Advanced Thermal Barrier Coatings for Operation in High Hydrogen

DOE UTSR Program Highlights: Advanced TBCs for gas turbine engines

Advanced Thermal Barrier Coatings for Operation in High Hydrogen Content Gas Turbines

PI : Sanjay Sampath

Contributors: Prof. C. M. Weyant, Dr. G. Dwivedi, V. Viswanathan

Dwivedi et al., JACerS, DOI: 10.1111/jace.13021

Viswanathan et al., JACerS, DOI: 10.1111/jace.13033

Dwivedi et al., JTST, DOI: 10.1007/S11666-014-0196-9

Viswanathan et al., JACerS, Under Review

DOE NETL UTSR

Contract #DE-FE0004771, 2010-2014

Program Manager: Dr. Briggs while

Research Highlights

Dwivedi et al., Adv. Mater. Process, 2013, 8(2), 49

Publication list


Multifunctional requirements of TBCs requires Multilayered architecture

Erosion, FOD, CMAS/Ash

Low-K material, Porosity,

Lower sintering rate

Remains complaint

Compatibility with Bondcoat

Mostly traditional TBC, High

toughness

Adequate roughness,

oxidation resistant (dense),

environmental effects

2

Multilayer,

Multifunctional

Coating

Solution

Challenges associated with

new gen TBCs

Layer-by-layer

optimization of

multilayered

TBCs is naturally

suited for Thermal

Spray


Bond coat processing and performance 3

Process selection and mapping strategy for

the density and oxidation optimization Initial performance evaluation

with HVOF bond coats

Stony Brook BC Commercial BC

Rene’ 80 substrate

300µm APS 8YSZ

FCT: 1100oC, 24 hrs cycle

Prior work conducted at Stony Brook has explored the process-property

relationship of bond-coat material deposited using Atmospheric plasma

Spray and High Velocity Oxy-fuel (HVOF) processes.

The selection of HVOF process was to obtain dense bond-coat

microstructure with minimum oxide inclusion (WOKA and DJ2700 in the

above figure).

Such coatings offer requisite oxidation resistance to the substrate.

For material selection, concurrent study was conducted at ORNL (coating

prepared by Stony Brook), on two different bond-coat chemistries- SM

AMDRY 386-2 (NiCoCrAlYHfSi) and SM XPT 449 (NiCoCrAlY). Results

suggested superior performance of AMDRY 386-2.

Therefore, the bond-coat study at Stony Brook, focused on the same

(AMDRY 386-2) material. The coating was deposited using liquid fuel

HVOF process.

Stony Brook’s BC

Poor adhesion

Commercial BC

Good adhesion

Initial results exhibited

significantly poor life of

TBC sprayed on Stony

Brook’s bond-coat,

compared to those sprayed

on commercially obtained

bond-coat.

Initial observations on failed specimens indicated that the poor FCT life in

Stony Brook’s bond-coat specimen was primarily due to lower bond-coat

roughness.

Ra = 3-4 µm Ra = 9-10 µm

Photographs of failed specimens

20µm Substrate

AMDRY 386-4

(Coarse powder) Roughness

Ra = ~ 8 – 9 µm

AMDRY 386-2

(Fine powder)

Layer-1: Oxidation

resistant dense layer

Layer-2: Roughness improvement

Two-layer bond-coat deposition: Improved roughness

NiCoCrAlYHfSi


Bond coat processing and performance 4

NiCoCrAlYHfSi

Two-layer bond-coat deposition: Improved roughness

Microstructure

20µm Substrate

AMDRY 386-4

(Coarse powder) Roughness

Ra = ~ 8 – 9 µm

AMDRY 386-2

(Fine powder)

Layer-1: Oxidation

resistant dense layer

Layer-2: Roughness improvement

FCT life

0

200

400

600

800

1000

To

p c

oat lif

e (

hrs

)


300µm APS 8YSZ


Single layer

Stony Brook’ BC Two layer

Stony Brook’ BC

Commercial

BC

Ni based Superalloy Substrate

Tw

o laye

r

Bo

nd

coa

t

Rough, Oxidation protection

strength/creep resistant

Erosion and CMAS Resistant

Low thermal conductivity

Phase stability

Sin

gle

laye

r ~100%

improvement

in FCT life

with two-layer

bondcoat

Processing strategies leveraged for the development of improved HVOF

bond coats

Based on process diagnostic and coating property measurements a two

layered HVOF coating was designed, fabricated and tested.

First layer would consist of a dense oxidation resistant coating comprising

of majority of the thickness while the second layer would serve to enhance

the surface roughness of the HVOF coatings

A significant improvement in FCT life obtained with the two layered bond

coat architecture compared with initial single layered coating

Processing strategies were utilized to provide significant performance

benefits for a chosen feedstock chemistry

Summary


5 Performance assessment of various APS YSZ architectures

D

50 µm 50 µm

DVC

unmolten

particles

vertical

cracks

A B

50 µm 50 µm

C

50 µm

interlamellar

pores

globular

pores microcracks

FC

T L

ife

(H

ou

rs)

0

200

400

600

800

1000

FCT Hours

FCT Hours: 784.0000

A B C D

Typical APS coatings

DVC

To

pco

at L

ife

(H

ou

rs)


Two layer Stony Brook’s BC


Five top coat architectures with decreasing porosity from A to DVC, were sprayed on

identical bond-coated superalloy substrates, and were subjected to FCT.

The FCT results indicated higher top coat life for coatings with lower porosity .

The measured elastic modulus and thermal conductivity of coatings also increases with

decreasing porosity.

The fracture toughness (measurement details are in next page) of coatings appears to be the

operating mechanism with higher toughness coatings resulting in higher FCT durability

0

10

20

30

40

50

60

0

1

2

3

4

Ela

stic M

odu

lus (

GP

a)

Fra

ctu

re T

oug

hne

ss (

MP

a

m)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

Th

erm

al C

ond

uctivity (

W/m

K)

A B C D DVC*

FCT life of coatings is higher with higher fracture toughness coatings. However, these coatings also have higher thermal

conductivity, which is not desirable for TBCs.

The high fracture toughness requirement only seems to be at the bond-coat and top-coat interface. Can we design a

multilayer top-coat with low overall thermal conductivity and high toughness at the interface

100µm 100µm

As-deposited TBC Failed (~600 hrs)


Microstructural dependence of fracture toughness of TBCs 6

GDZ

as-sprayed

YSZ

as-sprayed

GDZ

sintered

0 10 20 30 40 50 60 70 80

Indentation modulus, Eind (GPa)

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

Fra

ctu

re to

ug

hn

ess, K

IC (

MP

a-√

m)

YSZ

sintered

o high sintering rate

o high intrinsic

toughness

o low sintering

rate

o low intrinsic

toughness

FC

coarse

med E

highE

lowE

fine

150mm

med E

highE lowE

fine

PrecrackLoad, P Specimen

t

W

Wm

Notch

1. Fracture toughness of coatings sensitive to their processing conditions

2. Coatings with higher degree of porosity have both lower modulus and fracture

toughness

3. YSZ coatings shows better fracture toughness than Gd2Zr2O7 (GDZ) coatings.

4. With sintering, the improvement in fracture toughness of YSZ coatings is larger

than that of GDZ coatings

5. The map allows us to down select the coating material and processing

conditions to meet fracture toughness requirement for coating design 12 mm

39- 40 mm

20

mm

Free standing

coating

Max. Load for

fracture

21

4

13

StSPK mICIC

PIC - Maximum load at failure

- Poisson’s ratio

S - specimen width

Sm – moment arm

t - specimen thickness

- thickness correction factor

= 1-1.26(t/S)+2.4(t/S)exp(-pS/2t)

Test method KIC extraction

Double Torsion Technique


7 Revised TBC Architecture: Strategic approach for multi-functionality Functionally Optimized TBC with high fracture toughness interface layer

TB

C lifetim

e (

hours

)

RT Thermal conductivity (W/m-K)

0.6 0.8 1.0 1.2 1.4 1.6 400

600

800

1000

1200

1400

Single

layer TBCs

bi-layer

TBCs with

improved

durability

Superalloy Substrate

Overlay BC enhanced

roughness

Layer-1

Layer-2

High KIC TBC Layer

Porous TBC Low K

Low E

o Structural

Compliance

o Low Th.

Conductivity

o High fracture

toughness

Design Strategy

YSZ

as-sprayed

0 10 20 30 40 50 60 70 80

Indentation modulus, Eind (GPa)

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

Fra

ctu

re to

ug

hn

ess,

KIC

(MP

a-√

m) YSZ

sintered

FC

coarse

med E

highE

lowE

150mm

High Plasma

Power

Fine

Particles

Sintered for 24

hours

Processing condition selection

50µm

Representative microstructure

To

pco

at L

ife

(h

ou

rs)

0

200

400

600

800

1000

1200

1400

Co

nve

ntio

na

l

Single layer TBCs Bi-layer TBCs

Porous

100%

Int.

: H

igh

pla

sm

a k

W

Int.

: F

ine

po

wd

er

Int.

: 2

4 h

rs s

inte

rin

g @

12

00

oC

High KIC YSZ Layer

Porous YSZ

Bond coat

FCT Life: 24 hrs cycles @ 1100oC Failed specimen analysis

50 µm

epoxy

Bond coat

Spalled TBC

TGO

Classic TBC

spallation.

Failure mode

remains same

as conventional.

Coating with high

durability and low k

Simultaneous

optimization of

coating durability and

functionality

100% improvement

in life while

maintaining low

Thermal Conductivity

Summary


8


Layer-1

Layer-2

High KIC TBC Layer

Erosion ,CMAS Resistant

Low K ,E

Overlay BC enhanced

roughness

Porous TBC Low K

Low E

GDZ- DVC

Bondcoat

GDZ- Porous

YSZ-Dense

GDZ- DVC

Bondcoat

YSZ-Dense

GDZ- DVC

Bondcoat

YSZ- Porous

YSZ-Dense

GDZ- DVC

Bondcoat

YSZ- DVC

B D C E GDZ- Dense

Bondcoat

GDZ- Porous

YSZ- DVC

A

YSZ-DVC

GDZ-DVC

E

YSZ-Dense

YSZ-Porous

GDZ-DVC

C

YSZ-Dense

GDZ-DVC

D

YSZ-Dense

GDZ-Porous

GDZ-DVC

B

YSZ-DVC

GDZ-Porous

GDZ-Dense

A

0

300

600

900

1200

1500

1800

Sin

gle

laye

r

TB

C L

ife

tim

e (

hou

rs)

YSZ

based

coatings

YSZ-GDZ

multilayer coatings

A D C E B Bi-

laye

r: H

igh

KIC

in

terf

.

+ 200 hrs

Arc

hit

ectu

ral

des

ign

S

pra

ye

d

Mic

ros

tru

ctu

re

Molten ash

wicking

DVC GDZ

Lignite Ash

DVC architecture of GDZ material was evaluated.

Lignite ash was spread onto the topcoat (35

mg/cm2). Heat treatment: 1200oC for 24 hours.

The coating exhibited resistance to CMAS, and

the average penetration depth was ~ 25 microns

GD

Z D

VC

GD

Z F

ine p

ow

der

0

2

4

6

8

GD

Z P

oro

us -

1

GD

Z P

oro

us -

2

Po

rou

s Y

SZ

Ero

sio

n r

ate

(m

g/g

), 5

0µ

m (

D50)

Al 2

O3

50µm

GDZ DVC

Based on the

erosion data. DVC

architecture appears

to be promising

candidate for top

layer.

Multilayer TBC architevure: Strategic approach for multi-functionality Functionally Optimized TBC with high durability as well as erosion and CMAS resistance

Coating ‘C’ shows high FCT life, even better than

the Bi-layer YSZ, simultaneously addressing the

issues of CMAS and erosion.

High erosion resistance layer CMAS resistant layer Multilayer-multifunctional

architecture


0 20 40 60 80 100 0.6

0.8

1.0

1.2

1.4

1.6

Elastic Modulus ( GPa )

B

C

D

E

Engineered

Multilayered

coatings

As-s

pra

yed T

h. C

ond. (W

/m-K

)

ML

ML DVCs SL DVCs

A

S1 S2

GDZ

YSZ

ML: Multi Layer

SL: Single Layer

P-G

P-Y

Front Temp: ~ 1300oC

Subst Temp:~ 1100oC

Two-layer

HVOF bond

coat

Porous YSZ /

multilayer

TBC

35 mg/cm2

Lignite Ash

Substrate

100 um Bondcoat

YSZ

Gd2Zr2O7

`

Substrate

YSZ

Gd2Zr2O7

DVC

Multilayered TBC (C) lasted ~ 20 times more than

conventional YSZ TBC.

Lignite ash was adhered to the coating by baking it onto the

topcoat at 1000oC. Cycles with10 minutes heating and 2 minutes

air cooling were performed

Preliminary Gradient Testing

Multilayer TBC architevure: Strategic approach for multi-functionality Functionally Optimized TBC with high durability and low thermal conductivity

Specimen preparation Gradient test rig

Failed coatings microstructures

The microsturctures from this preliminary test suggest that the top

GDZ-DVC layer can mitigate CMAS penetration to the GDZ-YSZ

coating architecture.

The as-sprayed thermal

conductivity and elastic

modulus map shows

that the new multilayer

GDZ-YSZ coating

architecture (C) has

thermal conductivity as

low as conventional

YSZ coating. However

the elastic modulus is

somewhat higher due to

the inclusion of DVC

structure in the new

TBC design.

Functional property map

Ni based Superalloy Substrate

Mu

ltila

ye

r T

op

co

at

Bo

nd

co

at

Oxidation protection

strength/creep resistant

High fracture toughness layer

Sinter Resistant

Low Thermal conductivity

Erosion and CMAS Resistant

Low thermal conductivity

Phase stability

- e.g. Gd2Zr2O7

Pyrochlore

- High Y2O3-ZrO2.

Dense/DVC archit.

-Low K material

-Phase stability

sinter resistant comp.

- t‘ composition

- low defect density coat

Rough layer

Low porosity

oxid. Resist.

CoN

iCrA

lY (

+R

e)

Dense Gd2Zr2O7

Porous

YSZ/GDZ

Dense YSZ

Two layer

MCrAlY

bondcoat

Summary

Considering all the processing and materials

related challenges

9


10

Y1

•YSZ and GDZ process property relationships • Process Map development

• Toughness, Lignite ash penetration depth, erosion

Y2 •Rough bond coat process optimization with 40% increase in FCT life • Two layer Bond coats

Y3 • Bi-layer YSZ coating with two fold increase in FCT life, and maintaining low K • High toughness interface layer, Elastic energy model

Y4 •Multilayer YSZ-GDZ coating system • Enhanced life, Lignite ash penetration minimization, erosion resistance

Systematic progress over past four years

Dr. Briggs While, Program Manager

Prof. Toshio Nakamura, Stony Brook University

Dr. Curtis Johnson, Rtd. GE GRC

Prof. John Hutchinson, Harvard university

Drs. A. Shyam, E. Lara Curzio HTML, ORNL

Drs. B. Pint and A. Haynes, ORNL

Prof. Nitin Padture, Brown University

Gratefully Acknowledged


Layer-1

Layer-2

High KIC TBC Layer

Erosion ,CMAS

Resistant

Low K ,E

Overlay BC enhanced

roughness

Porous TBC Low K

Low E

Layer-2

Layer-1

Layer-3

Documents

Advanced Thermal Barrier Coatings for Operation in High Hydrogen