Effect of Water Vaporon Oxide Scale Microstructure

B. A. Pint, K. L. More and P. F. TortorelliMetals & Ceramics Division

Oak Ridge National LaboratoryOak Ridge, TN 37831-6156

Research sponsored by theMicroturbines Program

U. S. Department of Energy (DOE)contract DE-AC05-00OR22725 with UT-Battelle, LLC.

IntroductionDOEGoal - improved efficiency microturbines

Various design possibilities, for example:- increase pressure- increase system temperature

New materials needed!

TO MUFFLER/STACK

HRSG/WATER HEATER

ELECTRICITYGENERATOR

COMBUSTOR

FUELSYSTEM

COMPRESSOR TURBINE

AIR

AIR

RECUPERATOR

Motivation:Heat exchangers are essential for improvingthe efficiency of “Microturbines”

(>1MW natural gas-fired engines)

A primary surface recuperator (PSR) is the best method ofimproving engine efficiency in terms of size and cost.

A PSR uses metal foil (75-125µm thick) which must becreep- and corrosion-resistant.

With a limited volume of Cr or Al, the alloy foil must be verycorrosion resistant in order to achieve a long PSR lifetime.

Water vapor in the exhaust gas can accelerate the corrosionrate and is a critical concern for candidate materials

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Tota

l M

ass

Gai

n (m

g/cm

2)

0 5000 10000 15000 20000 25000

Time (h) at 900°C (1652°F)

AusteniticStainless

Steel(20wt%Cr)

4mil foil

253MA 4mil foil

FeCrAl+Ce 2mil foil

FeCrAl+Y 2O3 2mil foil

Haynes 214 2mil foil

Protective Surface Oxideshigher growth rate for Cr2O3 limits temperature

Growth rate of Cr2O3 is much higher than that for Al2O3

Thin foils which form Al2O3 have a distinct advantage at higher temps

(all contain some type of reactive element: Y, Ce, Zr etc.)

900°C, 1652°F

PM2000(ODS FeCrAl)

Haynes 214(NiCrAl+Y/Zr)

Alloys and ExperimentsAlloy Strengthening Oxidation

Type 321 precipitate 17%Cr(Fe-17Cr-11Ni-Ti)

Alloy 625 solid sol.+γ”+MCx... 21%Cr(Ni-21Cr-9Mo-4Fe-4Nb)

Haynes 214 γ/γ’ 4.5%Al(Ni-16Cr-4Al-3Fe+Zr/Y)

Plansee PM2000 oxide dispersion 5%Al(Fe-19Cr-5Al-0.4Ti+Y2O3)

Foil specimens in as-rolled surface finishThicker specimens polished to 0.3µm finish10vol% H2O added to dried air by water atomizer

Creep Rupture Life100MPa, 4-5mil foils

Data at 100MPa: PM2000 - 5mil, 625 & 214 - 4mil(high stress used to accelerate testing schedule)

PM2000 has the longest creep rupture life of the three foils214 somewhat better than 625

1

10

100

1000

10000

ruptu

re lif

e (

h)

700 800 900 1000 1100

Temperature (°C)

alloy 214

alloy 625PM2000

100 MPa

0

1

2

3

4

5

6

7

8

9

10

Spe

cim

en M

ass

Cha

nge

(mg/c

m2)

0 200 400 600 800 1000

Time (h) in 100h cycles

800°C H2O

800°C Air

700°C H2O

4 milType 321

Water vapor effects on 321SSProblems with type 321 stainless steel

Rapid attack with 10%H20 compared to laboratory air

No clear mechanism explains water vapor attackTest temperatures - above corrosion and creep limits for 321

700°C, 1292°F800°C, 1472°Fair+10vol%H2O

plan view

plan view

epoxy

10µm 1µm

10µm 0.5µm

Scale formed on 321SSSEM secondary electron images after 100h, 700°C

With water vapor: fewer and smaller nodules (rich in Mn)thinner oxide (?)

AIR

AIR+10%H2O

TEM of Scale on 321 Stainless Steelbulk alloy after 100h at 700°C

Differences with water vapor: (from first observations)- areas without nodules (arrows) are generally thinner with H2O-> is breakaway due to evaporation in these areas?

700°C, 1292°F

1µm

321

TEM Bright Field ImagesW plating

321 AIR

AIR+10%H2O

W plating

both cases: large (Cr, Fe,Mn)Ox particles

TEM of Scale on 321 Stainless SteelEDX maps of scale after 100h at 700°C

Nodules are Mn:Cr:Fe 4:2:1 from EDX spectraTi more associated with Cr in scale (W signal overlap)

also some Ti-rich particles

W plating

Fe, Cr oxide

321SS

STEM dark field image

AIR+10%H2O

Fe

0.25 m

Ni

Ti

Mn

Cr

-0.1

0

0.1

0.2

0.3

0.4

0.5

Spe

cim

en M

ass

Cha

nge

(mg/c

m2)

0 1000 2000 3000 4000 5000 6000

Time (h) in 100h cycles

800°C Air

800°C H2O700°C H2O

900°C H2O

rapid attack 100µm (4mil)Alloy 625

Oxidation Results: 625100h cycles at 700°-900°C on 100µm foi l

No rapid attack for water vapor observed (mass loss due to CrO 3 evap.)Higher Cr in 625 (21 vs. 17wt%) than 321Low mass gains at 700°C, very high mass gains at 900°CProbable use at ≈800°C depending on life requirements

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Spec

imen

Mas

s G

ain

(mg/c

m2)

0 1000 2000 3000 4000 5000

Time (h) in 100h cycles

800°C Air (4mil)800°C H2O (4mil)

900°C H2O (4 mil)

900°C air (2 mil)

50 and 100µm(2-4mil)

Haynes 214

Oxidation Results: 214100h cycles at 800°-900°C on 50-100µm foi l

800°C: low mass gain in air or air+water900°C: 500h cycles in air, 2mil material, lower transient mass gain

100h cycles in air+H2OMaximum temperature near 900°C with little observed H2O effect

Mass gain x 5.4 = scale in µm

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Spe

cim

en M

ass

Cha

nge

(mg/c

m2)

0 1000 2000 3000 4000 5000 6000

Time (h) in 100h cycles

700°C H2O

800°C H2O

900°C H2O

800°C air

800°C air

125µm (5 mil)PM2000

Oxidation Results: PM2000100h cycles at 700°-900°C on 125µm foi l

700°C: low mass gains800°C: variable results, somewhat lower rates in air + H2O900°C: large initial mass gain of some concernMaximum temperature also near 900°C depending on H2O effect

Mass gain x 5.4 = oxide in µm

Haynes 214

PM2000

epoxy

10µm 10µm

10µm

10µm

Metallography of Oxidation Productseffect of water vapor

Higher temperatures so that there is something to see

Both cases, metal-oxide interface appears rougher with water vapor

In some areas on PM2000 the scale appeared thicker with H2O

214100h

1100°C

AIR AIR+10%H2O

PM2000100h

1000°C

epoxy

Haynes 214

PM2000

PM2000

epoxy

epoxy

10µm 2µm

10µm 2µm

Scale formed on PM2000SEM backscattered electron images

after 100h, 1000°C

Without water vapor: Cr-rich outer layer (bright layer in BSE image)With water vapor: rougher and occasionally thicker oxide

rougher interface with protrusions (arrows)no Cr-rich layer observed

AIR

AIR+10%H2O

epoxy

epoxy epoxy

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 20 40 60 80 100 120

Mas

s C

hang

e (m

g/cm

2 )

Time (h)

PM2000, 1000°C, 10% H2O

PM2000, 1000°C

PM2000 Foil1000°C

0

0.1

0.2

0.3

0.4

0.5

0.6

0 20 40 60 80 100 120

Mas

s C

hang

e (m

g/cm

2 )

Time (h)

214, 1100°C, 10% H2O

214, 1100°C

214 Foil, 1000°C, 10% H

2O

OxidationKinetics

foil and 1mmthick specimens,air & air+10%H2O

Main effect of H2O

appear to be ontransient stageoxidation for aluminaforming alloys

TEM of Scale on Haynes 214bulk alloy after 100h at 1100°C

Some differences with water vapor:- thicker oxide- more voids in outer Ni-rich oxideDifficult to make TEM cross-section of scale on foil material

1100°C, 2012°F

500nm 214

alumina

TEM Bright Field Images

NiAl2O4

W plating

NiAl2O4

214

AIRAIR+10%H2O

TEM of Scale on PM2000bulk alloy after 100h at 1000°C

Some differences with water vapor:- no Cr-rich oxide at gas interface -> evaporated (?)- finer grain size (more diffusion paths leads to faster oxidation?)- no increase in void concentration (which contradicts a possible model)

1000°C, 1832°F

PM2000

alumina

TEM Bright Field Images

W platingFe, Cr oxide

PM2000

AIR AIR+10%H2O

TEM of Scale on PM2000bulk alloy after 500h at 1000°C

Longer-term observations with water vapor:- again no Cr-rich oxide at gas interface -> evaporation?- grain size looks similar to that in dry air- similar void concentration as observed in dry air

1000°C, 1832°F

PM2000

alumina

TEM Bright Field Images

W platingFe, Cr oxide

PM2000

AIR AIR+10%H2O

W plating

0.5 m

Oxidation Limited Temperature Predictions

Typical oxidation models give time to rapid(base metal) attack: “time to breakaway”Mechanistic models: supply vs. consumption rate

For 100µm foil, can’t tolerate large loss incross-sectional thickness

* More concerned with time to ≈20% metal consumption

For Cr2O3: on 625 at 800°C, predict life >50,000h

For Al2O3: less certainty about prediction >800°C

experimental results on 2mil foil:

PM2000/Haynes 214: ≈1000h life at 1050°C>22,000h life at 900°C

Some uncertainty regarding water vapor effecte.g. at 1100°C:10-40% lower life with 10%H2O

Acknowledgments

Research sponsored by the Microturbines Program,U. S. Department of Energy (DOE) contractDE-AC05-00OR22725 with UT-Battelle, LLC.

Experimental worked assisted at ORNLby:George GarnerMike HowellShawn Trent