Sage Journals: Discover world-class research

Abstract

The main aim of this study is to investigate the mechanical and structural properties of the interface layer (interlayer) between alumina and spinel materials after high temperature thermal treatment (1600°C). A microstructurally distinct layer with columnar grains of up to 50 μm length was detected by SEM. Mechanical properties, such as indentation hardness and elastic modulus of the interface layer, parent alumina and spinel parts, were measured and compared by nanoindentation method. According to the results, several microstructural factors including initial alumina raw powders, the intrinsic microstructural inhomogeneity of the interlayers, such as presence of porosities, unresolved hard alumina phase in the microstructures, size of columnar grains and position of grain boundaries were proposed to affect the mechanical properties. Different micromechanical results between the regions of the samples were also attributed to generation of residual compressive stress during grinding and polishing of the surfaces.

Keywords

Alumina Spinel Interface (interlayer)Nanoindentation Hardness Elastic modulus

Introduction

Bimaterials are composed of at least two layers or parts which are made of different materials or different compositions. They have functional properties, which depend on mechanical, electrical and magnetic properties of their components. Their application areas range from electronic packaging applications such as multilayer ceramic capacitors to thin film–substrate systems widely used in the microelectronics industry.^1–3

Cosintering process is to sinter two materials to one piece while they are in contact with each other; therefore, it can produce bimaterial with novel properties. Many researchers have studied cosintering process for different kinds of materials to understand and analyse the formation of mismatch stress due to difference in shrinkage strain of materials during cosintering and cooling processes.^4–9 The densification and the microstructural evolution during cosintering of alumina (Al₂O₃)–zirconia (Y-ZrO₂) and alumina–spinel (MgAl₂O₄) bimaterials, produced by copressing of powders were investigated in a previous study,⁷ where Yalamaç et al. focused on cosintering behaviours of oxide based bimaterials. Intermediate new spinel (MgAl₂O₄) phase forms when magnesia and alumina are heated together. The formation of new spinel phase has been studied from mainly diffusion perspective by investigating the diffusion behaviours of the components.^10–13 In another previous study,¹⁴ we investigated the microstructure of the interface layer (interlayer) between alumina and spinel materials after high temperature thermal treatment (1500°C). After cosintering of copressed alumina and spinel compacts, a microstructurally distinct interlayer with columnar grains of up to 40 μm length and 5 μm width was detected. Chemical composition of columnar spinel grains was slightly different from the parent spinel in bimaterials such that the concentrations of Mg in the columnar spinel region (13·5 wt-%) was lower than the stoichiometric amount of 16·9 wt-%.

On the other hand, analysis of micromechanical properties of the interface layer formed during cosintering of alumina–spinel compacts is very important to investigate the effect of production parameters and material types on diffusion control mechanism. Depth sensing micro- and nanoindentation techniques are widely used in the characterisation of mechanical behaviours especially hardness and elastic modulus of materials at small scales.¹⁵ Mechanical properties of materials can be determined directly from the indentation load and displacement measurements without the need to image the hardness impression.¹⁶ With high resolution testing equipment, this facilitates the measurement of properties at the micro- and nanometre scales.¹⁷ For this reason, the method has become a primary technique for determining the mechanical properties of thin films and small structural features.¹⁸

Hardness of materials is sensitive to indentation depth. Since hardness is accepted as an inherent material property, it should not vary with indentation load and size. However, investigations^19,20 have confirmed that hardness values of coatings and bulk materials are indentation size dependent especially at lower peak loads. When measured at small scales, the strength of many materials is greater than when tested in bulk form. An example of this phenomenon is the so called indentation size effect (ISE), which is manifested as an increase in hardness with decreasing indentation depth for conical indenters^21,22 and decreasing indenter radius for spherical indenters due to low peak load.^23,24 Mechanistically, the effect has been explained on the basis of Taylor hardening by the geometrically necessary dislocations to accommodate indentation induced plastic strain gradients.^21–25 Additionally, the difference between the heights before and after indentation is called a penetration depth and indicated as h_p, where the subscript ‘p’ stands for penetration. The ratio of h_max–h_p to h_max is called an elastic recovery rate (% ERR).²⁶

There are many studies in the literature on measurement of micromechanical properties of alumina and spinel ceramics by nanoindentation method. Oliver and Pharr¹⁶ tested indentation load–displacement behaviour of single crystal sapphire at maximum peak load of 120 mN with a Berkovich indenter. According to their study, the elastic modulus of the material was 441 GPa. In their other studies,^18,27 in addition to sapphire, they also measured elastic modulus and hardness of (100) single crystal spinel ceramics as 286 and 18·4 GPa respectively. Krell and Schadlich¹⁹ investigated depth sensing of submicrometre sintered alumina ceramics by nanoindentation at loads of 20–200 mN using Vickers hardness measurements. According to their results, Vickers hardness of alumina ceramics rises up to 30 GPa for dense microstructures and up to 25 GPa with 5% of residual porosity. They found that the indentation size effect and porosity are essentially the same as observed at larger loads (1000–5000 mN) for conventional measuring approaches. Twigg et al.²⁸ examined the initiation of surface damage under point loading in polycrystalline alumina materials using low load continuous depth sensing indentation equipment. Hot pressed alumina ceramics were impressed at a peak load of 1300 mN with a cube corner diamond indenter. They measured the mean value of elastic modulus and indentation hardness for pure alumina as 458 and 24·7 GPa respectively.

Elastic modulus of interface layer is an important parameter to understand and analyse the mechanical properties of whole structure. Nanoindentation techniques can provide well defined mechanical parameters such as elastic modulus of the interfacial zone between two different materials. The aim of this study is to analyse the mechanical properties of the interface layer formed by diffusion during cosintering of alumina–spinel compacts. Cosintering behaviour of alumina–spinel bimaterials, interface formation and the microstructural characterisation of the materials were studied in our previous studies.^7,14 In this study, mechanical properties of interlayer, especially hardness and elastic modulus, were tested as opposed to the above studies, which investigated the mechanism of microstructurally distinct interlayer formation.

Experimental

In this study, magnesium aluminate spinel (MgAl₂O₄) powder and two different commercially available submicrometre grained α-alumina powders were used for sintering process. One of the alumina powders (BMA15) and spinel powder (S30-CR) were Baikowski products, and the other alumina powder was an Almatis product (CT3000SG). Some physical and chemical properties of the powders are shown in Table 1.

Table 1

Measured physical and chemical properties of powders^38,39

		CT3000SG	BMA15	Spinel
d₅₀ by Sedigraph/μm		0·5	0·18	0·19
Specific surface area BET/m² g⁻¹		7·8	13·30	31
d_BET calculated from BET data/μm		0·193	0·113	0·055
Chemical analysis/ppm	Na	223	12	10
	Fe	105	6	10
	Si	204	10	20
	Ca	107	2	5
	Mg	241	…	…
	K	…	20	…

In order to achieve a good bond between alumina and spinel, the interlayer development must be controlled very well. Hence, two different sets of tests were planned, the first of which involved cosintering of alumina–spinel pairs of copressed pellets in order to see the effect of different types of alumina on the interface layer microstructure. The second one was run to collect information about mechanical properties of the interface layer and also parent alumina and spinel parts by nanoindentation test.

The green cylindrical (10 mm diameter) bimaterial samples were previously produced at a pressure of 250 MPa with single action uniaxial pressing mode. In the green compact production, the first powder was poured into die cavity and was settled down by tapping with a metal rod before the second powder was added and copressed together. Two different types of green compacts of bimaterials were produced based on alumina type (BMA15 or CT3000SG) by this way. Designation of bimaterials type is shown in Table 2. Prepared pellets were sintered in a high temperature furnace (HT17 Carbolite, England) at 1600°C. Sintering was performed with 4 h soaking or without soaking with 3·3°C min⁻¹ of heating rate. The sintered samples were cooled in furnace with 30°C min⁻¹ cooling rate.

Table 2

Designation of bimaterials type

Type of bimaterials	Materials combination	Sintering at 1600°C
Type of bimaterials	Materials combination	Without soaking	4 h soaking
Spinel–alumina B	Spinel+BMA15	SB-0	SB-4
Spinel–alumina A	Spinel+CT3000SG	SC-0	SC-4

In order to investigate the microstructures of the interlayer and components of the bimaterials after cosintering, sintered pellets were cut parallel to the cylindrical axis into two parts. Half of them were mounted into polyester resin before being ground and polished by conventional sample preparation methods. To reveal the morphology and microstructural alteration at the interlayer, the other half of bimaterials were ground and polished by finger before thermally etched at ∼100°C below the sintering temperature. Microstructures of the polished and thermally etched surfaces were observed by SEM (Quanta 250, FEI).

In this study all nanoindentation test samples were ultimately polished with 1 μm diamond paste to produce an optical finish before testing. The samples were analysed with an Ibis Berkovich Tip Nano Indenter produced by Fischer-Cripps Laboratories Pty Ltd (maximum indentation load up to 500 mN). Before nanoindentation tests, two calibration methods were applied to obtain accurate results. The first one is the calibration of the tip to scope travel distance, which is essential if accurately positioned indentations are required. This calibration is carried out by making a large indentation into the standard specimen and then translating over to the microscope position, finding the impression and recording the distance travelled. The second is the calibration of the results of nanoindentation machine. The machine was calibrated by performing a series of indentation tests on a standard fused silica (SiO₂) sample. The procedure is a very good way of verifying the performance and calibration of the instrument. If the hardness results are consistent with expectations, then the instrument is ready for use. After calibration procedure, the indentation tests were designed for an easier interpretation of the mechanical behaviours of materials. Four different loads of 300, 400, 425 and 450 mN were applied for five times at regular intervals of 25 μm, and 5 s dwell time was used at each peak load. A distance of at least three diagonal lengths between the centres of the indentations was allowed to avoid interaction between the workhardened regions and effects of the edge. In the depth sensing indentation measurements, the indentation depth h and the load P were recorded by a computer. The raw data were then used to construct the load–displacement plots.

Results and discussion

Microstructure of cosintered alumina–spinel Bi materials

Alumina and spinel copressed green compacts were cosintered at 1600°C at a heating rate of 3·3°C min⁻¹ with 4 h soaking or without soaking. An interlayer composed of new generation columnar spinel grains was observed between the alumina and spinel parents (Fig. 1). The thickness of interlayer was measured from SEM images, and the results are given in Table 3. According to the results, interlayer thickness of the bimaterials is nearly 50 μm, and width of columnar grains is ∼5 μm after sintering at 1600°C for 4 h soaking. There is slight difference between the interlayer thicknesses of bimaterials depending on the type of alumina. The results also showed that the length of interlayer was directly proportional to the square root of soaking time in accord with previous studies of the author.^7,14

General view of interlayer microstructure of a spinel–alumina A (SC-0), b spinel–alumina B (SB-0) bimaterials cosintered at 1600°C without soaking, c spinel–alumina A (SC-4) and d spinel–alumina B (SB-4) bimaterials cosintered at 1600°C for 4 h soaking

Table 3

Thickness measurements of interlayers

	Spinel–alumina A	Spinel–alumina B
1600°C, without soaking	(SC-0): 8·00 μm	(SB-0): 7·33 μm
1600°C for 4 h soaking	(SC-4): 53·00 μm	(SB-4): 42·50 μm

Micromechanical properties of cosintered alumina–spinel bimaterials

Micromechanical properties such as hardness and elastic modulus of interlayer, parent alumina and spinel parts of samples were measured by nanoindentation method. In order to accurately measure the mechanical properties of interlayer, 4 h soaked SB-4 and SC-4 coded samples that had a thicker interface were tested by nanoindentation technique. Four different loads were applied to determine the effects of indentation load and size on micromechanical properties of materials, depending on the microstructure. According to the applied indentation testing procedure in this study, five impresses for each load were applied to obtain minimum standard deviation in results. For example, Fig. 2a and b shows the indentation mapping of samples under different applied loads for alumina parts SC-4 and SB-4 respectively. When the applied load was increased from 300 to 450 mN, deformation surface area of samples increased, as seen in mapping images. The interface region of bimaterials could be also detected easily from the unetched sample parts by optical lens of the nanoindenter; therefore, all the related indentation loads were applied to the centre of the interface region and parallel through the layer.

General view of nanoindentation test indents at different loads on alumina parts of a spinel–alumina A (SC-4) and b spinel–alumina B (SB-4) bimaterials cosintered at 1600°C for 4 h soaking

The average loading–unloading (load–displacement and indentation curve) curves of the alumina, spinel and interlayer regions of SB-4 and SC-4 under 425 mN applied force are shown in Fig. 3a and b respectively. In Fig. 3, the applied load was a function of the displacement (elastic and plastic) of the indenter with respect to the initial position of the surface. Predictably, two different samples, three different regions for each sample and four different indentation loads give lots of indentation curves and mapping images. These graphs are not given here for the sake of brevity. Thus, only average loading–unloading curves of SB-4 and SC-4 sample regions at 425 mN applied load and the representative impress images (plastically deformed region) taken at the same load are presented in Figs. 3 and 4.

Average nanoindentation curves of a SB-4 and b SC-4 at 425 mN applied force

Images (SEM) of indentation impress for a SB-4-alumina, b SB-4-interlayer, c SB-4-spinel, d SC-4-alumina, e SC-4-interlayer and f SC-4-spinel under 425 mN applied force

When the indentation curves of SB-4 sample were analysed (in Fig. 3a), maximum indentation depth of alumina part of the sample was found smaller than interlayer and spinel parts. According to this result, under the same applied indentation load, alumina region of sintered samples had the highest elastic modulus and hardness, comparing with interlayer and spinel, as expected. The indentation curves of SC-4 sample are given in Fig. 3b. The curve of alumina part showed nearly the same behaviour with interlayer and spinel parts.

In order to avoid any influences of cracking on the hardness and elastic modulus results, the shape of the indents and surroundings were investigated by conventional light optical and scanning electron microscopes (in Figs. 2 and 4). From some of the indent micrographs, it appears that few microcracks like scratch formations are present around the indents. These are thought to originate from the surface finishing, as the applied peak loads (300–450 mN) were considerably lower than the minimum load for the fracture of spinel and alumina by Vickers indentation, which is normally >1 N.

Additionally, indentation results of interlayers are displayed in Figs. 5–8 in details. Figure 5 shows the indentation depth results of interlayer of SB-4 and SC-4 samples. According to the results, when the indentation force was increased from 300 to 450 mN, maximum and minimum indentation depths of samples increased. It means that the impression areas in interlayers of the sample increased due to larger plastic deformation under different applied indentation forces. However, maximum average indentation depth of SC-4 interlayer was measured as 1·17 μm at 450 mN applied force. Reasonably, indentation depth must be increased by increasing applied force for the same material. The reason of this inconsistency might be localised measurement due to the presence of rarely unresolved hard alumina phase in interlayer region. Besides, the width of the columnar grains is nearly the same as the impress size and some porosities are also present in the interlayer. Therefore, some of the indentations might have fallen on the grain boundaries or porosities during the indentaiton experiments.

Indentation depths of interlayer of SC-4 and SB-4 for different applied loads

The other deformation characteristics of the interlayers were ERRs (%ERR) under different applied forces (in Fig. 6). The theory of %ERR uses elastic and plastic work of samples, determined by typical nanoindentation load–displacement curves and maximum–minimum indentation depths. The indentation curves of SB-4 and SC-4 under 425 mN applied force are shown in Fig. 3; the curves had two parts: elastic and plastic regions. When the applied force was unloaded, elastically deformed depth recovered and plastically deformed depth appeared. From these results, ERRs of interlayers were calculated and displayed in Fig. 6. According to the results, ERRs of both interlayers of SB-4 and SC-4 had slight scattering and did not change systematically by applied force. Elastic recovery rate depends on elastic modulus, hardness and chemical composition of materials. If the material has high elastic modulus and hardness values, then high value of %ERR is expected for the material. On the other hand, some deviations from general trend were detected due to the same factors as the indentation depth results.

Elastic recovery rates (ERRs) of interlayer of SC-4 and SB-4 for different applied loads

Figures 7 and 8 represent the indentation hardness and elastic modulus of the samples. The error bars represent standard deviation of the means, often called the standard error for each load from five indentation results. According to the results, when the indentation force was increased from 300 to 450 mN, indentation hardness of SB-4 and SC-4 alumina regions decreased from 42 to 32 GPa and from 40 to 32 GPa (in Fig. 7), and elastic modulus of SB-4 and SC-4 alumina regions decreased from 592 to 547 GPa and from 593 to 517 GPa respectively (in Fig. 8). In these samples, different types of alumina were used as explained in Table 2. The results are close to each other at low indentation forces, but they are quite different at high forces (450 mN). Same type of spinel powder was used for the production of SB-4 and SC-4 bimaterails. Hardness and elastic modulus of spinel regions were also compared in Figs. 7 and 8. As the indentation force was augmented from 300 to 450 mN, indentation hardness of SB-4 and SC-4 spinel regions decreased from 26 to 25 GPa and from 26 to 23 GPa (in Fig. 7), and elastic modulus of SB-4 and SC-4 spinel regions decreased from 368 to 349 GPa and from 347 to 318 GPa respectively (in Fig. 8). Indentation hardness and elastic modulus variations of spinel regions of the samples showed different characteristic from depth sensing analysis. For the same materials, slightly different values were measured. Although alumina and spinel parts of the sample had regular microstructure and composition, these regions have some porosities and some different amount of residual stresses due to grinding and polishing processes.

Indentation hardness variations of alumina, interlayer and spinel parts of a SB-4 and b SC-4 under 300, 400, 425 and 450 mN applied forces

Elastic modulus of alumina, interlayer and spinel parts of a SB-4 and b SC-4 under 300, 400, 425 and 450 mN applied forces

One of the most important influences on the results from nanoindentation testing is the presence of residual stress in the specimen surface. This usually occurs in thin films, but may also result from grinding and polishing of bulk specimens. It is generally observed that the contact stiffness increases with increasing compressive residual stress. Studies on ceramic grinding have shown that the residually stressed layer is subjected to compression near the surface and tension underneath.²⁹ Although a quantitative analysis relating the effect of the existence of such a residually stressed surface on the hardness measurements is lacking in this work, it is still suspected that the presence of residual stress may affect the values of E and H.

The interlayer micromechanical properties were also compared in the Figs. 7 and 8. Indentation hardness of SB-4 and SC-4 interlayer decreased from 25 to 23 GPa and from 24 to 22 GPa respectively (in Fig. 7). The elastic modulus of interlayer of SB-4 and SC-4 decreased from 473 to 437 GPa and 383 to 344 GPa by increasing the applied force from 300 to 450 mN respectively (in Fig. 8).

Consequently, according to the results of Figs. 7 and 8, indentation hardness and elastic modulus of the interlayer of SB-4 was higher than the interlayer of SC-4. As explained the in characterisation part of the study, SB-4 included spinel+alumina BMA15 materials combination, whereas SC-4 had spinel+alumina CT3000SG. Difference between raw materials (type of alumina) resulted in different interlayer characteristics such as elastic modulus, hardness and ERR, impress area, etc., after sintering at 1600°C for 4 h soaking. The reason for this observation might be due to physical and chemical properties of raw powders such as presence of impurities in CT3000SG. These impurities are known to inhibit grain growth in the parent phases.

Moreover, indentation hardness and elastic modulus of the interlayer and spinel parts of bimaterials were also evaluated and compared. Hardness values of spinel parts are slightly higher than interlayer, but elastic modulus results show the opposite situation. Thus, the results of interlayer are higher than spinel region. Interlayer properties might be expected in between alumina and spinel parts. According to the indentation theory, elastic modulus and hardness of the regions were calculated from load–displacements curves. For instance, E and H values at 425 mN were calculated from Fig. 3. The elastic moduli of materials are calculated from 70% of unloading part of indentation curves, contact area and ratio of dP/dh. On the other hand, indentation hardnesses of materials were measured form apparent area of impress or final (penetration) depth when unloading procedure finished. Each part of the bimaterials shows different slopes on unloading curves and has different penetration depths h_p. Therefore, in this study, elastic modulus is not considered to be an increasing function of hardness.

According to the results of interlayers, maximum and minimum indentation depths increased, whereas indentation hardness and elastic modulus of interlayer decreased with increasing indentation force. Therefore, mechanical properties of the samples showed changes depending on different factors. The first one might be the indentation size effect (mechanical response of applied force). The second one might be the chemical composition difference. According to the previous study,¹⁴ chemical composition of columnar spinel grains was slightly different from the parent spinel. The third one might be the microstructural characteristics of the interlayer such as presence of porosity, unresolved alumina phase and grain boundary positions that significantly affect the results. Gong et al.,³⁰ examined the effect of microstructural inhomogeneity on the reproducibility of the nanoindentation data. They found that the results of nanohardness, and the Young's modulus of high purity, fine grained alumina ceramic were somewhat poor. Large scatters were observed in the results of H and E. These experimental findings revealed that microstructural inhomogeneity may play an important role in the material response to nanoindentation, and the responses should be treated as the local, rather than the bulk mechanical properties of the test.

As a result, indentation size effect, chemical composition and microstructure of interlayer formed by the diffusion controlled reaction between alumina and spinel at 1600°C substantially affected the elastic modulus and hardness under the same indentation force.

There are many ISE models to propose for describing the variation of the indentation size with the applied test load. The concepts that are thought to provide a basis for the occurrence of ISE involve the variation of contact surface, friction between the surface and the indenter, microfracture processes, presence of residual surface stress, energy dissipation associated with the contact surfaces and variation in solid load bearing contact area inside the indentation volume. Most importantly, the minimum resistance on the surface model, the proportional specimen resistance model and the modified proportional specimen resistance model provided considerable insight in explaining the ISE.^31,32 The main aim of the study is to just characterise elastic modulus and hardness of the interlayer and compare with parent alumina and spinel parts by nanoindentation method. In this study, applied loads of indentation tests were selected in narrow range (300–450 mN). Therefore, from a statistical point of view, the narrow load range is not enough and suitable for applying any ISE models for the whole structure of the bimaterials. Applying wide range nanoindentation load (10–500 mN) to explore ISE models and consequently attempting to find out if these same models remain valid in explaining ISE for a given material from alumina side through interface to the spinel side are the aims of the future work.

In this study, applied indentation forces are in the intermediate range (300–450 mN), since in the literature, Krell and Schadlich¹⁹ applied very low load range (20–200 mN) and Twigg et al.²⁸ applied high load of 1300 mN. In the former, hardness of sintered alumina ceramics was measured by Vickers indentation test. They reported that the Vickers hardness of sintered submicrometre alumina rises up to ∼30 GPa for dense microstructures and up to ∼25 GPa with 5% of residual porosity. In the latter, elastic modulus of polycrystalline alumina materials were investigated using cube corner indenter and 4 μm maximum depth under 1–2 N applied force was obtained. In their study, they observed indentation hardness and elastic modulus as 24·7±37 GPa and 485±75 GPa respectively. In addition to these studies, several researches^31,33–36 also studied micromechanical properties such as elastic and plastic properties indentation curve analysis and fracture toughness of alumina. Mao et al.³⁷ investigated the nanoscale pop-in phenomena in polycrystalline α-Al₂O₃ and single crystal α-Al₂O₃ (001) by nanoindentation with Berkovich indenters. They found that different radii and loading rates have significant effects on the pop-in formation, stress distributions and dislocation nucleations. At the load independent hardness region, the range of H equals 27·5±2 GPa for α-Al₂O₃ (001), and the range of H is ∼30±3 GPa for polycrystalline α-Al₂O₃. It can be concluded that the H of α-Al₂O₃ (001) is slightly smaller than that of polycrystalline α-Al₂O₃ at the small size scale in their work. However, in this study, indentation hardness of alumina parts of bimaterials was measured between 30 and 42 GPa with the Berkovich indenter.

Pharr et al.^18,27 measured elastic modulus of 286 GPa for (001) single crystal spinel using a cube corner indenter. In this study, the values of E between 300 and 350 GPa for polycrystalline spinel part of bimaterials are in close agreement with the generally literature value for single crystal spinel of ∼286 GPa, for the (100) direction.^18,27 On the other hand, indentation hardness of spinel part is between 22 and 26 GPa, which is slightly higher than the literature result of 18·4 GPa for single crystal.¹⁸ The reason for the difference might be the contact area of indent and the indentation size effect under 300–450 mN applied loads. The observed micromechanical properties of samples under different applied loads varied in alumina and spinel compacts and their interlayer due to microstructural differences. As several researches^{15–28,31,33–36} indicated, indentation size, production and microstructure effects under low indentation loads altered the hardness and elastic modulus of ceramic based materials. In addition to these affects, the compression residual stress on the polished surfaces might affect the hardness and elastic modulus of bimaterials. Therefore, hardness and elastic modulus results are slightly higher than literature results in this study.

Conclusions

In this study, micromechanical properties of alumina, spinel and interlayer parts of the bimaterials (SB-4 and SC-4) were tested by nanoindentation method.

Alumina, spinel and interlayer regions of SB-4 bimaterial have slightly higher nanohardness values than the regions of SC-4. On the other hand, SB-4 bimaterial regions have considerably higher elastic modulus results than the regions of SC-4. Therefore, SB-4 bimaterial showed higher micromechanical properties than the bimaterial of SC-4. There might be some factors that resulted in different hardness and elastic moduli between the regions of bimaterials. One of the significant factors might be the difference between the initial alumina raw powders. Two different alumina powders used should significantly affect the results. In addition to this, interlayers have scattered minimum and maximum indentation depth results. These scatters may be attributed to the intrinsic microstructural inhomogeneity of these two interlayers such as presence of porosities, unresolved hard alumina phase in the microstructures, size of columnar grains and position of grain boundaries. Another reason might be the generation of residual compressive stress during grinding and polishing of the surfaces that influenced the results and probably somewhat raised the hardness and elastic modulus results. Generally, the elastic modulus and hardness results of experiments are slightly higher than literature results due to the same factors mentioned above. The phenomenon of ISE usually involves a decrease in the measured apparent hardness with increasing applied test load through whole structure of the bimaterials. Unfortunately, from the statistical point of view, the experimental narrow load range was not enough and suitable to apply any ISE models for the samples.

Although further study is needed to characterise the magnitude of residual stress on the polished surface and explore indentation size effect model for the whole bimaterial structure, it is hoped that this study can contribute to a greater understanding of mechanical properties of interlayer and parent alumina and spinel ceramic parts.

Future work will concentrate on measurement of indentation fracture resistance of interlayer under dynamic ultra microhardness test.

Footnotes

Acknowledgements

The authors are grateful to Baikowski and Almatis for providing alumina and spinel powders. This work was supported financially by the Research Fund of Celal Bayar University (grant no. BAP 2011-25).

References

Boonyongmaneerat

and Schuh

: ‘Contributions to the interfacial adhesion in co-sintered bilayers’, Metall. Mater. Trans. A, 2006, 37A, 1435–1442.

Simchi

, Rota

and Imgrund

: ‘An investigation on the sintering behavior of 316L and 17-4PH stainless steel powders for graded composites’, Mater. Sci. Eng. A, 2006, A424, 282–289.

Moreno

: ‘Colloidal processing of ceramics and composites’, Adv. Appl. Ceram., 2012, 111, 246–253.

Cai

, Green

and Messing

: ‘Constrained densification of alumina/zirconia hybrid laminates. I: experimental observations of processing defects’, J. Am. Ceram. Soc., 1997, 39, 1929–1939.

Lannutti

, Deis

, Kong

and Phillips

: ‘Density gradient evolution during dry pressing’, Am. Ceram. Soc. Bull., 1997, 76, 53–58.

Ravi

and Green

: ‘Sintering stresses and distortion produced by density differences in bi-layer structures’, J. Eur. Ceram. Soc., 2006, 26, 17–25.

Carry

, Yalamaç

and Akkurt

: ‘Co-sintering behaviors of oxide based bi-materials’, in ‘Advances in sintering science and technology: ceramic transactions’, (ed. E. Olevsky and R. Bordia), 307–320; 2010, Westerville, OH, American Ceramic Society.

Götschel

, Gutbrod

, Travitzky

, Roosen

and Greil

: ‘Processing of preceramic paper and ceramic green tape derived multilayer structures’, Adv. Appl. Ceram., 2013, 112, 358–365.

Padovano

, Badini

, Biamino

, Pavese

, Yang

and Fino

: ‘Pressureless sintering of ZrB2–SiC composite laminates using boron and carbon as sintering aids’, Adv. Appl. Ceram., 2013, 112, 478–486.

10.

Navias

: ‘Preparation and properties of spinel made by vapor transport and diffusion in the system MgO–Al₂O₃’, J. Am. Ceram. Soc., 1961, 44, 434–446.

11.

Watson

and Price

: ‘Kinetics of the reaction MgO + Al₂O₃ –> MgAl₂O₄ and Al–Mg interdiffusion in spinel at 1200 to 2000°C and 1·0 to 4·0 GPa’, Geochim. Cosmochim. Acta, 2002, 66, 2123–2138.

12.

Zhang

, Debroy

and Seetharaman

: ‘Interdiffusion in the MgO–Al₂O₃ spinel with or without some dopants’, Metall. Mater. Trans. A, 1996, 27A, 2105–2114.

13.

Carter

: ‘Mechanism of solid state reaction between magnesium oxide and aluminum oxide and between magnesium oxide and ferric oxide’, J. Am. Ceram. Soc., 1961, 44, 116–120.

14.

Yalamac

, Carry

and Akkurt

: ‘Microstructural development of interface layers between co-sintered alumina and spinel compacts’, J. Eur. Ceram. Soc., 2011, 31, 1649–1659.

15.

Oliver

and Pharr

: ‘Measurement of hardness and elastic modulus by instrumented indentation: advances in understanding and refinements to methodology’, J. Mater. Res., 2004, 19, 3–20.

16.

Oliver

and Pharr

: ‘Improved technique for determining hardness and elastic modulus using load and displacement sensing’, J. Mater. Res., 1992, 7, 1564–1583.

17.

Pethica

, Hutchings

and Oliver

: ‘Hardness measurement at penetration depths as small as 20 nm’, Philos. Mag. A, 1983, 48A, 593–606.

18.

Pharr

: ‘Measurement of mechanical properties by ultra-low load indentation’, Mater. Sci. Eng. A, 1998, 253, 151–159.

19.

Krell

and Schädlich

: ‘Nanoindentation hardness of submicrometer alumina ceramics’, Mater. Sci. Eng. A, 2001, A307, 172–181.

20.

Uzun

, Kölemen

, Çelebi

and Güçlü

: ‘Modulus and hardness evaluation of polycrystalline superconductors by dynamic microindentation technique’, J. Eur. Ceram. Soc., 2005, 25, 969–977.

21.

Nix

and Gao

: ‘Indentation size effects in crystalline materials: a law for strain gradient plasticity’, J. Mech. Phys. Solids, 1998, 46, 441–425.

22.

Gao

, Huang

, Nix

and Hutchinson

: ‘Mechanism-based strain gradient plasticity. I: theory’, J. Mech. Phys. Solids, 1999, 47, 1239–1263.

23.

Swadener

, George

and Pharr

: ‘The correlation of the indentation size effect measured with indenters of various shapes’, J. Mech. Phys. Solids, 2002, 50, 681–694.

24.

, Huang

, Pharr

and Hwang

: ‘The indentation size effect in the spherical indentation of iridium: a study via the conventional theory of mechanism-based strain gradient plasticity’, Int. J. Plast., 2006, 22, 1265–1286.

25.

Shim

, Bei

, George

and Pharr

: ‘A different type of indentation size effect’, Scr. Mater. 2008, 59, 1095–1098.

26.

Park

, Kim

, Cho

, Oh

, Kwak

and Lee

: ‘Nano- and micro-indentation of photo-patterned spacers’, Proc. ASID-06, New Delhi, India, October 2006, 299–301.

27.

Harding

, Oliver

and Pharr

: ‘Cracking during nanoindentation and its use in the measurement of fracture toughness’, Mater. Res. Soc. Symp. Proc., 1995, 356–663.

28.

Twigg

, Riley

, Roberts

, Road

and Ox

: ‘Nanoindentation investigation of micro-fracture wear mechanisms in polycrystalline alumina’, J. Mater. Sci., 2002, 37, 845–853.

29.

Johnson-Walls

, Evans

, Marshall

and James MR : ‘Residual stresses in machined ceramic surfaces’, J. Am. Ceram. Soc., 1986, 69, 44–47.

30.

Gong

, Peng

and Miao

: ‘Analysis of the nanoindentation load–displacement curves measured on high-purity fine-grained alumina’, J. Eur. Ceram. Soc., 2005, 25, 649–654.

31.

Chakraborty

, Dey

, Mukhopadhyay

, Joshi

, Rav

, Mandal

, Bysakh

, Biswas

and Gupta

: ‘Indentation size effect of alumina ceramic shocked at 12GPa’, Int. J. Refract. Met. Hard Mater., 2012, 33, 22–32.

32.

Peng

, Gong

and Miao

: ‘On the description of indentation size effect in hardness testing for ceramics: analysis of the nanoindentation data’, J. Eur. Ceram. Soc., 2004, 24, 2193–2201.

33.

Yuan

and Hayashi

: ‘Influence of the grain size of the alumina coating on crack initiation in indentation’, Wear, 1999, 225–229, 83–89.

34.

Venkataraman

and Krishnamurthy

: ‘Evaluation of fracture toughness of as plasma sprayed alumina–13 wt.% titania coatings by micro-indentation techniques’, J. Eur. Ceram. Soc., 2006, 26, 3075–3081.

35.

Huang

, Binner

JGP

, Vaidhyanathan

and Todd

: ‘Quantitative analysis of the residual stress and dislocation density distributions around indentations in alumina and zirconia toughened alumina (ZTA) ceramics’, J. Eur. Ceram. Soc., 2014, 34, 753–763.

36.

Miyazaki

and Yoshizawa

: ‘Refined measurements of indentation fracture resistance of alumina using powerful optical microscopy’, Ceram. Int., 2014, 40, 2777–2783.

37.

Mao

, Shen

and Lu

: ‘Deformation behavior and mechanical properties of polycrystalline and single crystal alumina during nanoindentation’, Scr. Mater., 2011, 65, 127–130.

38.

Baikowski. http://www.baikowski.com (accessed 10 January 2014).

39.

Almatis. http://www.almatis.com (accessed 10 January 2014).