Interface-Related Damage Evolution in Air-Plasma-Sprayed Thermal Barrier Coatingsfuller/PRESENTATIONS/ASME... · 2004. 5. 4. · expanded to service market in 1999 first stage of

Edwin R. Fuller, Jr.,Mark R. Locatelli, and Ravi Kacker

National Institute of Standards and TechnologyGaithersburg , MD 20899-8522, U.S.A.

<[email protected]>

Interface-Related Damage Evolution in Air-Plasma-Sprayed

Thermal Barrier Coatings

Symposium on Durability and Damage Toleranceof Heterogeneous Material Systems

2003 ASME Int‘l Mechanical Engineering CongressWashington, DC - November 17, 2003

GE’s 9H Gas TurbineGE’s 9H Gas Turbine

Combined-Cycle PerformanceNet Output: 480 MW CompressorNet Efficiency: 60% Pressure Ratio: 23:1Firing Temperature: 2600ºF / 1430ºC Air Flow: 1510 lbs/sec

Materials Science & Engineering Laboratory


•In production since 1997 on 7FA classcoated multiple parts on H class in 1998expanded to service market in 1999

first stage of industrial gas turbine

•Air Plasma-Spray (APS) process used for ZrO2 top coat• bulk temperature reduction ( > 75°C)

significantly increases creep life

•Vacuum Plasma-Spray (VPS) orHigh Velocity Oxy-Fuel (HVOF) process used for MCrAlY bond coat

protection of substrate alloy from oxidation and hot corrosion

Thermal Barrier Coatings (TBC’s)on Gas Turbine Buckets


Two-layer structure:ceramic top coat: (ZrO2 + Y2O3) thermal barriermetallic bond coat: MCrAlY oxidation protection

Types of Thermal Barrier Coatingsand Deposition Processes

Two deposition processes:air plasma spray (APS) & physical vapor deposition (PVD)

EB-PVD TBC’sAdvanced APS TBCConventional APS TBC

Air Plasma Sprayed TBC’s


Industrial Needsfor TBC Life Modeling

• Translate laboratory results to enginelife models and life prediction

• Develop microstructural failure models to guide development of improvedmaterials and processing techniques

“… microstructurally-based models are needed,”

“…can be qualitative or quantitative”


Micromechanical Damage ModelFreborg et al. proposed a failure scenario based

on the stress reversal above asperities with TGO growth.

G. C. Chang, W. Phucharoen & R. A. Miller, "Behavior of thermal barrier coatings For advanced gas-turbine blades," Surface & Coatings Technology, 30 [1]: 13-28 (1987).A. M. Freborg, B. L. Ferguson, W. J. Brindley, G. J. Petrus, "Modeling oxidation induced stresses in thermal barrier coatings," Mat Sci Eng A-Struct 245 [2]: 182-190 (1998).C.-H. Hsueh & E. R. Fuller, Jr., "Residual stresses in thermal barrier coatings: effects of interface asperity curvature/height and oxide thickness," Mat. Sci. Eng. A-Struct 283 [1-2]: 46-55 (2000).J. Rösler, M. Bäker & M. Volgmann, "Stress state and failure mechanisms of thermal barrier coatings: role of creep in thermally grown oxide," Acta Mater., 49 [18]: 3659-3670 (2001). K. Sfar, J. Aktaa & D. Munz, “Numerical investigation of residual stress fields and crack behavior in TBC systems,” Mat. Sci. Eng. A-Struct 333 [1-2]: 351-360 (2002).


Spallation Failure Mechanismfor Thermal Barrier Coating

Metal

Ceramic TBC: Stress Free at Temperature

cooling

Tension

Compression

Metal

Substrate: Mechanically Loaded

High-Temperature OperationStresses in TBC-Substrate System:

Result from mechanical loading, thermal expansion mismatch, & TGO growth

Stress relaxation time in TBC is short compared to engine operation time, but long compared to engine cool-down time

TGO

TGO

heating

Low-Temperature Shutdown


Spallation Failure Mechanismfor Thermal Barrier Coating

Ceramic

Tension

Compressionlarge inter-facial flaw

TBC: Stress Free at Temperature

Substrate: Mechanically Loaded

cooling

Coating Failure: when pieces of the top coat spall

Damage accumulates near metal-ceramic interface due to mechanical, thermal expansion, & TGO growth stresses

When damage produces a critical-size crack, the top coat locally buckles and spalls, due to large in-plane stresses

TGO

Metal

Metal

TGO

High-Temperature Operation

heating

Low-Temperature Shutdown

Air Plasma Spray TBC on HVOF CoNiCrAlY100 cycles

50 µm

740 cycles

50 µm

0 cycles

50 µm

350 cycles

50 µmCourtesy of Jim Ruud, GE CR&D


Residual StressesAbove Asperities on Cooling

Tensile Normal Residual Stress

top coat

bond coat

NiCrAlY bond coat

air-plasma-sprayed 8 wt% Y2O3 partially

stabilized ZrO2 topcoat

René N5 substrate (not shown)(Compressive In-Plane

Residual Stress)


Cooling fromstrain-freetemperature

Residual Stresses from Thermal Misfit Strains

YSZ

Bond Coat

< CTE =10.0 ppm/K >

< CTE =15.2 ppm/K >Bond Coat

YSZ

tens

ion

com

pres

sion

Compressive NormalResidual Stress

(Compressive In-Plane Residual Stress)


Stress Reversal with TGO Growth

top coat

bondcoat

TGO

NiCrAlY bond coat

air-plasma-sprayed 8 wt% Y2O3 partially

stabilized ZrO2 topcoat

René N5 substrate (not shown)

α-Al2O3 thermally grown oxide scale


Residual Stresses from Thermal Misfit Strains

Alumina< CTE =8.0 ppm/K >

YSZ< CTE =10.0 ppm/K >

Alumina

YSZ

com

pres

sion

tens

ion

Cooling fromstrain-freetemperature


This Process Can Be Modeled With Three Concentric

Spherical Shells

C.-H. Hsueh & E. R. Fuller, Jr.,Scripta Mater., 42 (2000) 781.

YSZ CTE = 10.0

CTE = 8.0Bond Coat

CTE = 15.2

TGO


Effective CTE of the Inner Two ShellsEf

fect

ive

CTE

(10-6

/K)

Normalized TGO Thickness, (tTGO / RBC)

CTE of YSZ

Effective CTE ofBond Coat and TGO

CTE of Bond Coat


Residual Radial Interfacial Stress

Res

idua

l Rad

ial I

nter

face

Stre

ss (M

Pa)

Normalized TGO Thickness, (tTGO / RBC)

RYSZ / RBC = 20

3 spheres

Bond Coat - TGO interface

TGO - YSZ interface

RYSZ / RBC = 5RYSZ / RBC = 20

RYSZ / RBC = 5


Failure Scenario based onStress Reversal above Asperities

with TGO Growth

Tensile Residual Stresstop coat

bond coat

Compressive Residual Stresstop coat

bondcoat

TGO

A. M. Freborg, B. L. Ferguson, W. J. Brindley, G. J. Petrus, "Modeling oxidation induced stresses in thermal barrier coatings," Mat Sci Eng A-Struct 245 [2]: 182-190 (1998).


Fracture Mechanics ModelDetermine crack stability from an appropriate set of fracture mechanics expressions:

∫=a

TGOI dxtyxK0

),;,( etc.geometry,asperity σ)axG(

σ dx

σ dxxasperity tension

tTGO < tcritical

y

( )∫−

=1

021

12 ζζσζπ

π daaKIaGriffith crack:ax

=ζ


Crack Stability for Tensile Stresses

σ dx/w

KIC

KI(a/w)

x/w and a/w

σ(x/w)

x/wa/wa/w


Fracture Mechanics ModelDetermine crack stability from an appropriate set of fracture mechanics expressions:

∫=a

cTGOI dxtyxK ),;,( etc.geometry,asperity σ)

ac,

axG(

Fazil Erdogan, “On the stress distribution in plates with collinear cuts under arbitrary loads,”in Proceedings of Fourth U.S. National Congress of Applied Mechanics, pp. 547-553 (1962).

y

crack closure due tocompressive stresses

σ dx

σ dxxasperity compression

tTGO > tcritical

ac

( )∫=1

),(2

ac

daGcK ac

cIa ζζσζπ

π( )∫=1

),(2

ac

daGcK ac

cIc ζζσζπ

π

ax

=ζ

Crack Stabilityfor Compressive Stresses

σ dx/w

KIC

KI(a/w)

x/w, c/w, and a/w

σ(x/w)

KIclosure(c/w, a/w)

crack closure

x/wc/wa/w

c/w

a/w


Crack Stabilityfor Compressive Stresses

σ dx/w

KI(a/w)

x/w, c/w, and a/w

σ(x/w)

crack closure x/wa/w

KI(a/w)neglecting closure



Three-Parameter Roughness Model2w

tan(Ψ)

h2H

RvRp

Ψ

1

wΗ

(Rp / Rv)

after D. R. Clarke and W. Pompe, “Critical Radius for Interface Separation of a Compressively Stressed Film from a Rough Surface,” Acta mater., 47 [6], 1749-1756 (1999).


Three-Parameter Roughness Model

(Rp+ Rv) = w / sin(Ψ)h = w cot(Ψ)

peak-to-valley amplitude2H = w tan(Ψ/2)

h2H

RvRpΨ

tan(Ψ)1

valley-to-peak curvature ratio: (Rp/Rv)

wΗ

(Rp / Rv)

2wwavelength


Factorial Experimental Design

twelve (12) simulated interface microstructures

wavelength, w

amplitude, H

curvatureratio: Rp/Rv

60 µm

90 µm

45 µm

20 µm

0.5 1.0 2.0


Only the roughness parameters H/w & Rp/Rv have a significant effect on the stress distribution

Residual Stress as a Functionof Microstructure: R p/R v

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0 0.2 0.4 0.6 0.8 1

Ratio = 0.5 Ratio = 2.0

Mod

e I S

tres

s (G

Pa)

x/w

Ratio = 2.0

RpRv

Ratio = 0.5

2w

2H

σyy


Residual Stress versus Oxide Thickness

-0.1

-0.05

0

0.05

0.1

0.15

-100 -50 0 50 100

No Oxide 2 µm Oxide 8 µm Oxide 12 µm Oxide

Position (µm)

t = 00 µm t = 04 µm t = 08 µm t = 12 µm


Fracture Mechanics Results

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0 0.2 0.4 0.6 0.8 1

90-45-2 (w-H-Ratio)

t = 0t = 6t = 12 t = 18

Mod

e I S

tres

s (G

Pa)

x/w

0

0.2

0.4

0.6

0.8

1

0 0.2 0.4 0.6 0.8 1

90-45-2

KI

(MPa

•m1/

2 )

a/w

t = 0t = 6t = 12 t = 18


Influences of YSZ Microcrack Sintering

0

0.2

0.4

0.6

0.8

1

0 0.2 0.4 0.6 0.8 1

90-45-1

0 oxide 6 oxide 12 oxide 18 oxide

KI (M

Pa•m

1/2 )

a/W 0

0.2

0.4

0.6

0.8

1

0 0.2 0.4 0.6 0.8 1

Stiffened YSZ90-45-1

Stiff 0 oxide Stiff 6 oxide Stiff 12 oxide Stiff 18 oxide

KI (M

Pa•m

1/2 )

a/W

Modeling Real Microstructures

normal stress: σ yy

-1.5 GPa +1.5 GPa0


Interface-RelatedDamage Evolution in APS TBC’s

SUMMARY:A damage evolution mechanism proposed by Freborg et al. was quantitatively analyzed with a fracture mechanics weight function approach Reversal in residual stress distribution above interface asperities drives damage evolutionMicrostructural variables studied with a three parameter roughness model:

Wavelength AmplitudePeak Sharpness TGO Thickness


SUMMARY:

Interface-RelatedDamage Evolution in APS TBC’s

Crack closure enhances the crack-tip stress intensity factorHowever, calculated KI-fields are still below the expected threshold for crack growthTop coat sintering enhances damage evolutionSeveral other factors are under investigation, e.g., top coat sintering, TGO growth strain, bond coat creep, crack path


Abstract

INTERFACE-RELATED DAMAGE EVOLUTIONIN AIR-PLASMA-SPRAYED THERMAL BARRIER COATINGS

Edwin R. Fuller, Jr.,* Mark R. Locatelli, and Ravi KackerNational Institute of Standards and TechnologyGaithersburg, Maryland 20899-8520, U.S.A.

<[email protected]>

Spallation of air-plasma-sprayed (APS) thermal barrier coatings (TBC’s) typically stems from the damage that accumulates near the metal-ceramic and ceramic-ceramic interfaces in these coatings. Damage evolution is driven by stresses perpendicular to the interface that result from rough interfaces in combination with thermal-expansion-anisotropy and oxide-growth strains. These stresses are incorporated into a fracture-mechanics weight-function formalism to quantify the driving forces for crack growth near the interface. Residual stresses as a function of interfacial structure are derived for both periodic and random structures, and are used to derive crack-driving, stress-intensity-factor fields as a function of the interfacial, thermally grown oxide (TGO) thickness, and other microstructural parameters. Residual stresses and associated stress intensity factors are presented for both model and real interfaces, attempting to identify critical microstructural features for predicting damage evolution, and hence, reliability of TBC's.


Interface-Related Damage Evolution in Air-Plasma-Sprayed Thermal Barrier Coatingsfuller/PRESENTATIONS/ASME... · 2004. 5. 4. · expanded to service market in 1999 first stage of

Documents