Materials Science and Engineering A 509 (2009) 63–68

Contents lists available at ScienceDirect

Materials Science and Engineering A

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

Microstructure and mechanical properties of ZrO2/NiCr functionally gradedmaterials

Xin Jin, Linzhi Wu ∗, Yuguo Sun, Licheng GuoCenter for Composite Materials, Harbin Institute of Technology, Harbin 150080, PR China

a r t i c l e i n f o

Article history:Received 9 November 2008Received in revised form 24 December 2008Accepted 26 January 2009

Keywords:Functionally graded materials

a b s t r a c t

The microstructure and mechanical properties of ZrO2/NiCr functionally graded materials (FGMs) fabri-cated by powder metallurgy are investigated experimentally. Microscopic examination exhibits that thematerial composition and microstructure of the FGMs vary gradually. The distributions of mechanicalproperties in the FGMs are obtained from the mechanical testing of homogeneous composite samples(non-FGM) with different volume fractions of ZrO2. The experimental results show that the distributionsof mechanical properties strongly depend on the variation of microstructure. It is found that hardness

MicrostructureMechanical propertyInterface

increases and ductility decreases with the increase of ZrO2, which is attributed to the variation of thematrix phase from the metal to the ceramic. Bending strength and elastic modulus decrease as the vol-ume fraction of ZrO2 increase from 0% to 40%, however, increase as the volume fraction of ZrO2 increase

are m

1

mvcttoltct(taF

sphtZtm

0d

from 50% to 100%. These

. Introduction

Functionally graded materials (FGMs) offer an advantageousean of combining different materials and providing a spatial

ariation in composition and properties [1]. FGMs consisting oferamics and metals are promising candidates for the future highemperature applications. Ceramic component in the FGMs offershermal barrier effects and protects the metal from corrosion andxidation, and the FGMs are toughened and strengthened by metal-ic component [2]. The investigation of mechanical response ofhe FGMs is significant to the optimal design and fabricating pro-ess. The varying mechanical properties are usually determined byhe micromechanical method [3–5] or the finite element methodFEM) [6,7]. However, few experiments [8–10] have been conductedo investigate the relationship between the mechanical propertiesnd the material composition as well as the microstructure in theGMs.

Due to the high mechanical and thermal properties of the con-tituent materials, the ZrO2/NiCr FGMs can exhibit good serviceerformance under some severe environments, such as super-igh temperature and great temperature gradient [11,12]. However,

he experimental investigation on the mechanical response of therO2/NiCr FGMs is rare. Zhu et al. [13] conducted the mechanicalesting on a six-layered ZrO2/NiCr FGM sample, and found that the

echanical properties of the FGMs exhibit various gradient distri-

∗ Corresponding author. Tel.: +86 451 86402376; fax: +86 451 86402386.E-mail addresses: pzauzn@126.com (X. Jin), wlz@hit.edu.cn (L.Z. Wu).

921-5093/$ – see front matter. Crown Copyright © 2009 Published by Elsevier B.V. All rioi:10.1016/j.msea.2009.01.066

ainly caused by the weakly bonded ceramic/metal interface.Crown Copyright © 2009 Published by Elsevier B.V. All rights reserved.

butions with the composition change. Using powder metallurgy, Jinet al. [14] fabricated the ZrO2/NiCr FGMs with the volume fraction ofNiCr increasing from 0% to 50%. They found that the elastic modulusdecreases and the fracture toughness increases with the increase ofNiCr. But further research is needed to fully understand the influ-ences of the material composition and varying microstructure onthe mechanical response of the ZrO2/NiCr FGMs.

In this paper, an eleven-layered ZrO2/NiCr FGM sample isfabricated by powder metallurgy, and the microstructural char-acteristics of the FGMs are examined by the optical microscope.The distributions of mechanical properties in the FGMs are directlyobtained by the mechanical testing conducted on the non-FGMsamples. Finally, scanning electron microscope (SEM) is performedon the fracture surface to analyze the dominant failure mechanismsof the FGMs.

2. Materials and experimental procedures

2.1. Materials

Commercially available ZrO2 powder and Ni-20 wt.% Cr alloy(NiCr) powder are used as the raw powders. The characteristicsof the two raw powders are listed in Table 1. In this paper, aneleven-layered ZrO2/NiCr FGM sample is fabricated by powder met-

allurgy, as shown in Fig. 1. The ZrO2 and NiCr powders are mixedwith different volume ratios according to the design. The mix-tures are stacked, layer by layer, and compacted by cold-pressingat ∼30 MPa in a steel die to form a disk-shaped green compact. Tomaintain the designed composition distribution, the stacked layer is

ghts reserved.

64 X. Jin et al. / Materials Science and Engineering A 509 (2009) 63–68

Table 1Raw powders characteristics.

Materials Particles size (�m) Purity (%) Manufacture

NiCr <45 >98 Shanghai Institute of Ceramics,Shanghai, China

ZrO2 1.5 >99.9 General Research Institute forNonferrous Metals, Beijing,China

Fig. 1. Composition distribution model of the ZrO2/NiCr FGMs.

Fig. 3. Microstrcture of the ZrO2/NiCr FGM

Fig. 2. Schematics of the three-point bending test. (Dimensions in mm).

pre-compacted at a lower pressure before stacking the next layer.Then the green compact is sintered under the hot-pressing condi-tion of 1300 ◦C and 5 MPa for 1.5 h. With the same process, elevennon-FGM samples with different volume fractions of ZrO2 are alsofabricated. For microstructure inspection and mechanical testing,the samples are cut by a diamond saw, and their surfaces are groundand polished.

2.2. Mechanical testing

The distribution of Vickers hardness in the FGMs is directlydetermined by indenting with a load of 10 kgf on each layer ofthe FGM sample. The distributions of bending strength and elas-tic modulus in the FGMs are obtained from three-point bendingtest conducted on the non-FGM samples, as shown in Fig. 2. The

three-point bending test is conducted on an Instron 5569 universaltesting machine under the displacement control condition at therates of 0.5 mm/min. For each material composition, five samplesare tested.

s fabricated by powder metallurgy.

X. Jin et al. / Materials Science and Engineering A 509 (2009) 63–68 65

2

mmtm

3

3

irtndtvtoicmid

Fig. 4. Relative densities of the FGM and non-FGM samples.

.3. Materials characterization

Optical microscopy (Olympus-SZX12) is used to characterize theicrostructure of the FGMs. The density of the sintered samples iseasured by Archimedes’ method. For the samples fractured in the

esting, the fracture surfaces are examined by a scanning electronicroscope (SEM, Philips-XH3).

. Results

.1. Microstructure

The microstructure of the sintered ZrO2/NiCr FGMs is presentedn Fig. 3, where the white and gray phases are NiCr and ZrO2,espectively. These micrographs show a good gradual composi-ional variation in the FGMs. The obvious macroscopic interfaces areot observed even on the interfaces (marked ‘�’ in Fig. 3) betweenifferent layers where the material composition jumps. Fig. 3 illus-rates a typical variation of the microstructure in the FGMs. Thearying microstructure is characterized by the gradual variation ofhe matrix phase from the metal to the ceramic with the increasef ZrO2. In the metal-rich side, the ceramic particles are dispersed

n the metal matrix. With the increase of ZrO2, the clusters of theeramic are formed and their further growths result in the for-ation of a network structure. Then, the network of the metal

s gradually diminished and turns into the isolated metal particleispersed in the ceramic matrix in the ceramic-rich side.

Fig. 6. Load-deflection curves of the non-FGMs with the volume fractio

Fig. 5. Relationship between the Vickers hardness and the volume fraction of ZrO2.

On the other hand, the hot-pressing process is a more effectivemethod to produce the dense samples. The relative densities of boththe FGM and non-FGM samples are shown in Fig. 4. The FGM sampleexhibits a low porosity level and its relative density is 97.2%. Therelative densities of all non-FGM samples are more than 94.5%.

3.2. Mechanical properties

Fig. 5 illustrates the relationship between Vickers hardness andmaterial composition in the ZrO2/NiCr FGMs. A strong dependenceof the hardness on the volume fraction of ZrO2 can be observed.With the increase of ZrO2, the hardness increases monotonicallyfrom 1.79 to 11.14 GPa.

The typical load-deflection curves of all non-FGM samples areshown in Fig. 6. It can be found from these curves that the volumefraction of ZrO2 has significant effect on the ductility of the non-FGM samples. For the samples with the volume fraction of ZrO2less than 40%, the ductile property deteriorates gradually with theincrease of ZrO2, as shown in Fig. 6(a). But the samples with thevolume fraction of ZrO2 more than 50% fail in a brittle fracture mode,as shown in Fig. 6(b).

Fig. 7 shows the variation of bending strength with materialcomposition in the FGMs. As the volume fraction of ZrO increases

2from 0% to 40%, the bending strength decreases gradually from606.7 to 263.3 MPa. However, the bending strength increases from260.0 to 538.7 MPa as the volume fraction of ZrO2 increases from50% to 100%.

n of ZrO2 changing (a) from 0% to 40%, and (b) from 50% to 100%.

66 X. Jin et al. / Materials Science and En

F

otgit

o7iwgTvmbcgmfbtI

F

ig. 7. Relationship between the bending strength and the volume fraction of ZrO2.

The elastic modulus is calculated from the initial linear portionf the load-deflection curves (Fig. 6). The resultant distribution ofhe elastic modulus is shown in Fig. 8. The elastic modulus decreasesradually from 194.2 to 145.1 GPa as the volume fraction of ZrO2ncreases from 0% to 40%, but increases from 119.5 to 201.0 GPa ashe volume fraction of ZrO2 increases from 50% to 100%.

Figs. 9(a)–(d) show the SEM micrographs of the fracture surfacesf the selected non-FGM samples containing 10%, 30%, 50% and0% ZrO2, respectively. The fractograph of the 10% ZrO2 compos-te is characterized by the dimple morphology in the metal matrix,

hich is the characteristic of a ductile fracture process: nucleation,rowth and coalescence of the microvoids, as shown in Fig. 9(a).he morphology of fracture surface of the 30% ZrO2 composite isery complex, as shown in Fig. 9(b). In this case, the dimples in theetal matrix, the clusters of the ceramic particles and the voids

etween the metal matrix and the clusters of the ceramic particlesan be observed on the fracture surface. The characteristic of fracto-raph of the 30% ZrO2 composite is due to the discontinuity of theetal matrix caused by the clusters of the ceramic particles. The

ractographs of the 50% and 70% ZrO2 composites are characterizedy the debonded metal particles and their traces in the brittle frac-ure surface of the ceramic matrix, as shown in Figs. 9(c) and (d).n addition, none broken metal particles are observed in the frac-

ig. 8. Relationship between the elastic modulus and the volume fraction of ZrO2.

gineering A 509 (2009) 63–68

ture surfaces of these samples. These indicate the ceramic/metalinterfacial strength is relatively low.

4. Discussions

4.1. Influence of microstructure on the hardness

The influence of the varying microstructure on the hardnessof the FGM sample is obvious, as shown in Fig. 5. The hardnessincreases slowly as the volume fraction of ZrO2 increases from 0%to 30%. It is for the reason that the amount of the ceramic is too lowto form the rigid skeleton of the hard ceramic. In these layers, themicrostructure is characterized by the ceramic particles dispersedthe metal matrix, as shown in Fig. 3. As the volume fraction of ZrO2increases from 40% to 60%, the slope of the hardness curve changesand the hardness increases substantially. This is mainly caused bythe gradual variation of the microstructure in the FGMs. In these lay-ers, the matrix phase varies from the soft metal to the hard ceramic,as shown in Fig. 3.

4.2. Influence of microstructure on the ductility

With the increase of ZrO2, the mechanical behaviors of the non-FGM samples vary from the elastic-plastic type to the elastic type(Fig. 6), and the failure modes change from the ductile type to thebrittle type (Fig. 9). These results demonstrate that the ductility ofthe non-FGM samples deteriorates gradually with the increase ofZrO2. This is believed to be the result of the variation of the matrixphase from the metal with good plastic property to the ceramicwith poor plastic property. The samples of the NiCr alloy and the10% ZrO2 composite exhibit a good ductile property. The large defor-mation can be observed on the tensile side of the sample during thetesting. The NiCr alloy and the 10% ZrO2 composite do not fail untilbeing deflected by 3.37 and 1.41 mm, respectively. The reduction inthe ductility of the 10% ZrO2 composite is attributed to the pres-ence of the hard ceramic which can constrain the plastic flow of theductile metal matrix. The non-FGM samples containing 20%, 30%and 40% ZrO2 fail as deflected by 0.54, 0.34 and 0.32 mm, respec-tively. Obvious plastic deformation does not occur on the tensileside of the sample. For these samples, another reason causing thedeterioration of the ductility is the continuity of the metal matrixinterfered by the clusters of the ceramic particles, as shown in Fig. 3and Fig. 9(b). When the matrix phase varies from the metal to theceramic, the non-FGM samples mainly exhibit brittle behavior. Thenon-FGM samples with the volume fraction of ZrO2 more than 50%fail in a brittle fracture mode, as shown in Fig. 6(b).

4.3. Influence of microstructure on the bending strength

The distribution characteristic of the bending strength in theFGMs (Fig. 7) can be attributed to the microstructural effects includ-ing the porosity, the ceramic/metal interface and the distributionsof the two components. The pure ZrO2 sample has the highestporosity level among all non-FGM samples, but does not exhibitthe lowest bending strength. This implies that the porosity doesnot have prominent effect on the bending strength of the FGMs.

As the volume fraction of ZrO2 increases from 0% to 40%, thebending strength decreases, which can be explained by the pres-ence of the weakly bonded ceramic/metal interface. It can be foundfrom Figs. 9(a) and (b) that there exist many voids between themetal matrix and the ceramic particles, which is caused by the

debonding of the ceramic/metal interface during the testing. Thistype of interface forms as a result of a mechanical interlocking pro-duced by the shrinkage of the metal matrix around the ceramicparticles during the cooling from the sintering temperature. Theincrease of ceramic particles can lead to the increase of the total area

X. Jin et al. / Materials Science and Engineering A 509 (2009) 63–68 67

F ing: (a(

obpmtwntdAth

4

Zctoo(tithtt

ig. 9. SEM micrographs of fracture surface of the selected non-FGM samples includd) the 70% ZrO2 composite.

f the weakly bonded interface, which results in the decrease of theending strength. In addition, the ceramic particles with smallerarticle size are more susceptible to fill into the space of metalatrix and form the clusters, as shown in Fig. 9(b). The clusters of

he ceramic particles can destruct the continuity of metal matrix,hich may have negative effect on the bending strength. For theon-FGM samples with the volume fraction of ZrO2 more than 50%,he fractographs of them are characterized by the metal particlesebonded from the ceramic matrix, as shown in Figs. 9(c) and (d).s the volume fraction of the ZrO2 increases from 50% to 100%, the

otal area of weakly bonded ceramic/metal interface decreases, andence, the bending strength increases.

.4. Influence of microstructure on the elastic modulus

For the ZrO2/NiCr FGMs, the elastic moduli of the NiCr alloy andrO2 are 194 and 201 GPa, respectively. According to the microme-hanical models, such as Mori-Tanaka [3,4] or self-consistent [5],he predicted elastic moduli of the non-FGM samples may not varybviously due to the elastic modulus of ZrO2 nearly equal to thatf NiCr alloy. However, it is found from the experimental resultsFig. 8) that the minimum elastic modulus 119.5 GPa is 40.7% lesshan the maximum one 201 GPa. The difference between the exper-

mental and theoretical results in the elastic modulus is attributedo the weakly bonded ceramic/metal interface. Tan et al. [15,16]ave mentioned that the interfacial debonding can cause the reduc-ion in the elastic modulus of the composites as compared withhose with perfect bonded interface. Tsui et al. [17] investigated the

) the 10% ZrO2 composite; (b) the 30% ZrO2 composite; (c) the 50% ZrO2 composite;

influence of the weakly bonded interface on the elastic modulusof particle-reinforced composites by using FEM. They pointed outthat the reduction of the elastic modulus became significant withthe increase of the particles. For the non-FGM samples with the vol-ume fraction of ZrO2 varying from 10% to 40%, the microstructure ofthese samples can be considered as the ceramic particles dispersedin the continuous metal matrix. Hence, the elastic moduli of thesesamples decrease with the increase of ZrO2. However, for the non-FGM samples with the volume fraction of ZrO2 varying from 50%to 90%, the microstructure of these samples is characterized by themetal particles dispersed in the continuous ceramic matrix. There-fore the elastic moduli of these samples increase with the increaseof ZrO2. Due to the highest particle content of the 50% ZrO2 compos-ite among all non-FGM composites, it exhibits the minimum valuein the elastic moduli.

5. Conclusions

There is a gradual variation of the microstructure in theZrO2/NiCr FGMs, in which the matrix phase varies from the metalto the ceramic and the inclusion phase varies from the ceramic tothe metal.

With the increase of ZrO2, the hardness increases monotonously

and the ductility decreases gradually. The hardness and ductility ofthe ZrO2/NiCr FGMs are affected by the varying microstructure inwhich the matrix phase varies from the metal to the ceramic. Theceramic matrix can provide the rigid skeleton and constrain theplastic flow of the metal.

6 nd En

tiTidap

tsoF

A

N1I

[[[

[

8 X. Jin et al. / Materials Science a

The bending strength and elastic modulus firstly decrease ashe volume fraction of ZrO2 increase from 0% to 50%, and thenncrease as the volume fraction of ZrO2 increase from 50% to 100%.hese are mainly affected by the weakly bonded ceramic/metalnterface. The voids between the metal and the ceramic, theebonded metal particles and their traces in the ceramic matrixre observed in the SEM fractographs of the non-FGM sam-les.

In addition, the clusters of the ceramic particles also have a nega-ive influence on the ductility and bending strength of the non-FGMamples. And the porosity does not seem to have an obvious effectn the distributions of the mechanical properties in the ZrO2/NiCrGMs.

cknowledgements

The authors are grateful for the financial support by Nationalatural Science Foundation of China (10432030, 10872056 and0672049) and National Science Foundation for Excellent Youngnvestigators (10325208).

[

[[[

gineering A 509 (2009) 63–68

References

[1] A. Kawasaki, R. Watanabe, J. Jpn. Inst. Met. 51 (1987) 525–529.[2] S. Uemura, Mater. Sci. Forum 423 (2003) 1–10.[3] G. Dvorak, J. Zuiker, Compos. Eng. 4 (1994) 19–35.[4] M. Dong, P. Le�le, U. Weber, S. Schmauder, Mater. Sci. Forum 308 (1999)

1000–1005.[5] P. Le�le, M. Dong, S. Schmauder, Compos. Mater. Sci. 15 (1999) 455–465.[6] M. Grujicic, Y. Zhang, Mater. Sci. Eng. A 251 (1998) 64–76.[7] V. Cannillo, T. Manfredini, M. Montorsi, C. Siligardi, A. Sola, J. Eur. Ceram. Soc.

26 (2006) 3067–3073.[8] R. Rodríguez-Castro, R.C. Wetherhold, M.H. Kelestemur, Mater. Sci. Eng. A 323

(2002) 445–456.[9] V. Cannillo, L. Lusvarghi, C. Siligardi, A. Sola, J. Eur. Ceram. Soc. 27 (2007)

1935–1943.10] K. Tohgo, K. Araki, Y. Shimamura, Key. Eng. Mater. 345 (2007) 497–500.11] V. Teixeira, Key. Eng. Mater. 230 (2002) 335–338.12] H. Hamatani, N. Shimoda, S. Kitaguchi, Sci. Technol. Adv. Mater. 4 (2003)

197–203.13] J. Zhu, Z. Lai, Z. Yin, J. Jeon, S. Lee, Mater. Chem. Phys. 68 (2001) 130–135.

14] X. Jin, W. Xu, L.Z. Wu, Proceedings of the ICHMM-2008, Huangshan, China, 2008,

pp. 1254–1258.15] H. Tan, Y. Huang, C. Liu, P.H. Geubelle, Int. J. Plast. 21 (2005) 1890–1918.16] H. Tan, C. Liu, Y. Huang, P.H. Geubelle, J. Mech. Phys. Solids 53 (2005) 1892–1917.17] C.P. Tsui, D.Z. Chen, C.Y. Tang, P.S. Uskokovic, J.P. Fan, X.L. Xie, Compos. Sci.

Technol. 66 (2006) 1521–1531.