Ma

Aa

b

a

ARRA

KMCT

1

cTamphaFrNys[

tbsWhs

TT

0d

Materials Science and Engineering A 527 (2010) 6870–6878

Contents lists available at ScienceDirect

Materials Science and Engineering A

journa l homepage: www.e lsev ier .com/ locate /msea

icrowave and conventional sintering of 90W–7Ni–3Cu alloys with premixednd prealloyed binder phase

vijit Mondala, Anish Upadhyayaa,∗, Dinesh Agrawalb

Department of Materials Science and Engineering, Indian Institute of Technology Kanpur, Kanpur 208016, IndiaThe Pennsylvania State University, University Park, PA 16802, USA

r t i c l e i n f o

rticle history:eceived 9 April 2010eceived in revised form 8 June 2010

a b s t r a c t

The present study investigates the possibility of consolidating premixed 90W–7Ni–3Cu alloy – desig-nated as 90W–PM (Ni–Cu) – through microwave sintering. An attempt has been made to compare the

ccepted 22 July 2010

eywords:icrowave sintering

results between microwave and conventionally sintered samples. This study also compares the sinteringbehavior of 90W–7Ni–3Cu with prealloyed 90W–PA (Ni–Cu) in both conventional as well as microwavefurnace at various temperatures. The comparative analysis is based on the sintered density, densifica-tion parameter, hardness and microstructures of the samples. The present investigation also includes thevariation of matrix composition as a function of temperature by EPMA analysis. The results show thatmicrowave sintering requires about 75% less processing time than required by conventional method and

ical a

onventional sinteringungsten heavy alloys still provides better phys

. Introduction

Tungsten heavy alloy (WHA) is a group of two-phaseomposites, based on W–Ni–Cu and W–Ni–Fe alloys.ungsten–nickel–copper alloys are widely used for ordinancepplication, electrical contacts of switches, radiation shielding,ass balances, etc. In general the tungsten heavy alloys have been

rocessed through powder metallurgy route since 1930s [1]. Theseeavy alloys contain mainly pure tungsten as principal phase inssociation with a binder phase containing transition metals (Ni,e, Cu, Co) [2]. In W–Ni–Cu alloys, normally the nickel-to-copperatio ranges from 3:2 to 4:1. Price et al. [3] were the first to proposei–Cu as the binder for tungsten heavy alloys. Over the last severalears, these alloys have been extensively investigated for den-ification mechanism, microstructural evolution and properties4–7].

The effect of tungsten and copper powder size variation onhe sintered properties of W–Ni–Cu heavy alloys was carried outy Srikanth and Upadhyaya [8,9]. The effect of composition and

intering temperature on the densification and microstructure of

–Ni–Cu heavy alloys was studied by Ramakrishnan and Upad-yaya [10]. Kuzmic [11] proposed that rapid cooling from theintering temperature prevents the formation of brittle phase in

∗ Corresponding author at: Materials Science and Engineering, Indian Institute ofechnology Kanpur, P.O. IIT Kanpur, Kanpur 208016, UP, India.el.: +91 512 2597672; fax: +91 512 2597505.

E-mail address: anishu@iitk.ac.in (A. Upadhyaya).

921-5093/$ – see front matter © 2010 Elsevier B.V. All rights reserved.oi:10.1016/j.msea.2010.07.074

nd mechanical properties.© 2010 Elsevier B.V. All rights reserved.

order to obtain good mechanical properties. Ariel et al. [12] corre-lated the mechanical properties of W–Ni–Cu system with sinteredmicrostructure. Their study showed that the mechanical propertiesare a function of mean free path between tungsten grains, volumefraction of tungsten grains and the contiguity of tungsten spheroids.

The solubility of tungsten in the liquid binder plays a domi-nant role in determining the mechanical properties of the sinteredW–Ni–Cu alloys. The solubility of tungsten in copper is negli-gible. Even at temperatures as high as 1350 ◦C only 0.04 at.% oftungsten goes in solution with copper [13]. In contrast, tungstenexhibits appreciable solubility (up to 40 wt.%) in nickel. It is there-fore possible to tailor the tungsten solubility, wetting and thedihedral angle and thereby, the accompanying sintering responseand properties of the system by varying the Ni:Cu ratio [10,14].Nowadays, W–Ni–Cu alloys are also being consolidated by employ-ing prealloyed powders [15]. Use of prealloyed powder improveshomogeneity. The homogenization process accelerates sinteringand promotes densification. The alloy formation during sinteringdecreases the material viscosity and, hence, stimulates materialflow under the action of capillary forces [16].

Despite widespread application, difficulties still exist in themanufacture of liquid phase sintered tungsten heavy alloys. Inorder to avoid thermal shock, processing of tungsten heavy alloys inconventional furnace involves heating at a slower rate (<10 ◦C/min)and with isothermal holds at intermittent temperatures. This not

only increases the process time, but also results in significantmicrostructural coarsening during sintering, leading to the degra-dation of mechanical properties. This problem is further aggravatedwhen the initial powder size is extremely fine. Hence, it is envis-

A. Mondal et al. / Materials Science and Engineering A 527 (2010) 6870–6878 6871

Table 1Powder characteristics of as-received elemental W, Ni, Cu and prealloyed Ni–Cu powder.

Property W Ni Cu 69Ni–31Cu

Supplier Widia Inco A Cu Powder International, LLC AmetekProcessing technique Chemical reduction Carbonyl process Gas atomization Gas atomizationPowder shape Irregular Spiky Spherical Spherical

attpidti

irittos[hc

tthumeiRamtm

aiah

2

dtNccp(csgdu

Powder size (�m)D10 2.0 3.8D50 4.2 11.0D90 6.2 31.8

ged that a fast heating rate would mitigate this problem. One ofhe techniques to achieve fast and relatively uniform sintering ishrough microwaves [17]. The W–Ni–Cu alloys are usually pre-ared by mixing the constituent powders. One of the challenges

n the processing of these alloys is to ensure homogenous meltistribution [15]. Nickel and copper are known to interdiffuse atemperatures higher than 900 ◦C. However, in case of rapid sinter-ng in microwave this may not be readily feasible.

Microwave heating is much more uniform at a rapid rate result-ng in reduction of processing time and energy consumption. Alsoapid heating leads to finer microstructure enhancing the mechan-cal properties. Clark and Sutton [18] cited many other benefits ofhe process such as precise and controlled heating, environmen-ally friendly etc. Microwave heating is a very sensitive functionf the material being processed and depends on several factors,uch as sample size, as-pressed density, its mass and geometry19,20]. Though there have been attempts to explain microwaveeating of metal powders, still there is not yet any consensus on aomprehensive theory to explain the mechanism [21].

Applicability of microwave sintering to metals was ignored dueo the fact that they reflect microwaves. Roy et al. [22] reportedhat particulate metals can be heated rapidly in microwaves. Thisas led to the use of microwaves to consolidate a range of partic-late metals and alloys [23–25]. While researchers have reportedicrostructural refinement due to rapid heating in microwaves, the

ffect of microwave sintering on the microstructural homogene-ty and mechanical properties seem to be system specific [23–25].esearchers have also attempted to consolidate refractory materi-ls such as pure W, W–Cu, W–Ni–Fe and W–Ni–Cu [26–33] usingicrowave energy and reported significant reduction in process

ime, elimination of brittle intermetallic formation and superiorechanical properties.This paper reports the sintering behavior of premixed and pre-

lloyed 90W–7Ni–3Cu powders in both, microwave as well asn a conventional (radiatively heated) furnace. The comparativenalysis is based on the sintered density, densification parameter,ardness and microstructures of the samples.

. Experimental procedure

The as-received powders were characterized for their size, sizeistribution and morphology. Table 1 summarizes the charac-eristics of the as-received elemental W, Ni, Cu and prealloyedi–Cu powder used for this study. For investigating the densifi-ation response and compaction studies of various compositions,ylindrical pellets (diameter: 16 mm and height 8 mm) wereressed at 200 MPa using a uniaxial semi-automatic hydraulic pressmodel: CTM 50, supplier: FIE, India) of 50 T capacity. To facilitate

ompaction and subsequent removal of compacted samples, zinc-tearate powder was applied as die-wall lubricant. The as-pressedreen compact density varied from 56% to 58% of the theoreticalensity of the alloy, where the theoretical density was calculatedsing the inverse rule of mixtures.

19.4 8.247.0 17.396.3 40.4

To study the densification behavior; the green (as-pressed) com-pacts were sintered using conventional and microwave furnace.The experimental details have been mentioned elsewhere [32].

The sintered density was obtained by both dimensional mea-surements as well as Archimedes’ density measurement technique.To compare the densification response of various compositions, thesintered densities were normalized with respect to the theoreticaldensity. To take into account the influence of the initial as-presseddensity, the compact sinterability was also expressed in terms ofdensification parameter which is calculated as follows:

densification parameter= sintered density − green densitytheoretical density − green density

(1)

An Anter 1161 V vertical dilatometer was used for measuringaxial shrinkage and shrinkage rate during sintering with constantheating rate of premixed and prealloyed 90W–7Ni–3Cu compacts.It measures dimensional changes over the entire sintering cyclewith precision of 1 �m. The dilatometery studies were performedat constant heating and cooling rate of 10 ◦C/min. The sinteredsamples were wet polished in a manual polisher (model: LunnMajor, supplier: Struers, Denmark) using a series of 6 �m, 3 �m and1 �m diamond paste, followed by cloth polishing using a 0.04 �mcolloidal SiO2 suspension. The scanning electron micrographs of as-polished samples were obtained by a scanning electron microscope(model: FEI quanta, Netherlands) in the secondary electron (SE)mode. A quantitative analysis was also carried out on selected spec-imens using EPMA equipment (model: JXA-8600 SX Super-Probe,supplier: JEOL, Japan). Phase determination and phase evolution, ifany were studied for all the samples using an X-ray Diffractometer(model: Rich. Seifert & Co., GmbH & Co., KG, Germany).

Bulk hardness measurements were performed on polished sur-faces of sintered cylindrical compacts at a load of 5 kg using Vickershardness tester (model: V100-C1, supplier: Leco, Japan). The loadwas applied for 30 s. Micro hardness tester (model: 8299, supplier:Leitz, Germany) was used to evaluate the hardness of the matrixphase. The load applied on matrix phase during the micro hard-ness measurement was 15 g. The bulk hardness of the compactand the micro hardness of the matrix phase for each sample werean average of 5 readings taken at different locations on respec-tive phases. Transverse rupture strength (TRS) measurements ofpremixed and prealloyed samples were performed following theprocedure described in MPIF Standard 41 [34].

3. Results and discussion

3.1. Heating response and densification of W–Ni–Cu alloy

Fig. 1 compares the thermal profiles of 90W–7Ni–3Cu compacts(prepared using premixed and prealloyed Ni–Cu binder) in a con-

ventional and microwave furnace. It is evident from the figure thatW–Ni–Cu alloys couple with microwaves and undergo rapid heat-ing. In case of conventional furnace, in order to ensure uniformheating, the compacts heating rate was restricted to 5 ◦C/min andisothermal holds were provided at intermittent temperatures. In

6872 A. Mondal et al. / Materials Science and Engineering A 527 (2010) 6870–6878

F ◦

tpp

cnac(trrboaistmwcht

Fig. 2. Photograph of 90W–7Ni–3Cu compacts conventionally sintered at varioustemperatures ranging from 1200 ◦C to 1450 ◦C.

Fa1

ig. 1. Thermal profiles of 90W–7Ni–3Cu alloys heated up to 1450 C in conven-ional and microwave furnace. The heating profile of W–Ni–Fe alloys prepared usingremixed and prealloyed Ni–Cu binder is shown separately in case of microwaverocessing.

ontrast, the overall heating rate achieved in the microwave fur-ace was 22 ◦C/min. Taking into consideration the slow heating ratend isothermal holds associated with conventional sintering, theompacts were consolidated through microwaves with significant∼75%) reduction in the processing time. Figs. 2 and 3 show the pho-ograph of the 90W–7Ni–3Cu alloys sintered in the temperatureange of 1200–1450 ◦C in conventional and microwave furnace,espectively. Note that irrespective of the heating mode and theinder chemistry, none of the 90W–7Ni–3Cu compacts distortedr developed any cracks during sintering even up to temperaturess high as 1450 ◦C. It is interesting to note that despite a fast heat-ng rate, no macro cracking was observed in any of the microwaveintered samples. This is attributed to the volumetric heating ofhe compacts, a unique feature of microwave heating. From the

icrowave heating profile (Fig. 1), it is interesting to note that alloysith premixed Ni–Cu binder heat at slightly faster rate than those

ontaining prealloyed matrix. Elsewhere, Sethi et al. [23] and Upad-yaya and Sethi [35] too have reported similar heating response ofhe premixed and prealloyed Cu–12Sn bronze in microwave fur-

ig. 4. Effect of binder state (premixed vs. prealloyed), heating mode and sintering tempelloys. For the sake of comparison, the density and densification parameter data points300 ◦C and 1400 ◦C are also superimposed.

Fig. 3. Photograph of the 90W–7Ni–3Cu compacts prepared using (a) prealloyedand (b) premixed Ni–Cu binder and sintered in a microwave furnace at varioustemperatures.

nace. It can therefore be inferred that the chemical state of thealloying constituents influences the microwave coupling and theresultant heating. It is well recognized that microwave heating

is material dependent. Peelamedu et al. [36] and Roy et al. [37]demonstrated in multi-phase systems, the phases heat up at dif-ferent rates and lead to local an isothermal heating effect. Throughmodel experiments in systems, such as Y2O3–Fe3O4, BaCO3–Fe3O4,and NiO–Al2O3, Peelamedu et al. [36] demonstrated that the hotter

rature on the (a) sintered density and (b) densification parameter of 90W–7Ni–3Cu(denoted by symbol �) for the same composition (in premixed state) sintered at

A. Mondal et al. / Materials Science and Engineering A 527 (2010) 6870–6878 6873

Table 2Composition as a function of temperature and heating mode of the matrix phase of liquid phase sintered 90W–7Ni–3Cu alloys prepared by (a) premixed and (b) prealloyedpowders.

(a) Premixed powders

Temperature (◦C) Heating mode Elemental content (wt.%)

W Ni Cu

1200 Conventional 19.13 ± 1.93 53.15 ± 2.71 25.69 ± 0.92Microwave 18.43 ± 0.37 55.03 ± 0.85 24.18 ± 0.61

1250 Conventional – – –Microwave 20.41 ± 0.72 56.16 ± 0.89 20.77 ± 0.52

1300 Conventional 18.66 ± 0.42 56.04 ± 0.68 23.98 ± 0.88Microwave 24.29 ± 4.65 54.01 ± 1.84 18.37 ± 5.72

1450 Conventional 22.06 ± 4.66 57.03 ± 1.02 20.27 ± 5.36Microwave 23.57 ± 3.37 59.36 ± 0.35 14.41 ± 3.09

(b) Prealloyed powders

Temperature (◦C) Heating mode Elemental content (wt.%)

W Ni Cu

1200Conventional 29.25 ± 0.73 53.62 ± 1.00 15.59 ± 0.13Microwave 27.26 ± 0.96 54.42 ± 0.65 15.20 ± 0.40

1250Conventional – – –Microwave 25.12 ± 0.57 55.68 ± 0.78 16.32 ± 0.55

1300Conventional 26.37 ± 0.38 56.36 ± 0.39 15.36 ± 0.22

.73 ±

.25 ±

.34 ±

siciwc

cbFworstSeawit

ptsetbrrtKdtsw

Microwave 28

1450Conventional 23Microwave 19

pecies diffuse rapidly into the relatively colder ones and correlatedt with the very rapid diffusion rates observed in microwave pro-essed compacts. It is likely that a similar mechanism is also activen the W heavy alloys system as well. The presence of such effect

ill result in densification enhancement in the microwave sinteredomponents.

To test the above hypothesis, the densification responses of theompacts consolidated at various temperatures was evaluated foroth the heating modes and is summarized in Fig. 4a and b. Fromig. 4a, it is evident that the density of the compacts increasesith increasing sintering temperature. Since the green density

f all the compacts were the same, the densification parameteresults (Fig. 4b) closely follow the same trend as sintered den-ity. Note that with the exception of 1200 ◦C, for all other sinteringemperatures, microwave sintering results in better densification.ince factors that contribute to enhanced diffusivity are known tonhance densification during sintering, it can be inferred that thetomic mass transport is higher in case of microwave sintering. Thisill also result into less grain coarsening during microwave sinter-

ng. Consequently, the finer particle size is retained at the sinteringemperature which results in higher grain boundary diffusivity.

From the sintered density and the densification parameterlots shown in Fig. 4, it is evident that for both heating modeshe densification increases with increasing temperature. For theake of comparison the sintered density and densification param-ter results reported by Ramakrishnan and Upadhyaya [10] onhe similar alloy composition sintered at 1300 ◦C and 1400 ◦C haseen superimposed on Fig. 4a and b. Note that the densificationesults from this study on 90W–PM (Ni–Cu) are similar to thoseeported by Ramakrishnan and Upadhyaya [10]. Based on the sin-ering studies on a range of W–Ni–Cu alloys, Makipiritti [38] and

othari [6] have confirmed that densification in this system isiffusion-controlled. Kothari [6] and Crowson [39] have shownhat increasing sintering temperature results in enhanced diffu-ivity and lower activation energy in W–Ni–Cu alloys. The sameill hold good for the present investigation as well and thereby

18.0 50.34 ± 6.99 17.08 ± 10.63

7.56 56.96 ± 2.3 18.80 ± 5.603.09 57.97 ± 2.71 20.62 ± 5.80

activate densification. Besides enhanced diffusivity at high tem-perature, another factor that contributes to densification is thesolid-solubility of tungsten in the Ni–Cu matrix [38–40]. In one ofthe very early works, Prill et al. [41] based on the sintering studies of99W–1(Ni–Cu) alloys showed that the dissolution of the tungstengrains in the matrix and their subsequent reprecipitation from thesupersaturated matrix on to the larger tungsten is a major con-tributor to densification. To validate this, the effect of sinteringtemperature, heating mode and the powder type on the matrixcomposition was evaluated using EPMA. The corresponding resultsfor 90W–PM (Ni–Cu) and 90W–PA (Ni–Cu) compacts are summa-rized in Table 2a and b, respectively. From Table 2a, it is obviousthat higher sintering temperature leads to higher tungsten solubil-ity in the matrix for the samples prepared from premixed powder.The matrix composition analysis is similar to those reported by Sail-land et al. [42] and Xiong et al. [43] for the sintered W–Ni–Cu alloys.On the other hand for the prealloyed compositions (Table 2b) thetungsten solubility does not follow any specific trend as a functionof temperature. The complete explanation for this observation isnot readily available. It is hypothesized that this trend is possiblydue to the difference in diffusivity of tungsten in prealloyed Ni–Cumatrix.

3.2. Dilatometric evaluation of 90W–7Ni–3Cu alloy

For better understanding the phenomenology of the densifi-cation of the W–Ni–Cu compacts the dimensional changes weremeasured in situ during sintering using dilatometry. However,owing to the instrument’s limitations, dilatometric evaluationcould be performed through conventional heating mode only.Fig. 5a and b plots the dilatometric data of shrinkage and shrinkage

rate vs. temperature for the as-pressed 90W–7Ni–3Cu compactsprepared using premixed and prealloyed binder. From the figures,it is evident that both compacts undergo gradual densificationduring heating till 1280 ◦C. Subsequent heating beyond 1280 ◦Cresults in sudden increase in the axial shrinkage and shrinkage rate.

6874 A. Mondal et al. / Materials Science and Engineering A 527 (2010) 6870–6878

Fau

TmrawWl9uotgt1Wi[

cd9blh

ig. 5. Dilatometer plots showing the effect of Ni–Cu binder state (premixed vs. pre-lloyed) on the (a) shrinkage and (b) shrinkage rate of 90W–7Ni–3Cu alloys heatedp to 1400 ◦C at a constant heating rate of 10 ◦C/min in reducing atmosphere.

his spurt in the densification is attributed to the onset of binderelting which results in densification through capillary-induced

earrangement and solution-reprecipitation. The rapid shrink-ge resulting from liquid phase sintering sustains until 1400 ◦Cherein nearly full densification (98%) was achieved. Elsewhere,u et al. [44] and Martin et al. [15] too have reported simi-

ar dilatometric curves for 80W–14Ni–6Cu alloys and prealloyed0W–6.9Ni–3.1Fe, respectively. For the range of compositionssed, both researchers [15,44] have demonstrated that the onsetf melt formation in the W–Ni–Cu system with Ni and Cu inhe ratio 7:3 occurs at 1282 ◦C. From the binary phase dia-ram [45] and the experiments [44] the solidus and the liquidusemperatures of the 70Ni–30Cu matrix was determined to be345 ◦C and 1380 ◦C, respectively. However, in case of ternary–Ni–Cu system, due to substantial solubility of the tungsten

n the matrix the solidus and liquidus temperatures are lowered42].

From the dilatometric results (Fig. 5a and b), it can be dis-erned that the 90W–PA (Ni–Cu) alloy exhibits slightly higherensification particularly at lower temperatures as compared to the

0W–PM (Ni–Cu) compact. This trend is similar to that reportedy Huang et al. [46] who investigated the influence of preal-

oyed Ni–Fe–Mo binder on the sintering behavior of tungsteneavy alloys. They [46] noted that prealloying results in a more

Fig. 6. XRD plots showing the effect of heating mode and sintering temperature onthe phase evolution of 90W–7Ni–3Cu alloys prepared using (a) premixed and (b)prealloyed Ni–Cu binder.

homogenous microstructure and lowers the interfacial energy ofthe tungsten (solid)–matrix (liquid) phase providing a higher den-sity compact. The analyses of matrix composition of the W–Ni–Cucompacts in the present study (Table 2a and b) – particularly, theones sintered at 1200 ◦C – are in accordance with this hypothe-sis. At 1200 ◦C, the dissolved tungsten amount in 90W–PM (Ni–Cu)is lower than its prealloyed counterpart. Consequently, the lat-ter exhibits higher densification (Fig. 4b). For compacts heated orsintered at higher temperatures, the densification in both the com-pacts is not too different and is attributed to homogenization ofthe premixed matrix through interdiffusion. It is well known thatCu and Ni have mutual intersolubility [45]. Using X-ray analysis,Heckel et al. [47] have experimentally shown that significant inter-diffusion and homogenization through in situ alloying occurs in theNi–52 at.% Cu compact after sintering at 950 ◦C for 1 h. Since theCu:Ni ratio in the present system (90W–7Ni–3Cu) is much lower,hence, it is quite reasonable to assume that the premixed Ni–Cupowders will get completely homogenized during conventionalsintering.

3.3. Microstructural evolution and properties

Fig. 6a and b compares the effect of heating mode andsintering temperature on the phase evolution in 90W–PM

A. Mondal et al. / Materials Science and Engineering A 527 (2010) 6870–6878 6875

F binda

(Xtnti

ig. 7. SEM micrographs of 90W–7Ni–3Cu alloys prepared using prealloyed Ni–Cund (d) 1450 ◦C, respectively.

Ni–Cu) and 90W–PA (Ni–Cu) alloys, respectively. For all

RD experiments, the sample size and other test parame-

ers were kept constant. It is therefore rather surprising toote that in case of microwave sintered compacts some ofhe peaks do not even register on the diffractogram. Sim-lar observations have also been reported by Padmavathi

Fig. 8. SEM micrographs of (a) premixed and (b) prealloyed 90W–7Ni–3Cu alloy

er and consolidated in a microwave furnace at (a) 1200 ◦C, (b) 1250 ◦C, (c) 1300 ◦C

[48] in case of microwave sintering of various aluminium

alloys.

The reduction in the intensity indicates that the microwavesintered compacts have less perfect crystals and contain moredefects as compared to their conventionally sintered counter-parts. Recently, many researchers, [37,49–51] have reported that

s sintered at 1300 ◦C in conventional (left) and microwave (right) furnace.

6876 A. Mondal et al. / Materials Science and Engineering A 527 (2010) 6870–6878

alloy

esckim

tbtsti(npmtacttttpsk

mamsi

Fig. 9. SEM micrographs of (a) premixed and (b) prealloyed 90W–7Ni–3Cu

xposure to microwave heating leads to decrystallization andometimes even amorphization in some of the ceramic and semi-onducting systems. The defects within a crystalline structure arenown to activate sintering [52]. Hence, it may be possible thatn the present system too the defects generated on exposure to

icrowaves contribute to densification enhancement.The phenomenology of microstructural evolution in conven-

ionally sintered W–Ni–Cu alloys has been extensively investigatedy several researchers [3,10,15,40]. In the present study, the atten-ion was focused on the microstructural response of microwaveintered W–Ni–Cu system and its comparison with that obtainedhrough conventional sintering. Fig. 7a–d shows the effect of sinter-ng temperature on the representative microstructure of 90W–PANi–Cu) alloy consolidated in a 2.45 GHz multimode microwave fur-ace. It is quite evident that higher temperature not only results inore elimination but is also accompanied by redistribution of theatrix and coarsening of tungsten grains. It is worth noting that

he microstructure of the compacts sintered in solid-state (1200 ◦Cnd 1250 ◦C) is drastically different from the liquid phase sinteredounterparts (Fig. 7c and d). In case of compacts sintered at 1200 ◦C,he binder phase is inhomogenously distributed (Fig. 7a) and theungsten grains are faceted. In contrast, at 1250 ◦C (Fig. 7b), whilehe W-grains still remain faceted, the binder is more uniformly dis-ributed. The well rounded and coarser grains associated with liquidhase sintering (Fig. 7c and d) are associated with higher solid-olubility (Table 2) and relatively faster solution-reprecipitationinetics.

Fig. 8a and b compares the effect of heating mode on the

icrostructures of the 90W–PM (Ni–Cu) and 90W–PA (Ni–Cu)

lloys, respectively, sintered at 1300 ◦C. The correspondingicrostructures of compacts liquid phase sintered at 1450 ◦C are

hown in Fig. 9a and b. From Figs. 8 and 9, it can be discerned thatrrespective of the sintering temperature and the heating mode,

s sintered at 1450 ◦C in conventional (left) and microwave (right) furnace.

the prealloyed sintered compacts have slightly higher coarseningthan their premixed counterparts. This can be attributed to a morehomogenous microstructure and better spreading of the melt dueto the relatively lower tungsten (solid)–matrix (liquid) interfacialenergy in case of alloys prepared using prealloyed binder [46]. Morerecently, Martin et al. [15] showed that the Ni–Cu additive phasein 90W–6.9Ni–3.1Cu alloy with prealloyed matrix is more effec-tively distributed within the tungsten grains during sintering andcontributes not only to densification but also leads to enhancedtungsten grain coarsening as compared to premixed alloys.

One of the intriguing observations in the present study is thedramatic difference between the microstructure of W–Ni–Cu com-pacts sintered in conventional and microwave furnace (Fig. 8a andb). In case of conventionally sintered compacts, the microstructureat 1300 ◦C is quite similar to that obtained through solid-state sin-tering. Some of the earlier researchers [10,53,54] too have reportedsimilar findings at this temperature for premixed 90W–7Ni–3Cucompacts. The larger tungsten grain size in microwave sinteredW–Ni–Cu compacts at 1300 ◦C is contrary to the observations invarious systems that a fast heating rate will result in lower grain size[30,55,56]. However, similar contrary results have recently beenreported by Zaspalis et al. [57] in microwave sintered Mn–Zn fer-rites and by Kim et al. [58] in spark plasma sintered alumina. Theexact explanation of such observations is still lacking and furtherdetailed experiments need to be carried out in future to elucidatethis better. In contrast to 1300 ◦C sintered compacts, both con-ventional as well as microwave sintered 90W–7Ni–3Cu alloys (at1450 ◦C) exhibit well-defined liquid phase sintered microstructure

with rounded tungsten grains interspersed in the Ni–Cu matrix(Fig. 9a and b). Table 3 summarizes the effect of heating modeon the average tungsten grain size in premixed and prealloyed90W–7Ni–3Cu alloys liquid phase sintered at 1450 ◦C. From Fig. 9and Table 3, it can be inferred that irrespective of the state of ini-

A. Mondal et al. / Materials Science and Engineering A 527 (2010) 6870–6878 6877

Table 3Effect of heating mode on the average tungsten grain size in 90W–7Ni–3Cu alloysprepared using premixed and prealloyed Ni–Cu binder and sintered at 1450 ◦C.

Composition Heating mode W grain size (�m)

90W–PM(Ni–Cu)

Conventional 23 ± 10Microwave 21 ± 7

90W–PA(Ni–Cu)

Conventional 26 ± 8Microwave 20 ± 7

Table 4Effect of heating mode and sintering temperature on the bulk hardness (in kgf/mm2)of 90W–7Ni–3Cu alloys prepared using premixed and prealloyed binder.

Heating mode 90W–PM (7Ni–3Cu)a 90W–PA (7Ni–3Cu)

1300 ◦C 1450 ◦C 1300 ◦C 1450 ◦C

Conventional 276 ± 21 301 ± 16 290 ± 19 330 ± 13Microwave 309 ± 17 328 ± 13 330 ± 18 350 ± 12

ta

tsahs

iiooabapaecifissosbro1spta1tf

TEs

a As a note of comparison in one of the previous reports [10] the bulk hardness ofhe same composition has been listed as 175VHN and 220VHN for compacts sinteredt 1300 ◦C and 1400 ◦C, respectively.

ial binder chemistry (premixed vs. prealloyed), the microwaveintered compacts have relatively more refined microstructures compared to those consolidated in a conventional (radiativelyeated) furnace. Elsewhere, Upadhyaya et al. [30] has reportedimilar findings in liquid phase sintered W–Ni–Fe system.

The differences in the sintered microstructure of alloys consol-dated through conventional and microwave furnace is bound tonfluence the mechanical properties. Table 4 summarizes the effectf heating mode and powder chemistry (prealloyed vs. premixed)n the bulk hardness of 90W–7Ni–3Cu alloys sintered at 1300 ◦Cnd 1450 ◦C, respectively. To provide a basis of comparison, theulk hardness of premixed 90W–7Ni–3Cu alloy sintered at 1300 ◦Cnd 1400 ◦C reported by Ramakrishnan and Upadhyaya [10] is alsorovided in Table 4. It is remarkable to note that the hardnesschieved in the present study is far superior to those reported inarlier work [10,59]. From Table 4, it is worth noting that for allonditions, microwave sintering results in up to 13% improvementn the bulk hardness. This can be attributed to the increased densi-cation and a more refined microstructure in case of microwaveintered alloys. Previous report from Upadhyaya et al. [30] hashown that the hardness of tungsten grains per se does dependn the heating mode. However, on account of the varying tungstenolid-solubility, the hardness of the matrix phase is expected toe dependent upon the heating mode. To test this, Table 5 summa-izes the corresponding matrix hardness measurements performedn compacts consolidated in both the heating modes at 1300 ◦C and450 ◦C. It is interesting to observe that in case of compacts con-olidated at 1300 ◦C, the matrix hardness is higher for microwaverocessed compacts (Table 5). This can be correlated to the higherungsten solubility observed in microwave sintered alloys (Table 2and b). In contrast, the trend is reversed for compacts sintered at

450 ◦C. In case of prealloyed compacts, this can be correlated tohe higher W-solubility in conventionally sintered alloys. However,urther tests need to be performed to gain detailed insight for this.

able 5ffect of heating mode on the micro hardness (in kgf/mm2) of the matrix phase inintered 90W–7Ni–3Cu alloys.

Heating mode 90W–PM (7Ni–3Cu) 90W–PA (7Ni–3Cu)

1300 ◦C 1450 ◦C 1300 ◦C 1450 ◦C

Conventional 260 ± 20 333 ± 13 250 ± 25 320 ± 16Microwave 290 ± 23 312 ± 15 280 ± 21 290 ± 12

Fig. 10. Effect of heating mode and sintering temperature on the transverse rupturestrength of 90W–7Ni–3Cu alloys prepared using premixed and prealloyed binders.

Fig. 10 shows the effect of heating mode and powder chem-istry on the transverse rupture strength of the 90W–7Ni–3Cu alloysintered at 1300 ◦C and 1450 ◦C. One of the very interesting observa-tions is that microwave sintered compacts have higher transverserupture strength as compared to those sintered conventionally. Theenhancement of strength is as high as up to threefolds (∼211%) incase of alloys processed at 1300 ◦C. This can be correlated to a lesscontiguous microstructure consisting of rounded tungsten grainsthat result in alloys microwave sintered at 1300 ◦C (Fig. 8). Else-where, Chermant and Osterstock [60] have demonstrated similarinfluence of microstructure on the fracture toughness of WC–Cosystem. The transverse rupture strength of the conventionally sin-tered 90W–PM (Ni–Cu) system is similar to that reported earlierby Kalina et al. [61]. Another observation that can be inferred fromFig. 10 is that the transverse rupture strength of alloys preparedwith prealloyed Ni–Cu is marginally higher than those preparedusing premixed binder. This could be due to better homogeneity ofthe alloying additives in case of the prealloyed system.

4. Conclusions

90W–7Ni–3Cu alloys, compacts were prepared using bothpremixed and prealloyed Ni–Cu binder and were sintered attemperatures ranging between 1200 ◦C and 1450 ◦C. As com-pared to conventional sintering, both premixed and prealloyed90W–7Ni–3Cu alloys were microwave sintered in significantly(∼75%) less time. In spite of higher heating rate in microwave sin-tering, no micro- or macro-cracking was observed in all microwavesintered samples. Dilatometric analysis of the 90W–7Ni–3Cu alloyat slower heating rates (<10 ◦C/min) reveals that – unlike the W–Cusystem – the W–Ni–Cu system exhibits significant solubility oftungsten in the matrix. Consequently, a major portion of the overalldensification in this alloy occurs prior to melt formation. Due to theinter-diffusivity between nickel and copper, by the time the sinter-ing temperature was attained, both Cu and Ni form a homogeneoussolid solution. Hence, the sintering response of the 90W–7Ni–3Cualloy was per se independent of the initial matrix type (premixedvs. prealloyed). An interesting finding in the present study was the

observation of the heating mode effect on the microstructure of thecompacts sintered at 1300 ◦C. While the conventionally processedalloy had a typical solid-state sintered structure, the microwaveprocessed samples had a well-developed microstructure, typicallyassociated with liquid phase sintering. Consequently, as compared

6 and En

t9(

A

DPn

R

[[

[[

[

[

[[[[

[

[

[[[[[

[

[

[

[[

[[

[

[[[[[

[

[[

[[

[

[[

[

[

[

[

[[

[

[[

[

878 A. Mondal et al. / Materials Science

o conventional sintering, microwave sintering resulted in about% improvement in hardness (to ∼350VHN) and about threefold∼211%) enhancement in flexural strength to around 1625 MPa.

cknowledgement

This collaborative research was done as a part of the Center forevelopment of Metal–Ceramic Composites through Microwaverocessing which was funded by the Indo-US Science and Tech-ology Forum (IUSSTF), New Delhi.

eferences

[1] S.G. Caldwell, in: P.W. Lee, R. Lacocca (Eds.), Powder Metal Technologies andApplications, vol. 7, Materials Park, OH, USA, 1998, pp. 914–921.

[2] A. Upadhyaya, Mater. Chem. Phys. 67 (2001) 101–110.[3] G.H.S. Price, C.J. Smithells, S.V. Williams, J. Inst. Met. 62 (1938) 239–264.[4] A.L. Prill, Trans. AIME 230 (1964) 769–772.[5] J.H. Brophy, H.W. Hayden, J. Wulff, Trans. TMS-AIME 224 (1962) 797–803.[6] N.C. Kothari, J. Less Common Met. 13 (1967) 457–468.[7] R. Makarov, O.K. Teodorovich, I.N. Fruntsevich, Sov. Powder Metall. Met. Ceram.

4 (1965) 554–559.[8] V. Srikanth, G.S. Upadhyaya, Int. J. Ref. Hard Met. 2 (1983) 49–54.[9] V. Srikanth, G.S. Upadhyaya, Powder Technol. 39 (1984) 61–65.10] K.N. Ramakrishnan, G.S. Upadhyaya, J. Mater. Sci. Lett. 9 (1990) 456–459.11] J.F. Kuzmic, in: H.H. Hausner (Ed.), Modern Development in Powder Metallurgy,

vol. 3, Plenum Press, New York, 1966, pp. 166–171.12] E. Ariel, J. Batra, D. Brandon, Powder Metall. Int. 5 (1973) 125–128.13] V.N. Eremenko, R.V. Minakova, M.M. Churakov, Sov. Powder Metall. Met. Ceram.

15 (1976) 283–286.14] Y.V. Naidich, I.A. Lavrinenko, V.A. Evdokimov, Sov. Powder Metall. Met. Ceram.

16 (1977) 276–280.15] J.M. Martin, J.L. Johnson, R.M. German, F. Castro, in: E. Daver, C.J. Trombino

(Eds.), Advances in Powder Metallurgy and Particulate Materials, vol. 1, MetalPowder Industries Federation, Princeton, NJ, 2006, pp. 5.43–5.57.

16] A.P. Savitskii, Sci. Sinter. 37 (2005) 3–17.17] K.J. Rao, P.D. Ramesh, Mater. Bull. Sci. 18 (1995) 447–465.18] D.E. Clark, W.H. Sutton, Ann. Rev. Mater. Sci. 26 (1996) 299–331.19] R.M. Hutcheon, M.S. De Jong, F.P. Adams, Microw.: Theor. Appl. Mater. Proc. III,

Ceram. Trans., Am. Ceram. Soc. 59 (1995) 215.20] A. Mondal, D. Agrawal, A. Upadhyaya, J. Microw. Power. Electromagn. Energy

(JMPEE) 43 (2009) 5–10.21] M.W. Porada, Microw.: Theor. Appl. Mater. Proc. III, Ceram. Trans., Am. Ceram.

Soc. 80 (1997) 153.22] R. Roy, D.K. Agrawal, J.P. Cheng, S. Gedevanishvili, Nature 399 (1999) 668–670.23] G. Sethi, A. Upadhyaya, D. Agrawal, Sci. Sinter. 35 (2003) 49–65.24] R.M. Anklekar, D.K. Agrawal, R. Roy, Powder Metall. 44 (2001) 355–362.25] K. Saitou, Scripta Mater. 54 (2006) 875–879.

26] M. Jain, G. Skandan, K. Martin, K. Cho, B. Klotz, R. Dowding, D. Kapoor, D. Agar-

wal, J. Chang, Int. J. Powder Metall. 42 (2006) 45–50.27] G. Prabhu, A. Charkraborty, B. Sarma, Int. J. Ref. Met. Hard Mater. 27 (2009)

545–548.28] S.D. Luo, J.H. Yi, Y.L. Guo, Y.D. Peng, L.Y. Li, J.M. Ran, J. Alloys Compd. 473 (2009)

L5–L9.

[[

[[

gineering A 527 (2010) 6870–6878

29] A. Mondal, A. Upadhyaya, D. Agrawal, J. Microw. Power Electromagn. Energy(JMPEE) 43 (2009) 11–16.

30] A. Upadhyaya, S.K. Tiwari, P. Mishra, Scripta Mater. 56 (2007) 5–8.31] A. Mondal, D. Agrawal, A. Upadhyaya, J. Microw. Power Electromagn. Energy

(JMPEE) 44 (2010).32] A. Mondal, D. Agrawal, A. Upadhyaya, Mater. Sci. Technol. (2008) 2502–2515.33] A. Mondal, Modeling of microwave heating of particulate metals and its appli-

cation in sintering of tungsten-based alloys, Ph.D. Thesis, The Indian Instituteof Technology, Kanpur, 2009.

34] MPIF Standard 41, Determination of transverse rupture strength of powdermetallurgy materials, standard test methods for metal powders and powdermetallurgy products, MPIF, Princeton, NJ, USA, 2002, pp. 55–57.

35] A. Upadhyaya, G. Sethi, Scripta Mater. 56 (2007) 469–472.36] R.D. Peelamedu, R. Roy, D. Agrawal, Mater. Res. Bull. 36 (2001) 2723–2739.37] R. Roy, Y. Fang, J. Cheng, D.K. Agrawal, J. Am. Ceram. Soc. 88 (2005) 1640–1642.38] S. Makipiritti, Powder Metall. 4 (1961) 97–111.39] A. Crowson, Modern Developments in Powder Metallurgy, vol. 15, Metal Pow-

der Industries Federation, Princeton, NJ, 1985, pp. 507–520.40] G. Sethi, S.J. Park, J.L. Johnson, R.M. German, Int. J. Ref. Met. Hard Mater. 27

(2009) 688–695.41] A.L. Prill, H.W. Hayden, J.H. Brophy, Trans. AIME 233 (1965) 1960–1964.42] F. Sailland, J.M. Chaix, C.H. Allibert, Colloquium on Controlling the Properties

of Powder Metallurgy Parts Through Their Microstructures, Societe Francaisede Metallurgie, Paris, 1990, pp. 4.1–4.6.

43] H. Xiong, L. Zhang, Q. Shen, R. Yuan, J. Mater. Sci. Technol. 15 (1999) 94–96.44] Y. Wu, R.M. German, B. Marx, R. Bollina, M. Bell, Mater. Sci. Eng. A344 (2003)

158–167.45] P.R. Subramaniam, D.J. Chakrabarti, D.E. Laughlin (Eds.), Monograph Series on

Alloy Phase Diagram-10: Phase Diagram of Binary Copper Alloys, ASM Interna-tional, Materials Park, OH, 1994, pp. 276–286.

46] J.H. Huang, G.A. Zhou, C.Q. Zhu, S.Q. Zhang, H.Y. Lai, Mater. Lett. 23 (1995) 47–53.47] R.W. Heckel, R.D. Lanam, R.A. Tanzilli, in: J.S. Hirschhorn, K.H. Roll (Eds.),

Advanced Experimental Techniques in Powder Metallurgy, Plenum Press, NewYork, 1970, pp. 139–188.

48] C. Padmavathi, Liquid phase sintering of 2712, 6711 and 7775 aluminum alloysand their properties, Ph.D. Thesis, The Indian Institute of Technology, Kanpur,2009.

49] R. Peelamedu, R. Roy, L. Hurtt, D. Agrawal, A.W. Fliflet, D. Lewis III, R.W. Bruce,Mater. Chem. Phys. 88 (2004) 119–129.

50] D.C. Dube, M. Fu, D. Agrawal, R. Roy, A. Santra, Mater. Res. Innovat. 12 (2008)119–122.

51] J. Cheng, D. Agrawal, Y. Zhang, R. Roy, A.K. Santra, Synthesis of amorphous Si-Gealloys using microwave energy, J. Alloys Compd. 491 (2010) 517–521.

52] W. Schatt, E. Friedrich, Powder Metall. 28 (1985) 140–144.53] A. Upadhyaya, Distortion in liquid phase sintered tungsten heavy alloys, M.S.

Thesis, The Pennsylvania State University, University Park, PA, 1996.54] A. Upadhyaya, J.L. Johnson, R.M. German, Tungsten and Refractory Metals-1995,

Metal Powder Industries Federation, Princeton, NJ, 1995, p. 18.55] C. Zhou, J. Yi, S. Luo, Y. Peng, L. Li, G. Chen, J. Alloys Compd. 482 (2009) L6–L8.56] K. Rödiger, K. Dreyer, T. Gerdes, M. Willert-Porada, Int. J. Ref. Met. Hard Mater.

16 (1998) 409–416.57] V.T. Zaspalis, S. Sklari, M. Kolenbrander, J. Magn. Mater. 310 (2007) 28–36.

58] B.N. Kim, K. Hiraga, K. Morita, H. Yoshida, J. Eur. Ceram. Soc. 29 (2009) 323–327.59] G.S. Upadhyaya, V. Srikanth, Modern Developments in Powder Metallurgy, vol.

17, Metal Powder Industries Federation, Princeton, NJ, 1985, pp. 51–75.60] J.L. Chermant, F. Osterstock, J. Mater. Sci. 11 (1976) 1939–1951.61] M.M. Kalina, Z.G. Soboleva, V.N. Gorokhov, E.M. Rabinovich, Powder Metall.

Met. Ceram. 29 (1990) 406–409.