SPS Ni Cr Al Y Manuscript Jay

1    

High Temperature Microstructure and Microhardness Evolution in Dense NiCrAlY

Bulk Material Fabricated by Spark Plasma Sintering

Jianrong Song a, b, Kaka Ma a, Ying Li a, Lianmeng Zhang b, Julie M. Schoenung a*

*Corresponding author: [email protected]

a Department of Chemical Engineering and Materials Science,

University of California-Davis;

One Shields Avenue; Davis, CA 95616, USA

b Department of Materials Science and Engineering, Wuhan University of Technology,

122 Luoshi Rd.; Wuhan, Hubei 430070, China

Abstract

NiCrAlY is commonly used as the bond coat material in thermal barrier coatings (TBCs).

Detailed investigation into the microstructural evolution at high temperatures and the

consequent changes in microhardness in NiCrAlY bond coats is challenging due to the

complexity in microstructure and composition caused by thermal spray processes and the

presence of a superalloy substrate and a ceramic top coat in the TBC. In an effort to

simplify the system, fully dense NiCrAlY bulk samples were prepared by spark plasma

sintering (SPS) at 1000 °C. The microstructure, phase composition and microhardness of

the samples, in the as-sintered condition and select heat-treated-in-argon-and-quenched

conditions (1000 °C, 10 h & 50 h; 1050 °C, 10 h & 50 h; 1100 °C, 10 h & 50 h), were

studied by means of SEM, EDAX, HR-TEM and microhardness testing. It was found that

2    

the microstructure and phase compositions changed significantly after heat treatment.

Additionally, the average microhardness of the samples decreased as the heat-treatment

temperature and time increased, except for in the samples heat treated at 1000 °C for 50 h,

1050 °C for 10 h and 1050 °C for 50 h. To clarify this anomalous behavior,

microhardness was carefully measured within select grains. Results indicate that a

Cr,O-rich phase (specifically, Cr3O, as determined by HR-TEM), which appeared within

the β-NiAl grains after these three heat-treatment conditions but subsequently

disappeared from the β-NiAl grains after heat treatment at 1100 °C, strengthened the

overall microstructure.

Key words: thermal barrier coating, NiCrAlY, spark plasma sintering, phase

transformation, microhardness,

3    

1. Introduction

The degradation of turbine blades during service is one of the major concerns in material

science research within the field of energy production. Applying thermal barrier coatings

(TBCs) on the surfaces of blades is an effective method for protecting the materials from

hot corrosion and high temperature oxidation. It is well established that the performance

of TBCs is influenced by the properties of the bond coat layer, which lies between the top

coat and the substrate [1]. NiCrAlY alloys are one of the most important bond coat

materials used for protecting the substrates [2, 3]. Besides exhibiting high hardness, good

adhesion and good oxidation resistance, NiCrAlY bond coat materials can be tailor-made

in a wide range of compositions that correspond to the demands of the working

conditions and substrate materials [4]. The microstructure in NiCrAlY bond coats

consists mainly of two phases, the β-NiAl intermetallic and the γ-Ni(Cr) solid solution. In

some particular compositions, α-Cr and γ’-Ni3Al precipitates are also present within the

β-NiAl and γ-Ni(Cr) grains [5]. After thermal spray, Cr can also be present as a Cr-oxide

phase [6 - 8]. The Y generally separates out into Ni,Y-rich intermetallics [9].

During high temperature applications, the microstructure, phase composition, and

mechanical properties of NiCrAlY bond coats play a significant role in the durability of

the TBCs [10]. The relationship between structure and properties in these coatings has

been studied to some extent through the application of micro-tensile sample preparation

and testing, even at elevated temperatures [11-14]. The method is challenging, however,

because the bond coat layer is only on the order of ~150-200 µm thick [15] and it is

sandwiched between the ceramic top coat and the superalloy substrate. Moreover, this

method does not provide the opportunity to lock-in the high temperature microstructure

4    

for further characterization. In addition to these challenges, the bond coat microstructure

is inherently complex because of the thermal spray process by which it is usually applied

(such as air plasma spray, APS, or high-velocity oxy-fuel, HVOF), which can result in

the formation of splat boundaries, internal oxidation and other microstructural

inhomogeneities [16]. In order to overcome these challenges associated with

characterizing the bond coat layer itself, equivalent bulk materials are produced in the

current study through the application of spark plasma sintering (SPS).

SPS is a novel sintering technique characterized by its higher heating rate, lower sintering

temperature and shorter dwell time than conventional pressure-assisted sintering [17, 18].

The powder is heated by a direct-pulsed current, which flows through punches, die and

the processed material. The Joule heat due to the electric current rapidly leads to the

consolidation of the powder particles under pressure [19]. The SPS technique was first

applied to the NiCrAlY system in an effort to study the creep behavior of this system [20].

In recent years, SPS has been further applied to fabricate NiCrAlY coatings directly on a

substrate, resulting in coatings that exhibit better adhesion and antioxygenic properties

[21-23]. The relationships between microstructure and mechanical behavior of

SPS-processed NiCrAlY, specifically after exposure to elevated temperatures, has not,

however, been previously studied. Thus, the objectives of the current study are to

characterize the microstructural changes that occur in this material at elevated

temperatures and to quantify the consequential changes in mechanical behavior.

2. Experimental procedures

2.1 Materials

Commercially available, gas atomized NiCrAlY powder (Ni-343 alloy) with an average

5    

particle size of 12 µm and an average grain size of 1 µm was used as the feedstock

material. The chemical composition of the material was tested by chemical analysis, the

results of which are provided in Table 1.

2.2 Spark plasma sintering (SPS) process

The feedstock powder was loaded into a graphite die with a 20 mm inside diameter, and

pre-pressed under 25 MPa of uniaxial pressure using a hydrostatic press, to increase the

green density prior to sintering. In order to minimize the diffusion of carbon from the

graphite die into the sample, the inner wall of the die and inside faces of the plungers

were coated with a ~50-70 µm thick layer of BN prior to loading the NiCrAlY powder. A

schematic of the graphite die is shown in Fig. 1. Temperature 1 (at the die) and

Temperature 2 (at the top plunger) were measured by infrared pyrometers to obtain a true

value of temperature during the sintering process. The pre-assembled graphite die was

put into the chamber of the SPS device (SPS 825S system fabricated by DR. Sinter Inc.)

and pressed with 30 MPa of uniaxial pressure. This load and a vacuum of 4 Pa in the

chamber were applied throughout the sintering process. The details of the heating cycle

are as follows.

Room temperature 600 °C 1000 °C 1000 °C

room temperature

2.3 Heat treatment

The as-sintered samples were polished with SiC abrasive paper before heat treatment in

order to remove the carburized layer and BN layer on the surfaces of the sample. The

heat treatment process was performed in a muffle furnace, which has a maximum

6    

operating temperature of 1200 °C. The temperature control was performed using a

thermocouple in the middle of the furnace. Throughout the heat-treatment process, the

samples were sealed in stainless steel tubes filled with argon to avoid oxidation reactions.

The heat treatments were carried out at 1000 °C for 10 h, 1000 °C for 50 h, 1050 °C for

10 h, 1050 °C for 50 h, 1100 °C for 10 h and 1100 °C for 50 h. The temperatures studied

here were chosen according to the extreme service temperatures of NiCrAlY bond coats

in TBCs [24]. After heat treatment, the samples were quenched in water so that the high

temperature microstructure could be locked in.

2.4 Microhardness tests

The surfaces of the samples were polished with diamond polishing fluid (6 µm, 3 µm, and

1 µm) before conducting the microhardness tests. The hardness measurements were

performed using a microhardness tester (Micromet-2004 fabricated by Buehler Inc.). The

Micromet-2004 is a force-driven static measuring device for micro-indentation, and

equipped with a Vickers diamond pyramid indenter. The indenter is driven into, and

withdrawn from, the surface under a preset load. In this study, a load of 10 g was applied

for what we refer to as the average microhardness tests. The diagonal lengths of the

indentations under this test parameter were approximately 60-80 µm, which covers

enough area to get the average microhardness of material derived from the multiple

phases present in the microstructure. A minimal load of 0.01 g was applied for what we

call the localized phase-specific microhardness tests, which were conducted on individual

grains of the γ-Ni(Cr) phase and β-NiAl phase, respectively, on samples that had been

heat-treated. Instead of using the raw data generated from the microhardness tester, we

carefully measured the length of the indentation diagonals with scanning electron

7    

microscopy (SEM) so as to achieve higher accuracy. Five indentations were measured in

each sample to calculate the average values and standard deviation. The corresponding

microhardness values were calculated according to the equation below [25]:

HV=1.854F/D12

where HV is the Vickers hardness, F is the applied load, and D1 is the average diagonal

length of the indentations.

2.5 Characterization techniques

The microstructure and phase composition of the samples was investigated by using

micrographic examination and microprobe analysis performed with a Phillips FEI

XL30-SFEG scanning electron microscope (SEM), an Energy Dispersive X-ray

spectroscopy (EDAX) – ZAF Quantification (Standardless) system, and a JEOL

JEM-2500 high-resolution transmission microscope (HR-TEM). The average grain sizes

of select phases within the microstructure were calculated from SEM images through the

application of the image analysis software Image-Pro® 5.0.

3. Results

3.1 Spark Plasma Sintering (SPS) stages

Fully dense NiCrAlY bulk samples were successfully sintered by SPS at 1000 °C for 3

minutes under 30 MPa axial pressure. Fig. 2 shows the changes in electric current,

temperature and displacement that occur during the SPS process. Because temperatures

below 567 °C are not detectable by infrared pyrometry, only the temperature range above

567 °C is shown in Fig. 2. According to the displacement curve (noting that positive

displacement represents shrinkage), the SPS process can be divided into three stages,

consisting of: (A) preheating (room temperature-593 °C), (B) shrinkage (593-980 °C),

8    

and (C) thermal expansion (980-1001 °C).

3.2 Phase transformation and microstructure

The phases associated with the different microconstituents were identified by EDAX

mapping analysis (shown in Fig. 3): γ-Ni(Cr) (light gray), β-NiAl (dark gray), a Ni,Y-rich

phase (white, labeled as Ni-Y), and a Cr,O-rich phase (black spots, labeled as Cr-O).

These phases are consistent with conventional, thermal-sprayed NiCrAlY alloys, as

described in the Introduction. Although previously observed in our recent studies [6, 8],

the Cr-O phase has not previously been well characterized. Therefore, HR-TEM is

applied here to clarify the phase present in these regions. A representative HR-TEM

bright field (BF) image with the corresponding selected area electron diffraction pattern

is shown in Fig. 5. The Cr-O phase is thus identified as the cubic Cr3O phase. The higher

magnification image highlights the [2 1 0] crystal plane of a Cr3O grain.

Representative microstructures of the bulk NiCrAlY samples in the as-sintered condition

and after select heat-treatment conditions are provided in Figs. 3 and 4. Neither pores nor

cracks are observed in these samples. Significant grain growth and microstructure

evolution are observed after heat treatment. After heat treatment at 1000 °C/ 50 h, 1050

°C/ 10 h and 1050 °C/ 50 h, the β-phase grains have grown preferentially and have

engulfed the Cr-O phase and some of the Ni-Y phase. After heat treatment at 1100 °C/ 10

h, the Cr-O phase has disappeared from the microstructure. Longer exposure at 1100 °C

resulted in the apparent disappearance of the Ni-Y phase and further grain growth (Fig.

3b). Results from the image analysis on the β-phase and Cr-O grains after select heat

treatments (Fig. 6) indicate that, whereas the β-phase grains exhibited noticeable growth

with increased thermal exposure (increasing from 0.9 µm to 4.5 µm), the Cr-O grains

9    

remained fairly stable (increasing from 0.45 µm to only 0.75 µm).

Changes in the relative elemental compositions in the different phase regions were

confirmed by comparative EDAX measurements, as shown in Fig. 7, using an average of

five separate spots for each measurement in the histogram. For improved clarity, only the

values for Al and Cr are shown; the balance would be predominantly Ni. Given the

uncertainty associated with this measurement technique, we can qualitatively note that

the Al content increased and the Cr content decreased in the β-phase, as the thermal

exposure conditions became more severe, while the opposite trends were observed in the

γ-phase.

3.3 Microhardness

Fig. 8 shows the microhardness results and representative images of indentation

morphology. The average microhardness decreased significantly after heat treatment at

1000°C/ 10 h; then the average microhardness increased after heat treatment at 1000°C/

50 h, 1050 °C/ 10 h and 1050 °C/ 50 h, and decreased again after heat treatment at

1100°C/ 10 h and 1100°C/ 50 h (Fig. 8a). The localized phase-specific microhardness in

the γ-phase decreased as the heat-treatment temperature and holding time increased. The

localized phase-specific microhardness in the β-phase increased after 1000 °C/ 50 h, then

decreased rapidly after 1100 °C/ 10 h. The representative indentation from the average

microhardness test taken on the as-SPSed sample is a rhombus with a 62.9 µm diagonal

(Fig. 8b); numerous grains were deformed as a result of this indentation. The indentations

from the localized phase-specific microhardness tests taken on the sample heat treated at

1100 °C/ 10 h are located within only one grain of β- or γ-phase, respectively, as shown

in Figs. 7c and 7d.

10    

4. Discussion

4.1 Densification process

As shown in Fig. 2, temperature was more important than applied current during the

sintering process. The decrease in displacement during stage (A) demonstrates that there

was an expansion phenomenon owing to the thermal expansion of the powder and the

graphite die. The displacement began to increase with increasing temperature at 593 °C,

the beginning of stage (B), which indicates that the sample started to shrink at this

temperature. The shrinkage in stage (B) is approximately 42.5%, which is attributed to

the densification of the sample. Subsequently, the displacement began to decrease at 980

°C, because of the formation of a pore-free structure in the sample at this temperature.

The fully dense material exhibited a thermal expansion of 1.2% during stage (C).

4.2 Microstructure and microhardness

It is apparent from the images in Figs. 3 and 4 that the microstructure and phase

composition changed during the heat treatment process. Specifically, the microstructure

of the sample heat-treated at 1000 °C/ 10 h (the same temperature as the sintering),

shown in Fig. 4b, was significantly different from that of the as-SPSed sample (Fig. 4a),

with a growth in the γ-phase, β-phase, Cr-O and Ni-Y grains. After heat treatment at

1000°C/ 50 h, 1050 °C/ 10 h and 1050°C/ 50 h, the microstructure is further modified

with the Cr-O phase now observed within the β-phase grains instead of adjoining the

γ-phase matrix grains (compare Figs. 4c - 4e to Fig. 4b). This change in microstructure

can be attributed to the change in grain growth behavior, as illustrated in Fig. 6. Within

the heat-treatment temperature range of 1000 °C - 1050 °C, the β-phase grains grew

faster (with an increase in average grain size of over 180%) than the Cr-O grains (with a

11    

corresponding increase of only ~20%) leading to the Cr-O phase being engulfed within

the β-phase. As a result of this change in microstructure, the material exhibits an increase

in hardness despite an increase in average grain size [26]. It is interesting to note this

change in average microhardness illustrated in Fig. 8a for the 1000 °C/50 h versus the

1000 °C/10 h samples. These results can be explained by considering the localized

phase-specific microhardness data presented in Fig. 8a, which indicate that the presence

of the Cr-O phase within the β-phase grains leads to a substantial increase in β-phase

hardness. The small, uniform Cr-O phase grains act as second-phase dispersoids, which

provide for the strengthening effect [27]. This strengthening effect can be important in

application and during use, yet has never been previously demonstrated in NiCrAlY bond

coat materials. As expected, once the Cr-O phase is within the β-phase, further thermal

exposure, which leads to further grain growth (Fig. 6), leads to a decrease in both average

and localized phase-specific hardness values (Fig. 8a).

Further changes in the microstructure are observed after samples are heat treated at 1100

°C (Fig. 4f). Specifically, the Cr-O phase disappeared from within the β-phase, and, after

50 h at 1100 °C, the Ni-Y grains have also disappeared. The Cr-O phase consisted of

Cr3O (as identified in Fig. 5), which is a phase that forms in low oxygen conditions but is

unstable at elevated temperatures [28, 29]. As illustrated in Fig. 4f and 3b, the Cr-O

phase no longer exists in the samples exposed to the higher temperature heat treatments at

1100 °C. This change in microstructure results in a significant reduction in both the

localized phase-specific hardness of the β-phase grains and the average hardness (Fig. 8a),

because the dispersion strengthening effect from the Cr-O phase within the β-phase no

longer exists. The results of EDAX analysis (Fig. 7) provide evidence that after

12    

decomposition of the Cr-O phase at elevated temperature, the Cr has diffused into the

γ-phase. These results indicate that the Cr content in the β-phase grains decreased slightly

as the heat-treatment temperature was increased from 1050 °C to 1100 °C, while at the

same time, the Cr content in the γ-phase grains increased. After heat treatment at 1100

°C/ 50 h, Ni-Y grains can no longer be observed, yet some Ni-Y phase appears at the

interfaces between the β- and γ-phase regions (see Fig. 3b and arrows highlighted in Fig.

4f). This is not an unexpected phenomenon since the severity of these thermal exposure

conditions promotes diffusion and minimization of grain boundary interfaces.

4. Conclusions

Bulk NiCrAlY samples, fabricated at 1000 °C by SPS, were investigated to identify the

evolution in microstructure and hardness at elevated temperatures. There were three

stages in the sintering process, and the second stage from 593 °C ~ 980 °C was the

crucial stage for the densification of the samples. There were four phases in the

as-sintered sample: γ, β, a Ni,Y-rich phase and a Cr,O-rich phase. After heat treatment,

the composition and microstructure changed: at 1000 °C/10 h, grains grew; at 1000 °C/50

h, 1050 °C/10 h and 1050 °C/50 h, the interface between the β- and γ-phase migrated as

the β-phase grains grew and engulfed the Cr,O-rich phase and some of the Ni,Y-rich

phase; at 1100 °C/10 h, the Cr,O-rich phase (identified as Cr3O through HR-TEM)

disappeared as the Cr diffused into the γ-phase; at 1100 °C/50 h, the Y diffused from the

Ni,Y-rich grains into the grain boundaries and interfaces. As expected, the microhardness

decreased with increasing severity of the heat-treatment conditions, except for the

samples heated at 1000 °C/ 50 h, 1050 °C/ 10 h and 1050 °C/ 50 h. This anomalous

13    

behavior was explained on the basis of the dispersoid strengthening capability of the

Cr,O-rich grains that are located within the β-phase grains after these heat-treatment

conditions.

Acknowledgements

Financial support was provided by the U.S. Office of Naval Research (ONR) through

Grant N00014-06-1-0506 and the China Scholarship Council (CSC) through the

Government-sponsored Overseas Study Program.

14    

References

[1] M.C. Mayoral, J.M. Andrés, M.T. Bona, V. Higuera, F.J. Belzunce, Surf. Coat. Technol. 202 (2008) 1816-1824.

[2] D. Toma, W. Brandl, U. Koster, Surf. Coat. Technol. 120-121 (1999) 8-15.

[3] L. Ajdelsztajn, J.A. Picas, G.E. Kim, F.L. Bastian, J.M. Schoenung, V. Provenzanoa, Mater. Sci. Eng. A 338 (2002) 33-43.

[4] E. Lugscheider , C. Herbst, L. Zhao, Surf. Coat. Technol. 108-109 (1998) 16-23.

[5] T. Taylor, A. Ashary, A. Bolcavage, A. Feuerstein, J. Knapp, N. Hitchman, J. Therm, Spray Technol. 17 (2008) 199-213.

[6] K. Ma, J.M. Schoenung, Philos. Mag. Lett. 90 (2010) 739–751.

[7] F. Tang, L, Ajdelsztajn, J.M. Schoenung, Scr. Mater. 51 (2004) 25-29.

[8] K. Ma, X, Tang, J.M. Schoenung, Acta Mater. submitted.

[9] B. Baufeld, M. Bartscha, P. BroImage, M. Schmückera, Mater. Sci. Eng. A 384 (2004) 162-171.

[10] B. Saeedi, A. Sabour, A. Ebadi, A.M. Khoddami, J. Mater. Sci. Technol. 25 (2009) 499-507.

[11] K.J. Hemker, B.G. Mendis, C. Eberl, Mater. Sci. Eng. A 483-484 (2008) 727-730.

[12] C. Eberl, D. S. Gianola, K.J. Hemker, Exp. Mech. 50 (2010) 85–97.

[13] D.S. Gianola, S. Van Petegem, M. Legros, S. Brandstetter, H. Van Swygenhoven K.J. Hemker, 54 (2006) 2253-2263.

[14] M. Funk , K. Ma , J.M. Schoenung , M. Goeken , K.J. Hemker, Metall. Mater. Trans. A, submitted

[15] B.S. Sidhu, D. Puri, S. Prakash, Mater. Sci. Eng. A 368 (2004) 149-158.

[16] B.G. Mendis, B. Tryon, T.M. Pollock, K.J. Hemker, Surf. Coat. Technol. 201 (2006) 3918-3925.

[17] W. Li, L. Gao, Biomater. 24 (2003) 937-940.

[18] R. Licheri, R. Orrù, C. Musa, G. Cao, Mater. Lett. 62 (2008) 432-435.

[19] M. Cabibbo, C. Paternoster, R. Cecchini, A. Fabrizi, A. Molinari, S. Libardi, Mater. Sci. Eng. A 496 (2008) 121-132.

15    

[20] L. Ajdelsztajn, D. Hulbert, A. Mukherjee, J.M. Schoenung, Surf. Coat. Technol. 201 (2007) 9462-9467.

[21] D. Monceau, D. Oquab, C. Estournes, M. Boidot, S. Selezneff, Y. Thebault, Y. Cadoret, Surf. Coat. Technol. 204 (2009) 771-778.

[22] D. Oquab, C. Estournes, D. Monceau, Adv. Eng. Mater. 5 (2007) 413-417.

[23] D. Oquab, D. Monceau, Y. Thebault C. Estournes, Mater. Sci. Forum 595-598 (2008) 143-150.

[24] U. Martin, M. Jerenz, U. Mühle, H. Oettel, Mater. Sci. Eng. A 319-321 (2001) 388-392.

[25] N. P. Slavina, Meas. Tech. 5 (1962) 830-832.

[26] H. Conrad, J. Narayan, Scr. Mater. 42 (2000) 1025-1030

[27] F. Sun, S. Kim, C. Cao, Y. Lee, M. Yan, Mater. Sci. Eng. A 329-331 (2002) 661-667.

[28] N. Schonberg, Acta Chemica Scandinavica, 8 (1954) 221.

[29] P.V. Kiryukhantsev-Korneev, J.F. Pierson, J.P Baur, E.A. Levashov, D.V. Shtansky, 3rd France-Russia Seminar, (2007) 11-14.

16    

Table 1. Chemical composition of feedstock powder

Ni-343 alloy

Element Composition (wt %)

Ni 68.483

Cr 21.5

Al 8.91

Y 1.01

Fe 0.049

O 0.029

C 0.014

N 0.005

17    

Fig. 1. Schematic of Spark Plasma Sintering (SPS) set-up; not drawn to scale.

Fig. 2. Spark Plasma Sintering (SPS) process data changes in temperature, displacement,

and current.

Fig. 3. EDAX mapping analysis of polished cross sections of bulk SPS-sintered NiCrAlY

samples heat treated at: (a) 1000 °C/ 10 h; (b) 1100 °C/ 50 h.

Fig. 4. SEM images of the polished cross sections of bulk SPS-sintered NiCrAlY samples:

(a) as-sintered; and heat treated at: (b) 1000 °C/ 10 h; (c) 1000 °C/ 50 h; (d) 1050 °C/ 10

h; (e) 1050 °C/ 50 h; (f) 1100 °C/ 10 h.

Fig. 5. HR-TEM image of NiCrAlY bulk samples heat treated at 1050 °C/ 10 h, with

higher magnification image and selected area electron diffraction image of the Cr-O

phase, hereby identified as Cr3O.

Fig. 6. Characterization of grain growth in the β- and Cr-O phases after heat treatment.

Fig. 7. The relative change in elemental composition within the β-phase (Spot 1) and the

γ-phase (Spot 2) within bulk, heat-treated SPS-sintered NiCrAlY samples. Inset

micrographs are from samples heat treated at 1100 °C/ 50 h.

Fig. 8. (a) Microhardness values of bulk SPS-sintered NiCrAlY samples heat treated at

various conditions, where the average microhardness value is determined with a load of F

= 9.8 N and the localized phase-specific values are determined with a load of F = 0.0098

N; representative images of indentations used to determine: (b) average microhardness

values; (c) localized phase-specific microhardness value for β-phase; and (d) localized,

phase-specific microhardness value for γ-phase.

1

2

3

(b)

2 µm

β  

γ  

β  

γ  

γ  

NiL Crk

YL Ok Alk

NiL Crk

YL Ok Alk

(a)

2 µm

β  Cr-­‐O  

Ni-­‐Y  

γ  

4

(a)

Cr-O

Ni-Y

2 µm

β  

γ  

Cr-O

(b)

2 µm

  γ   γ  

γ  

β  Ni-Y  

(c)

2 µm

β  

β  

β  

γ  

γ  

γ  

Cr-O  

Ni-Y  

Cr-O

γ  

β  

γ  

γ  

Ni-­‐Y  (d)

2 µm Ni-­‐Y  

Cr-O

Ni-­‐Y  

Ni-­‐Y  

β  

β  

(e)

2 µm

γ  

β  

γ  

γ  

Ni-­‐Y  

β  

Cr-O  

2 µm

(f)

Ni-­‐Y  

Ni-­‐Y  

β  

β  

β  

β  β  

γ  

γ  

γ  

γ  

5

Cr-O  

Cr-O  

Cr-O  

β  

β  

β  

1 1 0  2 1 0  

2 0 0  

120 nm (Beam diameter)  

6

7

2  

5 µm

1  5 µm

γ   γ  

β  β  

8

a

d

2 µm

F  =  0.0098  N  

D1  =  6.56  μm  

HV  =  428.1  MPa  

Heat  treated  at  1100°C  /10  h  

b As-­‐sintered  sample  

F  =  9.8  N  

D1  =  62.9  μm  

HV  =  468.1  MPa  20 µm

c

2 µm

F  =  0.0098  N    

D1  =7.20  μm  

HV  =  358.3  MPa  

Heat  treated  at  1100°C/10  h  

γ  

γ  

γ  γ  

γ  

β  

β  

β  

β