Upload
keo-dinh-chuot
View
217
Download
0
Embed Size (px)
Citation preview
8/18/2019 Hot corrosion in GT
1/83
UNCLASSIFIED
AD NUMBER
LIMITATION CHANGES
TO:
FROM:
AUTHORITY
THIS PAGE IS UNCLASSIFIED
AD870745
Approved for public release; distribution isunlimited.
Distribution authorized to U.S. Gov't. agenciesand their contractors; Administrative/Operational Use; MAY 1970. Otherrequests shall be referred to Office of the ofthe Director of Defense Research and
Engineering, Washington, DC 20301.
ODDRE ltr 19 Oct 1970
8/18/2019 Hot corrosion in GT
2/83
NMAB-2^0
MAY
^70*
:
Hot
Corrosion
in
Gas
Turbines
...
Mechanisms
Alloy & Coating
Development
Environmental
Effects
Evaluation
A
Repor t
o f
|h e
\
NATIONAL
MATERIALS
'ADVISORY OARD
j.
ATIONAL
REStARCH
COUNCIL
N iTIONAL
ACADEMY
OF
SCIENCES-NATIONAL
ACADEMY
OF
ENGINEER1NC
8/18/2019 Hot corrosion in GT
3/83
«CESSIMMor
_
C F S T I
HIT SC3TIM
Q
O O C U F F
E C T I O «
U N A H H O W C E D
J U S T I F I C A r i W I
B Y
DISTRIBUTIV
/AVAIUBIl in T Ü E
H I S T .
AVAIL
4 / l (
NATIONAL ATERIALS
DVISORY OARD
DIVISION
F
NGINEERING
-
NATIONAL ESEARCH
OUNCIL
Chairman
iiam J. Harris,
Jr.
sident of Research
m
nd
T
8/18/2019 Hot corrosion in GT
4/83
HOT
CORROSION
N
GAS
TURBINES
' MECHANISMS
ALLOY COATING
DEVELOPMENT
*' ENVIRONMENTAL
EFFECTS
' EVALUATION
REPORT
OF
THE
COMMITTEE ON
ATMOSPHERIC
DETERIORATION OF SUPERALLOYS
NATIONAL
MATERIALS
ADVISORY
BOARD
Division of Engineering
-
National
Research
Council
This document is
subject
to special
export
controls nd
each ransmittal
o
oreign
governments
or foreign
nationals
may
e
made only with
prior
approval
of
he
Office
of
he
Director
of Defense
Research
and
Engineering
Chief, Materials Division,
OADCT) .
Publication
NMAB-260
National
Academy
of Sciences National
Academy
o f
Engineering
Washington,
. .
May
1970
8/18/2019 Hot corrosion in GT
5/83
This
report
is
one
of a series in studies undertaken
by
he
National
Mater'
s
Advisory
Board
for th e National Academy of Sciences and th e
National
Academy
of
Engineering in
partial
execut ion
of
work under
Contract
No. A-49-083
OSA-
3131
with
th e
Department
o f
Defense .
A s
a
part
of th e National Research Counci l , he National Materials Advisory
Board
performs
tudy,
evaluation,
or
advisory
funct ions
hrough
groups
com-
posed
of
individuals selected from academic, governmental, nd
industrial
sources
for
their
competence
or
interest
in
th e
subject
under
consideration.
Members
of
these groups serve as ndividuals contributing
their personal
know-
ledge
and
judgment s
and
no t
as
representatives of
an y
organization
in
which
they are emp loyed or with
which
they
may be associated.
The
quantitative data publ i shed in this report are
in tended
only o illustrate he
scope
and substance of
information considered
in
th e
study,
nd
should
not
be
used for an y other purpose,
uch
as n specifications
or
in design, nless so
stated.
No
portion
of this
report
may
e
republished
without
prior
approval
of th e National
Materials
Advisory
Board.
Copies of
this
report
are no t
available
from
th e
Clearinghouse for Federal
Scientif ic and Technical
Information CFSTI).
Qualif ied requesters
m ay pply
th rough
Defense
Documen ta t ion
Center.
Those
who
do
no t
use
he
Defense
Documen ta t ion Center may pply o
th e
National
Materials
Advisory
Board.
ii
8/18/2019 Hot corrosion in GT
6/83
NATIONAL
MATERIALS ADVISORY
BOARD
COMMITTEE ON ATMOSPHERIC DETERIORATION OF SUPERALLOYS
Chairman: Dr.
Robert I.
affee, enior
Fellow
and
Associate Manager, Materials
Science
and
Technology, Department
o f
Physics,
Battelle Memorial
Institute,
05
King
Avenue,
Columbus,
Ohio
3201.
Members:
Mr.
Charles C.
Clark, n
Charge, Aerospace Power, The Inter-
national Nickel Company, nc., 7 Wall Street,
ew
York,
ew
York 10005.
Dr. Matthew
J.
Donachie,
Jr.,
Development Metallurgist, MDL-J
Building,
Pratt
& Whitney Aircraft, East
Hartford,
Conn.
06108.
Dr. William R. reeman, Jr., Director, Materials Laboratories,
AVCO
Lycoming
Division,
50
S.
Main
Street,
Stratford,
Conn.
6497.
Dr. William
Hagel,
Head,
Metallurgy Division,
Denver
Research
Institute,
University
o f
Denver, University
Park,
Denver,
Colo.
0210.
Mr.
aul
E.
Hamilton,
ection Chief,
nvestigations Development
Metallurgy, Aircraft
Engine
Operations, Allison Division, General
Motors
Company,
. . ox
894,
ndianapolis,
ndiana
6206.
Dr.
Robert A .
Rapp,
Professor, Metallurgical
Engineering,
College
of Engineering,
h io
State
University, Columbus,
Ohio 43210.
Dr.
Alan
U.
Seybolt,
Metallurgy Ceramics
Laboratory,
General
Electric Company,
& D
Center,
P.O.
Box
8, Schnectady,
ew
York
12301.
Dr.
William
E.
Young,
Manager, Combustion, Corrosion
& MHD
Section,
Heat Transfer Power
Generator
R&D ,
Research
and
Development
Center,
Westinghouse Electric
Corporation,
Churchill
Borough,
Beulah
Road,
ittsburgh,
ennsylvania
15235.
iii
8/18/2019 Hot corrosion in GT
7/83
Liaison
Representatives:
Dr.
Hubert
B.
robsc,
Head, Oxidation
and
Refractory
Compounds
Section, ASA
Lewis
Research
Center,
1000
Brookpark
Road,
Cleveland, h io
44135.
Mr. Milton
Levy, Army
Materials
Mechanics
Research
Center,
Watertown, Massachusetts 02172.
Mr.
Henry
Morrow, . . Army Aviation Materiel Laboratories,
Fort
Eustis, Virginia
23604.
Mr.
George
J.
Danek,
Jr., Naval
Ship
Research and Development
Center,
Annapolis, Maryland
21402.
Mr. rving Machlin, Materials
Engineering, AIR-52031B,
Naval
Air
Systems
Command, Department of
the Navy,
Washington,
D .
.
20360.
Mr. Norman
M .
Geyer, Air Force
Materials
Laboratory, MAMP )
25-53136,
Wright-Patterson Air
Force
Base,
Ohio
45433,
Mr. ohn
C .
Barrett, DDR&E ,
Office
of
Assistant
Director
for
Chemical
Technology,
Room 3D117,
The
Pentagon,
Washington,
D. . 0310.
NMAB
STAFF:
Dr.
oseph R.
ane,
taff
Metallurgist,
National Materials
Advisory
Board,
Division
of
Engineering, National Research
Council, NAS-NAE, 101
Constitution
Avenue, . W.,
Washington,
. .
0418.
iv
8/18/2019 Hot corrosion in GT
8/83
SUBGROUPS
of the
COMMITTEE ON ATMOSPHERIC DETERIORATION OF SUPERALLOYS
Mechanisms
Dr. Alan
U.
eybolt,
Chairman
Dr.
William
Hagel
Dr. Robert
A. Rapp
Alloy Coating Development
Mr. Charles
C . Clark, Chairman
Dr.
Matthew J.
Donachie,
Jr.
Mr. aul E. Hamilton
Environmental Effects
Dr.
William
E.
Young,
Chairman
Dr. William R. reeman
Aided
by Contributions From
Mr.
Norman
Bornstein, United Aircraft
Research Center
Mr.
.
A . DeCrescente,
United Aircraft
Research Center
Mr.
rving Machlin,
Naval
Air Systems
Command
Mr. Warren Kentz,
AVCO
Lycoming
Div.
Miss
E. J.
MacNair,
Ministry
of
Defense,
Bath, England
D .
R.
Carlisle,
Rolls
Royce
Industrial,
Ansty,
England
Henry
L.
Morrow, . S. Army Aviation
Materials
Laboratory
Jeremy
J. Walters,
VCO Lycoming
Irving
Machlin,
Naval Air
Systems
Command
R.
M .
Shir'mer,
hillips
Petroleum
Alan
U.
eybolt,
General
Electric
G. . Danek, Naval Ship Research and
Development
Laboratory
Hubert
B. Probst, NASA
Lewis
Research
Center
W. T.
Reid,
Battelle
Memorial
Institute
W. E. Somgrs,
Pujlic
Service
Electric
and Gas Company
G.
.
Widers
urn,
Philadelphia
Electric
Company
Evaluation
Dr.
William
R.
reeman, Chairman
Mr. aul E. Hamilton
Mr. George J. Danek,
Naval
Ship
Research
and Development
Laboratory
8/18/2019 Hot corrosion in GT
9/83
CONTENTS
Abstract ii
I.
INTRODUCTION
H .
MECHANISMS
A .
ot
Corrosion Attack
B.
igh-Temperature
Oxidation
1
IE. MATERIAL AND
COATING
DEVELOPMENT
4
IV.
ENVIRONMENTAL EFFECTS 0
V.
EVALUATION 9
A .
Recommendations
9
VI.
SUMMARY OF
CONCLUSIONS
AND RECOMMENDATIONS
1
Appendices
I.
Recommendations
from
Coating Systems for
Gas Turbine Engines 4
1 1 .
Round Robin
Hot
Corrosion
Testing
Program 7
vi
8/18/2019 Hot corrosion in GT
10/83
ABSTRACT
Hot
corrosion (sulfidation)
n
gas
turbine
engines
has
become
a
major
problem
because
of
the
increased
use of
al loys
lo w
in
chromium and th e operation
in
environments
containing
alkali
metal
salts,
especially
near sea water.
The
mechanism
of
attack
is
understood
to
some
extent,
ut
more
work
is
needed. It
is
clear
that
sodium
and other
alkali
metal
salts are
i nvo lved .
Sodium
sulfate
is ngested
with
the
combustion
air
or
formed
from
the
sulfur
in
th e
fuel,
and reacts with the metal oxide scale
acting
as n
Na20
sink to
form
a
comp lex
sodium
salt. After scale
breakdown,
sodium sulfate
may
attack
the
underlying
metal, orming
sulfides. The
presence of NaCl in the
gas
nd
a
l iquid
salt film
seem to
be
required
for accelerated attack.
Coatings,
l loy modification, nd
additives
to
the
fuel
all
help
alleviate the
immediate
problem.
Al loy development,
dispersion
hardened and
fiber
strengthened
alloys,
nd
rare
earth additions to
superalloys m ay offer longer-range
solutions.
Target
performance
properties
are
needed
to
focus
the
future
research
and
development
work.
Sulfur
can
enter a
gas turbine
from
the
fuel, nd
chloride
and
sulfate
salts
from
th e
air.
There
is
little
prospect
for
removal
of
these
contaminants
to such
a
degree
as
to
eliminate
th e
problem.
However ,
rom
10
to
75% of
sea salt
can
be
removed
from
the air
intake
to retard
the
attack.
Data on
tolerable
levels
of con-
taminants are
needed.
Engine
testing
is still
th e only
reliable
evaluation
method
to
check
im-
provements
in
the
hot
corrosion
resistance
of
materials.
Test
rigs are common ly
used
fo r
a preliminary
evaluation of
hot
corrosion
resistance. Reproducibility
n
test
results
from
test
rigs
n
different laboratories
s very poor,
nd
further effort
toward
test
rig
standardization,
correlation
of
the data
between
different
tests,
nd
interpolation
of tests results nd
extension
to
different
testing
conditions, all are
needed.
Specific
recommendations for attacking th e hot
corrosion
problem are
detailed in
th e
report.
vil
8/18/2019 Hot corrosion in GT
11/83
I.
INTRODUCTION
The Problem
A s temperatures in
th e
turbine section
of
gas turbine
engines
have
increased,
n
increasing
amoun t
of
environmental
attack
has been
encountered
in
th e
superalloy
parts.
Operational
personnel
generally
have ascribed the problem
to
hot corrosion
or
sulfidation,
whereas
the
real
problem
m ay
have
been
excessive
oxidation
du e
to
overtemperature. For whatever cause,
he
results of this
attack
penalizes engine
performance
by
a
severe
restriction
in
operating
life.
The major
limitation
on
turbine
inlet temperature
is the
strength
of
the
turbine
rotor-blade and
stator-vane
materials.
There
have
been
major
improve-
ments in
the
high-temperature
strengths
of
these
materials,
primarily
through
th e
use of strong casting alloys,
ardened
by
a substantial
vo lume
fraction
of
the
coherent
y' precipitate,
Ni,
Co)
AI,
Ti).
However , he major
means
for
achiev-
in g
higher
gas
temperatures in
the
turbine
is
through the use of air cool ing ,
where-
by a
small
percentage
of
the
inlet
air
is diverted
through
cooling
passages
in
the
ho t
turbine
blading
an d vanes.
Through
this means,
he
temperature
of these
parts
is
owered
about 400-500
o
F
222-288
0
C) below
the gas
temperature
to
the
poin t
where
their strength capability is sufficient.
Up to th e
present time,
orrosion
and
oxidation
resistance have no t
been design considerations in
aircraft
gas
turbine
material selection
other
than
that
the
materials
resist th e
environment,
hich
superalloys
ontaining abou t
20%
chromium
do adequate ly
up to
the
temperature
at
which strength
is
maintained,
abou t
leSO'F (899
0
C ). However , he
development of the
stronger
gas
turbine
alloys
was
facilitated
by
a
reduction
in
chromium
content
to abou t 10%
chromium,
to
permit
an increased
amoun t
of
y' hardening
from
Increased contents
of
aluminum
8/18/2019 Hot corrosion in GT
12/83
and titanium.
The
high
aluminum
content maintained
th e
oxidation
resistance
but
not sulfidation resistance
equivalent
to th e higher chromium content.
Hot
corrosion
is an
accelerated oxidation
in
the
presence
of
NaCl and
Na
SO
usually
due
to operation
in marine
environments.
It
also
includes
attack
by
lead
vanadium
and
other fuel
contaminants.
The reduced oxidation
and
hot
corrosion
resistance
has
been
combated
by
coating th e
hot componen t s
with
aluminide-type
coatings.
The
coating
procedure entails
additional
expense and
a
problem with
reliability
compared with
th e former situation
where
uncoa ted
alloys
with adequate oxidation
and
corrosion
resistance
were
operated
at
lower
gas temperatures.
The Commi t t ee
The Department of
Defense ,
ecognizing th e
seriousness
of
th e
ho t
corrosion
problem, sked
th e
National
Research Counc i l to organize a
committee
to review th e problem. The Committee, which
was
assembled
by th e
National
Materials Adviso ry
Board,
s shown on
page
iii.
Members
of
th e Committee,
act ing
as
individuals
and not
as representatives of their employers,
ontributed
their services.
The
Committee
reflected
varying
viewpoints
ranging from
basic
researchers
to
those of
stationary
and aircraft
gas
turbine manufacturers.
Method
of
Operation
After
initially considering their individual positions
and
viewpoints on
th e
problem,
he
Committee
divided their
consideration into
four
areas:
mechanisms, materials
deve lopment ,
nvironmental
effects,
nd testing and
evaluation,
each
of
which
was assigned to a
subgroup.
The remainder of this
report
comprises th e
findings
of
th e
four
subgroups,
ol lowed
by
overall conclu-
sions
and
recommendations.
8/18/2019 Hot corrosion in GT
13/83
II.
MECHANISMS
Introduction
In this entire report, mphasis has been
placed on hot corrosion or
sulfidation of superailoys,
s opposed
to
their oxidation,
ecause
this
ype of
reaction
is
a
matter
o f
the
most immediate
concern and
study in
current
engine
service.
The
subject of
hot corrosion is discussed first.
It is
generally
agreed that
the basic corrosion
attack occurring
in
hot
corrosion or sulfidation is
oxidation which
occurs in
a
very rapid or
near-
catastrophic
manner
because
of the
disruption
caused
by the existence o f a salt
film of the normally protective scale. In addition,
ulfides
form
inside
the sur-
face grains
and
sometimes
deep along grain
boundaries
o
cause
serious
degrada-
tion
o f structural integrity of the alloy.
As
has
been pointed
out
in many publications see, or
example,
he
recent
ASTM
Symposium
on this
subject),
he
immediate
agent
responsible
for
hot corrosion in aircraft
engines
s
Na SO which forms from chemical reactions
taking place between NaCl
n sea air and sulfur in the luel, s
well
as from
the
appreciable
content
of Na„SO, n sea air. There is also evidence that NaCl-Na^SO,
2
4
4
mixtures
are
more
corrosive
han
Na
SO
alone.
While
the
same
type
of
hot
corrosion
attack
has been
observed
in
stainless steels
subjected
to
SO
atmos-
pheres ttention here s focused
on the
problem
caused
by Na SO
in
aircraft
jet
engines and
marine
gas urbines.
2
DeCrescente and
Bornstein
have shown that Na^SO, must be present
2 4
on
the attacked surface as a condensed
phase,
rdinarily
a liquid,
o
cause hot
corrosion.
a
SO apor
did
not
cause hot corrosion. or
this
reason,
most
investigations
have
found
that
for
pressures
near
atmospheric
his
accelerated
8/18/2019 Hot corrosion in GT
14/83
attack
occurs
over
th e
temperature
range
of
about
HOO-ISOOT.
In
qualitative
agreement
with
th e effect
of pressure
n
th e
dew po in t
of
Na
SO he temperature
range
of attack
is increased
with
higher
total
pressures.
There has been
comparatively
ittle
work
aimed
at
understanding
th e
mechanism
of oxidation as
altered
by
th e
presence
of
contaminants uch
as salts
like
Na
SO
NaCl , or
low-melting
oxides uch
as V O
These
materials can
form
l iquid
slags
and
hence
cause
catastrophic
oxidation
by
virtue
of
removal
of
protective scales. This
field of
high
temperature
corrosion
has
recently been
3
reviewed by Hancock, In
this review,
nearly
all th e investigations
reported
were
concerned with examination of field-returned
samples
or
with
attempts
in
th e
laboratory to
make
high-temperature
corrosion tests
which
essentially
dupli-
cated field conditions. Very
little work
aimed
at elucidating
mechanisms
has
been
reported,
particularly
in
simple systems amenable
to
scientific analysis.
A .
Hot
Corrosion
At tack
Characteristics
The
alloys of main concern
here
are used
as first-
or second-stage
blades and
vanes
in
aircraft
engines,
nd
are
th e
nickel-base
superalloys con-
ta ining
typically
6-22%
Cr,
-26%
Co, -10% of either
Ta, o or W, -6%
A l,
0-5%
Ti, nd
small fractional
percentages of carbon and other
elements. A
sample showing
severe hot corrosion attack exhibits a heavy surface
layer
of
oxides,
nd
just
below
this
layer,
xides
intermingled
with
alloy-depleted
nickel
matrix, nd below
this
level, hromium
sulfide
particles
which
m ay
contain
appreciable percentages
of
Ni,
o, A l,
Ti, nd refractory
metals.
A s
corrosion
proceeds,
he
chromium
sulfides
are converted
to
oxide
with
th e
sulfur
atoms
t hus released diffusing more deeply
into th e alloy to form more sulfides. The
alloy
matrix
surrounding th e
chromium
sulfides
is
depleted
in
chromium
to
an
appreciable
extent and would be
expected
to
oxidize
more
readily
than
th e
original
8/18/2019 Hot corrosion in GT
15/83
alloy.
However ,
his
effect
is considered to
be
of less
importance
than
th e
rapid
oxidation attack proceeding at
th e
oxide/alloy
interface because
th e oxide
scale
fails
to act
as
a diffusion barrier, s would occur in normal
air oxidation. Chro-
mium sulfides
both
in grains and in
grain boundaries cause loss
of structural
integrity,
nd
in cases where
such sulfides
are
formed deeply along grain bound-
aries,
severe
reduction
in
mechanical
properties results. Such
reductions in
mechanical properties
are
particularly
severe
in
blades,
which
are
highly
stressed
as compared to vanes.
Some recent
ho t
corrosion experiments
on
a
variety
of
alloys
appear
Q
to
confirm
th e
original
suggestion of
Simons
et
al
hat
th e
catastrophic
nature
of th e corrosion attack
invo lves
rapid
oxidat ion of
Ni S
Ni
eutectic
liquid.
This
liquid
usua l ly
is short-lived,
nd
may not
be
visible
in
an
alloy cooled
to room
temperature because of
th e displacement reaction
(7)
see
below).
On th e
other
9
hand,
Bornstein
an d
DeCrescente
have
found
that
oxidation
attack
of
several
different
alloys in th e presence
of NaNO
salt
closely
simulates
hat for NaSO
attack. The conclusion reached is
that
sulfur
and
alloy
depletion are
not he
important
aspects
of
hot
corrosion,
ompared
to
th e presence
of
sodium
com-
pounds.
Hot-corrosion resistance
can be
achieved
by
preventing
attack
by
th e
salt
of
th e
initial oxide
layer on
th e
alloy.
This
salt-oxide reaction
will
be
considered
first.
Scale
-
Salt Interaction
Since
Na
SO
is a very
stable
compound
there is
very little
thermal
decomposition
of
Na
SO
according
to
th e
fo l lowing reaction:
Na
2
S0
4
= a
2
0 + S0
2
+i
2
. 1)
8/18/2019 Hot corrosion in GT
16/83
However, f there is available an
Na O
sink which
effectively
retains
an
Na
activity
well below unity which
is assumed by equation 1) , hen Na O
can
continually
decompose. Formation
of
complex oxides
involving
Na
O and
oxides
normally
present
in the oxide
scale,
uch
as Cr O Al
iO
r
WO
an
4
serve
as sinks
for
the
Na„0
by such
reactions as:
2
2Na
2
S0
4
Cr
2
0
3
|0
- »
2N a
2
Cr0
4
2S0
2
2)
Experimental
corroboration of
Na
O-oxide
interaction
has
been observed
by
5
Bergman
and Kaufman
who
found
X-ray
diffraction patterns
for
sodium ungstate,
sodium
tantalate,
nd
sodium
titanate in corrosion products on hot-corroded
superalloys.
imilarly
an
NiO
scale breakdown by
Na SO would
occur
according
to
the
following
reaction:
Na SO + 2NiO
Na NiO
+
SO
3)
2 4
6
NaNiO
has been
reported
by
Bornstein.
Alternatively
NiO
may
break
down
d
according
to
the
following reaction:
Na SO +
NiO (N a O in
NiO)
SO + ̂ O 4)
Z r
A
combination of both
reactions, 3) and 4) , may
be
operative.
That
NiO
can
1
dissolve some
Na
O
has been suggested by
Danek , and Quets
and Dresher.
2
The
actual mechanism of
scale breakdown
is
probably
more complex than
that
suggested, ut it seems
obvious hat, or an understanding of
the
hot
corrosion
mechanism,
he
area
of
salt/oxide
interactions
should
be
carefully
investigated.
Although
they
have
not as yet
developed
a
detailed
theory,
DeCrescente
7
and Bornstein indicate
from
their
work on molten
salts
EMF measurements and
SO
evolution)
hat
they
have evidence hat the presence
of
oxides
such
as
Cr
O
3
on
an
alloy
reduces
the
Na
O
concentration by
formation
of
Na CrO
and
thus
2
4
lessens
hot corrosion attack.
8/18/2019 Hot corrosion in GT
17/83
Alloy-Salt
Interaction
Once
th e
oxide scale
has
been
penetrated, he
next
stage
of
attack
is
reaction
between
condensed
Na^SO,
or
Na„SO
NaCl
and
th e alloy itself. Various
2 4 4
investigators have
suggested
reactions based
on th e early
work
of
Simons who
suggested
a
reaction
equivalent
to
th e
tcilowing:
4M
+
Na
SO
=
a
O
+
3MO
+
MS.
5)
4
However ,
Quets
and
Dresher
have shown by
th e
use
of Pourbaix-EUingham
diagrams
that
pure
Na
O,
which
has
been considered
as
an
important
product
of
th e
hot
corrosion
attack
by Na SO
cannot
exist
at
unit
activity
with
th e
products
Ni
S
iS
or
C rS
as
would
be
required
by
equat ion
5).
For
this reason,
quation
(5) must
be modified for both th e
case
of
M =
Ni
or
M
=
Cr.
If
th e Na„SO. reacts
2 4
with
the metal
to form
a
ternary sodium
oxide
and
binary sulfide,
a modified
reaction
becomes
possible:
xM+NaSO, =Na„MO
+M S
6)
2
4
y (x-y)
In support
of
this
mechanism,
he
stabilities
of
th e
ternary sodium
oxides
extend
to
much higher
values
of
P
than
those for
pure
Na
O.
This mechanism
is
very similar to
th e
one
proposed
for
th e
oxide
scale
breakdown.
However , ince th e reaction
occurs
at
th e metal-scale inter-
face,
ts
oxygen
potential
is
set
by
that
interface.
Dur ing
sulifidization,
he
oxygen potential at th e metal-scale interface should
be approximately th e same
as
for
a
Ni-NiO
interface
because
of
th e chromium
depletion
in
th e
alloy.
For
th e formation
of
a
ternary
sodium
salt,
he
sulfur
potential is
high enough to
allow
th e formation
of nickel
sulfides.
However ,
ue
to th e displacement reaction
Cr,. ..
+NiS
rS
+
Ni,
„ 7)
(in
al loy)
in
al loy)
NiS
shou ld
be reduced
by
th e matrix
chromium
to
form
the
more stable chromium
sulfide.
But
if
th e
chromium
content
in
th e
al loy
s
rather
low,
eaction
7)
does
8/18/2019 Hot corrosion in GT
18/83
not
occur
apidly
or
completely
enough
to
reduce all
th e
NiS.
Therefore, n
severe cases
of hot corrosion,
iS actual ly
Ni on cool ing
to
room
tempera-
O It
ture) m ay be left
in
th e
structure.
Effect
of
Al loy Compos i t ion*
A s
might be expected for a class of alloys with a very
complex and
widely
variable composition, he individual
effects
of al loy components
on
re-
sistance
to
ho t
corrosion
is
no t
unambiguous .
Certain
generalizations
can
be
made, such as
that resistant alloys
usually
contain fairly high chromium
con-
tents,
around 15% Cr
or
higher.
There appears
to
be
agreement that
molybdenum
is
harmful,
bu t th e behavior
of
other
al loying
elements
s
ess
consistent, nd
therefore not clearly
established. Better understanding of compositional
effects
migh t
be
obtained
after
some of
th e
scale-Na
SO
and
alloy-Na
SO interactions
are
further
elucidated.
Research for
Defining
Mechanisms
The most important
areas
for
additional
research are
th e
salt/scale
and salt/alloy interactions
n th e temperature
range
of hot. corrosion, nd oxygen/
al loy
nteractions
at
higher
temperatures. Ano the r
area
which
needs
attention
is th e
stabilities
thermodynamics)
of
ternary
odium
oxides.
Water
vapor,
product
of
combus t i on , as
no t
been
seriously
con-
sidered
thus far
as corrosion
variable,
but
it
seems possible
that
water vapor
in th e
eng ine
might be
significant in view of
th e
ability
of
water vapor to cause
oxidation at
eng ine
operation
temperatures. Further
study
is
ndicated.
7
The
suggestion
of
DeCrescente
and
Bornstein
hat
additions
of
cer-
tain oxides
affect th e Na SO
alloy
reaction
should be investigated
further,
s
this
obviously
would be
significant
with regard
to
th e
type
of
corrosion
mechanism
operating.
*A more
detailed discussion of
th e
effect of
al loy
composition
of ho t corrosion is
given
in
th e
next
section
on
Material
and
Coat ing
Deve lopment .
8/18/2019 Hot corrosion in GT
19/83
Since there is
th e
possibility of using vanadium-containing fuels in
gas
turbines associated
with
ship
propulsion,
o
0 contamination or
possibly contam-
2 5
ination
by other fuel
components
should be studied. Some
work
along
these lines
has
already
been
done,
as
shown
in
reference
3,
but as
in
th e
Na
SO contamina-
t ion case,
uch
more work
on
mechanisms remains o
be
done.
Closely
associated
with
ph enomena
already
mentioned
is
he
question
of sulfur
diffusion
a long
grain
boundaries
an d in th e grains of
nickel-base
super-
alloys. This solid-state
diffusion
process
is
of prime
importance
in
th e
formation
of
th e chromium sulfide inclusions so characteristic
of hot corrosion. One
aspect
that
is
of
considerable
practical
importance is
o
ascertain
those
conditions which
lead to excessively
deep
grain
boundary
penetration
by sulfur.
Like sulfur, oxy-
gen
tends
to
concentrate
in grain
boundaries and
to
form
internal
oxides
much
deeper
along the
al loy
grain
boundaries than
in
th e bulk
grains.
Research on Reducing
Hot
Corrosion
There appear
to
be
four
main
avenues
of
research
aimed
at
reducing
hot corrosion
attack.
Alloy
Composition
By comparatively
minor changes
in
al loy
composition,
t
has
been
found that
considerable improvement in
hot
corrosion
resistance is possible,
usually at a small sacrifice
in high-temperature
strength. However ,
s ncreased
demands are made
upon
th e
mechanical capabilities of
nickel-base
superalloys,
this road to
corrosion
resistance appears
less
feasible.
Coat ings
NiAl- type
coatings are
now
standard
in
military jet
engines,
where
th e
temperatures are
somewha t
higher
than
in commercial engines.
Through
coating,
considerable
improvement in
both oxidation resistance and hot
8/18/2019 Hot corrosion in GT
20/83
10
corrosion resistance is
secured.
However ,
ncreased
demands for
higher
metal
temperatures approaching
1800
o
F
982
0
C)), or
longer times between
engine over-
hauls,
nd for improved coating
reliability
make current coatings
less and less
acceptable.
Research activity
in th e
coating
field
shou ld
be
cont inued; hopeful ly
some
research
would
be
aimed
at
th e fundamental
aspects nvolved,
uch
as
measurement
of
diffusion
rates in moderately complex systems.
Rare Earth Addit ions
Fractional percentages of
rare earth metals, uch as cerium '
lanthanum
, nd
gadolinium
have
been found
to generally
improve hot
corrosion
resistance
in
binary
Ni-Cr
alloys
and in
superalloys.
Much
of this work
has been
carried out in
laboratory
tests, often of
th e
simple
crucible
type, but th e few
available
hot
corrosion
tests
in
hot-corrosion
rigs,
imulating
to
some
degree
actual
engine
tests,
ave shown
marked
improvement.
One
cannot
expect
that
rare
earth
additions
cou ld
add
more
than
about
1000
hours
of
life
to
jet
eng ine
blades and vanes.
The
principal
difficulty
with
th e
use
of rare earth additions
is
that a low melting phase
associated
with
th e
Ni
M/Ni
eutectic
where
M is a
rare
earth) occurs to
reduce
high-temperature
strength
properties.
Recent
laboratory
work,
owever ,
hows
hat
greater
improvements
in
corrosion
resistance
are
offered
by
rare
earths
when added in oxide
form.
If
verified,
his fact
m ay
mply
a
change
in
processing
techniques
hrough
which
rare
earths would
be
present
in
th e
al loy
as
a
well-dispersed
oxide
phase.
Mos t
important,
heoretical
rationali-
zation
and
understanding
of
th e
effect
of
rare
earth
additions
hou ld
be a t temp ced.
Fue l
Addit ives
Thisis
discussed
in
th e section
under
th e
heading
of
Environ-
mental Effects.
8/18/2019 Hot corrosion in GT
21/83
11
B.
High-Temperature
Oxidation
Imp roved
oxidation
resistance for
strong superalloys
is
desirable
in
blade applications. Higher
turbine inlet
temperatures
could be accommodated,
and
th e
requirement for coating
constricted
air-cooling
passages could be
avoided.
The problem
s
complicated
because
ew
strong
alloys,
uch s
B1900,
ave
excellent
oxidation
resistance
but suffer
hot
corrosion
attack.
Therefore, sep-
aration
of
th e
two
mode s of attack
is
not
realistic.
In
th e
past,
l loy
oxidation
i
has been a
research
area with an emphasis on
short-range
improvement of
exist-
in g
alloy
compositions,
with th e best theory and experiments in
alloy
oxidation
coming
from
other countries,
namely Germany,
ngland,
nd
France.
Sugges t ed
al loy
oxidation
research, with
only
short
bridges
to
th e improvement of commer-
cial
superalloys, ncludes:
a)
th e
use
of
radiotracers
and
h igh
resolution
techniques
to
establish
th e
role
of
rare
earth and
other
additions
ike
Si
and
Mn
on binary
and
ternary
alloy
oxidat ion;
b)
a
study
of th e role
of carbon
in dissolved
and combined
form
in
al loy
oxidation;
c) th e role
of th e common ly used
strengthening components,
W
and Mo,
n
alloy
oxidation particularly
for
Co-base
alloys;
d)
further attempts to study
th e
complex morpho logy
and
transport
in
th e metal/scale
interface
when combined
internal
oxidat ion
and
external scale formation are
occurring;
e) further
tudy
of
th e modes
of
failure
of
external Cr O
and
A l O
scales
and th e healing
processes;
f)
specific
studies
relating
th e
evaporation rate
of Cr
O to
th e
vapor pressures of th e several
vapor
species
for
th e
oxide;
g)
further
studies
of
the phases formed and
th e kinetics
of
internal
oxidation
of
alloys;
8/18/2019 Hot corrosion in GT
22/83
12
h)
studies
of
th e
high-temperature
mechanical
properties
of oxides
i.
e.,
creep)
and
th e
relation
of
these
properties to
scale
fracture;
i) preferential grain
boundaiy
attack
in pure
metals
nd
alloys
and
its
prevention;
j)
many
of
th e
more
fundamenta l aspects of
al loy
oxidat ion:
nucleation of oxides
on
clean
al loy
surfaces,
oping effect,
olu-
bilities and diffusivlties of
oxygen
in
nickel and cobalt, olubilities
and
diffusivlties
of
aliovalent
impurities
in
simple
oxides,
particu-
larly
CoO,
NiO, Cr
O and A l O.
8/18/2019 Hot corrosion in GT
23/83
13
REFERENCES
1.
Hot Corrosion Problems Associated with
Gas Turbines, ASTM
Specia l
Publication
No.
421
1967).
2. DeCrescente, . A .
nd
Bornstein,
.
S., Corrosion
,
4,
1968), 27 .
3.
ancock,
P., Corrosion
of Al loys
at High
Temperatures
in Atmospheres
Consisting
of
Fuel Combus t ion Products and
Associated
Impurities, Her
Majesty's
Stationery
Office, ondon, 1968).
4.
Quets,
J. M . nd
Dresher,
W.
H.,
ournal
of
Materials ASTM)
,
4,
(1969), 83.
5.
e r gman , aul
an d
Kaufman, Murray,
hompson
Engineering
Laboratory,
General
Electric
River
Works,
W.
ynn ,
Massachusetts
personal
com-
municat ion) .
6.
Bornstein,
.
S.,
nited
Aircraft Research Laboratories, . Hartford,
Connec t i cu t (personal
communicat ion) .
7.
DeCrescente,
.
A .
nd
Bornstein,
N. S., nited Aircraft Research
Laboratories,
. Hartford, onnect icut To
be
published).
8. imons , . L.,
rowning,
. V., nd Liebhafsky, . A., Corrosion, ,
(1955),
05.
9.
Bornstein,
. . nd DeCrescente, M. A., Trans. Met . oc. AIME, 45,
(1969), 947.
10.
eybol t ,
A . U., Trans. Met . oc. AIME, 42, 1968), 955.
11 . Wasielewski, . . Materials and
Processes
Laboratory,
General
Electric Company,
chenec tady,
N.Y., Report no. 68AEG430.
8/18/2019 Hot corrosion in GT
24/83
14
HI.
MATERIAL
AND
COATING
DEVELOPMENT
Introduction
Although
there are
many
types
of hot
corrosion
which
are
potentially
of interest
to industrial
gas
turbine
producers,
h is
portion of th e report
will
only consider those means
that
might lead to th e development of improved sulfi-
dation and
oxidation-resistant
alloys for turbines
operating
on diesel quality or
better
fuels.
This
section will
discuss
th e
current
status of
corrosion-resistant
cobalt-, nickel-, nd iron-base
al loy
developments,
ttempt
to
predict
what
realistic
improvements can
be expected in th e near
and
distant future,
nd
sug-
gest areas
for potentially fruitful deve lopmen t
activity,
A Committee on Coat ings of th e NMAB has
made recommendations
regarding
coatings
for
superalloys. We are
in agreement
with
these recommen-
dations. The pertinent recommendations
of th e
Coat ings
Committee are
con-
tained
in
Appendix
I.
For
this reason,
coatings
have
not
been
g iven
detailed
consideration
in
this
report
except
as
required
to
place
he
need for
coat ing
deve lopment in
perspective
with
the
need
for al loy
development.
Current
Status
In
th e
early
1960's, t was
apparent that th e
then-existing advanced
alloys did no t
possess
dequate
suifidation resistance
to
permit long-time opera-
t ion in
environments conducive
to
this type of
attack.
Hence ,
he
gas urbine
industry
devoted
considerable
attention
to
coat ing
deve lopment .
Alumin ide coat-
ings
were
found
to
provide
good corrosion
protection.
However , s overhaul
lives
were
extended,
he existing coatings broke down after
several thousand
8/18/2019 Hot corrosion in GT
25/83
15
hours
and
did
not provide adequate protection. This was
especially
true of those
alloys
developed for maximum
strength capability,
hich
required
compositions
of lower
chromium
content.
It soon
became
apparent
that
more inherent
sulfidation
resistance
was
required
of
turbine
blade
and
nozzle
guide
vane alloys.
The
Naval Ships Systems
Command
established
a
program
in
1964
with
th e
fol lowing
objectives:
1)
evelop a turbine
blade
alloy combin ing th e ho t corrosion
resistance
of
Udimet
500
in
1 %
sulfur diesel
fuel
combustion
products
with the
strength
and
ductility of alloy
713C
at
1600
o
F
100
hr rupture
life
at43,000psi).
2)
evelop a cobalt-base turbine nozzle alloy with th e 1900
C
F
strength of WI-52 100 hr
rupture
life at 10,
000
psi)
but
with improved
hot-corrosion
resistance.
These targets were accepted
by
th e metals industry, because, f met,
th e
alloy
would
offer
useful
properties
for all types
of turbines. Soon
several
alloys appeared,
uch
as MAR-M-421 Mart in Metals), d ime t 710
Specia l
Metals), N-738 The Ihtemational Nickel Company) ,
nd
more
recently,
AK-
M-432 (Martin Metals). * These
met
or came close to meeting th e
above
targets
for blades.
They
are currently being
evaluated by
he
gas
turbine
industry and
one, MAR-M-421,
s being
used in
production.
Several
cobalt-base
alloys
de-
veloped
in th e Navy-General Electric program meet
the
targets for a cobalt-base
nozzle material.
Extensive preliminary
studies
conducted by
several
investigators
on
simple
all
» y systems
suggested
that
chromium,
nd,
o
a lesser
extent,
itanium
and
cobalt
were
th e only
elements
that
contributed
in
a
beneficial
way to
sulfidation
Compos i t ions of all
alloys
mentioned
in
this
report
are
contained
in
Table
1.
8/18/2019 Hot corrosion in GT
26/83
16
W
ffl
<
S
tr -
io
IM
Iß
in
o i
^ 3
Ü
OI
I
I
I
in
o
o o
o
Tf
00
00 C O
t>
in
CD
t
o
o
o
o
o
o
o
o
J
b
—
^
o
1
r—i
1
00
e d
H
in
t>
00
h
o
oi
00
T— t
—
00
1
3
1
C O
00
eg
o
C «
in
C O
T ̂
05
O
d
N
Oi
CO
Oi
C O
w w
. .
.
|̂ „•
-•
in
N
Oi
5
00
CO
I
Oi
CO
I
M
—
O
CD
CO
00
o
d
in
0i
CD
O
OJ
Oi Oi
o
in
o
o
u
in
CO
tN
t^
O J
t>
s
^
2
* —
in
T;
o
eg
o
O -1 H H -1
u
N
o
C
a
o
.2
e
o
ü
to
o
o
o
o'
(M
ai
s
a
o
8/18/2019 Hot corrosion in GT
27/83
17
resistance of nickel-base alloys. Some elements,
uch as
tantalum,
were
bene-
ficial
solely
with regard
to
oxidation resistance.
Some ,
such as
molybdenum
and
a lum inum,
were
considered
detrimental
to
sulfidation
resistance. Other alloying
elements
were
generally considered
neutral
or
ineffective.
These
effects
are
no t
necessarily
true
for
more
complex
alloys. Within
the
last
two
years it
has
been
found
that
a
proper
balance
of
elements,
particularly
the
refractory
metals,
can
substantially improve
sulfidation
resistance
even at relatively
low
chromium
levels. For
example, preliminary
tests ndicate
that
a new
alloy, N-792
(12.7
Cr),
possesses
nearly
the
corrosion resistance
of
Udimet
500
(17.5
C r)
bu t
with
rough ly 50
C
F
ncreased temperature capability based
on
rupture
strength)
over alloy 713 C. It
is
still under
evaluation
and has yet
to be
proven
for
turbine
applications.
The rare
earths, as
will be
ment ioned
later, m ay
also
be useful.
While th e alloys currently
be ing
developed
offer
improved
high-
temperature
capability,
hey do
not
exhibit
a
corresponding improvement
in
intermediate
temperature
1000 - 1500
o
F)
strength.
Similarly, lloys developed
for
improved
sulfidation resistance
of ten
lack
good
high-temperature
oxidation
resistance,
nd,
ence ,
ay
have
to
be
coated.
Clearly, herefore,
he
al loy
developer
must give
more
consideration
to
achieving
an
optimum balance of
properties.
Smal l evolutionary improvements can be expected
to
occur
in
nickel-
and
cobalt-base
alloys,
but
th e
lack of clearly defined
requirements
and markets
for future needs
tends
o
limit
th e support
for development work by
th e al loy
industry.
An
industry-government
effort
is
needed
to def ine requirements in
terms
of strength,
corrosion
resistance, nd economics. Present generalities
such
as,
We need
an
alloy that
can
be
used
uncoated
in
place
of
th e
coated
alloy
XXX
which
we are
now
using, s n insufficient
base
upon
which
to
establish
an
objective
al loy deve lopment program.
This
will
be
particularly important in th e
deve lopment
of
fiber
composites
or
dispersion-hardened alloys.
These
materials
8/18/2019 Hot corrosion in GT
28/83
18
will
not
have th e same balance of
properties
with
which designers
are
familiar.
What
will
constitute
a satisfactory
olut ion
to
th e
hot-corrosion
problem
shou ld
be stated
in terms
commensurate
with
prudent
economics. This can
be
done only
when th e different
t ypes
of
gas turbines are
considered
individually , ecause an
alloy
adequate
for
one ype of service m ay be completely
inadequate
for another.
Some specific requirements of aircraft
turbine
builders are contained in
References
3,
4,
nd
5.
Alternate Approaches
A s
indicated, prior
efforts
to
improve
th e
hot-corrosion xesistance
of
turbine blade an d nozzle
guide
vane alloys
have
been largely
in th e area of
alloy-
in g element modifications and/or coating application.
Alternate
approaches o
th e
deve lopment
of turbine
blade
an d
guide vane materials with better corrosion
resistance/temperature capability
migh t
be possible if
new strengthening
mechanisms
were
evolved, which
would
permit
th e
use of
base systems
and/or
al loying elements
with
greater inherent
hot-corrosion resistance.
A
number
of
approaches have
been considered for use
with
nickel-
cobalt-,
or iron-base
turbine
blades an d
guide vanes. For the
purpose
of
this
report, n improvement was defined as
a
system
having :
c
Sulfidation
an d
oxidation
resistance
Udimet
500
0
Temperature-strength
capability
for
blades
>
N-100
0
Temperature-strength
capability
for
vane s
s WI-52
In
order
to
guide
establishment
of
future deve lopmen t programs
o
that
efforts will
be
concentrated
in
th e most fruitful
areas,
n assessment
was
made of
th e capability and probability of each
approach. Capability
was
considered
as an
indication
of th e theoretical
possibilities that
an
approach
offers for achieving
Improvement,
whereas Probability is
realistic
but
empirical assessment
of
th e
chances
for realizing that
capability.
The
8/18/2019 Hot corrosion in GT
29/83
19
considerations
were
made
in
keeping
with
current
technology and
a
realistic
extrapolation
of
that
capability
into
th e
near
future, .
e.,
975.
Results
of
this
estimation are summarized
in
Tables
an d
3
nd
are discussed
in
th e fo l lowing
paragraphs.
Dispersion Harden ing
Dispersion
hardening
is
effective
in strengthening
alloys
at
high
tem-
perature > 1600
o
F)
but
contributes little to low temperature
8/18/2019 Hot corrosion in GT
30/83
20
eg
W
p q
<
O T
«
W
o
K
v̂
U
«
<
^
O
< ;
K
OH
ro
O
H
«
a
K
H
O
S
o
ft
H
O
K
S
>
O
w
Z,
Q
£
i-l
o
P 5
OH
K
<
s
§
P M
S
̂
O
H
HH
G
K
HJ
<
«
>
<
>H
W
n
o
£
Q
2
̂
< P Q
tH
W
H
Ä
t-4 HH
HJ
P Q
P 5
D
H
<
^
Q
Ü
W
Ä
H
HH
<
a
H
0
H
z
C O
W
W
P Q
H
C Q |
o
o
in
*->
o
u ;
o
d
r—t
o
Z
p
1—
Al
Al
(1 )
ü
C
•pH
C O
I—
•pH
*p-t
c d
(0
a
Ü cd
P ;
a
c
J3
o
•1-4
g
^
o a
2
c
5
o
cd
cr t
u
TS
a «
S S 3
5
O Q
e d
8/18/2019 Hot corrosion in GT
31/83
21
W
p a
<
ft
O
o
w
ft
o
H
. - 1
ft
c
4-i
c rt
-*-J
0 9
A
f-t
rt
&
ft