Available online at www.sciencedirect.com

ICM11

Interfacial Strength Evaluation Technique for Thermal BarrierCoated Components by Using Indentation Method

Yasuhiro Yamazakia*, Shin-ichiro Kugaa and Murugesan Jayaprakashb

a Niigata Institute of Technology, Fujihasi 1719, Kashiwazaki 945-1195, Japanb Japan Nuclear Energy Safty Organization, Fujihashi 1719, Kashiwazaki 945-1195, Japan

Abstract

Thermal barrier coatings (TBCs) have been essential technologies to improve the performance and efficiency of advanced gasturbines, which are in service at extremely high temperatures. Interfacial strength is one of the important properties in TBCs, andit has been often evaluated by tensile method, e.g. the ASTM standard; ASTM C633. However, the applicability of tensilemethod is limited, due to its size dependence. To overcome this problem, a new method has been proposed in the previous studyto evaluate the interfacial strength using indentation method. In the present study, the interfacial strength of TBC which wassubjected to thermal cycle fatigue has been investigated using the newly proposed indentation method. And also, the applicabilityof the proposed indentation method to evaluate the interfacial strength of retired TBC vane has been investigated. The resultsshowed that; (1) Interfacial strength of the ceramic top-coat increased with the thermal cycles due to the sintering of ceramic top-coat. (2) TGO layer becomes thicker and micro cracks are initiated at the interface between the TGO layer and the bond-coatafter thermal cycle fatigue. (3) The residual stresses in the ceramic top-coat and the TGO layer are varied with the thermal cyclefatigue. (4) Interfacial strength of a retired TBC vane was almost comparable to that of the as-sprayed TBC specimen because thethermal damage of the retired TBC vane was not severe.

Key wards: Thermal barrier coating, Interfacial strength, Indentation method, Thermal fatigue damage, TBC compornent

1. Introduction

A typical TBC system for gas turbines consists of an oxidation-resistant metallic bond coat on a super-alloysubstrate, and a thermal insulating ceramic top-coat attached to the bond-coat. The most critical issue in reducing thedurability of TBCs is spallation of ceramics top-coat. Once this type of damage has been realized, hot sectioncomponents made of super-alloy substrate would be overheated, resulting in tragic accidents. Interfacial strength (i.e.adhesion strength) is a measure of resistance to spalling of ceramics top-coat and it has been often evaluated bytensile method, e.g. the ASTM standard; ASTM C633.However, the applicability of tensile method is limited, due to its size dependence. Especially, ASTM C633 can't

be applicable to evaluate the remaining adhesion strength of actual coated components, because of the requirement

* Corresponding author. Tel.: +81-257-22-8109; fax: +81-257-22-8109.E-mail address: axel@mce.niit.ac.jp.

doi:10.1016/j.proeng.2011.04.139

Procedia Engineering 10 (2011) 845–850

1877-7058 © 2011 Published by Elsevier Ltd. Selection and peer-review under responsibility of ICM11

Open access under CC BY-NC-ND license.

© 2011 Published by Elsevier Ltd. Selection and peer-review under responsibility of ICM11

Open access under CC BY-NC-ND license.

846 Yasuhiro Yamazaki et al. / Procedia Engineering 10 (2011) 845–850

of relatively large specimen. And also, the interfacial fracture toughness can't be evaluated by this method. Toovercome these problems, indentation tests have been used to evaluate adhesion strength of the coatings [1-7].However, for thick coatings it is not possible to use the indentation at the surface of the coating, as there is noenough plastic deformation and also the indent cannot penetrate into the substrate. These drawbacks have beenovercome by the usage of Interface indentation test [1-5]. In this method, Vickers indentation was performed overthe cross section of the coating sample. By the virtue of its simplicity and practicality this method has been selectedfor the evaluation of the interfacial strength of coatings in the present study. In the present study the interfacialstrength of an air plasma sprayed thermal barrier coatings which was subjected to thermal cycle fatigue has beeninvestigated using a newly proposed indentation method in the previous study with an instrumented indentationmachine. And then the applicability of the newly proposed indentation method to evaluate the interfacial strength ofa retired TBC vane has also been investigated.

2. Experimental procedure

The coating used in this study is air plasma sprayed thermal barrier coating. The substrate material used was Ni-base super-alloy (MarM247), with a thickness of 5mm. The powder used for the top-coating was yttria partiallystabilized zirconia ( METCO 204NS) and the powder used for bond-coating was CoNiCrAlY alloy (AMDRY9951).Thickness of the top-coat and the bond-coat were approximately 500 m and 100 m respectively. Bar specimenswith dimensions of 30 mm length, 10 mm width and 5.6 mm thickness were cut from the coated material andsubjected to thermal cycle fatigue. The thermal cycle fatigue test between 400 and 1000 C was carried out usingMuffle furnace. The thermal cycle consists of heating the specimen to 1000 C at a rate of 200 C/h, followed by adwell time of 1h at the maximum temperature (1000 C) and subsequent cooling to minimum temperature (400 C) ata rate of 200 C/h. After thermal fatigue tests, the TBC specimens were cut in the middle of width and mounted onthe test jig. The specimen surface (i.e. cross section of coating material) was polished to mirror surface after anepoxy was filled to the gap between the coating specimens.The indentation test was carried out on the polished surface of the specimen using an instrumented indentation

machine. The schematic of the instrumented indentation machine is shown in Fig. 1. An indent was directly made atthe interface on the polished surface of the specimen. Indentation load was applied by an electrical actuatorcontrolled by a personal computer. The indentation speed was 1.5 m/s and the indentation load was held at themaximum value for 10s. During the indentation test, the indentation load and the indentation depth were measuredby load cell and gap sensors, respectively. After indentation, the crack length and the diagonal length of theindentation were measured using SEM. The detailed procedures were mentioned in Ref. 5 and 6.

Fig. 1 Schematic illustration of the instrumented indentation test equipment used in this study and typical load - displacement curves during theindentation test for a cold-rolled steel.

The residual stresses in the ceramic top-coat and the thermal grown oxide (TGO) were also measured in thepresent study. The residual stress in the ceramic top-coat was measured by elution method. That is, the residualstress was evaluated from the changes in the deflection and the surface strain of the ceramic top-coat before and

Yasuhiro Yamazaki et al. / Procedia Engineering 10 (2011) 845–850 847

after elution. On the other hand, the local residual stress in the TGO layer was measured by the shift in the photo-stimulated luminescence spectra from Cr3+ under a green laser (wave length of 532nm) radiation. The measurementwas carried out at room temperature in ambient air using a specially designed micro-Raman spectrophotometer. Theresidual stress measurement was performed on cross-section of the specimen. The detailed procedures werementioned in Ref. 8 and 9.

3. Results and discussions

3.1. Microstructure and residual stress after thermal cycle fatigue

Typical microstructures near the interface before and after thermal cycle fatigue are shown in Fig. 2. Twodifferent type thermal grown oxides (TGOs), a continuous oxide layer and a wart-like oxide, were formed afterthermal cycle fatigue (see Fig.2 (d), for example). It was reveled from the previous investigations [10] that, thecontinuous layer is a pure alumina layer and the wart-like oxide is a complex oxide consisted of Ni, Cr and Co.Thicknesses of the two types of TGOs were measured from 10-20 SEM images by the image processing

technique. The thicknesses of TGOs increased with an increase in number of thermal cycles. No micro-crack wasobserved up to 300 thermal cycles. However, the micro-cracks were initiated in the TGO layer after 1000 thermalcycles (Fig. 2(e)). And, the delamination was occurred after 1540 thermal cycles (Fig. 2(f)).The relationship between residual stress generated in the ceramic top-coat (in-plane mean residual stress) and

thermal cycles are shown in Fig. 3(a). As seen from the figure, the residual stress in the ceramic top-coat was almostzero at 0 thermal cycles (As-spray). At the initial stage of thermal cycle, the residual stress in the ceramic top-coatincreased in compression with thermal cycles, and after reaching 100 thermal cycles, the residual stress becomesalmost constant (100 MPa in compression), i.e., after 100 thermal cycles, the residual stress in the ceramic top-coatdid not change with a further increase in thermal cycle. The relationship between residual stress in the TGO(compressive residual stress) and thermal cycles are shown in Fig.3 (b). As seen from the figure, the compressiveresidual stress in the TGO decreased with an increase in the thermal cycles. It has been reported in Ref. 9 that, thecompressive residual stress in the TGO increased after thermal exposure. The decrease of compressive residualstress in the TGO after thermal cycle fatigue might be due to the micro-cracks initiated at the interface between theTGO and the bond-coat as shown in Fig. 2(e).

(a) at 0 cycle (as-sprayed) (b) after 100 cycles (c) after 300 cycles

(d) after 1000 cycles (e) after 1000 cycles, microcrack (f) after 1540 cycles

Fig. 2 Typical cross-sectional microstructure near the interface after thermal cycle fatigue .

TC

BC

Sub.

TC

BC

Sub.

TC

BC

Sub.

TC

BC

Sub.

TC

BC

TC

BC

Continuous pure alumina layer

Wart-like complex oxides

Micro-crack

848 Yasuhiro Yamazaki et al. / Procedia Engineering 10 (2011) 845–850

0 100 200 300 400 500-200

-100

0

100

Number of thermal cycles [cycle]In-planemeanresidualstressinTC

res[MPa]

0 500 10000

500

1000

1500

Number of thermal cycles [cycle]

Residualcompressivestress

inTGO[MPa]

Fig. 3 (a) In-plane mean residual stress in the ceramic top-coat and (b) local residual compressive stress in the TGO layer as a function of thenumber of thermal cycles.

3.2. Apparent interfacial toughness of TBC after thermal cycle fatigue

Typical interfacial crack initiated by the indentation test is shown in Fig. 4 (a). It can be seen from Fig. 4(a) thecrack mainly propagated in the ceramic top coat near the interface. The propagated crack after 500 thermal cycles isshown in Fig. 4(b). It can be observed from the figure, even after the formation and growth of TGO layer also, thecrack propagates only in the ceramic top coat. The cracks propagated parallel to the interface from the corner of theindentation mark. The apparent interfacial fracture toughness for coatings can be obtained from the followingequation [4, 5].

Ic PEcaacK 2/321 /)}/({ (1)

where, EI is the apparent Young’s modulus and a is the half diagonal length of the impression. The apparentYoung’s modulus of interface, EI, is obtained from the indentation load-depth curve according to the ISO standard;ISO 14577. The 1 and 2 are material constants obtained from the comparison between the indentation tests resultsand other test results, e.g. the four-points bending test. The 1 and 2 for TBC are -0.00989 and 0.00987, respectively.The apparent interfacial fracture toughness of TBC specimens, Kc, which was evaluated by the indentation

method, is shown in Fig. 5 as a function of number of thermal cycles. It can be seen from Fig. 5, the interfacialfracture toughness (Kc) increased after thermal cycle fatigue. For comparison, the interfacial fracture toughnessobtained by the modified four-points bending test [10, 11] are also shown in Fig. 5 by using the relationshipGc=K2c/E. As seen from Fig. 5, at a given thermal cycle, the apparent interfacial fracture toughness evaluated by theindentation method was almost comparable to that obtained by the modified four-points bending test.

Fig. 4 Typical interfacial cracks initiated by the indentation test; (a) for 0 thermal cycles (As-sprayed) , (b) after 500 cycles

(a)(b)

(a) (b)

Yasuhiro Yamazaki et al. / Procedia Engineering 10 (2011) 845–850 849

0 100 200 300 400 5000

1

2

3

Number of thermal cycles

Apparentinterfacialtoughness,

Kc[MPa

m]

: by the indentation test: by the mod. 4-point bending test

Fig. 5 The apparent interfacial fracture toughness plotted as a function of the thermal cycles.

It has been reported that, when TBCs are subjected to thermal aging, sintering would occur, resulted in increaseof fracture toughness (KIc) after thermal aging [12, 13]. In the present study, during thermal cycle the sintering of theceramic top-coat might be occurred, resulted in increase of interfacial fracture toughness (Kc) of ceramic top-coatwith thermal cycle loading. However, many micro-cracks were initiated in the TGO after 1000 cycles of thermalcycle fatigue as shown in Fig.2. If the micro-cracks were initiated in TGO, the Kc measured by the indentation testmay be decreased.

3.3. Application to a retired TBC vane

The instrumented indentation method was applied to a retired TBC vane to evaluate interfacial fracture toughness.Small samples were cut from the pressure side and the suction side near the reading edge of the TBC vane as shownin Fig. 6(a). In the samples, there was no interfacial and segmentation crack, except few small cracks that initiatedfrom cooling hole. And, the thickness of the TGO was less than 0.5 m, which indicates that, the operating conditionof these parts were not severe, i.e., the maximum operating temperature was relatively low and/or the exposure timeat the high temperature was short. It can be clearly understood from these observations that, these parts weresubjected to very little thermal damage. Figure 6(b) shows the apparent interfacial fracture toughness of a retiredTBC vane which was evaluated using the proposed method. For comparison the Kc of as-sprayed TBC specimen isalso shown in Fig. 6(b). It can be inferred from Fig. 6(b) that, using the proposed indentation method, the interfacialstrength of the actual TBC vane can be evaluated with small scatter. Furthermore, from Fig. 6(b), it can be observedthat the interfacial strengths of pressure side and suctions side of the retired TBC vane were almost comparable tothat of the as-sprayed TBC specimen.

0

1

2

3

Interfacialfracturetoughness,

Kc[MPa

m]

Suctionside

Pressureside Test Piece

(As-sprayed)Retired TBC Vane

54 16

Number: The number of indentation tests

Fig. 6 (a) The schematic illustration of the evaluated position in a retired TBC vane, (b) the apparent interfacial fracture toughness of a retiredTBC vane compared with the as-sprayed TBC specimen.

Sample of suction side

Sample of pressure side

(a)

(b)

850 Yasuhiro Yamazaki et al. / Procedia Engineering 10 (2011) 845–850

4. Conclusions

In the present study, the thermal fatigue behavior of an air plasma sprayed thermal barrier coating wasinvestigated. In addition to that, the residual interfacial strengths of thermal cycled TBC specimens as well as aretired TBC vane were also evaluated by means of an instrumented indentation machine. The main conclusions aresummarized as follows.(1) The apparent interfacial fracture toughness of TBC specimens increased after the thermal cycle fatigue due to

the sintering of ceramic top coating. (2) The thermal grown oxide (TGO) layer becomes thicker and the microcracks were initiated at the interface between the TGO layer and the bond-coat after thermal cycle fatigue. (3) Theresidual stresses in the ceramic top-coat and the TGO layer are varied with the thermal cycle fatigue. The in-planeresidual stress in the ceramic top-coat becomes compressive and it hardly changes with the thermal cycle. On theother hand, the compressive residual stress in the TGO decreased with the thermal fatigue cycles due to the initiationof micro crack at the interface between the TGO and the bond-coat. (4) The apparent interfacial strength of a retiredTBC vane can be evaluated by the instrumented indentation method. The interfacial strength of a retired TBC vanewas almost comparable to that of the as-sprayed TBC specimen, as the thermal damage of the retired TBC vanewasn't severe.

Acknowledgements

The authors would like to thank Research Foundation for the Electrotechnology of Chubu for their financialsupport of these endeavors. Also the authors would like to thank Prof. S. Zhu and Mr. T. Osaki of Fukuoka Instituteof Technology for their help in residual stress measurement.

References

[1] Maschio, R.D., Sglavo, V.M., Mattivi, L., Bertamini, L and Sturlese, S.: Indentation Method for Fracture Resistance Determination ofMetal/Ceramic Interface in Thick TBCs, Journal of Thermal Spray Technology 1994, 3, 51-56.

[2] Chicot, D., Démarécaux, P. and Lesage, J.: Apparent Interface Toughness of Substrate and Coating Couples from Indentation Tests, ThinSolid Films 1996, 283, 151-157.

[3] Drory, M.D. and Hutchinson, J.W.: Measurement of the adhesion of a brittle film on a ductile substrate by indentation, Proceedings of RoyalSociety London A 1996, 452, 2319-2341.

[4] Yamazaki, Y and Kuga, S., An engineering study of interfacil strength evaluation for thermal spray coatingsby indentation method, Journal ofJapan Thermal Spraying Society 2010, 47, 106-111 (in Japanese).

[5] Yamazaki, Y and Kuga, S., Interfacial Fracture Toughness Evaluation of Thermal Barrier Coatings by Instrumented Indentation method,Proceedings of the 1st International Conference Determination of Mechanical Properties of Materials by Small Punch and other miniaturetesting techniques, Ostrava, Czech Republic, 2010, 194-200.

[6] Bartsch, M., et al.: Interfacial fracture toughness measurement of thick ceramic coatings by indentation, Key Engineering Materials 2005,290, 183-190.

[7] Vasinonta, A. and Beuth, J.L.: Measurment of interfacial toughness in thermal barrier coating systems by indentation, Engineering FractureMechanics 2001, 68, 843-860.

[8] Yamazaki, Y., Fukanuma, H. and Ohno, N., Effect of bond coat on the residual adhesion strength of a plasma sprayed TBC system afterthermal cycle fatigue, Journal of Solid Mechanics and Materials Engineering 2010, 4, 196-207.

[9] Zhu, S., Fukuda, K. and Osaki, T., Residual stress in TGO and Interfacial damage in thermal barrier coating after thermal exposure and cyclicindentation, Journal of Solid Mechanics and Materials Engineering 2010, 4, 244-251.

[10] Yamazaki, Y., Kinebuchi, T., Fukanuma H. and Ohno, N. , Influences of the Interface Roughness and the Bond Coat Spray Method on theAdhesion Strength in an Air Plasma Sprayed Thermal Barrier Coatings, Journal of the Society of Materials Science, Japan 2008, 57, 596-602 (in Japanese).

[11] Yamazaki, Y., Schmidt A. and Scholz, A., The determination of the delamination resistance in thermal barrier coating system by four-pointbending tests, Surface and Coatings Technology 2006, 201, 744-754.

[12] Yamazaki, Y., Kinebuchi, T., Fukanuma, H., Ohno, N. and Kaise, K., Deformation and fracture behavior in the free-standing APS-TBC,Key Engineering Materials 2007, 353-358, 1935-1938.

[13] Yamazaki, Y. Correlation between the mechanical properties and splat microstructures of an APSed TBC, Journal of Solid Mechanics andMaterials Engineering 2008, 1275-1286.