precision Hot Processing of Metals, Harbin Institute of Technology, Harbin 150001, ChinaSchool of Materials Science and Engineering, Harbin Institute of Technology, Harbin 150001, China

⁎ Corresponding authors at: P.O. Box 435, No. 92 WesChina.

E-mail addresses: hagongda@hit.edu.cn (Y. Zong), d.b.

http://dx.doi.org/10.1016/j.matchar.2014.07.0241044-5803/© 2014 Elsevier Inc. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 12 May 2014Received in revised form 17 July 2014Accepted 24 July 2014Available online 25 July 2014

Keywords:Metal matrix composites (MMCs)Compressive deformationMicrostructuresMechanical propertiesFractureHot workability

The compressive behavior of Al18B4O33w/2024Al composites fabricated by squeeze casting was investigatedunder low and elevated temperature. Microstructure shows that the compression exerts a significant effect onwhisker fracture and rotation. The theory of synergistic effects caused by different strengthening mechanismsis used to predict the yield strength. Experiments show that compressive yield strength of composites improvesby 47% comparedwith those of 2024Al at 623 K and agrees relatively well theoretical value. The compressive de-formation depends onmatrixmainly at lower temperature and themain failuremode is shear fracture. Addition-ally, fracturemechanisms are investigated further through fracture surface analysis. During hot compression, thepredominated softening mechanisms also include dynamic recrystallization and strain softening except for dy-namic recovery, which corresponds well with the shape of flow curves, microstructural observation and changeof activation energy. Lastly, the optimumprocess parameters are determined to be about 0.1 s−1 and 723 K basedonDynamicMaterialModel and validated bymicrostructure evolution. Experiments show that the strain rate hasa mixed effect on whisker fracture.

© 2014 Elsevier Inc. All rights reserved.

1. Introduction

Metal matrix composites (MMCs) have aroused enormous concernduring the past several decades, and aluminum matrix composites areregarded as promising materials because of the high specific modulus,high specific strength, excellent dimensional stability and goodwear re-sistance [1–6]. Therefore, they have been considered to be applied inmany fields, such as automotive and aerospace.

Aluminumboratewhiskers (Al18B4O33w) are regarded as a potentialkind of whiskers to reinforce aluminummatrix with the advantages ofexcellent mechanical qualities and low cost in comparison with others,which can broaden the application in industry further [6–9]. As the per-formance of the composites depends on the aluminummatrix to a greatextent, 2024Al alloy is widely considered to the promising matrixmaterial because of its high strength, age hardenability and good work-ability [10,11]. Due to the existence of magnesium and copper in the2024Al alloy, the interfacial reaction products [12,13] and intermetallics[14,15] can be formed in Al18B4O33w/2024Al comparing to Al18B4O33w/pure Al, which will have an influence on the mechanical properties.

Generally, the strengtheningmechanisms of aluminummatrix com-posites can be investigated by different models due to the addition ofthe reinforcement [16,17]. The improvement of composites can be

t Dazhi Street, Harbin 150001,

shan@gmail.com (D. Shan).

considered in perspective of the effective load transfer, enhanced dislo-cations triggered by thermal and geometrical mismatch, Orowanstrengthening and so on [18–20]. However, the strength improves atthe cost of reduction in plasticity. Fortunately, they show preferableplasticity when formed at elevated temperature [21]. In view of this, se-ries of hot forming methods have been adopted, such as hot extrusion,hot rolling, hot compression and isothermal forging [6,22–27]. Factually,many defects still exist after hot forming, such as surface cracking andwhisker fracture. Considering the widespread application in differentfields, the workability analysis of aluminummatrix composites is indis-pensable. Since the forming quality is associated with the many param-eters, researchers have conducted the experimental evaluation inconjunction with finite element methods [22]. Especially, as for hotcompression, the processing map and kinetic analysis are effectiveways to optimize the processing parameters to guarantee the serviceperformance of composites [26,27].

Recently, some tests concerning the tensile behavior and thermalstability of Al18B4O33w/2024Al with different coatings have been stud-ied [28,29], and some researchers have demonstrated that the laser sur-face treatment can be adopted to improve corrosion properties of thecomposites [14]. However, researches related to compressive behaviorof Al18B4O33w/2024Al which is vital to explore the mechanical proper-ties, fracture behavior, softening mechanism and suitable process pa-rameters during hot compression have not been analyzed in detail.

In the present work, the compressive behavior of Al18B4O33w/2024Al under low and elevated temperature was investigated, and

85C. Yang et al. / Materials Characterization 96 (2014) 84–92

compared with the 2024Al alloy. The strengthening mechanism andsurface fracture analysis were analyzed correspondingly. Particularly,the softening mechanisms of composites during hot compressionwere studied, the hot workability of composites was analyzed throughkinetic analysis and processing map based on Dynamic MaterialModel in combination with microstructure evolution additionally.

2. Experimental

Al18B4O33wwhiskers with the length of 30–100 μm and diameter of0.5–2.5 μm were used to reinforce 2024Al alloy by the squeeze castingtechnique. The volume fraction (Vf) of whiskers was 20%. The whiskerswere distributed uniformly and oriented randomly in the 2024Al alloy.As for thematrix, themain chemical composition of 2024Al alloy exceptAl was 4.79% Cu, 1.49% Mg, 0.611% Mn, and 0.245% Fe. In the process ofthe squeeze casting, the preheating temperature of moldwas 773 K andthematrix alloy was poured at 1073 K. After fabrication, the compositesare solution-heated at 763 K and quenched down to the roomtemperature.

The compressive experiments were performed on Gleeble-1500thermal–mechanical simulator at room temperature, 473 K andelevated temperature (623 K to 773 K at 50 K intervals) witha strain rate range between 0.001 and 1 s−1 in a vacuum. Thecompressive specimens were 8 mm × 12 mm and compressive reduc-tion is 60% for hot compression. Specimens were firstly heated for3 min and then compressed. Graphite was used as lubricant to reduce

Fig. 1.Micrographs of (a) SEM, (b) TEM image of as-cast Al18B4O33w/2024A

the friction and the composites were quenched by water aftercompression.

The microstructure of the composites and fractographs during com-pression were observed by a Quanta 200FEG and Helios Nanolab600iscanning electron microscope with energy-dispersive spectroscopy(EDS). These graphs concerning microstructure evolution are takenfrom the section parallel to the direction of compression. The interfacebetweenwhiskers and 2024Al and specificmicrostructurewere investi-gated using a Tecnai G2 F30 transmission electron microscope with anaccelerating voltage of 120 kV. Meanwhile, the samples for TEM obser-vation were thinned by ion milling.

3. Results and discussion

3.1. Microstructure of composites

Fig. 1 shows themicrographs of as-castAl18B4O33w/2024Al and com-pressed composites. It is clear that the whiskers break seriously duringthe fabrication with regard to the original materials. Large amount ofhard phase CuAl2 can be also observed, as proved by energy-dispersivespectroscopy (EDS) analysis (Fig. 1a). Except for hard phase CuAl2, afew reaction products MgAl2O4 can be found between 2024Al alloyand whiskers (Fig. 1b). It can be concluded that compression exerts asignificant effect on whisker fracture and rotation according to Fig. 1c.Many microcracks can be found around the clusters of broken whisker.In addition, whisker fracture is more serious around CuAl2.

l and (c) SEM image of compressed composites at 623 K and 0.1 s−1.

86 C. Yang et al. / Materials Characterization 96 (2014) 84–92

3.2. Strengthening mechanism

Compression test results of composites and matrix at various tem-peratures concerning the yield strength and compressive fracture strainare summarized in Table 1.

It is clear that the compressive yield strengthσcom and fracture strainεcom of composites are affected slightly by temperature change below473 K. When the temperature reaches 623 K, σcom decreases drasticallyand εcom increases sharply. 2024Al alloy presents similar characteristicwith better plasticity. Under the same parameters, the yield strengthof composites is higher than that of 2024Al alloy and improves by 47%at 623 K. Particularly, the strengthening effect at 623 K is more obviousthan that at room temperature and 473 K, which can be regarded as anillustration that the compressive deformation of composites depends onthe matrix at lower temperature while the composites show increasingcompatible deformation between the matrix and whisker at highertemperature.

The strengthening mechanisms for Al18B4O33w/2024Al can bediscussed from following aspects [18–20,30]. (I) Load transfer. The ef-fective load transfer from matrix to whisker is the main strengtheningmechanism for composites. Generally, the transfer of load dependsmainly on the volume fraction of reinforcement and interface betweenmatrix andwhisker, which can be described by shear lagmodel approx-imately. In order to predict the efficiency of load transfer and thestrength of discontinuous reinforced composites accurately, some re-searchers have developed the model taking the orientation effect ofshort fiber into account [31]. The improvement of compressive yieldstrength of short fiber reinforced composites can be expressed as thefollowing equation concerning the load transfer.

Δσ load ¼ 12σmV f s cos2θþ 3π−4

3π

� �1þ 1=sð Þsin2θ

� �ð1Þ

where θ is the angle between the orientation and compression direction(average value is π/4), σm is the yield strength of 2024Al alloy, the vol-ume fraction of whisker (Vf) is 20%, s is the average aspect ratio of whis-kers after fabrication and the value is 8.

(II) High density dislocation caused by thermal mismatch and geo-metrically necessary dislocations. Considering the fact that the matrixand whisker are not in equilibrium state when the composites arecooled from higher temperature to lower temperature, the mismatchin the coefficients of thermal expansion (CTE) between matrix andwhiskers induces the high-density dislocation in matrix and can beexpressed by [32]:

ρt ¼BV f αm−α f

� �ΔT

b 1−V f

� �t

ð2Þ

where B is a geometrical constant=10 forwhisker, b is 2.86nm for Bur-gers vector [18],αm andαf are the thermal expansion coefficients ofma-trix andwhisker (23.6 × 10−6 and 2.6 × 10−6 K−1 respectively), t is theminimum size of whisker (1.5 μm), ΔT is the difference between com-pression and annealing temperature.

Table 1Compression test results of composites andmatrix at various temperatures concerning theyield strength and compressive fracture strain.

Temperature (K) Room temperature 473 623

σcom (MPa) 394.8 364.8 169.7σal (MPa) 372.2 334.4 115.2εcom 0.24 0.28 0.6εal 0.36 0.38 N0.6

On the other hand, except for the thermal mismatch, the discordantdeformation betweenwhiskers andmatrixwould result in geometrical-ly necessary dislocations as well, which can be given as [18–20]:

ρg ¼ 8V f εpbt

ð3Þ

where εp is the plastic strain and can be taken as 0.2%.Then the improvement of yield strength based on dislocation model

can be calculated as:

ΔσD ¼ λGmbffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiρt þ ρg

p ð4Þ

where λ is a constant = 1.25 and Gm is 28.3 GPa for the shear modulus.(III) Orowan strengthening. The existence of the whiskers inhibits

the movement of dislocation and the dislocations are inclined to bend,which gives rise to the strengthening of composites. The followingequation can be use to describe the contribution concerning thisstrengthening mechanism [33].

Δσorowan ¼ 0:13Gmb

t 12 =V f

� �13−1

� � lnt2b

ð5Þ

Considering the fact that the synergistic effects and interactions be-tween different mechanisms, the compressive yield strength accordingto the theory of Goh [30] can be expressed as:

Δσ ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiΔσ load

2 þ ΔσD2 þ Δσorowen

2q

ð6Þ

Substituting Eqs. (1), (4) and (5) and into Eq. (6), the theoreticalyield strength of the composites can be calculated to be 553 MPa(room temperature), 494MPa (473 K) and 186MPa (623 K) respective-ly. The larger gap between the theoretical yield strength and experi-mental values and at low temperature can further demonstrate thatthe deformation depends mainly on matrix. The theoretical yield coin-cides relatively well with experimental values at 623 K, and the differ-ence can be attributed to softening caused by whisker rotation [6].Certainly, the existence of MgAl2O4 should be taken into accountwhile the assumption that the interface is perfect in model.

3.3. Fracture surface analysis

Fig. 2 shows the macromorphology of composites at different tem-peratures after compression (0.1 s−1). It can be seen that macroscopicfailure of composites is shear fracture at room temperature and 473 K.Many longitudinal surface cracks can be observed on the surface at623 K and the quality of compressed sample is qualified at 723 K.

The fractographs of composites at different temperatures are pre-sented in Fig. 3. Because of the less deformation of whisker comparingto matrix and the poor compatibility between them at room tempera-ture and 473 K, the deformation of composites was controlled mainlyby matrix. The composites generate shear deformation and the com-pressed sample cracks along 45° with respect to the compressive direc-tion in priority. The characteristic of groove with many microvoids

Fig. 2. The macro photo of Al18B4O33w/2024Al at different temperatures aftercompression.

Fig. 3. SEM images of the fractographs at (a) (b) room temperature, (c) 473 K and (d–f) 623 K.

87C. Yang et al. / Materials Characterization 96 (2014) 84–92

which is contributed to themotion of whisker to a certain extent can beclearly found at room temperature, as shown in Fig. 3a. Factually, due tothe existence of the particle CuAl2, the matrix in vicinity of the CuAl2 islikely to generate the crack (Fig. 3b). Certainly, the particle CuAl2 is in-clined to crack caused by high local stress in turn. When the tempera-ture reaches 473 K, the proportion of the region representing the flowlocalization increases and thematrix has to accommodate the deforma-tion between the region of plasticity and non-plasticity. Consequently,damage would be caused by plasticity instability [27,34]. This can beproved by the secondary crack in Fig. 3c. The occurrence of localized

plasticity signifies the improvement of the plastic deformation capacity,nevertheless, the whole deformation process shows typical characteris-tic of brittle fracture.

The compressive fracture of composites at 623 K and 0.1 s−1 belongsto ductile fracturewith lots of dimples, as presented in Fig. 3d. The com-pressive fracture of composite is related to: (I)whisker fracture. Becauseof the obviousmatrix–whisker andwhisker–whisker interactions underlarge degree of deformation, the degree of whisker fracture is seriousand many cavities nucleate around the whisker clusters (Fig. 1c). (II)Whisker–matrix decohesion. Somemicrovoidswhich are representative

88 C. Yang et al. / Materials Characterization 96 (2014) 84–92

of the decohesion between whisker and matrix can be observed and il-lustrated by black arrow in Fig. 3e. This is partly due to the fact thatthe rise of temperaturemakes the interfacial bonding strength decrease.(III) Crack initiation around the interface. The stress concentrationaround the interface caused by the mismatch in coefficients of thermalexpansion (CTE) and discordant deformation cannot be counteractedby matrix flow totally, which leads to the result that the stress is closeto or more than ultimate compressive strength of matrix, then thecrack is initiated in the matrix around interface (Fig. 3e). The interfaceis also inclined to generate cracks because of the brittle productMgAl2O4 [28]. (IV) Crack initiation around CuAl2. Taking account intothe high dislocation density caused by CuAl2, the region around hardparticle CuAl2 can be the origin of cracks as well (see Fig. 3f). Corre-spondingly, whisker fracture is more serious around the hard particle-rich region. (IV) Matrix ductile failure. Discontinuous flow localizationwould help initiate and propagate microcracks in matrix directly (seeFig. 3f). Considering the fact that the matrix undergoes large plastic de-formation, the fracture shows the obvious characteristic of matrix duc-tile failure. In addition, it is worth noting that the tear ridges whichcan be defined as smaller and shallower dimples (Fig. 3f) denote thatthe rate of crack propagation is relatively fast. Generally, the fracture re-sistance of composites would increase to a degree because the probabil-ity of interaction between crack and whisker is increased through crackbridging and crack deflection. The model can be expressed as [35,36]:

ϕ ¼ GIC

Gmatrix∝s ð7Þ

where GIC and Gmatrix are the critical energy release rates of fracture re-sistance with and without reinforced short fiber respectively. The im-provement of fracture toughness is positively related to length–

Fig. 4. Stress–strain curves of (a) 2024Al alloy and (b) Al18B4O33w/2024Al at different temperadifferent strain rates (T = 673 K).

diameter ratio. Therefore, the broken whisker after large compressionwould not help control the expansion of cracks effectively. On theother hand, the microcracks caused by the interface, CuAl2 and whiskerfracture would link together and provide channel for crack propagationto the whole matrix.

3.4. True stress–strain curves and softening behavior

The flow stresses of both 2024Al and Al18B4O33w/2024Al increasegradually with decreasing temperature and increasing strain rate, aspresented in Fig. 4. After the initial stage of work hardening, thestress–strain curves of composites exhibit a gradual decline with in-creasing strain while those of 2024Al approximately enter into steadyconditions.

As for 2024Al, the true stress and strain curves suggest that the pre-dominated softeningmechanism is dynamic recovery (DRV) during hotcompression, corresponding to the fact that possibility of occurrence ofdynamic recrystallization (DRX) is reduced as the stacking fault energyof 2024Al alloy is relatively high. As for the composites, the dynamic re-covery is also an essential part and subgrain structure indicative of DRVcan be observed in Fig. 5a. The flow stresses of composites decrease rap-idly with the increasing strain after reaching the peak, which is general-ly regarded as strain softening caused bywhisker rotation [6]. Factually,the more reasonable explanation is the combined effect of strainsoftening and dynamic recrystallization. The pile-up of dislocationsand increase of the dislocation density caused by the existence ofwhisker promote the nucleation of DRX. In Fig. 5b, the high dislocationdensity can be observed in composites, especially the region aroundthe interface. Moreover, some equal-axial grains would prove theoccurrence of DRX. To sum up, the softening mechanisms of the

tures (ε̇= 0.1 s−1); stress–strain curves of (c) 2024Al alloy and (d) Al18B4O33w/2024Al at

Fig. 5. TEMmicrograph of Al18B4O33w/2024Al composites compressed at 0.1 s−1 and 723 K showing (a) subgrain morphology and (b) dynamic recrystallization.

89C. Yang et al. / Materials Characterization 96 (2014) 84–92

Al18B4O33w/2024Al are the superposition of three aspects: dynamic re-covery, dynamic recrystallization and strain softening.

The relationships between peak flow stress and temperature of theAl18B4O33w/2024Al and 2024Al alloy are shown in Fig. 6a. The peakflow stress of the composites is much higher than that of 2024Al alloy,and the difference between them decreases with increasing tempera-ture, indicating that the softening effect of composites is more evident.As for the composites, except for the increasing softening effect in ma-trix caused by DRX and whisker rotation, the bonding strength andthe ability to transfer load between matrix and whiskers decrease asthe deformation temperature increases, which gives rise to the declineof difference.

The peak flow stress of the composites and matrix is positively cor-related with strain rate, as shown in Fig. 6b. Nevertheless, the strainrate has mixed effect on the difference of the peak flow stress betweenthem. The difference is biggest at 1 s−1 and this can be explained as fol-lows. Thewhiskers are inclined to break rather than rotate at high strainrate as the rotation of the whisker requires enough time and thegeometrical dislocations improve with increasing strain rate, and thenthe discordant deformation between the whisker and matrix makesthe flow stress increase promptly. Unexpectedly, the difference at0.001 s−1 is bigger than that at 0.01 s−1 and 0.1 s−1. This is due to seri-ous whisker fracture caused by matrix–whisker and whisker–whiskerinteractions for a long time at 0.001 s−1, and the deformation resistanceof composites increases as a consequence.

Fig. 6. Peak flow stress of the composites and 2024Al alloy

3.5. Kinetic analysis and processing maps with microstructure evolution

As two important kinetic parameters, the strain-rate sensitivity coef-ficientm and apparent activation energy Q are widely used to describethe intrinsic deformation behavior of metal and MMCs [37,38]. Basedon the standard kinetic rate equation, Sellars and Tegart [39,40] pro-posed the hyperbolic sine law containing the strain rate, temperatureand activation energy to predict the flow stress of hot deformation be-havior. This Arrhenius equation is given by:

ε̇¼ A sinh ασð Þ½ �n exp −Q=RTð Þ ð8Þ

where σ is the flow stress,ε̇ is the strain rate, T is the compressive tem-perature. R is the ideal gas constant and α (MPa−1) is a material con-stant independent of temperature. In addition, comparing to thedifferent types of Arrhenius equation (power law and exponentiallaw), Q and m can be given as follows [38]:

m ¼ ∂ ln sinh ασð Þð Þ∂ ln ε̇

� �T

ð9Þ

Q ¼ R∂ ln sinh ασð Þð Þ

∂ 1=Tð Þ� �

ε̇

∂ ln ε̇∂ ln sinh ασð Þð Þ

� �T: ð10Þ

as a function of (a) temperature and (b) strain rate.

Fig. 7. Dependence of apparent activation energy of Al18B4O33w/2024Al on the tempera-ture and strain rate.

90 C. Yang et al. / Materials Characterization 96 (2014) 84–92

After calculation, the dependence of apparent activation energy ofAl18B4O33w/2024Al on temperature and strain rate is illustrated inFig. 7. The activation energy of composites is higher than the self-diffusion activation energy of Al (142 kJ·mol) at different deformationconditions. Generally, the existence of reinforcedwhisker, interfacial re-action product and hard phase CuAl2 would make composites formeddifficultly. The value of Q drops by 25% from 623 K to 673 K, whichcan be regarded as the change in the deformation mechanism. Whenthe temperature reaches 673 K, the deformation mechanism of thema-trix would be converted from the dynamic recovery into combined in-teraction of dynamic recovery and dynamic recrystallization, as aresult, the occurrence of DRX in composites can be further proved. TheQ of composites at 773 K is approximately close to the self-diffusion acti-vation energy of Al and the deformationmechanismof composites is con-trolled by diffusion primarily. In addition, the values of Q at 0.001 s−1 arehigher than those at 0.01–0.1 s−1 correspondingly. Except for seriouswhisker fracture caused by interaction of whiskers and discordant

Fig. 8.Processingmapof Al18B4O33w/2024Al at strain of 0.6withmicrostructure evolution after co(d) 723 K, 0.1 s−1; (f) 773 K, 0.1 s−1.

deformation at relatively low strain rate, there would be enough timefor microvoids and cracks to nucleate, which is consistent with the for-mation ofwedge cracking at higher temperatures and lower strain rates[40,41]. Consequently, the values of Q at 0.001 s−1 are relatively higher.

According to the theory of Dynamic Material Model first presentedby Prasad and Sasidhara [27,34,41–43], the dissipated power duringhot deformation includes the heat generated in the hot deformationand the power contributed to microstructure evolution. Consideringthe difference between the characteristic of non-linear and an ideal lin-ear dissipation, the efficiency of power dissipation η can expressed as:

η ¼ 2m= mþ 1ð Þ: ð11Þ

Except for the power dissipationmap, inabilitymap is also necessaryfor processing map. The criterion developed by Prasad is used to obtainthe inability map, which can be mathematically presented as:

ξ ε̇ ¼ ∂ ln m= mþ 1ð Þ½ �

∂ ln ε̇ þmb0 ð12Þ

where ξ(̇ε) is the dimensionless parameter.When the value of ξ(̇ε) is neg-ative, that mean series of deformation defects will occur. Thus, the re-gion that exhibits flow instability must try to be avoided.

Fig. 8 shows the processingmap of composites at a strain of 0.6 withthemicrostructure evolution (Fig. 8a–e) at different domains. The shad-ed areas indicate the regions of instability and the power dissipation ef-ficiency is listed aswell. Since the strengthening effect of the compositesis positively related to length–diameter ratio, the microstructure evolu-tion concerning whisker fracture is used to validate the optimum pro-cess parameters for hot compression further. Additionally, the degreeof whisker fracture can be judged according to the change of whiskerlength measured in Fig. 9. The whisker fracture after hot compressionis negatively related to temperature except at 773 K. The strain ratealso hasmixed effect on the whisker fracture. In view of this, the micro-structure evolution concerning whisker fracture agrees well with theanalysis of true stress–strain curves and change of apparent activationenergy.

mpression at different parameters: (a) 623K, 0.001 s−1; (b) 623K, 1 s−1; (c) 673K, 0.1 s−1;

Fig. 9.Whisker length of the composites as a function of temperature and strain rate.

91C. Yang et al. / Materials Characterization 96 (2014) 84–92

Based on Figs. 8 and 9, the hot workability of Al18B4O33w/2024Alcomposites can be analyzed in detail. The power dissipation efficiencyrises with increasing temperature and reaches to maximum at about0.1 s−1 without considering the effect of temperature. The lower tem-perature and higher strain rate would bring about the flow instability.Adiabatic shear bands are estimated to form because of limited timeto dissipate heat at high strain rate [27]. Due to the discordant deforma-tion between whisker and matrix and high deformation resistance, thedegree of whisker fracture is most serious under this condition. Lowerstrain rate is not conducive to hot forming either for the reason thatthe power dissipation efficiency is relatively low. In this region,microvoids and cracks are possible to nucleate for a long time. Especial-ly, the whisker fracture caused by matrix–whisker and whisker–whis-ker interactions cannot be neglected, as shown in Fig. 8a. Though thecomposites compressed at 773 Khave high power dissipation efficiency,the over-burning of 2024Al alloy is likely to occur at this condition andthe microstructure damage such as burnt structure would lead to theshortening ofwhiskers (Fig. 8e), whichgives rise to the reduction inme-chanical properties. The occurrence of precipitate dissolution at hightemperature would also reduce the workability [40]. Generally, highertemperature (723 K) and intermediate strain rate (0.1 s−1) with highpower dissipation efficiency is suitable for hot deformation. Under thiscondition, whisker fracture is not serious and the apparent activationenergy is relatively low. Moreover, the occurrence of DRX analyzed pre-viously will promote the workability for composites by softening thematrix and optimizingmicrostructure. In conclusion, the optimum pro-cess parameters of the composites are about 0.1 s−1 and 723 K.

4. Conclusions

According to the analysis of the compressive behavior ofAl18B4O33w/2024Al under low and elevated temperature, the brief con-clusions are listed as follows:

1. Compressive deformation of composites depends on matrix mainlyat lower temperature. Though whisker rotation and interfacialreaction lower the yield strength slightly, the experimental value at623 K agrees relatively well with theoretical strength calculatedbased on the theory of synergistic effects caused by differentstrengthening mechanisms.

2. Shear fracturewould be themain failuremode at lower temperature.The compressive fracture belongs to ductile fracture at 623 K and thefailure modes are related to matrix ductile failure, whisker fracture,whisker–matrix decohesion, interfacial cracking, and crack initiationaround CuAl2.

3. The predominated softeningmechanism of 2024Al is dynamic recov-ery while the softening mechanisms of composites are dynamic re-covery, dynamic recrystallization and strain softening during hot

compression. Additionally, the softening effect of composites ismore evident as the temperature increases.

4. Activation energy of composites during hot compression is greaterthan the activation energy of self-diffusion and decreases withthe rise of the temperature. In particular, the sharp drops of Q from623 K to 673 K can prove the occurrence the DRX.

5. Whisker fracture after hot compression is negatively related to tem-perature except at 773 K and strain rate also has mixed effect on it,which agreeswell with the change of stress–strain curves and activa-tion energy. In addition, the optimumprocess parameters of compos-ites are about 0.1 s−1 and 723 K based on processing map andmicrostructure evolution.

Acknowledgments

This work was supported by the National Natural Science Founda-tion of China (No. 51275132) and the Postdoctoral Science-ResearchFoundation of Heilongjiang Province (No. LBH-Q13055).

References

[1] S.C. Tjong, Z.Y. Ma, Microstructural and mechanical characteristics of in situ metalmatrix composites, Mater. Sci. Eng. 29 (2000) 49–113.

[2] L.M. Tham, M. Gupta, L. Cheng, Effect of limited matrix-reinforcement interfacial re-action on enhancing the mechanical properties of aluminium–silicon carbide com-posites, Acta Mater. 49 (2001) 3243–3253.

[3] Z.H. Melgarejo, P.J. Resto, D.S. Stone, O.M. Suárez, Study of particle–matrix interac-tion in Al/AlB2 composite via nanoindentation, Mater. Charact. 61 (2010) 135–140.

[4] F. Rotundo, L. Ceschini, A. Morri, T.S. Jun, A.M. Korsunsky, Mechanical and micro-structural characterization of 2124Al/25 vol.%SiCp joints obtained by linear frictionwelding (LFW), Compos. A: Appl. Sci. Manuf. 41 (2010) 1028–1037.

[5] L. Ceschini, G. Minak, A. Morri, Forging of the AA2618/20 vol.% Al2O3p composite:effects on microstructure and tensile properties, Compos. Sci. Technol. 69 (2009)1783–1789.

[6] Z.J. Li, W.D. Fei, L.D. Wang, Hot compressive deformation behavior of aluminumma-trix composite reinforced by Al18B4O33 whisker coated with Bi2O3, Mater. Sci. Eng. A432 (2006) 275–280.

[7] K. Suganuma, T. Fujita, N. Suzuki, K. Niihara, Aluminium composites reinforced witha new aluminium borate whisker, J. Mater. Sci. Lett. 9 (1990) 633–635.

[8] L.J. Yao, G. Sasaki, H. Fukunaga, Reactivity of aluminum borate whisker reinforcedaluminum alloys, Mater. Sci. Eng. A 225 (1997) 59–68.

[9] S.J. Zhu, T. Iizuka, Fatigue behavior of Al18B4O33 whisker-framework reinforced Almatrix composites at high temperatures, Compos. Sci. Technol. 63 (2003) 265–271.

[10] C. Luo, X. Zhou, G.E. Thompson, A.E. Hughes, Observations of intergranular corrosionin AA2024-T351: the influence of grain stored energy, Corros. Sci. 61 (2012) 35–44.

[11] C.K.S. Moy, M. Weiss, J.H. Xia, G. Sha, S.P. Ringer, G. Ranzi, Influence of heat treat-ment on the microstructure, texture and formability of 2024 aluminium alloy,Mater. Sci. Eng. A 552 (2012) 48–60.

[12] L.J. Yao, H. Fukunaga, TEM study on the interfacial reaction of Al18B4O33/Al compos-ites, Scr. Mater. 36 (1997) 1267–1271.

[13] J. Pan, G. Sasaki, L.J. Yao, M. Yoshida, H. Fukunaga, Interfacial reaction and its effecton strength of Al18B4O33/Al composites, Mater. Sci. Technol. 15 (1999) 1044–1048.

[14] J. Hu, P.L. Wu, L.C. Kong, G. Liu, The effect of YAG laser surface treatment on corro-sion resistance of Al18B4O33w/2024Al composite, Mater. Lett. 61 (2007) 5181–5183.

[15] A. Albiter, C.A. León, R.A.L. Drew, E. Bedolla, Microstructure and heat-treatment re-sponse of Al-2024/TiC composites, Mater. Sci. Eng. A 289 (2000) 109–115.

[16] M. Taya, Strengthening mechanisms of metal matrix composites, Mater. Trans. JIM32 (1991) 1–19.

[17] M. Taya, T. Mori, Dislocations punched-out around a short fiber in a short fibermetalmatrix composite subjected to uniform temperature change, Acta Metall. 35 (1987)155–162.

[18] F. Tang, E.I. Anderson, T.G. Herold, H. Prask, Pure Al matrix composites produced byvacuum hot pressing: tensile properties and strengthening mechanisms, Mater. Sci.Eng. A 383 (2004) 362–373.

[19] M. Cabibbo, S. Spigarelli, A TEM quantitative evaluation of strengthening in anMg–REalloy reinforced with SiC, Mater. Charact. 62 (2011) 959–969.

[20] Z. Trojanová, Z. Drozd, S. Kúdela, Z. Száraz, P. Lukáč, Strengthening in Mg–Li matrixcomposites, Compos. Sci. Technol. 67 (2007) 1965–1973.

[21] H. Zhang, Y.S. He, L.X. Li, Tensile deformation and fracture behavior of spray-deposition 7075/15SiCp aluminum matrix composite sheet at elevated tempera-tures, Mater. Charact. 59 (2008) 1078–1082.

[22] E.M. Herba, H.J. McQueen, Influence of particulate reinforcements on 6061materialsin extrusion modeling, Mater. Sci. Eng. A 372 (2004) 1–14.

[23] X. Jiang, M. Galano, F. Audebert, Extrusion textures in Al, 6061 alloy and 6061/SiCpnanocomposites, Mater. Charact. 88 (2014) 111–118.

[24] A. El-Sabbagh, M. Soliman, M. Taha, H. Palkowski, Hot rolling behaviour of stir-castAl 6061 and Al 6082 alloys–SiC fine particulates reinforced composites, J. Mater. Pro-cess. Technol. 212 (2011) 497–508.

[25] P. Cavaliere, Isothermal forging of AA2618 reinforced with 20% of alumina particles,Compos. A: Appl. Sci. Manuf. 35 (2004) 619–629.

92 C. Yang et al. / Materials Characterization 96 (2014) 84–92

[26] B.L. Xiao, J.Z. Fan, X.F. Tian, W.Y. Zhang, L.K. Shi, Hot deformation and processingmap of 15%SiCp/2009 Al composite, J. Mater. Sci. 40 (2005) 5757–5762.

[27] S. Ramanathan, R. Karthikeyan, M. Gupta, Development of processing maps for Al/SiCpcomposite using fuzzy logic, J. Mater. Process. Technol. 183 (2007) 104–110.

[28] H.Y. Yue, E.J. Guo, W.D. Fei, Z.M. Yu, L.P. Wang, Effects of Cr2O3 coating of whiskerson the tensile properties and thermal stability of aluminum borate whiskers rein-forced 2024Al composite, Mater. Sci. Eng. A 533 (2012) 33–37.

[29] H.Y. Yue, E.J. Guo, W.D. Fei, L.P. Wang, Tensile properties and thermal stability ofZnO-coated aluminum borate whiskers reinforced 2024Al composite, Mater. Sci.Eng. A 528 (2011) 2407–2411.

[30] C.S. Goh, J. Wei, L.C. Lee, M. Gupta, Properties and deformation behaviour ofMg–Y2O3 nanocomposites, Acta Mater. 55 (2007) 5115–5121.

[31] H.J. Ryu, S.I. Cha, S.H. Hong, Generalized shear-lag model for load transfer in SiC/Almetal-matrix composites, J. Mater. Res. 18 (2003) 2851–2858.

[32] R.J. Arsenault, N. Shi, Dislocation generation due to differences in coefficients ofthermal expansion, Mater. Sci. Eng. 81 (1986) 175–187.

[33] Z. Zhang, D.L. Chen, Consideration of Orowan strengthening effect in particulate re-inforced metal matrix nanocomposites: a model for predicting their yield strength,Scr. Mater. 54 (2006) 1321–1326.

[34] S.V.S.N. Murty, B.N. Rao, B.P. Kashyap, Instability criteria for hot deformation of ma-terials, Int. Mater. Rev. 45 (2000) 15–26.

[35] K. Badrinarayanan, A.L. McKelvey, K.T. Venkateswara Rao, R.O. Ritchie, Fracture andfatigue-crack growth behavior in ductile-phase toughened molybdenum disilicide:

effects of niobium wire vs particulate reinforcements, Metall. Mater. Trans. A 27A(1996) 3781–3792.

[36] Y. Qiao, X. Kong, Unstable crack advance across a regular array of short fibers inbrittle matrix, Compos. Sci. Technol. 64 (2004) 711–717.

[37] A. Galiyev, R. Kaibyshev, G. Gottstein, Correlation of plastic deformation and dynam-ic recrystallization in magnesium alloy ZK60, Acta Mater. 49 (2001) 1199–1207.

[38] V. Senthilkumar, A. Balaji, R. Narayanasamy, Analysis of hot deformation behavior ofAl 5083–TiC nanocomposite using constitutive and dynamic material models,Mater. Des. 37 (2012) 102–110.

[39] C.M. Sellars, W.J. McTegart, On the mechanism of hot deformation, Acta Metall. 14(1966) 1136–1138.

[40] A. Patel, S. Das, B.K. Prasad, Compressive deformation behaviour of Al alloy (2014)–10 wt.% SiCp composite: effects of strain rates and temperatures, Mater. Sci. Eng. A530 (2011) 225–232.

[41] Y.V.R.K. Prasad, T. Seshacharyulu, Modeling of hot deformation for microstructuralcontrol, Int. Mater. Rev. 43 (1998) 243–258.

[42] Y.V.R.K. Prasad, S. Sasidhara (Eds.), Hot Working Guide: A Compendium of Process-ing Maps, ASM International, Materials Park (OH), 1997.

[43] Y.V.R.K. Prasad, K.P. Rao, M. Gupta, Hot workability and deformation mechanisms inMg/nano-Al2O3 composite, Compos. Sci. Technol. 69 (2009) 1070–1076.