Protective Coatings For Turbine Blades

2002 ASM International. All Rights Reserved.
Protective Coatings for Turbine Blades (#06738G)
Protective Coatings
for Turbine Blades
Y. Tamarin
Materials Park, Ohio 440730002

www.asminternational.org

Copyright 2002
by
ASM International
All rights reserved
No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise, without the written permission of the
copyright owner.
First printing, September 2002
Great care is taken in the compilation and production of this book, but it should be made clear that NO
WARRANTIES, EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, ARE GIVEN IN CONNECTION
WITH THIS PUBLICATION. Although this information is believed to be accurate by ASM, ASM cannot
guarantee that favorable results will be obtained from the use of this publication alone. This publication is
intended for use by persons having technical skill, at their sole discretion and risk. Since the conditions of
product or material use are outside of ASMs control, ASM assumes no liability or obligation in connection
with any use of this information. No claim of any kind, whether as to products or information in this publication,
and whether or not based on negligence, shall be greater in amount than the purchase price of this product or
publication in respect of which damages are claimed. THE REMEDY HEREBY PROVIDED SHALL BE
THE EXCLUSIVE AND SOLE REMEDY OF BUYER, AND IN NO EVENT SHALL EITHER PARTY BE
LIABLE FOR SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES WHETHER OR NOT CAUSED
BY OR RESULTING FROM THE NEGLIGENCE OF SUCH PARTY. As with any material, evaluation of
the material under end-use conditions prior to specification is essential. Therefore, specific testing under actual
conditions is recommended.
Nothing contained in this book shall be construed as a grant of any right of manufacture, sale, use, or reproduction, in connection with any method, process, apparatus, product, composition, or system, whether or not
covered by letters patent, copyright, or trademark, and nothing contained in this book shall be construed as a
defense against any alleged infringement of letters patent, copyright, or trademark, or as a defense against
liability for such infringement.
Comments, criticisms, and suggestions are invited, and should be forwarded to ASM International.
Prepared under the direction of the ASM International Technical Book Committee (20012002), Charles A.
Parker, Chair
ASM International staff who worked on this project included Steve Lampman, Manager of Book Acquisitons;
Bonnie Sanders, Manager of Production; Carol Terman, Production Project Manager; and Scott Henry,
Assistant Director of Reference Publications.
Library of Congress Cataloging-in-Publication Data
Tamarin, Y.
Protective coatings for turbine blades / Y. Tamarin.
p. cm.
Includes bibliographical references and index.
1. Aircraft gas-turbinesBlades. 2. Protective coatings. 3.
Gas-turbinesMaterials. I. ASM International. II. Title.
TL709.5.B6 T36 2002
629.134353dc21
2002027690
ISBN: 0871707594
SAN: 2047586
ASM International
Materials Park, OH 440730002
Printed in the United States of America
Cover: Test blades (shiny) in a low-pressure turbine. Source: Advances in Turbine Materials, Design, and Manufacturing,
Proceedings of the Fourth International Charles Parsons Turbine Conference, The institute of Materials, 1997

Contents
Foreword....................................................................................................................... v
Preface......................................................................................................................... vi
Chapter 1: Introduction.................................................................................................... 1
Chapter 2: Choosing Optimum Coatings for Modern Aircraft Engine Turbine Blades................. 5
Conditions of Turbine Blade Operation ............................................................... 5
Requirements Imposed on Turbine Blade Coatings................................................ 8
Principles of Choosing Coatings for Aircraft Engine Turbine Blades ........................ 8
Causes of Coating Failures on Aircraft-Engine Turbine Blades.............................. 10
Chapter 3: Technological Processes for Deposition of Protective Coatings to Turbine Blades .... 25
Diffusion Coatings ........................................................................................ 25
Overlay Coatings .......................................................................................... 38
Chapter 4: Phase Composition and Structure of Coatings on Superalloys ............................... 55
Phase Composition and Structure of Diffusion Coatings....................................... 55
Phase Composition and Structure of Ni-Cr-Al, Ni-Co-Cr-Al, and Co-Ni-Cr-Al
Alloys and Overlay Coatings Made of Them...................................................... 69
Phase Composition and Structure of Overlay Coatings......................................... 71
Chapter 5: Phase and Structural Changes in Coatings during High-Temperature Tests .............. 79
Changes of Phase Composition and Structure in Diffusion Coatings....................... 80
Changes of Phase Composition and Structure in Overlay Coatings at HighTemperature Tests ......................................................................................... 87
Chapter 6: Turbine Blade Coating Protective Properties ...................................................... 97
Protective-Properties Evaluation Methods .......................................................... 97
Heat Resistance of Aluminides and Alloys for Overlay Coatings ..........................100
Heat Resistance of Coated Superalloys.............................................................106
Resistance of Overlay-Coated Alloys to Hot Corrosion .......................................109
Resistance of Coated Superalloys to Hot Corrosion ............................................113
Chapter 7: The Effect of Protective Coatings on the Mechanical Properties of Superalloys .......119
Thermal Expansion Coefficients and Elasticity Modulus of Coating Alloys ............120
Mechanical Properties of Coating Alloys ..........................................................124
Thermal Stresses in Superalloy Coatings ..........................................................128
Effect of Coatings on High-Temperature Strength of Superalloys..........................134
iii

iv / Protective Coatings for Turbine Blades
Effect of Coatings on Fatigue Strength of Superalloys ........................................138

Effect of Coatings on Thermal Fatigue of Superalloys ........................................142
Coating Effect on Thermomechanical Fatigue of Superalloys ...............................150
Chapter 8: Electron Beam Thermal Barrier Coatings..........................................................161
Ceramics for EB Evaporation .........................................................................162
Main Features of TBC Deposition Technique ....................................................165
Thermophysical Properties of Condensed Ceramics ............................................175
TBC Ceramic Layer Durability .......................................................................180
Chapter 9: Some Principles of Strength Designing for Turbine-Blade Protective Coatings ........195
Calculation of Stresses and Strains in Coatings..................................................195
Thermal Barrier Coatings ..............................................................................203
Appendix....................................................................................................................211
Index .........................................................................................................................217
iv
ASM International is the society for materials
engineers and scientists, a worldwide network

dedicated to advancing industry, technology, and
applications of metals and materials.
ASM International, Materials Park, Ohio, USA
This publication is copyright ASM International. All rights reserved.
Publication title
Product code
Protective Coatings for Turbine Blades
06738G
To order products from ASM International:

Online Visit www.asminternational.org/bookstore
Telephone 1-800-336-5152 (US) or 1-440-338-5151 (Outside US)
Fax 1-440-338-4634
Mail
Customer Service, ASM International

9639 Kinsman Rd, Materials Park, Ohio 44073-0002, USA
Email CustomerService@asminternational.org
American Technical Publishers Ltd.

27-29 Knowl Piece, Wilbury Way, Hitchin Hertfordshire SG4 0SX,
In Europe United Kingdom
Telephone: 01462 437933 (account holders), 01462 431525 (credit card)
www.ameritech.co.uk
Neutrino Inc.
In Japan Takahashi Bldg., 44-3 Fuda 1-chome, Chofu-Shi, Tokyo 182 Japan
Telephone: 81 (0) 424 84 5550
Terms of Use. This publication is being made available in PDF format as a benefit to members and
customers of ASM International. You may download and print a copy of this publication for your
personal use only. Other use and distribution is prohibited without the express written permission of
ASM International.
No warranties, express or implied, including, without limitation, warranties of merchantability or
fitness for a particular purpose, are given in connection with this publication. Although this
information is believed to be accurate by ASM, ASM cannot guarantee that favorable results will be
obtained from the use of this publication alone. This publication is intended for use by persons having
technical skill, at their sole discretion and risk. Since the conditions of product or material use are
outside of ASM's control, ASM assumes no liability or obligation in connection with any use of this
information. As with any material, evaluation of the material under end-use conditions prior to
specification is essential. Therefore, specific testing under actual conditions is recommended.
Nothing contained in this publication shall be construed as a grant of any right of manufacture, sale,
use, or reproduction, in connection with any method, process, apparatus, product, composition, or
system, whether or not covered by letters patent, copyright, or trademark, and nothing contained in this
publication shall be construed as a defense against any alleged infringement of letters patent,
copyright, or trademark, or as a defense against liability for such infringement.

Y. Tamarin, p1-3
DOI: 10.1361/pctb2002p001
Copyright 2002 ASM International

All rights reserved.
CHAPTER 1
Introduction
THE DAMAGE TO blade surfaces caused by
oxidation and hot corrosion results in considerable deterioration of the mechanical properties
of blades and shortens their service lives. That
is why protection of turbine blade surfaces
against damage has been a common problem.
Solving this problem is critical for designing
cost-effective and reliable aircraft engines and
stationary gas turbine units. The idea to apply a
layer with protective properties to the surface of
a nickel superalloy was first practiced in the
1960s and found a wide application area. Since
then, aircraft engine turbine blades with protective coatings have been used.
In the 1960s, extensive research into the properties of various coatings demonstrated that diffusion aluminide coatings had the best protective properties on turbine blades made of nickel
superalloys. Owing to their properties, such as
high oxidation resistance, high stability when
deposited on nickel superalloys, and a favorable
combination of physical and mechanical properties, the diffusion aluminide coatings have
been the predominant type of aircraft engine turbine blade coatings used for several decades.
The simplicity of the technique used in forming
diffusion coatings on the blade surfaces in the
powder mixtures containing free or bonded aluminum and aluminum halides contributed to
their most common use. Translation of diffusion
aluminide coatings in a commercial practice and
the work on improvement of their protective
properties and further development of their deposition techniques started concurrently.
This further development was the result of the
need for improved properties of protective coatings in the media containing sulfur compounds.
Diffusion aluminide coatings with the increased
chromium content in their outer layer (chromizing-aluminizing method) had found application in the protection of aircraft engine blades
suffering from hot corrosion. Alloying aluminide coatings with silicon (silicification in combination with aluminizing) turned out to be one
more technique that allowed the improvement of
coating resistance to hot corrosion and oxidation. Since the 1970s, both modifications of the
coatings have been used for protection of aircraft
engine turbine blades.
The studies aimed at modifying diffusion aluminide coatings are ongoing. Some of the coatings designed (e.g., Pt-Al) are used widely in
aircraft engines and stationary gas turbine units.
Designing new-generation engines with increased inlet gas temperatures resulted in the
shortening of the surface lives of the blades protected with diffusion coatings. Low ductility of
diffusion coatings and high stresses generated in
cooled blades cause the formation of numerous
thermal fatigue cracks in the outer layer of diffusion coatings on the blade surfaces. In updating diffusion coatings, designers seek to improve primarily the resistance of the coatings to
thermal fatigue. The problem of improving the
coating property in question arises from the limited technological feasibility of changing its
composition, structure, and consequently, its
physical and mechanical properties.
The studies aimed at increasing ductility of
diffusion aluminide coatings resulted in the
gradual decrease of the aluminum contents of
the coatings from 34 to 36%, typical of the coatings used in the 1960s, down to 20 to 24%, typical of the coatings used since the 1970s. This
trend, however, resulted in a decrease in oxidation resistance.
High labor consumption and expenditure of
energy on the formation of diffusion aluminide
coatings on turbine blades in powder mixtures
and the need for the use of halides in the technological process stimulate the seeking of alternative techniques for forming diffusion coatings
2 / Protective Coatings for Turbine Blades
on turbine blades. Coating deposition from

slurry allowed considerable improvement of the
process conditions under which the coatings
were formed on the turbine blades. The method
has experienced a wide application area in aircraft engine manufacturing. Further technique
improvements resulted in the development of
various gas aluminizing methods. These methods allowed less employment of powder
mixtures and higher efficiency of gas transfer of
aluminum to blade surfaces. By now, the techniques of blade aluminizing that eliminate completely the need for covering the blades with
powder mixtures have been developed and are
enjoying application. In this case, aluminum
transfer to the blade surface occurs by circulation of gaseous aluminum halides formed in a
special reaction zone of the unit.
A further advancement in the practice of diffusion coating deposition was the use of electricarc evaporation of aluminum alloys and their
transfer to the blade surface in the form of
plasma. In this case, the need for the use of halides and powder components is eliminated
completely, and it opens up the possibility for
the accurate control of diffusion coating thickness and its location on the blade surface. This
technique is used for forming the diffusion coatings modified with silicon, yttrium, and other
elements on turbine blades.
At the same time, it is noteworthy that all
technological processes of diffusion coating deposition used in turbine blades production form
the coatings with similar protective properties.
As a rule, the choice of a certain technique for
coating application depends on requirements for
its quality, technological equipment available,
and personnel skills.
Although diffusion coatings have been enjoying wide application, they do have a limitation:
They do not allow deliberate changes of their
properties to meet the requirements imposed by
specific blade service conditions. This restriction
exists because all diffusion coatings always are
based on Ni-Al and Ni3Al aluminides with typical combinations of physical-chemical and mechanical properties. This drawback follows from
the principle of diffusion coating formation due
to interaction of aluminum transferred to the surface from an aluminizing medium and the components of a nickel superalloy. It cannot be overcome by modernization of coating compositions
and techniques of their deposition. That is why,
as a following step in developing protective
coatings for turbine blades, it was quite natural
to apply some new principles of their formation,

such as deposition of the alloys, which compositions met complex requirements to protective
coatings for turbine blades most adequately.
Design of electron-beam, electric-arc, plasma
methods, and equipment for their realization
triggered the change to the new principles of development and application of protective coatings
to turbine blades. Since the 1980s, some variants
of overlay coatings have found application for
the protection of turbine blades. Unlike diffusion
coatings, overlay coatings admit the control of
their composition, thickness, and combination of
physical-chemical and mechanical properties
within wide ranges. It allows the choice of overlay coatings, which have the properties optimal
for certain engine types and their service conditions.
New principles of coating deposition opened
up new possibilities for purposive improvement
of coating compositions and variation of their
properties. However, a wide range of possible
variants of coatings compositions and techniques for their deposition presents the problem
of making an adequate choice in coatings for
certain engine types.
For this reason, development of calculation
methods for determining stresses and strains in
the coatings on turbine blades and thus their service lives under conditions of high thermal
stresses further advanced the work on overlay
coatings. To calculate the values in question, it
is necessary to have certain information on
chemical and mechanical properties of the coatings within a wide range of probable compositions.
Basically, a new stage of progress in the area
of turbine blades protection has been achieved
with designing thermal barrier coatings. The
outer layer is formed of stabilized zirconium oxide-base ceramics. This coating type allows not
only protection of a coating surface against corrosion damage as all previously designed coatings did, but it was also the first to allow protection of the structure of a superalloy in cooled
blades against changes caused by exposure to
high temperatures and, hence, against its softening.
An apparently simple idea of designing a ceramic layer on the blade surface called for extensive research and development for its realization that involved designing techniques for
production of special ceramics, their evaporation
with electron beam, special test methods, and
calculations of service lives for thermal barrier
coatings on turbine blades.
Introduction / 3
As is well known, metal structural materials

properties depend on their chemical compositions and structures. Control of metal materials
compositions and structures makes it possible to
produce a great variety of structural alloys with
required properties. Similarly, protective coatings properties depend on their chemical compositions and structures. And yet, the approaches worked out for structural materials
hardly remain valid for coatings because the relationships connecting chemical compositions,
structures, and properties in thin coating layers
do not reveal themselves clearly. Besides that, a
coating-superalloy system is a dynamic one.
The chemical and phase compositions of a coating, as well as its structure, vary continuously
under blade service conditions at elevated temperatures. The initial coating properties on the
blade and those it has at any moment of its
operation in the turbine may be quite different.
Since the first use of protective coatings on
the turbine blades of different designations, hundreds of papers that deal with their properties
have been published. However, to create a database covering the papers published and storing
the information required for the proper choice
of coatings and for calculations of the service
lives of coated blades, one should overcome the
problems of coatings identification and lack of
data on chemical and phase compositions and
structures of the coatings studied. No unified
standard methods for research into coatings
properties are available. In most cases it makes
impossible comparison and classification of the
test results obtained.
Special databases should be created to solve
the problems of description and systematization
of the steadily increasing number of protective
coatings for turbines. The databases should provide systematized, accumulated information on
the coatings properties studied and on the experience of their use in practice. Systematization
of the results presented by laboratories and
firms, which deal with the problems of turbine
blades protection in the form of databases, may
allow further simulation of changes in coatings

properties, as well as formalization of composition-structure-property relationships. Besides
that, the characteristics required for coated blade
service life calculations can be derived.
This book presents knowledge accumulated
by the author over many years of his work on
the research and development of diffusion and
overlay protective coatings for aircraft engine
turbine blades. The book includes detailed descriptions of some technological processes for
diffusion and overlay coatings deposition, together with compositions of the coatings used
for turbine blades protection, coatings structures
after their deposition on the blade surfaces, and
their changes during high-temperature testing.
The results of a number of tests for coatings and
samples of the alloys used for coating deposition
are discussed also. Some chapters deal with thermal barrier coatings, technological features of
their deposition, and test results, as well as with
the principles of calculations of stresses and
strains in the coatings on blades.
The composition-property regression equations derived on the basis of research into aluminide alloys of Ni-Cr-Al, Ni-Co-Cr-Al, and
Co-Ni-Cr-Al systems used for deposition of
overlay coatings are given in the Appendix.
They allow simulation of aluminide alloys properties. These simulations cover the compositions
of both as-deposited protective coatings and
those after their long-term operation on turbine
blades. In some cases the simulations obtained
may be used for assessment and prediction of
coatings properties. Many coatings dealt with in
this book are used currently for turbine blades
protection.
The information on technological processes
and coatings properties presented in the book
may be useful to the specialists who work in the
field of turbine blades protection, to the scientists interested in research into materials behavior at hot corrosion, and to the students whose
specialty is protection against hot corrosion.

Y. Tamarin, p5-23
DOI: 10.1361/pctb2002p005

CHAPTER 2
Choosing Optimum Coatings for

Modern Aircraft Engine Turbine
Blades
COATED TURBINE BLADES boast service
lives of 2 to 5 times those of their uncoated
counterparts. However, coating protection effects are manifested only if the coating has been
chosen correctly, both in terms of blade design
and engine running conditions.
Conditions of Turbine Blade Operation

The main parameters determining the operating conditions of turbine blade surfaces of aircraft gas-turbine engines are the gas temperature
at the turbine inlet; the pressure, speed, and composition of the gas flow; the stresses due to the
effect of the centrifugal and gas dynamic forces;
and the irregularity of the temperature field.
The gas temperature at a turbine entry is the
most important parameter determining the specific power, specific weight, and efficiency of an
aircraft engine. During the last 20 years, the
mean mass temperature of the gas at a turbine
entry increased from 1200 to 1300 K to 1700 to
1800 K. Together with the turbine temperature
growth, the pressure ratio of the compressor has
grown significantly as well; it has reached 20 to
30. The combined growth of the temperature and
the pressure ratio results in significantly more
intensive heat flows in turbines; they make 1 to
2 MW/m2 for modern engines.
One of the main properties of aircraft gas-turbine engines is a significant number of variable
conditions when the level of temperatures and
stresses differs significantly from those measured in stationary conditions. Statistical analysis shows that for aircraft engines, about 20% of
the flaws developed at operation are flaws resulting from the cyclic temperature variations
(Ref 1). Figure 2.1 shows the operating diagram
of an airliner for a typical medium-range flight.
Every 1.5 h flight has four corresponding cycles of condition variation; that is, 10,000 h of
engine life correspond to 27,000 cycles of
variation in flight conditions. Every time an engine reaches takeoff power, the nonuniformity
of the temperature field in the turbine blades increases significantly. Running the engine under
cruising conditions does not eliminate the effect
of nonuniform temperature and stress fields on
the blades. The number of varied powers for
warplane engines, especially fighters, is significantly greater than that for passenger aircraft engines.
The surface layers of aircraft-engine turbine
blades during operation contact the combustion
products of the aviation fuel. This is hydrocarbon-based (98 to 99%), with the remaining 1 to
2% including sulfur, nitrogen, oxygen compounds, and trace metals as well as their compounds. A widespread impurity in fuels is vanadium oxide (V2O5), whose content depends on
the deposit of the source oil. The fuel sulfur has
the form of its various compounds, mercaptanes,
sulfides, disulfides, and thiophenes. The total
sulfur content of an aviation fuel is in the range
of 0.05 to 0.25%.
Aviation fuels usually get impurities from the
environment and from the production process,
resulting from corrosion and rubber decomposition and so forth. Typical impurities of jet fuels
are as follows: C, H, S, N, Fe, Si, Sn, Ca, Mg,
Pb, Na, Ba, Ti, Ni, Al, Cu, Zn, Mn, and O.
The quantity and the composition of the impurities depend on local operation conditions, on
the fulfillment of requirements to fuel controls
and purification, and on aircraft fueling conditions (Ref 2 to 4).
In a combustion chamber, the process of fuel
combustion is realized at the temperature of
2273 to 2473 K. Hydrocarbon combustion is a
process of their oxidation by the oxygen in air,
resulting in carbon dioxide and water, which
may contain some carbon monoxide (CO), H2,
methane (CH4), and hard carbon particles. The
complete combustion of 1 kg of fuel requires
15 kg of air. The total factor of air excess in
combustion chambers compared to the amount
necessary for complete fuel combustion is 3.5 to
4.5; it may change in dependence of the flight
conditions. The specific fuel consumption for
modern engines provides 0.75 to 0.90 kg/kgf of
thrust.
The high-temperature gas flow at the turbine
inlet contains mainly oxygen, carbon dioxide,
water vapors, and some amounts of CO, H2,
CH4, hard carbon particles, compounds of sulfur, and metals present in fuels. Various chemical compounds may come into the gas flow
from the environment. These are sea salts from
flight over sea regions and various compounds
that are exhausts of industrial enterprise. As a
rule, they contain sulfur, salts of alkaline metals,
Fig. 2.1
Operating diagram of an airliner engine. Source: Ref 1
and other compounds, depending on the industry

situated in the flight area.
The temperature pattern of the gas flow at the
turbine inlet is highly inhomogeneous; the inhomogeneity is usually caused by the design of
a combustion chamber, by the design and layout
of burners, and by other factors determining the
flow-gas dynamics and the perfection of the fuel
combustion process. The measurements of the
gas temperature before the nozzle guide vanes
show that a difference between the extreme gas
temperatures at a nozzle block inlet may be 400
to 500 C.
The nozzle vanes of a high-pressure (HP) turbine situated just behind the combustion chamber are the most heated elements of an engine.
The temperature of the cooled nozzle vanes is
200 to 300 C lower than that of the gas flow,
but the thermal inhomogeneity of the gas at a
turbine inlet results in some vanes working at
temperatures differing significantly from the average ones. The temperature measurement of
vanes has shown, that the thermal pattern of the
vanes leading edges in the middle section differs
by 200 to 250 C. The thermal inhomogeneity
of the nozzle vanes results in significantly differing rates of flaw development in vanes situated in different parts of an engine relative to the
combustion chamber baffles and burners.
The gas temperature pattern inhomogeneity is
observed not only by the hot-duct circumfer-
Choosing Optimum Coatings for Modern Aircraft Engine Turbine Blades / 7
ence, but by height, too. The characteristic temperature curve by the vane height shows an area
of maximal temperature in the vane mid-part and
lower temperatures at end areas. Nevertheless,
other temperature patterns may also be observed. The greatest difference between the middle and peripheral areas of the vanes is 100 to
150 C. The temperature gradient value grows
at the transition to the gas high-temperature conditions.
The vane and blade thermal state depends on
the engine power; it changes from 450 to 500 C
to 1000 to 1100 C within a few seconds at the
engine transition from idle power to take-off
conditions. The temperature inhomogeneity of
the vane and blade surfaces increases significantly under the transient conditions of startup,
acceleration, and run. The maximal thermal
stresses of the vanes, especially at their surfaces,
within the coating areas are observed at these
very conditions. The computed data that characterize variations in cooled-vane temperatures
at the engine startup show that the difference in
the temperatures between the vane back and
trailing edges may reach 500 to 600 C. The
computation results on the thermal stresses corresponding to the above temperature patterns
show that their values attain 400 to 500 MPa
(Ref 5).
The analysis of the strains in a vane caused
by such thermal stresses reveals that some areas
of a vane are in a state of plastic flow. The level
of thermal stresses and strains in a vane surface
layer may exceed significantly the mean level of
the vane stresses due to the presence of the temperature gradients along the cooled-vane wall
section and due to different physical and mechanical properties of the protective coating and
superalloy. The accumulation of thermal variations at operation results in an accumulation of
residual strains, which, in their turn, result in
thermal fatigue cracks on the vane surfaces, first
on the leading and trailing edges. These cracks
are the most widespread flaws of vanes.
Analysis of the vane operating conditions
shows that the combined effect of high temperatures, gas dynamic forces, and thermal stresses
in vanes causes cracks due to thermal fatigue,
burnouts, and warping. The higher temperature
of the engine gas resulted in higher temperature
gradients in the vane surface layers, and the use
of cast nozzle blocks caused higher stiffness of
restraint of individual vanes. As a result, the
thermal fatigue cracks have become the most
widespread flaw of the vanes during tests and
operation, despite the use of efficient film methods for cooling.

One of the critical elements of an aircraft gasturbine engine is its turbine blade. It is its thermal state that determines the maximal admissible gas temperature in the turbine and the
reliability and life of the turbine and the engine
in whole. The blades are affected by centrifugal
and gas dynamic forces causing tension, bend,
and torsion of the blades. The high temperatures and stresses, the unstable conditions of
heating and loading, and the possibility of resonance vibrations make a blade one of the most
complicated elements of an engine.
In contrast to the nozzle vanes, the gas temperature field before the turbine blades is more
uniform. The high rotation speed of a blade results in a more homogeneous thermal state of
individual blades, if compared with the vanes.
The temperature gradient by the blade surface
depends on the gas-temperature curve by height
and by the specific operation of the blade cooling system. The highest temperature area is, as
a rule, in the upper third of a blade airfoil portion; this is explained by the need to place it
within the area with the lowest centrifugal loads.
The temperature distribution by the turbine second stage blades is even more homogeneous,
though the non-cooled, second-stage blades of
some engines have higher average temperatures
than the ones of the first stage.
The limitations due to high-temperature
strength (heat resistance) of the nickel-based superalloys and the necessity to improve the parameters of modern aircraft engines pushed the
design of sophisticated and efficient cooling systems; they permitted lower average temperatures
of a blade wall by 300 to 400 C, compared to
the gas temperature. The higher temperature and
compression rates in modern engines result in
greater temperature irregularities in the blades.
When the gas temperature increases by 100 C,
the temperature gradient on the surface of a
cooled blade increases by 40 to 50 C; the compression rate effect is similar. The temperature
pattern of the modern cooled blades is complex
enough: the areas of high temperature, 1000 to
1100 C, whose size is often 10 to 30 mm2, are
alternated by the average (900 to 950 C) and
low (750 to 800 C) temperature areas. The temperature gradient by the surface and wall section
of the blades causes high thermal stresses, reaching maximal magnitudes at engine operation under non-cruise conditions (Ref 6 to 8).
A blade is affected by tensile stresses caused
by centrifugal forces. Their levels are not usu-
ally higher than 200 to 250 MPa. A blade is also

affected by the bending stresses caused by gas
dynamic forces. These stresses are determined
by the difference in pressures at the concave
(pressure) and convex (back) surfaces of a blade.
The combined static stresses caused by the external forces affecting turbine blades are distributed irregularly along a blade. The highest tensions are observed near the blade root, the lowest
ones exist at the end of a blade airfoil portion.
The section with the least static strength security
margin or with the maximal exposure to flaws
due to the long-term static load is usually between the maximal stress and maximal temperature sections (Ref 9 to 11).
The vibration loads are also critical for the
blades; these loads result from the inhomogeneity of the gas flow hitting the turbine wheel
rim. The low-frequency vibrations are caused by
the flow thermal inhomogeneity due to the presence of burners and pillars; the high-frequency
ones result from the finite number of nozzle
vanes. The resonance vibrations are the most
widespread cause of the surface fatigue cracks
and rupture of the turbine blades. The level of
variable stresses in the blades reaches 80 to 100
MPa (Ref 12).
Requirements Imposed on Turbine

Blade Coatings
Any coating deposited on turbine blades (or
vanes) must offer protection within a specified
period of service life against destructive attacks
of high-temperature corrosion and erosion,
when the said components are exposed to a flow
of fuel combustion products containing aggressive ingredients and solid particulate matter.
With this in mind, the coating must meet the
following requirements:
It must withstand hot corrosion, oxidation,

and erosion when placed into a flow of gas
whose parameters are similar to those of turbine gases.
It must safely withstand the static and alternate stresses applied to the blade surface; to
this end the coating must have the requisite
combination of strength and ductility.
It must show good stability and not be destroyed by interaction with the substrate.
It must not degrade the blade material mechanical properties.
Even after a long service life, coated blades

and vanes must have better mechanical properties than their uncoated counterparts due to protecting the blade surfaces from damage and softening.
In addition to the requirements to metal coatings listed previously, the ceramic coatings
(thermal barrier coatings, or TBCs) must ensure
lower average wall temperatures in cooled
blades and protect superalloys against softening.
TBCs must level off the temperature over the
blade surface and reduce thermal stresses during
engine transient running.
The techniques involved in depositing protective coatings must guarantee that the coatings
have the required composition, thickness, and
structure, and, as such, the requisite set of physical-chemical and physical-mechanical characteristics. Of utmost importance is the need to
maintain these characteristics depositing protective coatings on blades under the commercial
conditions.
Three technological principles are applied to
form coatings on the surfaces of aircraft engine
turbine blades:
Chemical-thermal treatment (aluminizing) of

blade surfaces in the media, containing aluminum or aluminum with additives of other
elements (such as chromium, silicon, yttrium,
etc.). A diffusion aluminide coating is built
up as aluminum or its compounds interact
with the superalloy surfaces. At present, diffusion aluminide coatings are used to protect
80 to 90% of all aircraft turbine blades.
Depositing the overlay coatings by evaporating special alloys under vacuum and condensing vapor or plasma with coating components on blade surfaces
Forming coatings from powders by arc
plasma spraying
Principles of Choosing Coatings for

Aircraft Engine Turbine Blades
The requirement to attain highly protective
coatings (i.e., adequate resistance to hot corrosion and oxidation) conflicts with the demand to
preserve high mechanical properties in the
coating-superalloy system.
Until recently, engine tests have remained the
main method of examining the properties of and
choosing optimized coatings for blades. How-
ever, the high cost of systematic studies of coating protectiveness using this approach has fueled the desire to seek shorter equivalent trials.
Under these conditions, laboratory research into
physical-chemical and physical-mechanical
coating properties and their relation to blade service life facilitates significantly the selection of
optimal chemical compositions of the coatings,
reduces the number of their field tests, and
promises high cost efficiency of the study.
Shown in Fig. 2.2 are the key parameters to
consider in selection of coatings for turbine
blades. The starting points to be taken into account when choosing a coating of optimum life
expectancy include:
Experience in applying protective coatings on

blades running under similar conditions
Analysis of causes of coating damage on
blades after long-time tests or service operation
Presumable alterations in running conditions
of newly designed blades (changed parameters of gas flow, values of thermal stresses,
running conditions, etc.)
Results of research into physical-mechanical
properties of coatings and their effects on the
main properties of the superalloy from which
the blades are made (resistance to thermal fatigue, endurance, high-temperature strength)
Of prime significance for the protectiveness
of a coating is its resistance to hot corrosion and
oxidation in the temperature range of 600 to
Fig. 2.2
Scheme of the choice of a coating for aircraft engine

turbine blades
1200 C, that is typical of the running conditions

of modern-engine turbine blades. The mechanism of coating damage in this range can vary
depending on the temperature and composition
of gas in turbine, blade surface temperature, and
design. Coatings are chosen from the laboratory
results of coating tests for heat resistance and
hot corrosion.
Actual engine running conditions can result
in blade surfaces being exposed to temperatures
either below or higher than 900 C. Moreover,
the demands on coating life expectancy at temperatures above 900 C may conflict with life
expectancy requirements imposed on blades
working in corrosive media at temperatures below 900 C. When choosing coatings, due attention should be paid to possible application of
engines in coastal regions or in areas with air
heavily laden with industrial impurities. Should
the blade damage be mostly caused by hot corrosion, then coatings with high resistance to this
type of corrosion should be preferred. Alternatively, multilayer coatings can be used, given
that individual layers of such coatings offer high
resistance to oxidation and hot corrosion.
Coating thickness is undoubtedly a very important characteristic of its protectiveness; coating resistance to hot corrosion and oxidation is
directly proportional to the thickness of coating
used. It should be borne in mind, however, that
the thickness of coatings applied to turbine
blades are restricted, by both the probable decrease in mechanical properties of the blades
with thick coatings and by technological problems involved in depositing thick coatings. Design problems must also be taken into account
(e.g., the reduction of the open flow area in the
turbine).
As overlay coatings add to the mass of blades
they are applied to, hence, the centrifugal loads
on the blades and turbine disk grow. This must
be counted when thick coatings are used. Variable thickness coatings are most advantageous;
a thicker layer is deposited on the area that is
most likely to be attacked by corrosion. Other
areas are thinner-coated.
No correlation between coating thickness and
thermal fatigue resistance has been noted in
overlay coatings up to a thickness of 120 lm. A
positive correlation seems more likely in thicker
coatings prone to formation of various defects,
which, in turn, are conducive to the emergence
of thermal fatigue cracks. When using coatings
thicker than 120 lm, the blade fatigue limit must
be taken into consideration.
Coating composition and thickness can be determined once the required service life has been
specified. Once the coating composition has
been determined, one must then consider how
long the coating of given composition will stay
steadily on the blade surface after long exposure.
Coating protectiveness is limited by diffusion
into the alloy of those elements that control resistance to oxidation and hot corrosion. The diffusion progress can cause the decrease in aluminum, chromium, and cobalt contents of the
coating. It can call for the use of special methods
for retarding diffusion processes by introducing
barrier layers based on phases that are stable
with respect to both the coating and the alloy.
On the other hand, as coating elements diffuse
into the alloy, a zone of poorer heat resistance
is formed under the coating. It is a factor that,
in thin-walled blades, leads to reduction of the
margin of safety. No coating lacking the requisite stability can be used for turbine-blade protection.
The life of a coated turbine blade depends on
its mechanical characteristics and the stresses on
the surface. Thermal fatigue cracks have become
the commonest cause of service life reduction
and discarding of blades. Calculation of the
stresses and strains in turbine blade construction
is mandatory in coating selection for modern turbine engines.
Engine testing is the final step in coating selection and estimating useful life on turbine
blades. Defects in a coating on blade surfaces
may be caused by hot corrosion and oxidation,
low resistance of blade construction to thermal
Table 2.1 Alloy composition for overlay coatings deposited on turbine blades
Chemical composition, wt%
Alloy
Ni
Co
Fe
Cr
Al
Co20Cr12AlY
Base 1822 1113
0.20.6
Co23Cr12AlY
Base 2224 1113
0.20.6
Co26Cr9AlY
Base 2528 810
0.30.6
Co25Cr4AlY
Base
2426 3.55
0.30.6
Co30Cr6AlY
02 Base
2832 57
0.30.6
Co28Cr10FeY
Base 812 2630 0.1
Co22Ni23Cr12AlY 1825 Base 2224 1113
0.30.5
Co8Ni23Cr12AlY
610 Base 2224 1113
0.20.6
Ni20Co20Cr12AlY(a) Base 1822 1822 1113
0.20.6
Ni8Co20Cr12AlY(a) Base 610 1822 1113
0.20.6
Ni20Cr12AlY(a)
Base
1822 1113.5 0.20.6
Ni20Cr5AlY
Base
1822 46
0.20.6
Ni36Cr5AlY
Base 3538 46
0.20.5
Fe25Cr5AlY
Base 2326 46
0.20.5
Fe22Ni24Cr5AlY
2025 Base 2325 46
0.20.5
(a) Composition used in aviation. Source: Ref 13, 14
fatigue, poor high-temperature strength and endurance of the substrate, or by a combination of

these factors. In order to assess life expectancy
of coated blades in long-term tests, their temperature schedule must be as close to the actual
engine running conditions as possible. The matter is that in the course of engine-equivalent
quick tests, where the turbine blades are running
at higher than actual service temperatures, the
processes occurring on the blade surfaces, as
well as between the coating and superalloy, may
significantly differ from those occurring at lower
temperatures. Elevation of temperatures during
equivalent quick tests may be misleading due to
reduced condensation of aggressive components
from the gas flow on the blade surfaces and suggest longer-than-actual service lives. Alternatively, such tests may degrade coating resistance
owing to sharp acceleration of diffusion processes and rapid destruction of protective oxides.
Analyzing the results of engine tests pinpoints
the coating-failure mechanism and enables, if
necessary, appropriate modifications to the composition, thickness, and/or design of multilayer
coatings.
Chemical compositions of some alloys used
for coating deposition on turbine blades are presented in Table 2.1 (Ref 13, 14). It is noteworthy
that the coating composition depends on the currently used technological process of its deposition. It can differ from that of the initial alloy.
In this book, the coatings are marked with the
grades of the alloys used for their deposition.
Otherwise, when the coatings are marked according to their chemical compositions, the respective notes are supplied.
Chemical compositions of the materials used
for applying diffusion coating to aircraft turbine
blades are presented in Table 2.2 (Ref 15 to 21).
Chemical compositions of the superalloys used
for research into coating properties are presented
in Table 2.3 (Ref 22, 23).
Causes of Coating Failures on AircraftEngine Turbine Blades

The investigation of coating damage, detected
on the blades of engines after a long time in
service, is an important stage in the choice of an
optimum coating. Careful investigation of the
nature of damage and the sources of origin will
enable the designer to choose a coating that has
Table 2.2
Chemical compositions of the materials used for forming diffusion coatings
Method of coating
Grade
Chemical composition of alloy (mixture), % mass
Pack calorizing(a)
Pack calorizing(a)
Pack calorizing(a)
Pack calorizing under vacuum(b)
Slurry aluminizing(c)
Al
Al
Al
AlCr
Al
AlSi
Al
AlSiY
Granules of alloy of Fe(3550%)Al; 2% NH4Cl

Granules of alloy of Fe(7075%)Al; 2% NH4Cl
Powder 98%Al, 2% NH4Cl
Powder (78%Al; 3540%Cr; Al2O3); 0.30.4% NH4Cl-balance
Powder of (100% Al) in colloxiline solution of amilacetate
Powder of (Al550%Si) in colloxiline solution of amilacetate
Alloy: 100% Al
Alloy: Al base; 4.015.0% Si; 0.62.0% Y
Electric arc vacuum deposition(d)
(a) Ref 15, 16. (b) Ref 17, 18. (c) Ref 16, 19. (d) Ref 20, 21
Table 2.3
Superalloys used for aircraft turbine blades

Chemical composition(b), wt%
Alloy (a)
Cr
Ti
Mo
Re
Ta
Al
Co
Hf
Nb
Zr
Ni10CrWMoCo
JS6K
JS6U
VJL12U(c)
JS6F(c)
JS30
JS26(c)
JS32
MAR-M-200
MAR-M-002
CMSX-4
10.0
11.0
8.5
9.5
5.5
7.0
5.0
5.0
9.0
9.0
6.5
3.0
2.5
4.5
1.0
1.9
1.0
1.9
1.7
1.0
10.0
4.0
1.8
3.0
0.9
0.7
1.1
1.0
0.6
5.0
5.0
10.0
1.4
12.0
11.8
11.7
8.3
12.5
11.0
6.0
4.0
3.0
4.0
2.5
6.5
4.5
5.5
5.5
5.5
5.3
5.2
5.8
6.0
4.7
5.5
5.6
5.0
4.3
9.5
9.5
9.4
8.5
9.0
9.0
10.0
10.0
9.0
0.9
0.8
1.5
0.1
1.0
0.8
1.6
0.9
1.6
1.5
1.8
0.015
0.015
0.015
0.015
0.015
0.05
0.05
0.05
0.05
0.05
0.10
0.15
0.15
0.15
0.11
0.15
0.15
0.15
0.14
0.15
(a) JS alloys are designed in All-Russian Institute of Aviation Materials (VIAM). (b) Average values. (c) These alloys contain 1.0% V. Source: Ref 22, 23
Depth
of
high-temperature
corrosion
in
Ni20Cr12AlY overlay coating vs. test temperature at
testing for heat resistance (); 200 h hot corrosion (o), and engine
test, 100 h (D)
Fig. 2.3
the highest life under certain running conditions

and point out ways to improve existent coatings
and develop alternative systems.
Two temperature regimes, one at 600 to 850

C and one above 1050 C (Fig. 2.3), have been
singled out as the ones where coatings are subjected to intensive damage. This result was
achieved while analyzing aluminide-based coatings for their protectiveness. At 600 to 850 C,
hot corrosion was noted to develop in coatings
owing to the presence of aggressive compounds
constituted mostly of sulfur and vanadium. The
temperature boundaries and intensity of corrosion will depend on a number of factors: composition of protective coating, quantity of aggressive compounds on the surface, and duration
of tests. During engine tests, the intensity of hot
corrosion will be affected (besides temperature)
by the gas pressure in the turbine and by design
features of the blades and vanes which can facilitate or impede condensation of aggressive
compounds on different areas of the surface.
In the interval from 900 to 1000 C, only
modest oxidation has been noted due to the high
protective properties of oxides Al2O3 Cr2O3
forming on the surface of this type of coating.
In laboratory heat resistance tests in the temperature range of 900 to 950 C, the diffusion
aluminide-based coatings, 50 lm thick, with
aluminum content of 34 to 36%, have manifested their ability to protect superalloys from
oxidation during 10,000 h.
At temperatures above 1000 C, the rate of
coating oxidation rises. At the same time, the
diffusion processes between the coating and alloy speed up, resulting in the aluminum and
chromium content of the coating dropping appreciably and the rate of oxidation growing. A
diffusion coating 50 lm thick (34 to 36% Al)
has exhausted its protective ability in 300 to 500
h when tested in the laboratory at 1100 C for
heat resistance.
The trends that have been noted during laboratory tests are also observed on engine tests.
Hot corrosion develops when engines are run
continuously. The rate of coating damage is generally lower when testing coated blades in the
range of 700 to 850 C (in comparison with laboratory tests). This phenomenon is attributed to
lower amounts of aggressive compounds from
the gas flow in contact with the blade surfaces.
Oxidation processes at 900 to 950 C go on
turbine blades more intensively than during heat
resistance tests. Generally, coatings are oxidized
during high-temperature cycles of engine running. At such temperature, the life expectancy
of a diffusion coating 50 lm thick (34 to 36%
Al) spans between 6000 and 8000 h when the
engine is run at maximum load. However, life
expectancy dramatically drops when the temperature on blade surface is elevated up to 1050
to 1100 C. When the blades were rig-tested at
1050 C, their diffusion-type protective coating failed in 80 to 100 h during engine running
at high temperatures.
Hot Corrosion of Turbine Blades. Since
aviation fuels contain insignificant quantities of
sulfur and vanadium, they do not destroy coatings through hot corrosion below service lives
of 5,000 to 10,000 h. The environment is the
main source of aggressive components entering
the hot channel of aircraft engines. This applies,
first of all, to sea regions, where a diverse variety
of compounds contained in the seawater get into
the engine hot channel.
Another source of aggressive components is
the waste gas of industrial plants. The location
of test rigs or the region of engine operation may
be the cause of this mechanism of coating failure.
Investigations carried out have revealed two
modes of hot corrosion on turbine blades of aircraft engines. Those are (a) pit and (b) uniform
corrosion modes (Ref 16). Pit corrosion emerges
and progresses first of all on those areas of the

blade surface that work continuously at 620 to
760 C (Fig. 2.4). Such surfaces are located on
the blade airfoil near the fir-tree root and shroud
platform. As early as in the first dozen hours of
testing or operation, numerous pits may appear
on the said surfaces. These pits, filled with corrosion products and looking like small nodules,
gradually, as time goes on, spread towards the
blade central region.
Pit corrosion proceeds on most-lengthy working cycles of engine operation. As the analysis
of turbine work parameters has shown, the
higher the gas flow velocity, the lesser the corrosion damage. Variation of pressure and air surplus coefficient produces no significant effect on
blade corrosion. Of much greater significance
are the blade design features and the local aerodynamics. For the same temperatures on the suction and pressure surfaces, the pressure surface
will suffer more from pit corrosion than the suction surface where corrosion attack is much less
severe. The poorer the fuel quality, the more severe is the corrosion.
Research has been conducted into the microstructure of diffusion coatings in corrosion-inflicted areas on the blades removed from different engines, either rig-tested or service-run, in
diverse conditions. The research has proved that
the microstructure was the same in all cases
studied. The oxides penetrating into coating
outer zones are encircled with light edging, consisting of Ni3Al compound (Fig. 2.4b). The
longer the time of testing or running, the greater
the depth of penetration of oxides into the coating. This, however, continues only until inner
zones with high contents of chromium-base
phases are reached. At this point, the oxide penetration is lowered and the inner zone of the
coating retards corrosion penetration into basic
metal. The analysis of the chemical composition
of pit corrosion products has shown that, apart
from the coating elements, they contained from
5 to 10% V and from 0.5 to 1.0% S.
The higher the blade working temperatures,
the less the pit corrosion. Thus, no pit corrosion
was observed on turbine blades from engines
that had been running at 900 C for as long as
8,000 to 10,000 h.
Elevation of temperatures, higher compression rates, and greater gas flow velocities in engine turbines have altered the outer appearance
of corrosion-stricken blades (Fig. 2.5). This
should be attributed, first of all, to the elevation
of gas temperatures up to 1330 to 1400 K and
to the resulting changes in the composition, fluidity, and thermodynamic properties of aggressive compounds. The condensation of aggressive compounds is helped by a high temperature
gradient between gas flow and surface of cooled
blades. Higher gas flow pressure and velocity
enhance erosion effects upon the corrosion products by briskly sweeping the latter off the blade
surfaces.
Fig. 2.4
Instead of formation of individual corrosion

pits, uniform corrosion has been noted. The latter affects large areas of the blade surface, particularly the pressure surface and the leading
edge exposed to temperatures of 780 to 850 C.
Consisting of aluminum, chromium, and nickel
oxides, the corrosion products penetrate deeply
into the coating and the superalloy (Fig. 2.6).
A unique feature of hot corrosion is that a thin
(a) Pit corrosion location vs. blade surface temperature (service time around 4000 h); (b) Blade surface in the zone of pit
corrosion. 20; (c) Microstructure of diffusion coating in pit corrosion zone. 300. (Light phase around oxides is Ni3Al.)
Fig. 2.5
Fig. 2.6
Uniform corrosion zone location vs. blade surface temperature (test time around 3000 h).
Microstructure of JS6U-alloy blade surface under

layer of oxides. 5000
layer of metal with sulfur precipitates emerges

underneath the corrosion products in the hightemperature zone. Such microstructure was ob-
served in all blades stricken with hot corrosion.

As for vanadium, its content in corrosion products is either insignificant or nil.
The uniform corrosion is typical of continuous engine running. Uniform hot corrosion is
noted on engine blades running at high temperatures (gas temperature from 1650 to 1700 K).
This should be attributed to effective cooling
systems providing, on certain surfaces (e.g., in
the lower zone of the blade airfoil), favorable
temperature conditions for aggressive compounds to condense.
The choice of a coating to protect engine
blades from hot corrosion within the required
life period is generally based on the results of
laboratory tests conducted in line with unified
methods. Coating service life is understood as
time to coating failure caused by hot corrosion.
The correctness of coating choice and prediction
of its service lifespan is totally dependent on
how sound the selected laboratory methods are.
Accelerated tests may be fraught with such serious errors that the results of the tests cannot be
used to predict lives. Moreover, one cannot rule
out errors involved in comparative analysis of
the properties of different coatings.
Life expectancies of coatings can be predicted
from the results of tests described in Phase and
Structural Changes in Coatings during HighTemperature Tests, Chapter 5. The approximate life expectancy of a new coating in terms
of its corrosion-resistance can be inferred from
the rate failure of an existing engine-run coating

and the results of comparative hot-corrosion laboratory tests of both the existing and new coatings.
The suggested approach to inferring blade-life
expectancy is simplified and substantiated by
test experience in using the laboratory methods
and then applying the data generated in selecting
the most appropriate coatings for aircraft engines. The best way of improvement here (i.e.,
in estimating the coating life expectancies) is not
to sophisticate further the relevant calculations
but, rather, to select the right test methods and
to accumulate data on the properties of different
coatings. The problem of accumulation of data
on hot corrosion can be solved, at reasonable
cost, by using the alloys, imitating possible compositions of coatings, and applying statistical
methods of processing the test results.
Figure 2.7 illustrates the results of life-expectancy calculations for 100 lm thick coatings
tested at 850 C under laboratory conditions for
the Ni-Co-Cr-Al-Y alloys. The same data can
easily be obtained for other systems too. Additional investigations will be required to infer
from these data life expectancies for coated
blades.
Ceramic coatings provide new opportunities
to lengthen blade-life expectancy in a hot-corrosion environment. The effects of such coatings
can be dual:
Adequate protection of metal coating from
aggressive components with proportional
lengthening of life expectancy of a double-
Service life of Ni-Co-Cr-Al-Y coatings 100 lm thick

calculated from data on laboratory-tested Ni-Co-CrAl-Y alloys in GZT ash, 850 C, 200 h
Fig. 2.7
layer coating (ceramic-metal coating). In

practice, the pertinent solution may be to increase coating thickness.
The ceramic coating alters the thermal state
of blade surface conditions (elevates the surface temperature). This reduces the amount of
aggressive compounds condensing on the surface. It also enhances the fluidity of such aggressive compounds, thus facilitating their removal by gas flow.
This factor has been confirmed by tests. When
tested in the laboratory, the ceramic coatings had
no significant effect, though tests on power units
showed the increase in life expectancy of blades.
Further investigations and development of special ceramic compositions will be necessary to
bring ceramic coatings on a par with the requirements imposed on coatings for hot-corrosion
protection.
Oxidation of Turbine Blade Operating at
Temperatures above 1000 C. As the temperature of a blade surface increases from 750
to 850 C to 900 to 950 C, the life expectancy
of coatings rises. In this temperature interval,
diffusion coatings guarantee adequate protection
of aircraft-engine turbine blades within 6000 to
8000 h. Increasing blade surface temperature up
to 1050 to 1100 C results in appreciable decreases in diffusion coating life expectancy. At
such temperatures the coatings fail mostly due
to consumption of aluminum for the formation
of the protective oxide. Besides, the aluminum
content was reduced owing to interdiffusion of
aluminum and nickel between the coating and
superalloy. As Fig. 2.8 shows, a turbine blade,
after being tested for 100 h, displayed full oxidation of the coating (blade wall thereafter) in
30 to 40 h when exposed to a temperature of
1100 C.
Figure 2.9 illustrates oxidation rates in a diffusion coating deposited on the blade from superalloy JS6U. The coating has been built up by
slurry aluminizing and it had an aluminum reserve of 40 g/m2. After being tested for
100 h, the coating outer zone was found to be
fully destroyed at a narrow strip on the leading
edge 2 to 4 mm broad and 10 to 20 mm high
when exposed to the maximum temperature of
1050 C. The consumption rate of aluminum for
protective oxide formation in the leading edge
region was 0.5 g/m2 h. With the temperature
at the blade inlet edges elevated to 1100 C, the
diffusion coating with aluminum reserve of 60
Fig. 2.8
External appearance of a blade with diffusion coating after 100 h of testing
(1) Oxidation depth of diffusion and (2)

Ni20Cr12AlY overlay coatings on the leading edge
of a blade during testing
Fig. 2.9
to 70 g/m2 was found to be fully oxidized in 30

to 40 h of tests.
Figure 2.9 illustrates the depths to which oxidation has penetrated in the overlay coating
Ni20Cr12AlY, 120 lm thick, with aluminum reserve 200 g/m2. Measurements were taken on
the leading edges of turbine blades. When running under maximum loads, the temperature on
the leading edge was 1050 C. Test duration
is indicated as a summed value for all running.
In comparison with diffusion coatings, the overlay coatings had greater aluminum reserves,
which made the life expectancy 4 times as
long as that of diffusion coatings. The designed
average consumption of aluminum and chromium for protective oxide formation in
Ni20Cr12AlY coating varied within 0.5 to 0.6
g/m2 h.
As for the phase transformations and structural changes in the coating Ni20Cr12AlY occurring on the leading edge of the blade, they
are similar to those typical of laboratory tests for
heat resistance. After 140 h, a thin layer of csolid solution is formed on the coating surface,
controlling the kinetics of oxidation processes.
As test time gradually increases to 300 h, the csolid solution covers a greater zone and its
boundary spreads further into the coating. The
diffusion of aluminum, and to a much lesser extent chromium, from the coating central zones
toward the surface makes the coating composition change in the direction NiAl cNi3Al
cc. In this case 10 to 15% of aluminum
and chromium diffuse under the coating into the
zone of superalloy interaction.
As the Ni20Cr12AlY coating is becoming oxidized, the surface of a blade is deformed in a
rippled or folded fashion (Fig. 2.10). This
phenomenon can be attributed to gas flow pressure on the coating that is soft at these temperatures. Its deformation is affected by compression
stresses occurring in the coating at temperatures
above 1000 C.
Should the working temperatures on turbine
blades with the overlay coating Ni20Cr12AlY
be raised to 1100 to 1150 C, the coating protection capability reduces dramatically, provided that the described mechanism of coating
damage prevails. Figure 2.11 shows typical
structures of Ni20Cr12AlY coating on turbine
blades due to overheating from 1000 to 1150 C
in certain zones.
The oxidation mechanism remains the same
on all surface areas. Its nature can be explained
by consumption of certain elements (aluminum
standing the first) for oxide formation. The rate
of aluminum and chromium consumption varies

with the temperature level. While in zones at
1000 C, the coating has oxidized to an insignificant depth (Fig. 2.11a); in the zone at 1150
C, phase transformations associated with csolid solution formation have been completed.
Aluminum content there has been reduced to
3%, and the processes of internal oxidation have
started (Fig. 2.11d). Aluminum and chromium
levels consumed to form the interaction zone do
not exceed 10 to 15%, even at maximum temperatures.
The heat resistance of diffusion and overlay
coatings is discussed against the background of
laboratory tests, enabling protective properties
of different coatings to be compared, given that
correct methods and proper criteria have been
chosen. However, the test data, pertinent to specific changes of mass (g/m2) cannot be used in
calculations of coating life expectancies on
blades. For the same mechanism of oxidation,
the extent of blade damage from oxidation will
be many times that of laboratory samples.
Fig. 2.10
In laboratory tests and in tests on the blades,

the protective function is performed not by the
coating, but rather by the oxides that are formed
on its surface. In fact, the coating acts as a reserve that supplies aluminum and chromium for
the formation of oxides if they are damaged. In
this case it will be the aluminum reserve (mAl)
that emerges as a criterion for coating life expectancy (Ref 16). This has been substantiated
by data, which showed that all aluminum contained in a diffusion coating on superalloys is
eventually consumed for oxide formation. The
aluminum reserve criterion correlates well
with the results of both laboratory and field tests.
Aluminum reserve, mAl in a diffusion coating is found from the following function:
n
mAl K
Xi qi hi
i1
in g/m2 where Xi is aluminum mass fraction in

a definite zone of coating; qi is the density of
External appearance of Ni20Cr12AlY coating on a turbine blade after 500 h of testing. (a) Leading edge. (b) Pressure
surface. 10
that zone of coating; hi is the thickness of that

zone of coating; and K is the alloying factor.
Aluminum reserve in the coating equals the
increase of specific mass of samples at alitizing,
and it can be found out experimentally. The effect of alloying the diffusion coatings can be
counted by introducing the coefficient K ,
which for aluminized coatings equals unity (K
Microstructure of Ni20Cr12AlY coating in the blade zones with gradual increasing of working temperatures from (a) 1000
C to (d) 1150 C. 500. In (a), (b), and (c), a dark phase in the coating is c-solid solution; a light one is Ni3Al. In (d), a
dark phase in the coating is oxides; a light one is c-solid solution.
Fig. 2.11
1), while for the modified coatings alloyed

with chromium, silicon, and yttrium, it takes the
value KAl 1.11.6.
The aluminum reserve criterion can also be
used in comparing life expectancies of overlay
coatings, having the same or nearly the same
system of alloying. When determining aluminum reserve in Ni-(Co)-Cr-Al system, Cr concentration must be taken into account too. Other
elements must be counted by introducing suitable alloying factor K:
mAl (XAl 0.5 XCr) q h K
2
in g/m . The coefficient 0.5 for chromium mass

fraction XCr in this case was taken from the regression equation, interlinking the effect of these
elements (Al and Cr) on the oxidation kinetics
in Ni-Co-Cr-Al system alloys. The tests were
conducted at 1200 C.
When aluminum consumption for oxide formation on turbine blades is known, then the
coating life expectancy will be directly proportional to the aluminum reserve. The coating life
parameter at temperatures above 1000 C will
depend on the effects of interaction between the
coating and superalloy. In both laboratory and
engine tests, a part of the aluminum reserve is
lost in diffusion of aluminum and chromium
from the coating into the superalloy. In laboratory tests the losses maybe considerable compared with aluminum consumption for oxide
formation, but in testing on turbine blades the
oxide consumption will always exceed interaction losses.
Since interaction zone dimensions (between
diffusion/overlay coatings on the one hand and
nickel-base superalloys on the other) do not exceed 30 to 40 lm when tested/run for up to 1000
h at temperatures below 1000 C, these may be
disregarded.
Thermal Fatigue of Turbine Blades. As a
result of higher gas temperatures in aircraft engines and the development of turbine blades
with highly efficient cooling, the thermal
stresses in blades have risen. Thermal fatigue
cracks in blades have become the commonest
defect. It has been suggested that all thermal fatigue cracks occurring in turbine blades of different aircraft engines can be conventionally
classified into two types, proceeding from either
crack origins within the substrate or from thermal stress cracking in coatings (Ref 24):
Thermal fatigue cracks of the first type
Thermal fatigue cracks of the second type
(fragmentation)
Differences in the appearance and microstructure between the two types are as follows:
The first type of thermal fatigue cracks are

generally few. In the process of operation,
they penetrate deeply into the blade wall, reducing the thermal stresses arising between
certain zones of the surface and eventually
leading to blade failure.
The second type of thermal fatigue cracks
(fragmentation) can appear on blade surfaces
during the first hours of blade service life.
Usually these are numerous fine cracks, occasionally covering the full blade surface.
The depth of their penetration is usually limited to within the coating thickness; thereafter, the cracks either discontinue their growth
or continue at a very slow rate. The unique
feature of such cracks is that their distribution
density is very high; distances between adjacent cracks are comparable with the coating
thickness.
The research into the fragmentation phenomenon shows that its occurrence is due to the following:
High gradients of temperatures across the

blade wall section, the blades being provided
with highly efficient cooling systems
Increased transient running on the engine
Sharp changes of heating on transient running
(thermal shocks), during which the rate of
temperature rise and drop is measured in hundreds of degrees per second
Difference between the thermal expansion
coefficients (TECs) of coating and superalloy
Insufficient ductility of coatings in the area
where maximum strains occur
Fragmentation emerges as a result of the coating-superalloy system exposure to high thermal

stresses and strains. Illustrated in Fig. 2.12 is a
turbine blade made from the superalloy JS6F,
with a diffusion coating and after testing for 100
h. The entire pressure surface of the blade is covered with a net of the second type of thermal
fatigue cracks, penetrating not beyond the coating thickness.
The optimum way of removing fragmentation
is to reduce thermal stresses in the blade coating
to as low a level as possible. Bringing closer the
coefficients of thermal expansion of the coating
and superalloy can do this. While calculating the
cyclic service life of a coated blade, the designer
can predict the emergence of fragmentation
cracks and thereby choose coatings that offer the

best life expectancy.
The principles of a cyclic service life calculation for a coating deposited on a blade surface
are dealt with in Chapter 8, Strength Designing
of Turbine Blade Protective Coatings. The calculation is based on stresses and strains arising
in the coating due to centrifugal forces, bending
moments, nonuniformity of blade thermal conditions, and thermal stresses caused by different
TECs of the coating and the alloy protected.
In performing calculations, it must be borne
in mind that the composition of the coating and
the whole complex of its physical and mechanical characteristics are subject to continuous
changes with time. The rate of such changes is
higher when the temperature of the blade surface
rises. The coating cyclic life expectancy must be
calculated for the original and final coating compositions. Should the coating properties undergo
essential changes with time, interim calculations
may be necessary.
While protecting the blade from oxidation, the
coating simultaneously produces extra stresses
and strains on the blade surface. These stresses
and strains damage the coating and reduce its
cyclic service life. This reasoning holds only
with respect to an ideal blade, because in actual
conditions, when uncoated blades are exposed
Fig. 2.12
to vigorous oxidation, the resultant loss of blade

service life far exceeds the losses caused by the
coatings.
With high strains in blade coating (Deic
0.5%), the rate at which thermal fatigue cracks
emerge in the coating (fragmentation) may exceed
the rate of blade surface damage from oxidation.
In this case, the deposition of coatings would bring
about a shortened cyclic life of the blade.
Figure 2.13(a) and (b) illustrate the results of
cyclic tests conducted on engines whose turbine
blades are made from superalloys JS6F and
JS6U with diffusion and overlay coatings
Ni20Cr12AlY,
Ni20Co20Cr12AlY,
and
Co25Ni20Cr12AlY. The number of blades with
fragmentation grows with test time. After 10 to
50 h, all blades are covered with a network of
cracks. No cracks have been observed on the
blades with Ni20Cr12AlY coating.
Spalling of Ceramics on Turbine Blades
with Thermal Barrier Coatings. The development of thermal barrier coatings (TBC) offers
new horizons in further lengthening of turbine
blade life expectancy. Not only does the TBC
protect the blade surface from hot corrosion and
oxidation, it also offers thermal protection and
thus shields the superalloy from softening during high-temperature exposure. The TBC is basically a structure consisting of a bond coat and
(a) External appearance of a blade with fragmentation exposed by capillary flaw detection after 500 h of testing. (b)
Microstructure of a coating in a fragmentation zone. 500
an outer ceramic layer. The latter is based on

zirconium oxide and serves as the TBC main
element.
The properties of the TBC ceramic layer are
entirely dependent on the mode of its formation.
If the electron beam (EB) technique is applied
and if all process problems have been satisfactorily settled, the ceramic layer will have a specific columnar structure. The typical features of
such a structure are low tensile strength and low
elastic modulus in the direction normal to the
crystallite-growth axis. As thermal stresses and
strains arise, the ceramic layer freely breaks into
fragments. The individual fragments are sized
by such factors as stress level and the ceramic
layer thickness. Any strains in the blade cause
fragmentation. This property protects the ceramics from high stresses and spalling.
Life expectancy of a TBC working at temperatures above 1000 C is totally dependent on
the adhesion of the ceramic layer. The EB technique enables ceramic layers with original adhesion strength higher than 70 MPa to be obtained.
Number of (a) JS6F and (b) JS6U blades with fragmentation vs. time of testing. 1, Ni20Cr12AlY
coating; 2, aluminized coating; 3, Ni20Co20Cr12AlY coating; 4,
Co23Ni20Cr12AlY coating
Fig. 2.13
Depositing a TBC on the blade alters its characteristics. This must be noted when designing
blades with TBC. The thermal barrier effect DT
is controlled by the ceramic layer thickness, d; its
thermal conductivity, k; and the thermal flux
through the blade wall, Q such that D T
Q d/ k.
There are certain restrictions on the extent to
which ceramic thickness can be increased. In
view of its low tensile strength, the ceramic layer
is unable to carry mechanical loads produced by
centrifugal forces. Hence, extra loading is applied to the turbine blades (and disk). This dictates that it is better to make the ceramic layer
of variable thickness throughout blade surface.
Maximum thickness is needed in the areas where
most intensive thermal fluxes will be experienced or where the thermal field is nonuniform.
In other zones, the ceramic layer may have minimum thickness or be absent. The optimum
thickness of a ceramic layer on the blade varies
within 120 to 250 lm. At lower thickness, there
may be rapid oxidation of the ceramic-bond coat
boundary and poor adhesion strength. At greater
thickness, considerable stresses resulting in spallation are likely to occur.
The composition of a bond coat and its thickness are chosen as when using metal heat resistant coatings. Since TBCs are employed on
blades with high thermal stresses, preference
should be given to bond coats offering higher
resistance to thermal fatigue cracking. Once the
ceramic layer has spalled, the bond coat has to
protect the blade from oxidation.
The main cause of TBC failure is spallation
of the ceramic layer from the blade while being
exposed to higher temperatures for a long time
(Fig. 2.14). The spallation is assisted by tear-off
stresses occurring on the convex blade surfaces
when thermal stresses occur inside the ceramic
layer. These internal compression stresses reach
their maximum as the blade is cooled down to
room temperature.
The onset of TBC failure is betrayed by a network of microcracking (fragmentation) appearing on the surface of the blade. Fragmentation
of a ceramic layer does not indicate its failure.
The blades in such a state may be exploited for
quite a long time. Fragmentation appears on
some surface areas where tensile stresses can be
generated in the ceramic layer under the certain
engine operation conditions.
TBC life expectancy can be estimated by adhesion strength that has to be maintained
throughout the service life of the coating. According to this criterion, the adhesion strength
must exceed the shearing stress needed to separate the ceramic coating from the surface. Failure of the ceramic layer in the TBC is evidence
of a reduced adhesive strength on the ceramic-
Fig. 2.14
Blades with a spalled ceramic layer after 800 cycles of testing at 200 1100 C
bond coat boundary. Spallation of ceramics after

high-temperature tests takes place even in the
absence of stresses in the sample. The effect is
associated with the formation of various Al2O3type oxides on the ceramics-metal interface.
Such oxides reduce the adhesive strength and
alter the stress condition on the boundary because their physical and mechanical properties
differ from those of ceramic and metal layers in
TBCs.
All the above processes, leading to ceramic
layer spallation, can be allowed for by the integral characteristic of adhesion strength, ra,
found experimentally in relation to the temperature and duration of tests. The criteria applicable
to the determination of adhesive strength for a
ceramic layer on blades with TBC are discussed
in Chapter 8, Strength Designing of Turbine
Blade Protective Coatings.
To retain adhesive strength with time is the
most formidable challenge to further lengthening of TBC life expectancy. Two prerequisites,
however, must be fulfilled for this: to reduce oxygen diffusion mobility in the lattice of zirconium oxide and to prevent fragmentation of ceramics in high-temperature applications.
Benefits that TBCs provide include the opportunity to elevate the gas temperature at the
turbine inlet without changing the temperature
of the turbine blades. Most often, however,
TBCs are used to extend blade service life. Estimation of the TBC benefit is a difficult task.
The issue is that while comparing the blades
with and without TBC, some side effects occur
that affect the blade temperature (e.g., reduction
of perforated hole diameter). Analysis of plastic
strains in identical blades with and without
TBCs, tested in the engine under similar conditions revealed lesser elongation in blades with
TBC in comparison to their counterparts without
TBC (Fig. 2.15); that is, they demonstrate higher
high-temperature strength.
Use of TBCs allows higher temperatures in
engine turbines to be employed, though
ZrO2 Y2O3 oxide imposes significant restrictions on its use at the temperatures above 1150
C. These restrictions are due to high oxygendiffusion mobility in zirconium oxide. The development of new ceramics for high-temperature application is the most promising route to
further TBC improvement.
REFERENCES
Fig. 2.15
Elongation of blades with (1) Ni20Cr12AlY coating and with (2) TBC vs. service time
1. N.D. Kuznetsov, Problems of Thermocyclic

Strength of Gas Turbine Engines Elements,
Probl. Prochn., Vol 6, 1978, p 38
2. Y.B. Chertkov, K.F. Rybakov, and V.H.

Zrelov, Impurities and Purification Methods of Oil-based Fuels, Moscow, Khimiya,
1970
3. Ya.B. Chertkov and V.G. Spirkin, Use of
Jet Fuels in Aviation, Moscow, Transport,
1974
4. I.H. Shishkov and V.B. Belov, Aviation Lubricants, Fuels, and Special Fluids, Moscow, Transport, 1979
5. I.A. Birger, B.F. Schorr, and I.V. Demianushenko, Machine Elements, Moscow,
Mashinostroyeniye, 1975, 455 p
6. N.D. Kuznetsov, Designing the Strength of
a Long Life Gas Turbine Engine, Probl.
Prochn., 5, 1976, p 39
7. N.D. Kuznetsov, Strength of Gas Turbine
Engine Turbine Elements under Complex
Loads and Related Problems, Probl.
Prochn., 3, 1982, p 1014
8. M.Ia. Ivanov and V.P. Pochuev, Problems
of Designing High-Temperature Turbines
of Modern Aircraft Engines, Conversion in
Machine Building of Russia, 5, 2000, p 34
46
9. V.H. Abiants, Jet Engines, Moscow, Mashinostroyeniye, 1985
10. Ye.N. Bogomolov, Service Processes in
Cooled Turbines of Gas Turbine Engines
with Perforated Blades, Moscow, Mashinostroyeniye, 1987
11. S.Z. Kopelev, M.N. Galkin, A.A. Harin, and
I.V. Chevtchenko, Thermal and Hydraulic
Characteristics of Cooled Gas Turbine
Blades, Moscow, Mashinostroyeniye, 1993,
176 p
12. G.P. Dolgopolenko, M.D. Romanov, and
V.V. Gatin, Plane and Helicopter Gas Turbine Engines, Moscow, Mashinostroyeniye,
1983
13. D.H. Boon, Overlay Coatings for Improved
Oxidation/Corrosion Protection and Ductility for High-Temperature Applications,
Airco Temescal, May, 1977, 12 p
14. B.A. Movhan and I.S. Malachenko, Heat

Resistant Coatings Deposited in Vacuum,
Kiev, Naukova Dumka, 1983, 232 p
15. A.G. Andreeva, V.V. Terekova, and G.D.
Fomenko, Heat-Resistance Coatings of
Nickel-Base Alloys, High-Temperature
Coatings, Leningrad, Nauka, 1967, p 96
110
16. Y.A. Tamarin, Heat Resistant Diffusion
Coatings for Turbine Blades, Moscow,
Machinostroenie, 1978, 133 p
17. P. Galmiche, Applications en Construction
Airospatiale et Retomlees de Techniques
Thermochimiques ONERA (The Use of
Chemical Heat Treatment of ONERA in
Aircraft and Ground Constructions), Aircraft and Space, 1973, Vol 41, p 3342
18. P.T. Kolomytzev, Heat Resistant Diffusion
Coatings, Moscow, Metallurgiya, 1979,
272 p
19. T.V. Levchenko, V.I. Moroz, and L.P. Buiyanova, Protective Coatings on Metals,
Kiev, Naukova Dumka, 4, 1971, p 158164
20. S.A. Muboyadjyan, I.A. Pomelov, Y.A.
Tamarin, N.V. Zabrodina, and R.I. Beliakova, Aluminum-Base Alloy for Coatings,
Authors
Certificate
USSR
N1067847, 1983
21. E.B. Kathanov and Y.A. Tamarin, Protective Coatings: Effective Path of Increase of
Reliability of the Blades, Aviation Materials
on the Eve of 21st Century, Moscow,VIAM,
1994, p 296304
22. R.E. Halin, E.B. Kachanov, I.L. Svetlov,
and V.N. Toloriya, Single-Crystals of
Nickel Superalloys, Moscow, Mashinostroenie, 1997, 333 p
23. L.B. Getzov, Gas Turbine Components:
Materials and Strength, Moscow, Nedra,
1996, 590 p
24. Y.A. Tamarin, V.G. Sundyrin, and E.B. Kachanov, Gas Corrosion and Thermal Fatigue of Protective Coatings for Turbine
Blades, High Temperature Corrosion and
Protection, China, Liaoning Science and
Technology Publishing House, 1991, p
161166

Y. Tamarin, p25-54
DOI: 10.1361/pctb2002p025

CHAPTER 3
Technological Processes for

Deposition of Protective Coatings to
Turbine Blades
A GREAT VARIETY of techniques for deposition of protective coatings to aircraft turbine
blades have been designed. The use of each
technological process is warranted by a number
of reasons including requirements to the properties of protective coatings, production equipment available, personnel experience, and common practice. The preference of new techniques
to conventional ones is justified only in cases
where fundamental changes in the properties of
protective coatings are required. For example,
new techniques are necessary in the case of designing a new turbine. As a rule, superseding in
the field of technology involves considerable financial expenditure connected with the purchase
of equipment and materials and training of personnel.
given in percents by weight). Coatings alloyed

with chromium, silicon, and other elements can
be formed. The method of aluminizing in the
powder mixtures with activators offers such advantages as simple production process and
equipment, a wide temperature range for aluminizing (from 500 to 1200 C), highly even
thickness and homogeneity of the produced
coatings, and low production cost.
Powder mixtures for aluminide coating formation are composed of aluminum or its alloys
with iron, chromium, and other elements (as an
active component), aluminum oxide (as an inert
filler), and a halide activator (usually NH4Cl).
When the container with the saturating mixture
and aluminized parts is heated, the following reactions take place:
NH4Cl NH3 HCl
Diffusion Coatings
Application of Diffusion Coatings in
Powder Mixtures with Halide
Activators
The diffusion saturation of superalloys with
aluminum and with aluminum combined with
other elements (Al-Cr, Al-Si) from the powder
mixtures with halide activators is a widely used
method of aluminide coating formation on turbine blades of aircraft gas turbine engines. The
wide application of this method is justified by
the fact that different mixtures used for aluminizing ensure a wide range of aluminide coatings
with aluminum contents varying from 18 to 69%
Al (here and subsequently all compositions are
2NH3 2N 3H2
The activator decomposition products displace

the air from the container and reduce partly the
oxides on the metal powder surface and on the
aluminized alloy surface. Hydrogen chloride interacting with aluminum forms aluminum chlorides:
6HCl 2Al 2AlCl3 3H2
2AlCl3 Al 3AlCl2
AlCl3 2Al 3AlCl
These chlorides carry aluminum to the saturated

surface on which the disproportionation reac-
tions occur, resulting in the formation of the corresponding aluminides and the highest valence
chlorides:
3AlCl2 n Ni 2AlCl3 NinAl
3AlCl 2n Ni AlCl3 2NinAl
AlCl3 resulting from the reaction goes to the

powder mixture where it interacts with aluminum again. This interaction feature is that aluminizing does not involve the consumption of
chlorides. The equilibrium mixture of aluminum
chlorides at 1200 to 1700 K has the prevailing
AlCl2 content, the maximal amount of which is
formed at approximately 1400 K.
The aluminizing process occurs because of
the thermodynamic equilibrium shift in the gaseous mixture resulting from absorption of aluminum and its diffusion into the saturated metal.
The outcomes of aluminizing, however, such as
dimensions and compositions of the coatings
formed, depend not only on the amount of aluminum dichloride in the mixture but also on the
rate of its supply to the saturated surface, the rate
of the surface reaction, the rate of the reactionproduct discharge to the mixture, and the rate of
the diffusion processes in the coating being
formed.
The quantitative analysis of all these processes is extremely complicated; hence, for
practical purposes, the saturating mixtures are
usually characterized by the experimental relationships between the mixture composition and
the properties of the coating, such as its composition, thickness, and hardness.
Influence of NH4Cl activator content on thickness of

coatings forming on nickel-base superalloys. Aluminizing in mixture: Al-Fe alloy (50% Al)-NH4Cl at 950 C for 4
h. Source: Ref 1
Fig. 3.1
The amount of a halide activator added has a

strong effect on the outcome of aluminizing,
such as coating thickness and surface quality
characteristics (Fig. 3.1). When the amount of
NH4Cl increases from 0.1 to 2%, a gradual
thickening of the coatings formed on nickel-base
superalloys occurs at their aluminizing from the
powder mixture that contains an Al-Fe alloy
with 50% of Al as an active component. The
further increase in NH4Cl content up to 10% results in thinner coating due to a greater amount
of chlorides displaced from the container and,
respectively, to lower aluminum content of the
Al-Fe alloy.
This effect is particularly well seen at a repeated use of the powder mixtures with more
than 4% NH4Cl activator. When a mixture with
10% of NH4Cl is used ten times (the activator
has been introduced before every next process),
the thickness of the formed coatings is four
times lesser, while for 2% NH4Cl mixture, the
coating thickness is reduced by a factor of 1.8
only. Hence, the optimal activator content of the
powder mixture ensuring the thickest coatings at
an insignificant depletion of the mixture is in the
range of 1 to 3%. Other halide-containing compounds, such as NH4F, NaF, KF, (NH4)HF2,
NH4I, NH4Br, N2H2-2HCl can also be used as
activators at aluminizing of superalloys. The
mechanisms of their activating effects are similar.
The main parameter of the powder mixtures
used for aluminide coating manufacturing is
their activity determined by the amount of aluminum that the mixture feeds to the saturated
surface. This parameter determines the concentration of aluminum and other elements in the
coating, its thickness, and, hence, its protective
and other properties. The saturating mixtures
may be conventionally classified into high, average, and low activity ones on the basis of the
aluminide structure and composition in the coating outer layer.
A high-activity mixture is usually composed
of aluminum powder and inert filler. As the content of free aluminum increases, the activity of
such a mixture also increases, resulting in thickening the coatings formed on the surface of the
aluminized alloy (Fig. 3.2). The aluminum content of the coating outer layer exceeds 40%. The
mixtures with aluminum contents of 10 to 15%
have found use in the application of aluminide
coatings to turbine blades and other elements of
aircraft gas turbine engines.
The high activity of the powder mixtures containing free aluminum allows their use for alu-
Technological Processes for Deposition of Protective Coatings to Turbine Blades / 27
minizing at the temperatures below 900 C. If

the aluminum content of the mixture increases
up to 98%, the aluminizing temperature may be
lowered down to 500 to 600 C. The mixture
containing 98% aluminum powder and 2%
NH4Cl forms thick enough coatings with aluminum content of 66 to 69% on nickel-base superalloy surfaces within the temperature range
of 500 to 600 C (Fig. 3.3).
The powder mixtures are widely used to obtain protective aluminide coatings on superalloy
surfaces. They contain bound aluminum as a
component of aluminum-iron or aluminumchromium alloys. These alloys may be obtained
by presmelting followed by reducing to powder
with the required particle size or by sintering the
powders of their raw elements. In some cases
Influence of aluminum content of mixture of AlAl2O3-5% NH4Cl on the thickness of coatings forming on nickel-base superalloys at 900 oC
Fig. 3.2
Fig. 3.3
Kinetics of coating growth on nickel-base superalloys. Aluminizing in mixture: 98% Al and 2% NH4Cl
the saturating mixtures are composed of powders of individual elements and sintered at aluminizing.
The halide formation and interaction with the
saturated surface obey the previously mentioned
laws for powder mixtures containing bound aluminum. In the gaseous phase forming at aluminizing, however, the corresponding halides of
the elements added together with aluminum can
be found. The amount of halides at aluminizing
depends on the thermodynamic parameters of
their formation and on the amount of the element
fed. It is noteworthy that the thermodynamic potential varies significantly less when iron, chromium, and silicon halides form in the mixture
than when aluminum halides form; hence, their
amount in the gaseous phase of the mixtures
with more than 30% Al is not large. The interaction between the saturated surface and halides
of iron, chromium, silicon, and other elements
added to the mixture results in a formation of
the aluminide coatings alloyed with these elements.
If iron, chromium, nickel, cobalt, and some
other elements are introduced into the saturating
mixtures together with aluminum, the mixture
activity, the aluminum content of the resulting
coating, and the coating thickness decrease. For
example, an increase in the iron content of the
saturating Al-Fe alloy from 30 to 50% reduces
the thickness of the coatings formed on the JS6K
alloy at 950 C for 4 h from approximately 1000
to approximately 60 lm. In the coating, the reduction of the aluminum content is usually accompanied by the increase in the second-component content of the saturating alloy. As the
mixtures with iron, chromium, or nickel contents of more than 50% feature low activity, aluminizing is usually carried out at temperatures
above 1000 C to obtain the required thickness
of the coating.
For aluminizing at temperatures above 850
C, Fe-Al alloy-based mixtures with aluminum
content of 50% are convenient (Ref 1). The
outer layer of the coating formed when these
mixtures are used to saturate nickel-base alloys
has aluminum content of 32 to 36%. The low
rate of the coating thickening in these mixtures
as compared with that observed in the high-activity ones allows accurate enough control of the
thickness of the coatings formed within the temperature range of 850 to 1100 C currently used
for superalloy aluminizing. The kinetics of diffusion coating formation on the nickel-base alloys is shown in Fig. 3.4. The mixtures com-
posed of Al-Fe alloy with aluminum content of

50% feature high stability that is particularly
important for their commercial application.
The Al-Fe alloy-based mixtures with initial
aluminum content of 40% feature low activity.
The outer layers of the coatings obtained on
nickel-base superalloys at aluminizing from
such mixtures at 950 to 1200 C have an aluminum content of 30% and less. If the aluminum
content of Al-Fe alloy is more than 70%, such
mixtures feature high activity. They may be used
to saturate nickel-base superalloys at 600 C. In

this case, a 30 to 50 lm thick aluminide coating
with aluminum content of up to 60% is formed
within the period of 4 to 5 h.
The mixtures composed of aluminum and silicon powders analyzed elsewhere (Ref 2) have
somehow different properties. The interrelation
of the aluminum and silicon effects on the properties of the coatings has been studied using the
mixtures containing 50% (Al Si), 45% Al2O3,
and 5% NH4Cl. The addition of silicon to the
saturating mixture reduces the activity of the
mixture and the thickness of the coating formed
(Fig. 3.5), but this drop of activity is significantly less than that for the previously mentioned Al-Fe alloys. This is due to the absence
of chemical interaction between aluminum and
silicon and of intermetallide compound formation. At 1000 C, the increase in the silicon content above 50% results in coating thickening due
to the intensification of the siliconizing process.
The aluminide coatings alloyed with silicon may
be obtained at the aluminum/silicon (Al/Si) ratio of the saturating mixture within the range of
1 to 9 to 9 to 1. The kinetics of coating formation
on the JS6K alloy at its aluminizing from the
mixture of powders containing 10% Al 50%
Si, 38% Al2O3, and NH4Cl is shown in Fig. 3.6.
The method of aluminizing in powder
mixtures that contain halide activators allows the
application of protective aluminide coatings of
various compositions to turbine blades and other
Kinetics of coating growth on nickel-base superalloys. Aluminizing in mixture: 98% Al-Fe (50% Al)
alloy and 2% NH4Cl. Source: Ref 1
Fig. 3.4
Influence of aluminum and silicon contents of mixture on thickness of the coatings forming on nickelbase superalloys. Aluminizing in mixture: 50% (Al-Si), 45%
Al2O3, and 5% NH4Cl. Source: Ref 2
Fig. 3.5
Kinetics of coating growth on nickel-base superalloys. Aluminizing in mixture: (10% Al 50% Si),
38% Al2O3 and 2% NH4Cl
Fig. 3.6
parts within a wide temperature range of 500 to

1200 C. The initial considerations, which allow
the selection of the saturating mixture and conditions of aluminizing and diffusion annealing,
are the requirements to the protective coating
composition and thickness. The key problem is
the selection of the proper thermal conditions for
the coating formation. If the heating of the parts
in the container with the powder mixture and
their cooling after aluminizing are slow, the mechanical properties of the superalloy may
worsen significantly. Hence, if it is possible, aluminizing should be combined with heat treatment operations that do not require accelerated
cooling, mainly, aging.
When the high-activity mixtures, which allow
aluminizing in the temperature range of 500 to
800 C are used, diffusion annealing after aluminizing may be combined with any operation
of the alloy heat treatment. Sometimes at aluminizing in low activity mixtures, when the process temperature is above 1000 C, special heat
treatment is required to compensate for the reduction of alloy heat resistance after aluminizing.
Many different mixtures are known to be used
to obtain heat resistant coatings on superalloy
elements. The mixtures saturating such critical
elements as aircraft engine turbine blades should
meet some special requirements; their properties
should be considered from the viewpoint of
commercial application where aluminizing is
carried out in large containers (having room for
20 to 100 blades). The mixtures for turbineblade aluminizing should guarantee reproducibility of the coatings with the required composition and thickness, allow for the possibility of
their repeated use (provided that the thickness
and composition of the coatings in question do
not change), and guarantee the same initial class
of surface roughness of the finished blades.
The first requirement is usually met by selecting a suitable mixture for the aluminizing
temperature range in use and by using a highly
homogeneous mixture. Those are the features of
the mixtures composed of cast Al-Fe or Al-Cr
alloys. The ingots are crushed into 2 to 5 mm
granules. When the mixtures, which contain aluminum powder and oxides of aluminum, chromium, and other elements, are used, the initial
powders should be fine (less than 60 lm) and
thoroughly mixed. To increase stability of the
powders, they are annealed under H2 atmosphere.
The proper use of the activator is also very
important in obtaining coatings of the preset
composition, thickness, and high surface quality. The ammonium chloride (NH4Cl) should be
predried at 130 to 150 C, and then the crushed
activator should be distributed uniformly in the
mixture.
To produce high quality coatings, thoroughly
preparing the blade surfaces for aluminizing is
of particular importance. The surface of the elements should be cleaned from oxides and impurities. To improve the quality of coatings, the
blade surfaces may be treated with glass balls.
Before the blades are arranged in the containers,
they should be degreased.
The homogeneity of the coating composition
and the evenness of its thickness both depend on
the proper and regular arrangement of the blades
in the container for aluminizing. They should be
positioned in such a manner that the thickness
of the surrounding layer of the saturating mixture is not less than 25 to 30 mm. Reference
specimens are placed in the middle and near the
walls of the container. The excess of gases resulting from the activator decomposition should
be pumped out of the furnace space.
As a consequence of large-sized industrial
containers and low heat conduction of the saturating mixtures, the greatest share of the process
time (70 to 90%) is spent on heating the container to the required temperature and on cooling. The coatings form during both heating and
cooling; this should be accounted for when computing the aluminizing time. Hence, if large containers are used, it is more convenient to use the
low and average activity mixtures in order to get
the coatings of required parameters on the blades
situated in the center and near the walls of the
container with the least possible tolerances.
The amount of aluminum consumed when
forming the surface coatings is small enough
compared to its content of a mixture, but, when
a powder mixture is used more than once, the
reduction of its activity, the aluminum content
of the coatings, and the coating thickness is observed. There are two reasons for this. The first
is that the mixture is depleted of aluminum. It is
spent on the formation of gaseous aluminum
chlorides partly replaced from the container and
on the formation of the coating. The second reason is that mixture oxidation with the air coming
into the container while cooling oxidizes the
mixture. The oxidation is extremely intensive in
the mixtures composed of fine powders of aluminum, chromium, and other elements. In this
case, a protective atmosphere (Ar, H2) is created
in the container at its heating and cooling to
make possible a repeated use of the mixture.
The multiple use of the mixtures consisting of

98% Al-Fe alloy (50% Al) and 2% NH4Cl results in the aluminum content of the mixture of
up to 39% after the twentieth process. In this
case the aluminum content of the coating formed
drops from 36 to 24%, and the coatings themselves are about six times thinner than those obtained in the first process (Fig. 3.7). To use a
saturating mixture repeatedly and to obtain the
constant thickness and composition of the coatings, a certain mixture beneficiation process
should be commercially used; that is, a certain
amount of the mixture of its initial composition
should be added to the reused one. When Al-Fe
alloy-based mixtures are used, 5 to 25% of the
initial alloy is added after every process. The
particles equal to 1 mm or smaller are screened.
The control of the added alloy parameters, that
is, the mixture activity, allows the control of the
aluminum content of the coating within the
range of 24 to 36%.
The powder method of aluminide coating application to turbine blades has some serious
drawbacks. Preparation of powder mixtures, arrangement of the components in the container,
and their removal from the container after aluminizing are labor consuming. It is difficult to
protect some surfaces (e.g., blade roots) from
aluminizing. There is some risk that the blade
cooling channels will be fouled with the powder
mixture. Blade cooling after aluminizing is slow
enough. Considerable emission of halides to the
atmosphere and a large volume of powder
wastes should be subjected to special utilization.
Updating the powder method may eliminate
some of the aforementioned drawbacks. The
powder mixture may be placed into special cavities in the container. The parts are arranged in
Influence of repeated use of saturating mixture on

aluminum concentration in (1) Al-Fe alloy, (2) on
aluminum content of the outer layer of coatings, and (3) on coating thickness on nickel-base superalloys. Aluminizing in mixture:
98% Al-Fe alloy and 2% NH4Cl at 950 C for 4 h. Source: Ref 1
Fig. 3.7
the container so that a free circulation of the halides released from the mixture at heating is ensured. The airtight container is blown by argonhydrogen or evacuated to the vacuum of 0.65 Pa
before aluminizing to prevent oxidation of the
parts and the mixture at heating. The transfer of
aluminum halides becomes more efficient due to
their forced circulation in the container.
Aluminizing powder mixtures can be used for
forming diffusion coatings on the inner cavity
surfaces of the cooled blades. To form the protective coating on the intricate inner surface, it
is necessary to use the process that guarantees
the transfer of aluminum to this surface. It occurs when the inner cavity of the blade is filled
with the powder mixture. The powder mixtures
used for this purpose should be free-flowing
bulk materials both initially and after their use
for aluminizing of the inner cavity surfaces. Besides that, they should not cake and stick fast to
the blade surface.
The technological process for forming a diffusion coating on the inner cavity surface combines the elements of aluminizing under vacuum
and aluminizing with the use of halides. It includes the following main stages: preparing a
powder mixture; filling the blade inner cavity
with the mixture; aluminizing; cleaning the
blade inner cavity surface from the powder mixture; and inspecting the coating obtained. The
chromium-aluminizing mixture is prepared of
aluminum, chromium, and aluminum oxide
powders (Ref 3, 4). Powders of individual components are mixed in the following proportions:
aluminum, 7 to 8%, chromium, 35 to 40%; and
aluminum oxide, balance. The mixture prepared
is thoroughly intermixed and annealed in
vacuum at 1000 C for 3 to 4 h. During that
period of time, chromium and aluminum interact
and form a solid solution of aluminum in chromium and lower chromium aluminides. If sintered, the mixture is crushed and sifted through
a sieve.
To activate the mixture, it should be annealed
again after adding 0.3 to 0.4% aluminum chloride (AlCl3). After being well intermixed with
aluminum chloride, the mixture is placed into a
leakproof container from which air has been expelled by argon blowing-through. Mixture annealing is conducted for 2 to 3 h. During annealing, the formation of aluminum chlorides
occurs in the mixture. At cooling, they condense
on powder particle surfaces. The chloride compounds activate aluminum and chromium transfer from powder particles to the surface saturated.
Adequate preparation of the inner cavity of

the blade for coating is a stage of great importance. It should guarantee required evenness and
thickness of the coating on the inner blade surface. The matter is that these surfaces have imbedded oxides, ceramic inclusions (from cores),
various impurities, casting porosity, and so
forth. During preparation of the inner cavity for
aluminizing, all impurities should be removed
by abrasive blasting. Blasting in two directions
(from the blade root and from its trailing edge)
is considered to be good practice, thereby eliminating shaded zones where processing otherwise would have been of poor quality.
The operation of filling saturation mixture
should assure complete filling up of blade inner
spaces: no coating is formed in the spaces unfilled with mixture. A relatively high fineness of
the powder-mixture components (100 lm) and
its good flowability enable most intricately
shaped interiors of the blade to be fully filled.
Since the mixture retains its free-flowing properties unaffected after the saturation procedure
is finished, it is easy to remove the mixture from
the blade inner spaces completely. To prevent
mixture outpour from the cooling channels
while filling, the channel outlets, where they
open onto the surface, are sealed with a thick
organic-based suspension.
The process of aluminizing is carried out under (1 to 3) 101 Pa vacuum at 1000 to 1030
C for 5 to 6 h. Aluminum-to-surface transfer
occurs first through a gas phase and then through
a vapor phase. The thickness of the coating
formed will depend on mixture activity, process
temperature, and time, as well as on the ratio of
a certain area of the blade inner cavity surface
being aluminized and the amount of mixture on
this very area.
The process of diffusion saturation involves
the mixture depletion of aluminum, the reduction of its activity, and the drop of coating formation rate (Ref 3). In 8 h, when held at 1050
C, the aluminum content of the mixture near a
saturated surface will drop from 7 to 2%, and
the zone of changed composition will penetrate
into the mixture depth by 2 to 3 mm. As a result,
in some positions of the inner cavity where blade
spacing is 1 to 2 mm, the saturating mixture rapidly becomes depleted of aluminum, and its activity lowers. The process of saturation ceases.
This is why the blade inner surfaces that have
been aluminized will inevitably show great variation in coating thickness. Saturating for 5 to 6
h at 1000 to 1030 C will form a coating 10 to
30 lm thick, containing 18 to 22% Al. The fact

that coating thickness depends not only on time
and temperature, but also on ratio between the
surface area saturated and the volume of the
mixture on the surface (this ratio being characteristic of the blade design type) dictates individual selection of saturation conditions for a
certain blade design.
To improve the process of aluminizing in
powders containing halides, new techniques for
diffusion coating formation have been developed.
Application of Diffusion Coatings in

Circulating Gaseous Phase
The circulation technique uses chemical processes similar to those that occur during aluminizing from powder mixtures containing halides.
The difference is in eliminating the preparation
and use of powder mixture. The technique is
based on forced circulation of the gaseous medium that contains aluminum halides (AlCl3,
AlCl2) in the space between the source of aluminum (i.e., liquid aluminum) and the blades
being aluminized. The required gaseous medium
is created in a special reaction zone by passing
gaseous halide AlCl3 over melted aluminum surface having a temperature approximately 900 to
1000 C (Ref 5, 6).
A fan directs the gaseous compound (AlCl2)
formed in the reaction zone toward the aluminizing zone where the blades heated up to 900
to 1000 C are stationed. On their surfaces, reaction of disproportioning takes place. As aluminum reacts with the superalloy, a coating is
formed whose aluminum content and structure
are close to those obtained by aluminizing from
powder mixtures with halide activators. The
AlCl3 formed in the aluminizing zone is headed
for the reaction zone again. Thus, the aluminizing process proceeds with halides continuously
circulating over melted aluminum and the blades
being aluminized. Temperatures in the zones of
the unit are chosen such that optimum conditions are produced for the reactions in the course
of which AlCl2 is first created and then decomposed on the blade surfaces.
Besides temperature, the velocity and pattern
of motion of gaseous halides markedly affect the
kinetics of reactions. In order to obtain a uniform
flow of saturation components and to ensure
steady reaction, the velocity of halide flow must
correspond to a minimum enough to impart turbulence to the gaseous mixture.
The design of the unit for the circulation technique of aluminizing is shown in Fig. 3.8 (Ref
7). The blades are placed on the shelves in the
gas-tight muffle. A system ensures halide circulation in the muffle. It consists of a fan and a
few special baffles. Aluminum used in the circulation technique of aluminizing is in a special
bath in the lower part of the muffle. The halides
that are necessary for the process form in the
reactor and then head for the muffle. The halides
that leave the muffle condense in the special system for exhaust gas purification.
Diffusion coating growth kinetics in the circulation technique of aluminizing depends on
the process temperature and time. The relationship between the coating thickness and the alloy
composition is observed. Besides these wellknown factors, the coating thickness depends on
gas flow velocity. The thickest coating forms
when gas flow changes from laminar to turbulent. In the outer zone, aluminum content is 22
to 26%.
The main advantage of the circulation technique is the possibility of deposition of diffusion
coatings on inner space surfaces of cooled
blades. In comparison with the blade outer surfaces, the inner surfaces work at lower (900
o
C) temperatures. Besides, severe erosion and
corrosion do not affect them. Instead, however,

they require coatings with rather high ductility
because the intricate profile of their inner surfaces features numerous stress concentrators.
The main difficulty here is to ensure adequate
thickness and composition of coating throughout full inner space. This task can be accomplished by blowing through the inner space a
saturating gaseous medium, with the gas well
agitated during the procedure. In this way a uniform mass transfer of aluminum from the carrier
gas to the surface can be ensured.
Uniformity of mass transfer in the blade inner
space has been examined, revealing the dependence of this characteristic on gas flow velocity.
The most uniform mass transfer is observed
when blowing through the inner space from both
ends, that is, from its root and trailing edge sides.
The use of the gaseous medium containing aluminum iodides (AlI, AlI3) helps provide optimum conditions for aluminum mass transfer
(Ref 6).
Experimental studies of the processes involved
in saturating the insides of blades with aluminum
by employing the circulation technique have
shown the following tendencies. The process
temperature and fan rotation speeds (i.e., gas-carrier flow velocities) appear to have affected most
essentially the coating thickness. Coating evenness was satisfactory in all the experiments as
measured at four opposite points of blade interior.
When applying the circulation technique for aluminizing, the aluminum content of coating varies
within 16 to 25%, which usually meets the requirements imposed on the protective properties
of the blade inside. The circulation technique can
be used for depositing chromium-base diffusion
coatings as well as diffusion coatings alloyed
with chromium and silicon.
The advantages of the circulation technique
are ecological friendliness (as no powder mixture is required) and even coating thickness on
both inner and outer blade surfaces. The disadvantage of this technique is corrosion damage of
equipment parts contacting with halides.
Application of Diffusion Coatings Using

Slurry
The design of the unit for applying the diffusion coatings by the circulating technology of aluminizing. 1,
furnace; 2, lower cover; 3, upper cover; 4, gas-tight muffle; 5,
director baffle; 6, fan; 7, saturating metal (liquid aluminum in the
circulating method of aluminizing); 8, reactor for halide formation; 9, turbine blades; 10, halide condensation system; 11,
vacuum pump; 12, loading mechanism. Source: Ref 7
Fig. 3.8
The application of aluminizing slurries by

painting followed by diffusion annealing is
widely used for obtaining aluminide protective
coatings on turbine blades and vanes (Ref 8, 9).
The use of slurries has a series of advantages
over aluminizing from powder mixtures. They
are as follows: a shorter thermal cycle of coating

application due to quick heating and cooling of
the treated part; a possibility of local aluminizing, for example, saturation of a blade airfoil
portion when its root is finished; a possibility of
applying coatings to large-size parts, for example, combustion chambers. The use of slurries
increases significantly the output of aluminizing
and allows process automation. The slurry is
made of the saturating element powders, such as
Al, Al-Si, Al-Si-(Y, Ce, B), and so forth, with
the particle size of less than 40 lm (as its active
component) and of an organic binder. The
binder solutions are selected so they can ensure
a good suspension of the active component, a
quick drying of the slurry on the saturated surface, and an even layer with the required
strength to withstand next operations. A readyto-use slurry should be good for a long-term
storage without deterioration of its properties
and have a viscosity facilitating its application
by various methods, such as dipping, brushing,
and spraying. Diffusion annealing should ensure
a complete enough burnout or volatilization of
the organic component without deteriorating the
alloy and the resulting diffusion coatings.
Slurries manufactured on the base of aluminum powder (as an active component) and of
the solution of colloxiline in the mixture of amylacetate and dimetyloxolate (as an organic
binder) are widely used. Colloxiline with its low
flashpoint of 180 to 200 C has a high decomposition rate at diffusion annealing of the treated
parts. This permits spending less of the slurry
and reducing the nitrogen and carbon contents
of it. The process features an intensive release
of mainly carbon-containing compounds at the
flashpoint. As the treated specimens are further
heated, all the remaining organic fractions are
removed when the temperature of 650 C is
reached. At the same time a part of the powder
is oxidized by the active component, resulting
in a respective gradual increase in the slurry
weight. As colloxiline burns out, the active component is concentrated. The surface-tension
forces make the powder particles converge and
partly take out the volume of the burned organic
components. It allows metallic contacts between
the particles of the saturating elements and the
saturated surface. This condition is necessary for
a diffusion coating formation.
The presence of oxides on both aluminum
particles and the saturated surface is an obstacle
for the diffusion processes. Yet, as the elements
are heated to temperatures close to the aluminum
melting point, the diffusion mobility of the atoms increases and surpasses these barriers.
Vacuum is the most favorable condition to guarantee the contact formation. It facilitates the
evaporation of the impurities and the removal of
the adsorbed gases. The diffusion layers obtained under vacuum are of even depth and high
surface-finish quality.
At heating, the saturating elements enter into
reactions with oxide films on the nickel-base alloy surface. These reactions favor the formation
of even, high-quality coatings; that is why the
aluminizing process under an atmosphere reducing the oxides is undesirable. The liquid phase
is most favorable for reactions. It guarantees
high rates of diffusion processes. That is why
the best quality layers are obtained at saturating
with several elements (Al-Si, Al-Si-Cr) when
the slurry has a significant aluminum content.
At saturation with aluminum, the liquid phase
consists of Al-NiAl3 eutectics that spread over
the surface, favoring the dissolution of the nearest solid particles and the absorption of the saturating elements on the whole surface of the alloy. The dissolution of the solid phase in the
liquid one goes on to the point when the element
contents in the melt reach the level when solid
refractory compounds may form. Their formation begins at the metal-melt interface, resulting
in a gradual solidification of the whole volume
of the liquid phase. After solidification, a specific dendrite pattern appears on the surface. At
further holding, the diffusion processes continue
in the solid phase, but their rate is much lower.
The diffusion coating forms actually during 10
to 15 min of the diffusion annealing; further
holding thickens the coating. The interaction between the active component of the slurry and the
saturated surface improves with the increase in
the temperature of diffusion annealing; that is
why higher temperatures result in the formation
of the coatings with more even thickness and
high quality finish.
Thus the process of saturation with aluminum,
or aluminum together with silicon, chromium,
and other elements, using slurries is a variant of
the liquid-phase aluminizing method, where the
applied layer of the slurry limits the liquid-phase
volume. The results of saturation, that is, the aluminum content of the coating formed, its alloying element contents, and, hence, hardness, ductility, and other parameters, depend on the
thickness of the applied slurry layer, the ratio of
aluminum and other metal-constituent contents
in its active component, diffusion annealing
temperature, and time. The liquid phase on the

saturated surface ensures favorable conditions
for adsorption and diffusion processes, but it is
not obligatory for diffusion coating formation.
A diffusion layer may form at saturation from
the solid phase as well. In this case the process
is more favorable for the formation of intermetallide phases with high heat of formation.
The slurries used for turbine blade aluminizing are prepared of a solution of colloxiline in
the mixture of amylacetate and dimetyloxolate
(as an organic binder) and of aluminum powder
(as an active component). The aluminum powder and colloxiline solution are mixed in the ratio of 600 to 700 g of the powder per 1 L of the
organic binder. Then it is filtered through a sieve
with a mesh size of up to 40 lm. The slurry is
thoroughly mixed before application. Its viscosity should ensure the possibility of its spraying.
At the same time, its running off the saturated
surface should be prevented.
At aluminizing from slurry, particular attention should be paid to the preparation of the part
surface. It should be cleared from dirt, grease,
and oxide particles. If the surface is of a low
class finish, if it has pits, freckles, and other
flaws where the slurry cannot penetrate, or, on
the contrary, where it will be thicker than required, all this can result in either uncovered areas or areas with too thick a coating after annealing. Hence, the saturated surface preparation
determines to a great extent aluminizing parameters and final coating quality.
The evenness of the coating thickness after

annealing depends completely on the evenness
of the applied slurry layer. Taking into account
the strict requirements to the evenness of the
blade coating thickness, a special attention
should be paid to the technique of slurry application; it should be automated if possible.
After diffusion annealing, a thick layer of the
slurry (80 lm) and its sags may cause a thick
brittle coating on these areas, prone to cracking
at cyclic loads. When the slurry with a complex
active component is used, special attention
should be paid to uniform distribution of the active component elements. When the saturating
elements have noticeably different specific
weight values (Al-Ta, Al-Nb), special mixers
should be used. They should prevent heavy fractions from precipitation while the slurry is being
applied to the alloy.
The slurry is applied to the surfaces of the
blades arranged in special facilities preventing
its application to the blade roots or other elements not to be saturated. The slurry layer on
the blade surface is then air dried and visually
inspected. Then the parts are placed in special
fixtures in the container; slurry damage should
be avoided. The parts are annealed either in
vacuum or under argon or air atmosphere (if mechanical finishing follows). The annealing conditions are selected in accordance with the heat
treatment conditions of the saturated alloy. Annealing at 950 C for 4 to 6 h ensures good results. The thickness of the resulting aluminide
coating, h, is a linear function of the applied
slurry thickness, H; it may be approximately calculated from the following relationships:
h 0.7 H for annealing at 950 C for 4 to 6 h
h H for annealing at 1200 C for 1 to 2 h
Influence of diffusion annealing temperature on

hardness (lines 1 and 2), and aluminum content of
the coating outer layer (line 3) on nickel-base superalloys. Aluminizing from slurry: 100% Al as an active component. Annealing time: (1) 1 h; (2) 6 h; (3) 46 h. Source: Ref 9
Fig. 3.9
A thicker layer of slurry (100 lm) may spall

at diffusion annealing. The geometric dimensions of the parts change while saturated. The
increase factor equal to the quotient of the specimen linear dimensions by the coating thickness
varies in the range of 0.9 (at annealing at 650
C) to 0.3 (at annealing at 1000 C).
The annealing temperature effects the surface
finish. The best results have been obtained
within the range of 850 to 1100 C. However, at
these temperatures the surface roughness increases somehow in comparison with the initial
one (Ra 0.32 to 0.25 lm). If the annealing
temperature is increased from 700 to 1000 C,
the aluminum content of the coating outer layer
drops from 43 to 18%. The coating hardness

drops respectively (Fig. 3.9) (Ref 9).
An important advantage of the method of aluminizing from slurry is the ease of saturation
with several elements such as aluminum and
chromium, aluminum and silicon, aluminum
and boron, and so forth. At multicomponent saturation, an active component of the slurry consists of a mixture of powders of the elements
alloying the aluminide coating. The powders
with the particle size of 40 lm are thoroughly
mixed and added to the amylacetate solution.
Further stages of the aluminizing process are
similar to those analyzed in previous sections.
Chemical composition, thickness, and properties
of the coatings formed after annealing depend
on the selected ratios of aluminum and the second element of the active component. In this
case, the best coatings are obtained if the active
component contains a large amount of aluminum powder (50%).
In the case of simultaneous saturation with
aluminum and silicon (Al-Si) followed by diffusion annealing at 950 C, the thickness of the
coatings on nickel-base superalloys reduces with
the increase in the amount of silicon in the active
component of the slurry. At annealing temperatures above 1050 C, the coating thickness increases (Fig. 3.10) (Ref 2). It is due to intensification of the process of saturation with silicon
at high temperatures. Diffusion annealing at
temperatures of 950 to 1050 C with the silicon
content of the active component of the slurry of

5 to 50% produces the aluminide coatings alloyed with silicon. If the silicon content of the
slurry active component exceeds 50%, the resulting coatings are brittle.
The main disadvantage of this method is a
need for application of the slurry by hand. It
causes uneven coatings on a blade surface. The
permissible variation in thickness between the
blades of one batch is usually 20 to 50 lm, yet
the real variation may be much greater. The variations in the coating thickness may be caused
by the presence of oxides and large carbides on
a blade surface. They hinder the diffusion processes and result in uncoated areas. The slurries
used contain organic components and metal
powders of high explosion hazard. That is why
the use of well-known electrostatic units for applying the slurry to blade surfaces is a problem.
These shortcomings can be overcome by using
new physical techniques and special units for
application of diffusion coatings.
Application of Diffusion Coatings by

Electric Arc Technique under Vacuum
The starting considerations at development of
new aluminizing techniques are as follows: selecting a proper active medium containing the
saturating element; providing a reliable means
for transfer of the saturating element to the surface to be saturated; and activating the transfer
of the saturating element from its medium to the
surface to be saturated. The technological process should ensure controllable conditions under
which the transfer of aluminum and the formation of the required diffusion layer occur at the
highest possible rate.
All these requirements are met by ionized
plasma flows, which are produced by the electric
arc (EA) technique (Ref 1013). The unit using
EA for material evaporation can:
Influence of silicon content of the active component of (Al-Si) of slurry on thickness of resulting
coatings on nickel-base superalloys during diffusion annealing at
850 to 1200C for 2 h. Source: Ref 2
Fig. 3.10
Conduct ion cleaning up and activating the

surfaces of saturated parts by treating their
surfaces with plasma flows with the ion energy that exceeds 200 eV
Produce active plasma phase from the ions of
saturating elements and bring them to the surface under saturation
Transfer ions from plasma flow to the surface
to be saturated and maintain required boundary conditions throughout the entire process
of treatment
The new technology for forming the diffusion
aluminide-type coatings on turbine blades (Ref
14, 15) is based on use of the EA technique. The

plasma accelerator cathode is made of an aluminum-base alloy. In the electric arc burning between the cathode and the anode, aluminum and
other elements of the cathode are evaporated,
ionized, and transferred to the surfaces of the
parts to be coated in the form of an ion flow.
The nature of surface reactions depends on the
energy of the ions in plasma, their number, the
physical and chemical characteristics of plasma
and the materials that the plasma interacts with,
and on the interaction temperature.
Depending on the negative potential applied
to the part being coated, the energy of the ions
deposited may vary within a wide range from
tens to hundreds of electron volts. The features
of the processes occurring on the surface under
saturation depend on the energy of ions, their
density, and the temperature of the surface. If
the energy of aluminum ions exceeds 200 eV,
the saturated surface will be cleaned up from
impurities and thin oxide layers. At the ion energy in the range of 10 to 50 eV, aluminum ions
will be transferred to the surface to be saturated.
The mechanism and kinetics of subsequent processes depend on the speed of aluminum ions
transferred and on the rate of their absorption
by the surface, depending on the surface temperature.
The rate of ion transfer to the saturated surface, m, can be calculated from the equation involving arc current in vacuum, Iea, and transfer
coefficient, l, depending on the type of material
evaporated (Ref 12, 16).
For aluminum alloys, the transfer coefficient,
l, is equal to (4.5 to 5.0) 105 gram/Coulomb.
Under steady working conditions of the unit,
when arc current, Iea, does not vary (Iea constant) the mass of aluminum alloy, m, transferred
to the blade surface within a period of time, t,
can be expressed in the following form: M
K l Iea t in g/m2 where Kratio (1/m2), depending on the type of the unit and on the surface area saturated. The average rate of aluminum transfer from the cathode to the part surface
is equal to 1 to 2 g/m2 per minute. Aluminum
alloy deposition time required for application of
a diffusion coating 25 to 50 lm thick (its thickness is measured after annealing at 1000 C for
4 h) is 20 to 40 min.
The mass of aluminum transferred, m, is directly proportional to the process parameters. It
allows full automation of the aluminizing process and formation of the coating with required
thickness and chemical composition.
Three alternative processes are feasible on the

surface being saturated, proceeding from the
amount of aluminum, m, transferred to the surface per time unit, and amount of aluminum,
mAl, that can be absorbed by the surface in
diffusion processes.
The amount of aluminum transferred to the

surface to be saturated considerably exceeds
aluminum mass that can form a diffusion
coating, that is, m mAl. It is possible at high
rates of condensation and low temperatures,
which are not high enough to assure proper
kinetics of diffusion processes (below 640
C). In this case a coating formed on the surface features the composition of that of the
feedstock evaporated. At the surface temperature above 640 C a layer of liquid aluminum
alloy will form. The kinetics of its growth is
dependent on the rate of aluminum ion condensation.
surface to be saturated is equal to the amount
of aluminum that can be absorbed through
diffusion at a given temperature: m mAl. In
this case, an aluminum-base diffusion coating
forms on the surface. Its aluminum content
reaches its possible maximum. The coating
composition and kinetics of its thickening depend on the temperature and kinetics of diffusion processes.
surface to be saturated is less than the amount
that can be absorbed through diffusion processes: m mAl. In this case, the diffusion
coating forms on the surface. The kinetics of
its growth depends on the rates of aluminum
ion transfer to the surface to be coated.
Aluminide-type coatings can be alloyed with
different elements by making the cathode of aluminum alloys containing various elements. This
type of alloying is inapplicable in other currently
used aluminizing techniques.
The Al-Si system was selected as a base system for development of coating compositions.
Its choice was dictated by the research into alloying the diffusion aluminide coatings with silicon that was conducted earlier. Silicon improves casting characteristics of aluminum
alloys. It enables cathode to be cast at conventional aluminum foundries. To improve cyclic
heat resistance of diffusion coatings and, hence,
to prolong their service lives, the aluminum alloys for coatings and the coatings themselves
were alloyed with yttrium, cerium, and some

other elements.
The technological cycle of forming diffusion
coatings consists of the following sequential
phases. As-finished blades are subjected to
clearing up and degreasing all over their surfaces. Next, the blade airfoil is blasted first with
synthetic corundum with the grain size of 30 to
60 lm and then with compressed air. Thus, prepared blades are assembled in special fixtures
that are then placed inside the unit chamber. The
unit is pumped to vacuum of 1 102 Pa.
The aluminum alloy is evaporated in electric arc
discharge struck in the unit between the aluminum cathode and the anode and transferred to
the blade airfoil surface.
As a rule, the units do not provide for the
blade preheating systems. Thus, interaction between plasma and the surface being coated occurs according to the first of the alternative processes. In this case the mass of the material
transferred to the surface, m, is much larger than
the mass of aluminum, that has entered into
forming the diffusion coating, mAl. As a result,
a dense layer of the alloy with the same composition as that of the feedstock evaporated is
formed on the blade airfoil surface. Usually it
takes 20 to 40 min to treat the blade in the
plasma flow.
After the blades are coated with the aluminum
alloy to a required thickness, they are taken out
of the unit and annealed under vacuum. The annealing temperature and time depend on kinetics
of diffusion and phase transformations. They are
chosen to meet the requirements to the coating
composition and thickness. After diffusion annealing, the blades will have a dark-gray color.
No additional treatment of the surface is required. The diffusion coating thickness and the
aluminum content of its outer layer depend on
the thickness of the aluminum layer deposited
and on the annealing time and temperature. The
thickness of the diffusion coating in use is usually 25 to 50 lm. Its aluminum content is 20 to
26%.
The main advantages of the technique in
question are as follows: automation of the process of aluminum alloy deposition on the blade
surfaces; precise reproducibility of the coatings
with specified compositions and thickness; the
possibility of coating deposition on the certain
blade surfaces and areas and, therefore, no problems of preventing coating deposition on the
other areas; and ecological friendliness of the
technological processes as no powders or halides are used.
Inspection of Diffusion Coatings

Diffusion aluminide coatings protect the
blade surface against high temperature gas corrosion and erosion and determine to a great extent the blade life and reliability. That is why the
quality control of the coatings deposited is a
very important stage of the technological process of diffusion saturation.
The main parameters determining the coating
protective properties and serviceability are its
composition and thickness. The chemical composition of the aluminide coatings is characterized mainly by its aluminum content. In the case
of alloyed aluminide coatings, aluminum and alloying element contents characterize it. The
thickness inspection and the analysis of the aluminum content of the coating (as well as that of
the alloying element for multicomponent saturation) are of primary importance for testing
coating quality and protective properties.
It is noteworthy that as a rule, no composition
analysis of the coatings deposited is performed
in commercial practice. This is due to the complexity of such an analysis. Also, the methods
of the coating formation from powder mixtures
with halide activators, which are widely used in
industry guarantee stable aluminide coatings
with a constant aluminum content. The situation
changes when the aluminide coatings alloyed
with chromium, silicon, and other elements are
formed. To guarantee the presence of the alloying elements, which impart the required properties to aluminum coatings, especially in the
cases where the element ratio in the saturating
composition may change at application, methods and equipment for direct analysis of the
composition of the coating formed should be developed.
The coating thickness is determined by a metallographic method or by measuring the increase in the dimensions of the reference specimens saturated together with a batch of blades.
For example, at saturating nickel alloys from a
mixture containing 98% Al-Fe (50% Al) alloy
and 2% NH4CI the side increment B of the specimen and the coating thickness, h, are connected
by the following relationship: B 0.6 h.
When the slurries are used for aluminizing, a
direct connection between the specific mass (g/
m2) of the applied layer and the thickness of the
coating formed is observed. In this method, the
thickness of the slurry layer is inspected. As the
physical properties of the applied slurry layer
differ from those of the base alloy, nondestruc-
tive testing may be used. The applied layer

thickness inspection is a critical and necessary
operation, because if the layers are too thick, it
will result in thick and brittle aluminide coatings
prone to spalling at operation.
In cases where aluminizing is well perfected
under the industrial conditions and the technique
in use guarantees the coatings with the thickness
and composition varying in a narrow known
range, indirect methods of quality and thickness
inspection may be used. These are testing hardness, measuring specific weight gain (g/m2) at
aluminizing, and metallographic structure analysis. For aluminide coatings formed on nickel
superalloys at the temperature of 900 to 1100 C
the connection between the aluminum content of
the outer layer of the diffusion coating and its
hardness is observed. The outer-layer hardness
is tested on metallographic specimens used for
coating thickness measurements. The outerlayer hardness may vary for alloys with different
chemical composition. This also depends on the
medium used for aluminizing, which contains
alloying elements in addition to aluminum. In
these cases the calibration curves of hardness/
aluminum concentration are plotted for each alloy, saturating medium, and the conditions of
aluminizing.
For indirect analysis of the aluminum content
of the aluminide coatings on nickel superalloys,
the metallographic method (comparison of the
outer and inner layer thickness) may also be
used. At the aluminum content of 33 to 36% in
the outer layer, the ratio of the outer and inner
layer thickness is about 4. At 24 to 26%, the
thickness ratio drops to 2 to 2.5.
The value of the specific mass gain (Dm, g/
m2) of the specimen after aluminizing characterizes to a great extent composition, thickness,
and protective properties of a diffusion coating.
In fact it is a measure of an aluminum reserve
in the coating; that is, mass of aluminum to be
consumed on the formation of protective oxides
during the blade service life. The media of certain compositions show a direct connection between the specimen weight gain and the thickness of the diffusion coatings formed.
The inadmissible flaws of the diffusion coatings are as follows: uncoated areas of the blade
surface to be protected and cracks in a coating.
Coating discontinuity can be detected by visual
inspection of the blades annealed under oxidizing atmosphere. The uncoated area is covered
with an oxide layer of a specific color. The
cracks are detected by routine nondestructive
testing.
Overlay Coatings
The electron-beam (EB) and electric-arc (EA)
coating techniques are based on the fundamental
physical processes of electron-substance interaction, evaporation, formation of a vapor cloud
or plasma flow, and condensation of substance.
Since these physical processes have been
broadly discussed in the literature (Ref 1719),
to treat them here in detail is not intended.
Some Physical Principles of Material

Evaporation by Electron Beam
Electron beam evaporation is based on thermal effects that occur as the flow of electrons
interacts with the substance. The EB evaporator
consists basically of an EB gun generating a directed flow of electrons possessing necessary
energy and an evaporation unit containing the
substance to be evaporated. The power of the
electron beam, Pe, will depend on the gun accelerating voltage, Ue, and on beam current, Je:
Pe Ue Je
in kWt. The EB gun accelerating voltage is chosen to meet power and economy requirements.
Increasing the voltage will complicate the highvoltage system and make the unit more expensive. Usually the EB coaters used in industry are
rated for 18 to 20 kV.
As an EB strikes the surface, it penetrates into
the substance being evaporated, imparting to it
its power, Pe. The depth of electron penetration
depends on their energy and the density of the
substance evaporated. Normally the depth of
penetration varies within 1 to 5 lm. As the electrons interact with the solid substance, their kinetic energy is translated into thermal energy
and Roentgen (x-ray) radiation. The thermal energy is spent to heat, melt, and evaporate the
substance, but some part of it will be lost
through heat radiation of the substance heated
and also due to a heat flow entering the evaporator wall. The evaporator thermal efficiency is
determined by the relationship between that part
of the power consumed for evaporation and the
total power of the beam. The efficiency depends
on the evaporator design. When evaporated by
the EB technique, 99% of the substance evaporated consists of neutral atoms and their complexes.
The rate of evaporation from surface area unit
is dependent on the physical properties of the
substance evaporated and on the temperature of

its heating:
V 4.4 104 po M/Tv
in g/s where V is the rate of evaporation from

surface area unit; is the coefficient of evaporation; po is the equilibrium pressure of vapor of
substance evaporated at a given temperature; M
is the molecular mass of substance evaporated;
and Tv is the temperature of evaporation.
When an alloy of two and more components
is heated and evaporated by the EB technique,
the alloy components are evaporated at different
rates. The ratio between evaporation rates depends on the activity coefficients of each component, ca and cb, taking into account the interrelation between melted components; on the
molar concentration of melted components, Ma
and Mb, and on the vapor partial pressures of
each component, pa and pb:
ab
cb pb Ma
ca pa Mb
1/2
With ab 1, both components evaporate simultaneously, and condensate composition is

identical to that of the substance. With ab 1,
the component b will have higher concentration
in condensate than in original substance. And
conversely, when ab 1, component b will
have lower concentration in condensate than in
original substance. In the process of evaporation, the pool content is continuously changing
and, hence, the value of ab also varies.
The time of evaporation, taken by condensate
to become elementally similar to the original
substance, is known as the stabilization time.
This time will vary with substance composition,
electron gun power, and evaporator design. Difference between the rates of evaporation of various components in complex alloys imposes certain restrictions on the employment of EB
technique. Such elements as Hf, W, Ta, Re, and
Si do not evaporate simultaneously with Ni, Cr,
and Al from the Ni (Co)-Cr-Al-base system. The
same can be said about the ceramic materials.
As the substance is evaporated on the evaporator surface, the partial pressure of its vapor
near the surface reaches 100 to 300 Pa, and a
directed flux of vapor is formed whose density
can be expressed through spatial distribution
U():
U() U0 cosn()
Accordingly, condensed substance thickness at

various points of the surface running parallel to
the evaporator surface and located at a certain
distance away from it will correlate as:
h() h0 cosn3()
where h0 is the thickness of condensed substance

layer (overlay) on the surface running parallel to
the evaporator surface at the 0 point opposite to
evaporator; h() is the thickness of a condensed
substance layer on the surface running parallel
to the evaporator surface, at a point which is r
away from the 0 point; and n is assumed to take
values in excess of 1 and be dependent on the
evaporator characteristics and rate of evaporation.
A multi-crucible system of evaporators and
facilities presetting the required movement of
the components past the evaporator are used to
make coating more uniform in thickness.
Some Physical Principles of Material

Evaporation by Electric Arc
The EA deposition coater operates by creating
an ion flux from the material evaporated by an
EA burning in vacuum (Ref 1921). The EA
strikes between the two electrodes, which are the
cathode and the anode. Depending on arcing
conditions, one of the electrodes (the cathode)
suffers erosion, so arcing occurs in the plasma
consisting of the ions of this electrode. The most
commonly used vacuum-arc discharge is that
initiated between the integrated cathode and anode. In this case arcing occurs between the cathode spots formed on the cathode surface and the
anode, being evenly spread over the surface of
the latter. The number of cathode spots is proportional to the arc current, the spots chaotically
moving about the cathode surface. High density
of the current (106 to 107 A/cm2) and energy
flux (107 to 108 W/cm2) in a cathode spot result
in intensive evaporation of the material and in
the formation of plasma flux with high flux velocity (104 m/s). The average energy of the
ions in the plasma flux is high enough for them
to break through the difference in potentials between the cathode and the anode and to move in
any direction. The consensus is that cathode
spots are the sources of plasma with the degree
of ionization close to 100%.
The high ionic current, consisting of the cathode material ions, can be derived from the
plasma of the vacuum arc burning on the cathode surface without affecting the conditions of
arcing. Total ionic current generated by plasma

of the arc under vacuum comprises up to 7 to
10% of the discharge current. The amount of
material evaporated per second, m, is proportional to electric arc current, Iea, and the cathode
material transfer coefficient, l (Ref 10, 20): m
l Iea in g/s. The coefficient of transfer in a
flux will depend on physical properties of the
material evaporated (Table 3.1) (Ref 21, 22).
The important parameters of the evaporation
process are the temperature of the cathode spot
and the pressure of evaporated material vapor in
this spot. High pressure in the cathode spot is a
must for evaporating the required amount of
substance (Table 3.2) (Ref 23).
In the cathode spot, the material is evaporated,
spattered, and ionized. The amount of vapor,
ions, and microdroplets varies with the properties of material evaporated, the design of evaporators, and the electric arc current. The amount
Table 3.1 Coefficients of transfer in the flux for
various materials
Cathode material
Tboiling
Zn
Ag
Al
Cu
Cr
Ni
W
1180
2483
2740
2868
2956
3003
5973
l 105, g/s
K
K
K
K
K
K
K
2.15
1.50
4.50
1.15
0.40
1.00
0.62
Source: Ref 22, 23
Table 3.2
materials
Cathode
material
Cathode spot parameters for various

Temperature
in cathode
spot, K
Tboiling/Tevaporation
2770
5210
9550
2340
1.15
0.90
1.73
0.81
Cu
W
Ta
Sn
Vapor
pressure
in spot, Pa
4.05
2.02
96
6.07
104
104
105
104
Source: Ref 23
Table 3.3 Mass-spectrometric analysis data for

material evaporation
Cathode
material
Al
Ni
Ion content
Degree of
ionization
5060%
6070%
Source: Ref 24, 25
Ions
1
Al , Al , Al
Ni1, Ni2, Ni3
M1
M2
M3
60
65
38
33
2
2
of microdroplet phase can comprise 1 to 30%

(by mass); the droplet size can vary within the
range of a fraction of a micrometer to tens of
micrometers. The degree of ionization of evaporated products also depends on the material
evaporated and the design of the cathode. The
degree of ionization rises as materials of higher
melting points are evaporated (Table 3.3) (Ref
24, 25).
The degree of ionization will also be higher
when a Hall end-face plasma accelerator is used
as an evaporator. Underlying the operating principle of such an accelerator is the presence of a
spatial nonuniform external magnetic field that
is generated by a solenoid in the simplest case
(Ref 26, 27). Under currents of 100 to 500 A,
the concentration of charged particles will be
1012 /cm3, so magnetic fields of 10 to 50 G intensity will be necessary for the Hall acceleration to manifest effectively.
The Hall plasma accelerator made in the form
of a magnetic coil wound around the cathode
enhances ionization in the flux, accelerates ions,
and collimates ion flow along the accelerator
axis. The microdroplet phase is charged with
drifting electrons and localized on flux periphery. Ionic energy can be controlled from tens to
hundreds of electron volts by using a plasma accelerator and by altering the induction of magnetic current. The degree of plasma flux ionization in the accelerator and its homogeneity
will be much dependent on magnetic field magnitude. The greater the magnitude is, the higher
is the degree of plasma ionization and the lesser
the homogeneity of plasma flux due to a sharper
focusing of plasma.
Differences in the valence of ions, their
atomic masses, and the effects of interaction between the magnetic field of accelerating coil and
the electric field of plasma flux bring about separation of ions and change condensate (overlay)
composition in comparison with the composition of the evaporated cathode. The evaporator
design has an effect on composition similarity
between the material evaporated and the coating
deposited. For instance, the evaporation of
Ni20Cr12AlY alloy will result in aluminum
content of the condensate dropping by 10 to
30%. Aluminum ions separate across the plasma
flux area (i.e., across the height of the unit working zone). Separation of heavier ions (for example Hf, W) is much less pronounced (5%).
The comparison of the chemical compositions
of the cathode and the coating deposited in the
EA coater are given in Table 3.4 (Ref 28).
The accelerator working conditions must be

chosen to meet maximum ionization of the
plasma outgoing flux and fall within admissible
inhomogeneity of ion current density (in the last
analysis of the coating thickness) across the flux
area. In choosing the accelerator working conditions, one should consider, first of all, the relationship between the rate of coating deposition
and the distance from the workpiece to the cathode surface. These measurements are taken at
different radii with respect to the plasma flux
axis.
After the accelerator working conditions have
been chosen, the working conditions of depositing also must be chosen, proceeding from the
negative potential applied to the workpiece. Increasing negative potential and, hence, intensifying the energy of ions on the workpiece surface, will bring about re-evaporation and
slowdown of deposition buildup. At the voltage
above 300 V, workpiece surfaces undergo the
processes of cleaning. As the negative potential
applied to a workpiece rises, the coating composition diverges from that of cathode.
The processes of arc burning, evaporation of
cathode material, and condensation of plasma
flux must occur in a vacuum not worse than 1
102 Pa; otherwise, high plasma activity will
Table 3.4 Chemical composition similarity of
Co28Cr5AlY cathode and coating
Chemical composition, %
Object
Cr
Al
Cathode
Coating
2729
2932
4.06.0
2.03.5
0.40.5
0.10.2
Source: Ref 28
Table 3.5
Specification of the UE-type coater
Parameter
Total power of coater
Number of guns for evaporating substance
Power of one gun for evaporating material
Number of guns for heating blades
Power of one gun for blade heating
Type of guns for heating and evaporation
Gun operating voltage
Number of cylindrical evaporators, 70 mm in
diameter each
Distance from evaporator surface to the rotary
center of a holder with blades
Vacuum required for coating deposition
Number of chambers for loading of blades
Maximum blade length that can be coated
Rate of metal coating deposition
Rate of applying ceramic coatings
Value
250 kW
4
60 kW
2
40 kW
Flat-beamed
1820 kV
4 pieces
330 mm
4 102 Pa
2
400 mm
1012 lm/min
45 lm/min
entail uncontrollable pollution of workpiece surfaces with oxides, carbides, and nitrides.
Units for Coating Deposition

EB Coater. The design concepts are similar
for all EB units. Basic differences are in the design and power of the EB guns in use (flat-beam
or axial guns), the number and the shape of the
evaporators (the most commonly used ones are
cylindrical evaporators with the diameters of
70 mm), and blade position relative to evaporators. Other differences can include technological process monitoring, control systems, and
some special facilities available.
The units for EB coating (UE-175, UE-187,
and UE206) designed at Paton Electric Welding
Institute (Table 3.5) are used for depositing
overlay coatings on turbine blades by the EB
technique (Ref 29, 30).
The UE coater consists of three interconnected vacuum chambers (Fig. 3.11). These are
the main processing chamber, where deposition
takes place, and two other chambers to load the
blades. The main processing chamber is divided
into two sections, one of which accommodates
EB guns and the other contains evaporators.
A multi-crucible evaporator of the coater with
linearly arranged evaporators enables depositing
different alloys and depositing strengthening oxide phases that are uniformly introduced into the
matrix, multilayer coatings, thermal barrier ceramic coatings, and so forth. All the crucibles
are made of copper and cooled with running water. The evaporators are provided with special
mechanisms that continuously feed the materials
to be evaporated. Two movable, mounted copper shields in the main processing chamber
make it possible to separate blades from evaporators.
Connected to the upper wall of the main processing chamber is a chamber with two EB guns
to heat the blades in the process of deposition.
The main processing chamber communicates
with the two chambers for blade loading through
side walls. Each loading chamber incorporates a
mechanism to secure a holder with blades and a
rod to advance the holder with blades into the
main processing chamber and make the blades
rotate and reciprocate over the evaporators.
Blades are heated, and flat-beam EB guns
evaporate materials with a linear hot cathode
(Ref 29). In operation, the EB is controlled by
deviating and scanning along two coordinate
axes. Each electron gun has individual packs of
power feed, filament current control, and an
electromagnetic system of deviation and scanning.

The coater has one source of accelerating
voltage with the power levels of 250 kW. Feedstock for evaporation is loaded into the evapo-
rators via a hole in the processing chamber. The

evaporators are loaded with ingots of evaporated
alloys 70 mm in diameter and up to 400 mm
long or with ceramic pellets 70 mm in diameter.
EA Coater. The MAP, Duga, and other units
Diagram of central chamber of UE coater for electron-beam deposition. 1, coating deposition chamber; 2, electron-beam
guns for material evaporation; 3, electron-beam guns for blade heating; 4, crucibles with material evaporated; 5, chamber
for blade loading; 6, facilities for fixing and rotating blades. Source: Ref 30
Fig. 3.11
Diagram of MAP unit for electric-arc deposition. 1, evaporable cathode; 2, anode; 3, vacuum chamber; 4, electromagnetic
catch of cathode spots; 5, 6, cathode vertical movement mechanism; 7, turbine blades; 8, cathode shield; 9, cachion
electrode; 10, power supply of negative voltage to the blade; 11, magnetic coil; 12, electric arc striking mechanism; 13, unit for protection
from microarc striking; 14, source voltage. (B-magnetic field, Iea-arc current) Source: Ref 13
Fig. 3.12
(coaters) are designed for depositing metal overlays on turbine blades and other engine components by EA technique (Ref 12, 13, 31).
The MAP Unit. (Fig. 3.12 and Table 3.6) designed at All Russian Institute of Aviation Materials (Ref 13) consists of a vacuum chamber
(3) accommodating a device for generating
plasma flux from evaporated substance. The device is made up of a cylindrical cathode (1), an
anode (2), and electric arc striking mechanism
(12). By means of a special mechanism (5, 6),
the cathode is moved vertically with respect to
the cooled interior of a magnetic coil, the latter
fixing the burning arc ring zone on the cathode
Table 3.6
Specification of the MAP Unit
Parameter
Value
Diameter of coating deposition

chamber
Chamber volume
Vacuum in coating deposition
chamber
Total power
Cylindrical cathode dimensions
Arc voltage
Arc current
Voltage on blades
Number of stations for holders with
blades
Maximum height of components
coated
Maximum diameter of components
coated
Number of blades coated at a time
Rate of MCrAlY alloy deposition
onto revolving blades
Unit overall dimensions
1000 mm
0.6 m3
5 103 Pa
75 kW
Height, 340 mm; OD, 180
mm; ID, 140 mm
35 V
From 400 to 1200 A
From 10 to 300 V
24
surface. Components (7) to be coated are arranged around the cathode on special holders,
the latter revolving about their own axes. The
unit is handled from above through the vacuum
chamber cover.
Duga-90 Unit. Another type of the unit for
coating deposition by EA technique is the Duga90 (Ref 3133) (Table 3.7) designed at Machine
Research Problems Institute, Russian Academy
of Science. Eight end-plasma evaporators are
used in the unit (Fig. 3.13). They are arranged
in pairs on the opposite walls of the chamber for
coating deposition. The dimensions of the chamber for coating deposition are 1150 mm in height
and 1000 mm in diameter. The rate of coating
deposition is 0.1 to 0.2 lm/min when one pair
of evaporators is used, arc discharge current is
100 A, and planetary workpiece rotation takes
place. The flat cathodes used in the evaporators
have diameters of 100 mm.
The main feature of the Duga-90 unit is a high
level of ion flow control. It allows even coating
deposition along the height of the working zone
of 800 mm. The difference in the thickness of
200 mm
130 mm
From 12 to 96
0.30.6 lm/min
3000 4000 2800 mm
OD, outside diameter; ID, inside diameter. Source: Ref 12, 13
Table 3.7
Specification of the Duga-90 Unit
Parameter
Diameter of coating deposition
chamber
Vacuum in chamber
Total power
Flat cathode diameter
Number of cathodes
Arc current
Voltage on blades
Number of stations for holders with
blades
Maximum height of blades coated
Maximum diameter of blades coated
Number of blades coated at a time
Rate of MCrAlY-type alloy
deposition on to revolving blades
Unit overall dimensions
Source: Ref 32, 33
Value
1000 mm
5 104 Pa
55 kW
100 mm
8
From 100 to 200 A
From 10 to 1500 V
8
800 mm
340 mm
From 1 to 8
0.10.2 lm/min
2500 2000 2100 mm
Chamber for electric-arc coating deposition of

Duga-90 unit from inside. In the center there are
blades 500 mm high; electric-arc evaporators are in the side
walls. Source: Ref 33
Fig. 3.13
this zone is not more than 20%. This unit may

be used for coating deposition on the blades with
the length of 800 mm. The blades are arranged
in a planetary mechanism that ensures their rotation in the plasma flow. The number of the
blades in the operating zone depends on the
blade dimensions.
To control the process parameters, the unit is
equipped with facilities for measuring rates of
coating deposition using quartz resonators and
for process-temperature monitoring using infrared pyrometers. The obvious advantage of the
unit is the use of ion gas sources for blade surface clearing before coating deposition.
The unit features the possibility of applying
high voltage (up to 1500 V) to the processed
blades. At this voltage, effective clearing of
Table 3.8 Variation of contents of the alloying

elements in the process of Ni20Co20Cr12AlY alloy melting
Chemical composition, wt.%
Melting technique
Charge
VI
VIEBR
Ni
Co
Cr
Al
Base
Base
Base
20.0
19.7
19.8
22.5
22.7
22.5
11.0
11.1
10.2
0.50
0.25
0.18
VI, vacuum induction; EBR, electron beam remelting
Table 3.9 Gas contents in various zones of ingot

Ni20Co20Cr12AlY produced without extra deoxidation
Gas contents, wt.%
Melting
technique
O2
H2
VI
VI
VI EBR
VI EBR
0.0036
0.0020
0.0024
0.0026
0.0024
0.0027
0.0020
0.0020
0.00030
0.00022
0.00018
0.00038
Sample
cutout location
Ingot top
Ingot middle
Ingot top
Ingot middle
VI, vacuum induction; EBR, electron-beam remelting
Table 3.10 Gas contents in ingots produced by

VI and subjected to extra deoxidation with carbon and cerium
Gas content, wt.%
Melting stage
Melt
Introducing 0.02% C
Introducing 0.02% Ce
Introducing 11% Al
Finished ingot
VI, vacuum induction
O2
0.0025
0.0014
0.0011
0.0011
0.0010
0.0023
0.0012
0.0012
0.0012
0.0010
blade surfaces with plasma ions takes place. It

allows eliminating the preliminary surface preparation and ensures high adhesive strength at the
interface of the substrate and the deposited coating.
Basic Elements of the Overlay Coating

Technique
This technique of coating turbine blades includes the following basic processes: producing
ingots and cathodes for coatings; preparing
blade surfaces for coating; depositing coatings
by the EB or EA techniques; peening and heat
treating coated blades; and inspecting the quality
of coatings.
Producing Ingots and Cathodes for Coatings. The ingots and cathodes produced for
evaporation should have sound structure. After
being turned, surfaces must be free of poor
melted spots, discontinuities, nonmetallic inclusions, cracks, sinks, and pits. There must be no
spray of metal from the liquid pool when evaporating the feedstock ingots and cathodes in
compliance with the requirements set forth for
deposition coatings.
The alloys for coatings deposited by the EB
technique are melted by a duplex process, including vacuum-induction (VI) melting and
electron beam remelting (EBR). Only the VI
melting is used for making cathodes. Instead of
EBR, vacuum arc remelting (VAR) can be used,
meeting all the requirements to gas content of
the ingots and to their density. Only VI melting
is used for producing the cathodes for electric
arc evaporation.
During VI melting, the alloying elements are
introduced in the following succession: nickel,
cobalt, chromium, and aluminum. As the melting process approaches its completion, yttrium
(NiY, master alloy) is introduced into the melt.
Analysis of variations in alloying element contents during VI melting and subsequent processes has shown that they are within the limits
presented in Table 3.8.
Such variation of yttrium concentration
should be attributed to its oxidation during melting. The oxidation of yttrium in the melt can,
however, be reduced by additional deoxidation
of the melt before introducing yttrium into it.
Deoxidation can be effected by introducing carbon and cerium in the amount of 0.01 to 0.1%
either into the NiCr melt. Such a working cycle
enables obtaining the materials with minimum
content of gases and nonmetallic intrusions (Tables 3.9 and 3.10). Preliminary deoxidization of
the melt will reduce yttrium losses in melting by

20 to 25%.
Homogeneity of ingot chemical composition
over the ingot height (i.e., its macrosegregation
value) is undoubtedly very important for ensuring coating process stability. As studies have
shown, applying a two-stage process (VI
EBR) reduces macrosegregation to a low level
(Table 3.11).
Table 3.11
Macrosegregation in ingot
Chemical composition (a),
wt%
Cutout spot
Ingot, top
Center
Edge
Ingot, bottom
Center
Edge
Cr
Co
Al
14.45
13.22
14.42
12.42
14.58
14.12
14.42
14.25
29.84
25.93
29.84
26.27
30.33
28.36
30.33
28.28
8.19
7.84
8.12
7.80
8.15
8.16
8.22
8.38
0.08
0.035
0.12
0.036
0.12
0.041
0.08
0.038
(a) Top figure, vacuum-induction (VI) melting; bottom figure, VI melting

electron-beam remelting (VI EBR)
Influence of surface preparation methods on the

number of intercrystalline defects (leaders) in
coating: (a) sand blasting (Ra 57 lm); (b) vibropolishing (Ra
1.25 lm). Source: Ref 36
Fig. 3.14
Preparing Blade Surfaces for Coating. Preparing operations are conducted to make blade
surfaces activated, uniform in outer appearance,
and clear of organic and nonorganic impurities.
The quality of a finished coating to a great extent
depends on how carefully the surface has been
prepared. Preferable surface preparing treatments before coating include hydroabrasive
blasting, vibropolishing, rinsing, and degreasing
in aqueous and organic solutions.
The EB technique imposes stricter requirements to the preparedness of surfaces onto
which the evaporated material will be deposited.
This is dictated by intercrystalline defects of
structure (known as leaders) that tend to form
on areas running at greater than 20 to 30 degrees
with respect to the direction on which the vapor
flux comes. Such intercrystalline defects should
be attributed to the effects of shadowing microareas of surface with surface microirregularities (traces of machining, spalling of carbides,
etc.) and to a sharp decrease of overlay buildup
in such areas (Ref 34, 35).
The occurrence of intercrystalline defects is
assisted by rotation of blades in the vapor flux
when each part of surface at a certain moment
is presented to the vapor flux heading at an angle
greater than 20 to 30. Intercrystalline defects
formed in surface depressions are elongated in
parallel with the axis about which the workpiece
rotates in the vapor flux. Especially amenable to
intercrystalline defects are convex surfaces
(blade suction surfaces); conversely, concave
surfaces (blade pressure surfaces) are much less
vulnerable to the occurrence of such defects.
Investigating how the surface preparation
techniques influence the formation of intercrystalline defects revealed that the number of such
defects and their depth penetration into the coating correlate with roughness parameters. Particularly coarse defects occur in sandblasted areas whose roughness, Ra, is 5 to 10 lm and on
ground surfaces having rough notches elongated
in parallel with the workpiece rotary axis. The
defects grow through the full thickness of coating. The smaller the roughness, the fewer is the
number of defects per unit area and the lesser
the depth of the coating the defects penetrate.
When applying the machining methods, which
keep surface roughness within Ra 1.25 lm,
the number of intercrystalline defects in coatings
up to 100 lm thick will not be great, and the
defects will penetrate only through approximately 13 of coating thickness (Fig. 3.14) (Ref
36).
Thus, any machining method applied to machined blade surfaces to which coatings will be
subsequently applied by the EB technique must

ensure surface finish not worse than Ra 1.25
lm. Various types of mechanical processing can
be applied, the choice being dictated by the
available equipment, such as hydroabrasive processing, vibropolishing, and so forth.
Substantial rise of labor-intensity in further
machining aimed at greater decrease of surface
roughness is justifiable only if the coating working characteristics (thermal fatigue resistance in
particular) are improved in this way. To lower
the risk of intercrystalline defect occurrence in
blades with high thermal stresses, the surfaces
of such blades should be machined to R 0.63
lm. Such surface finish can be obtained by vibropolishing, as a result of which the surface
will acquire a smooth-bright homogeneous surface. Special bodies or granules are used to conduct the vibropolishing process. The vibropolishing procedure is followed by thorough rinsing
and degreasing of blades.
The surfaces of cast blades made of modern
superalloys have quite a number of voids and
Influence of blade surface radius and heating temperature on the number of intercrystalline defects
(leaders) in the coating over 400 lm length. Source: Ref 36
Fig. 3.15
Table 3.12 Effect of surface preparation on adhesion strength of Ni20Cr12AlY coating deposited by electric arc technique
Method of surface preparation
Manual polishing
Vibropolishing
Hydro-abrasive blasting
Abrasive blasting with corundum (5060 lm)
Source: Ref 37
Adhesion
strength, MPa
6
9
24
36
pipes whose size in cross section may reach 20

lm and coarse carbide inclusions. In the course
of the machining process, these defects are
partly preserved, the carbide inclusions having
a knack of spalling off. It is at these voids, pipes,
and spalled-off carbides where intercrystalline
defects form during coating deposition.
The defects can be minimized and surface
roughness simultaneously reduced down to Ra
0.32 lm with the following method of blade
surface preparation. First, the blade airfoil is
machined as described earlier. Then, on an appropriately prepared surface, a layer of
Ni20Cr6AlY coating 20 to 40 lm thick is deposited. After annealing in vacuum at 1000 to
1050 C, the blades are vibropolished. Thanks
to a high structural homogeneity and good
machineability of the coating the vibropolishing
will make the surface uniformly rough and not
coarser than Ra 0.32 lm throughout. It results
in a defectless coating possessing a good combination of characteristics when tested for endurance and thermal fatigue.
The number of intercrystalline defects in EBdeposited coatings will depend not only on how
carefully the surface has been prepared, but also
on the temperature of blade heating and on the
radii of curvature of the surfaces coated. Heating
blades above 1000 C will markedly reduce the
number of coating defects (Fig. 3.15) (Ref 36).
No defects occur, in fact, in the coatings deposited on concave surfaces of blades (pressure surfaces), yet their number grows with the decrease
of curvature radii on the leading and trailing
edges of blades.
The EA technique imposes less stringent requirements on the preparation of blade surfaces
on which coatings must be deposited. Though
the demands of clearing the surface of organic
and inorganic impurities remain the same, surface roughness may constitute Ra 5 to 7 lm.
Such surface finish can be obtained by processing the surface with synthetic corundum (with
grain size of 30 to 60 lm). When depositing
coating by this method, the turbine blade is not
heated excessively, the chief aim being to ensure
sufficiently high adhesion of coating and to prevent its spalling before high-temperature annealing. The relationship between the adhesion
strength of EA-deposited coating and surface
preparation technique is illustrated in Table 3.12
(Ref 37). All the measurements are carried out
before high-temperature annealing.
The basic preparatory operations include rinsing and degreasing of blade surfaces, such as
washing outside and inside surfaces with hot water, and rinsing outside and inside surfaces of
blades in a degreasing aqueous solution.
After washing and drying, the blades are
weighed on scales loaded into a clean and closed
container and conveyed to the coating deposition area. Note that the admissible period during
which workpieces should be kept in a clean and
closed metal container between the last preparatory operation and their placement into chamber for depositing and vacuuming must not exceed 10 h.
Base Stages of the Technological Cycle for
Coating Deposition by the EB Technique.
The cycle of coating deposition includes the following steps: loading the unit with the blades;
vacuum pumping to the working condition;
blade heating to the required temperature; coating deposition; and bleeding-in and unloading
the unit.
Once the unit is loaded with the blades and
the vacuum in the unit reaches 5 102 Pa,
the electron guns of the evaporator come into
operation. The number of evaporators working
depends on blade sizes, blade holder design, and
demands on the evenness of coating thickness.
The coating alloy, while evaporating, must
form near the workpiece surface a homogeneous
and uniformly dense cloud of vapor, ensuring
the required parameters of condensation (overlaying). An evenly thick coating with uniformly
distributed alloying elements can be ensured by
certain practical methods.
One of them is ensuring continuous evaporation by continuously feeding the feedstock
into a liquid pool. Meeting this requirement
makes the evaporation process steady in time.
Such time of stabilization depends on the chemical composition of the substance evaporated. It
is also dependent on the conditions of the evaporation process, such as specific power input,
size of liquid pool, and intensity of convection
mixing processes going in the pool.
The time necessary to stabilize the vapor is
determined by experiment, that is, by analyzing
the changes of condensate (coating) composition
in time. Research into the evaporation of Ni18.6%Cr-12.5%Al-0.3%Y alloy in a watercooled crucible with the diameter of 70 mm in
the UE unit (the dimensions are the same for all
the units of this type) has shown that chromium
and aluminum contents of the vapor stabilize
within 15 to 20 min. Further on, the stability of
condensate composition depends principally on
the stability of EB gun operation coupled with
uniform feed rate of the ingots into the evaporation zone, position of a liquid pool, and some
other factors.
It takes much more time to stabilize the yttrium content of vapor. During the first 40 to 60
minutes, its content of the vapor remains within
0.01 to 0.05% and only gradually, after 3 to 4 h
of evaporation, it will equalize the ingot level.
By that moment, yttrium content of the liquid
pool reaches 10 to 15% (Fig. 3.16). Introducing
yttrium into the pool at the initial stage of evaporation can accelerate vapor stabilization of the
yttrium content (or of that of some other refractory elements). The required quantity of yttrium
is determined from the volume of liquid pool
and yttrium content of it (the latter must be 10
to 15%), since such amount of yttrium will ensure its content of vapor within 0.1 to 0.3%. The
same goal can be achieved with the aid of a special start ingot whose composition includes 10
to 15% content of the alloy. It should be noted,
however, that increasing yttrium content of liquid pool would initiate the formation of the slag
consisting of yttrium oxides on the pool surface.
This decelerates the evaporation process and
causes splashing.
Raising the EB specific power brought to the
pool surface gives rise to higher evaporation
rates and speeds up stabilization of vapor elemental content. On the other hand, intensified
evaporation results in overheating and splashing
of liquid pool and in producing microdroplets
that drop on workpiece surfaces and degrade
coating quality. Lowering EB specific power
lengthens the time of vapor stabilization and
produces coatings of variable chemical composition within their thickness. The UE-type unit,
having acceleration voltage of 18 kV and EB
current of 1.8 to 2.2 A, will evaporate
Ni20Cr12AlY and Ni20Co20Cr12AlY alloys at
the rates of 30 to 50 g/min.
When an EB coating technique is used, vapor
condenses on heated blade surfaces. Structural
features of the coating to be obtained dictate the
choice of blade heating temperature in coating.
The metallic coating must be as solid as possible
and have a minimum number of intercrystalline
defects, whereas the ceramic coating must have
columnar structure whose porosity at crystallite
height is kept at the minimum level. Its crystallite size must be 0.5 to 1.0 lm.
There are two boundary temperatures, T1 and
T2, at which condensate structure undergoes
changes (Fig. 3.17) (Ref 38). In the first lowtemperature zone, at temperature below T1 (T1
Fig. 3.16
(a) Microstructure of Ni20Cr12AlY ingot bath after 20 deposition processes (ingot diameter, 70 mm) (500). (b) Distribution of elements in bath after 20 deposition processes
Fig. 3.17
Influence of temperature on the structure of overlay coatings. Source: Ref 38
Table 3.13 Liquidus-solidus temperatures and

crystallization intervals of Ni-Cr-Al and Ni-CoCr-Al alloys
Alloy
Ni10Co15Cr8Al
Ni30Co15Cr8Al
Ni10Co30Cr8Al
Ni30Co30Cr8Al
Ni10Co15Cr14Al
Ni30Co15Cr14Al
Ni10Co30Cr14Al
Ni30Co30Cr14Al
Ni10Co22Cr11AlY
Ni30Co22Cr11AlY
Ni20Co15Cr11AlY
Ni20Co20Cr11AlY
Ni22Cr11AlY
Ni14Cr3AlY
Ni18Cr5AlY
Ni16Cr9AlY
Ni21Cr9AlY
Ni18Cr16AlY
Ni19Cr12AlY
Tliquidus, oC
Tsolidus, oC
Dcryst, oC
1357
1368
1320
1356
1452
1481
1463
1457
1422
1390
1379
1370
1360
1424
1407
1363
1346
1482
1400
1280
1284
1240
1278
1256
1279
1320
1304
1275
1255
1274
1275
1293
1384
1336
1305
1293
1275
1283
77
84
80
78
196
202
143
153
147
135
105
95
67
40
71
58
53
207
117
Distribution of Ni20Cr12AlY coating thickness

across middle section of turbine blade. The blades
are stationary, fixed in a cassette.
Fig. 3.18
0.3 Tliquidus), the coating will embody separate

cone-shaped crystallites and quite a number of
micropores. The other temperature zone lies between T1 and T2 (T2 0.45 to 0.50 Tliquidus).
The coatings in this zone will have columnar
structure and be free of microporosity. Above
T2, the coatings will have equiaxed structure, the
formation associated with diffusion processes
that occur between separate crystallites. Clearcut division into structural zones is typical of
thick overlays of solid solutions of refined metals. In the two-phase systems, temperature T2
will be offset to a high-value range when highmelting phases (oxides, carbines, intermetallides, etc.) are involved.
When depositing metallic coating on turbine
blades, the temperature of blade heating should
exceed T2 by 200 to 250 C. This is necessary
to intensify the diffusion processes between the
coating being formed and the blade surface. Liquidus and solidus temperatures for certain alloys can be found in Table 3.13.
The regression equation, showing the relationship between liquidus and solidus temperature on the one hand and the alloying of the NiCo-Cr-Al system alloys on the other, is given in
the Appendix. The time necessary to heat blades
up to 850 to 950 C will vary with blade size
and design of holders on which blades are secured. In the coaters of UE-type it usually takes
10 to 15 min.
Blade movement over the evaporators is programmed to bring about the required spreading
of the coating over the blade surface. The unit
feature of great importance is the distance from
the surface of metal evaporated to the blades.
The amount of metal overlaid on blades will
vary as the inverse square of the distance from
the evaporator. Hence, the thickness of overlay
on the blade airfoil will vary as a function of
blade motion characteristics in the vapor cloud.
Thus, coating will inevitably be thicker on those
blade areas whose surfaces follow the circumference of maximum radius, approaching most
closely to evaporator. Figure 3.18 shows a typical coating spread over the turbine blades. The
coating has been deposited with the blades positively secured in relation to the holder rotary
axis.
Coating thickness on the pressure and suction
surfaces can be varied by altering the angle at
which the blade is secured on the holder and by
using various masking. To guarantee coating
evenness over the full blade airfoil surface, it is
necessary to use the holders, which allow blade
rotation about its own axis.
Maintaining the required temperature of blade

surfaces in the process of coating deposition is
undoubtedly of great importance when designing blade holders and their rotation parameters,
since the rotation of blades in the coater affect
temperature distribution over blade surfaces.
Due to breaking of vacuum, breakdowns in
high-voltage circuit, and so forth, the evaporation of material may cease when depositing coating on the blades. Ceasing of deposition will result in discontinuities in overlays and coating
buildups. If this is the case when depositing a
metallic coating, then, after the cause of failure
is eliminated, the deposition process can be resumed and completed. In the event of a ceramic
coating failure, the process must be stopped and
the part rejected.
The process of deposition being over, the
holder with blades is handled to the loading
chamber. The blades cool down there to 150 to
200 C. After the cooling is over, the chamber
is opened and the holder with blades is taken
out.
The duration of the coater continuous operation will depend on the consumption of material
evaporated. With 100 lm thick metallic coatings, the continuous operation time of the UEtype coater will be 10 to 15 h.
Base Stages of Technological Cycle for
Coatings Deposition by the Electric Arc
Technique. The cathode material should be
sound, without any cavities or other metallurgical defects. No cracks or other metallurgical
defects are permissible on the cathode surface.
When arrangements for fixing of the blades are
designed, the possible nonuniformity of coating
deposition along the height of the operating zone
should be considered. In MAP unit for the blade
with the height of 200 mm, nonuniformity of the
coating thickness may be such that in the central
zone the coating is up to twice as thick as in the
peripheral zone.
As a rule, all electric arc units are of batch
operation, as the cathode assembly and the
blades are situated in the same chamber. The
coating deposition cycle includes loading of the
parts into the unit; pumping to operating
vacuum; ion clearing of the blades; coating deposition; bleeding-in; and unloading of the unit.
After loading of the blades into the unit, the
unit is covered with a cap and the unit chamber
is pumped to the vacuum of 1 to 5 102 Pa.
Pumping time is usually 30 to 40 min. After the
required vacuum is reached, the process of ion
clearing of the blades in the cathode material
plasma is carried out. To carry out this process

of clearing, the electric arc is drawn between the
cathode and the anode. The negative voltage of
150 to 1000 V is applied to the parts. Under
these conditions ion current is 20 to 30 A. The
process of ion clearing of blade surfaces takes 3
to 5 min; its end is determined from ion current
stabilization on the blades. The clearing process
is accompanied by uncontrolled part heating. It
should be taken into consideration for thin-wall
parts, as their heating beyond the temperature of
500 C is not permissible.
When the stage of ion clearing is finished, the
negative potential on the parts is lowered to 5 to
15 V. The operating arc current and its stable
arcing is set by adjustment of electric parameters. No special blade heating typical of EB technique is carried out.
When the unit electric parameters are stable,
the cathode material evaporation rate and the
coating thickness are proportional to electric arc
current and time of coating deposition. In the
electric-arc unit the material utilization factor
depends on the number of blades and their arrangement in the chamber. It is usually 10 to
15%. After the process of coating deposition is
over, bleeding-in is allowed and the holders with
blades are removed from the unit and passed to
the further procedures of the technological process.
Peening Finished Overlay Coatings. Blades
with overlay coatings undergo some (thermal
and mechanical) finish treatment and processing. To obtain a higher coating-to-surface adhesion strength, the blades are annealed for 2 to
4 h under (1 to 4) 102 Pa vacuum at 950 to
1050 C. The choice of annealing temperature is
dictated by the need to preserve the mechanical
characteristics of a superalloy and also by the
formation of a diffusion zone of the alloy-coating boundary, thus assuring necessary strength
of that boundary.
The annealing operation is mandatory for EAdeposited coatings, as these have low adhesion
immediately after coating deposition. The EBdeposited coatings demonstrate fairly high adhesion, provided all relevant technological parameters are strictly adhered to. This is why such
coatings may be subjected to peening, obviating
the annealing stage (Table 3.14).
Peening EB-deposited coatings with 150 to
300 lm diameter micro-balls made of steel, superalloy, or glass is the next operation. Coatings
deposited by EA technique feature dense structure without intercrystalline defects. They do not
require peening with micro-balls. In the course

of peening, the coating is compacted due to high
compression stresses (up to 1100 MPa). During
peening, the intercrystalline defects, formed on
the blade surface irregularities, are squeezed.
Of utmost importance in peening is uniform cold
working all over the blade surface coated. In this
case, the micro-balls must hit the surface at an
incidence angle close to 90. Generally it takes
3 to 4 min to blast around a blade up to 120 mm
high. The surface roughness of the blades
peened with micro-balls must not be more than
1.25 lm.
Strain-hardening processing can involve other
techniques, such as vibropolishing, glassball
blasting and so forth. The chief requirement to
any technique chosen is to obtain uniform cold
working on all coated surfaces.
Peening produces appreciable compression
stresses in the coatings. In the event of poor adhesion of the coating, the latter may spall.
After peening is completed, the blades are degreased, washed, and sent to recrystallization
annealing. The temperature and time of such annealing are chosen so the superalloy preserves
its mechanical properties and recrystallization
processes in the coating are unimpeded. As a
rule, the annealing is carried out in (1 to 5)
102 Pa vacuum at 1000 to 1050 C for 2 to 4
h. Having passed peening and heat treatment, the
coatings will have dense, equiaxed, fine-grained
structure.
When necessary, the coated blades may be
subjected to additional heat treatment of aging
if it is required to guarantee the specified mechanical properties.
All blades are subject to mandatory checkcontrol operations to see whether the blades with
heat-resistant overlay coatings meet the requirements of the process and design documents (Ref
39).
Table 3.14
coating
Deposition
technique
EB
EB
EA
EA
Adhesion strength of Ni20Cr12AlY

Method of surface
preparation, operation
Adhesion
strength, MPa
Vibro-polishing, after deposition

Vibro-polishing, annealing 1050
o
C, 2 h
Abrasive blasting with corundum,
after deposition
Abrasive blasting with corundum,
annealing 1050 oC, 2 h
110 (a)
(a) Determined at tear-apart testing of brazed joints
130 (a)
36
130 (a)
Inspecting of Overlay Coatings. The quality

of overlay is dependent on the stability of technique used and on the following characteristics.
Thickness of coating and its even spreading
over the blade airfoil must be consistent with
preset specifications. The difference in coating thickness between batches of blades is tolerable within 10% or less.
Deposited coatings must be uniform in composition through a full depth. Batch-to-batch
elemental variation must be within the tolerances specified.
After being deposited and annealed, the coating must be strong enough to prevent spalling
in operation (for metallic coatings, radhesion
100 MPa; for ceramic ones, radhesion 70
MPa).
The surface of EB-deposited coating must
have surface roughness within Ra 1.25 lm
after all processes have been completed.
Voids, intercrystalline defects, and droplets
are objectionable both on the coating surface
and in its structure. No cracks whatsoever is
tolerable.
The coating quality will depend on the quality
of superalloy ingots and ceramic briquettes delivered, on how thoroughly the conditions of
blade preparation are met, and on how coated
blades are processed. To meet all the abovementioned requirements, mandatory controlling
operations are performed at all stages of the process (Ref 27).
All as-delivered ingots of superalloys must
have marks and a certificate with data on their
chemical composition and the results of testing
for splashing of the bath delivered. The ceramic
briquettes are checked for quality by chemical
and phase analyses of powder samples.
Prior to depositing coatings on blades with
prepared surfaces, the latter should be inspected
visually for consistency with standard reference
samples. Before securing blades to their holders,
the blades should be checked by weighing to
0.1 g accuracy.
Flat blank-samples of wrought superalloy are
loaded together with working blades into the
holder. Such blank samples must have their surfaces prepared like those of the working blades.
The EB UE-type unit controls the following
parameters while depositing coatings on the
blades:
Vacuum in the main process chamber

Voltage and current of the electron beams of
evaporating and preheating guns and geom-
etry of beam area on the surface of substance

evaporated and parameters of its scanning
Temperature of blades and their motion over
the evaporator
Level of pool of evaporated substance and
rate of its feed to evaporators
Rate of condensation (overlaying) of evaporated substance
Time of coating deposition
In the EA unit, the following parameters are
controlled in work:
Vacuum in the main process chamber

Voltage, arc current on cathode, and stability
of arc burning
Ionic current on blades
Coating deposition cycle time
All of these parameters should meet the requirements specified in the process documentation based on the results of test-operation during
coating development.
After the deposition process is over and the
blades are taken out of the coater, the coating is
inspected visually by comparing it with the reference sample. No droplets or any other defects
are tolerable.
The adhesion of coating is tested prior to (for
EB technique) or after (for EA technique) annealing of the blades. Flat blank samples coated
simultaneously with a batch of blades are tested.
The samples are bent over a mandrel of 3 mm
radius to the angle of 90. There must be no
spalling during their bending. Flaws in the coating should not be regarded as grounds for rejection. The same requirements hold for ceramic
coatings.
If the blank-samples show poor adhesion
while tested for bending, then all the blades from
the same batch are rejected and sent for coating
removal. After the surfaces have been prepared
anew, they are recoated.
Having passed peening, all blades are visually
inspected for detection of spalling and swelling.
Any blade with such a defect is rejected. Its coating is removed, and its surface is prepared anew
and then recoated.
Average thickness of blade coating is inferred
from weight gain correlation charts plotted for
each blade design version, such charts showing
the relationship of coating thickness to blade
mass gain after coating. Thickness of ceramic
coatings can be examined using various thickness gages.
Necessary chemical and spectroscopic techniques are used to analyze coating chemical
composition. Coating microstructure is inspected while running-in the working process,

introducing some alterations, and selectively for
some processes conducted within a month.
All process operations being finished, the
blades are flow checked.
REFERENCES
1. A.G. Andreeva, V.V. Terekova, and G.D.

Fomenko, Heat-Resistance Coatings of
Nickel-Base Alloys, High-Temperature
Coatings, Leningrad, Nauka, 1967, p 96
110
2. G.V. Zemskov and R.I. Kogan, Multicomponent Diffusion Saturation of Metals and
Alloys, Metalurgiya, 1978, 207 p
3. P.T. Kolomytzev, Heat Resistant Diffusion
Coatings, 1979, 272 p
4. N.V. Abraimov, High-Temperature Materials and Coatings for Gas Turbines, Mashinostroene, 1993, 336 p
5. B.N. Arzamasov, Chemical-Heat Treatment
of Metals in Activated Gaseous Media,
Mashinostroene, 1979, 224 p
6. B.N. Arzamasov, Processes and Equipment
for Ecologically Sound Chemical Heat
Treatment, Vestnik Machinostroeniy, N 5,
1996, p 2628
7. B.N. Arzamasov, A.K. Beliavskiy, L.M.
Mulakaev, and G.B. Stroganov, Circulating
Method of Diffusion Coatings Application,
Vestnik Machinostroeniy, N 11, 1991, p 43
48
8. Yu.A. Tamarin, Heat Resistant Diffusion
Coatings for Turbine Blades, Mashinostroene, 1978, 133 p
9. T.V. Levchenko, V.I. Moroz, and L.P. Buiyanova, Protective Coatings on Metals,
Naukova Dumka, 4, 1971, p 158164
10. T.I. Blinov, A.M. Dorodnov, V.Ye Minaychev, S.A. Muboyajan, et al., Vacuum High
Current Plasma Devices and their Use in
Microelectronics Production Equipment,
Obzory po elektronnoy tekhnikie, 7 (268)
and 8 (269), Moscow, CNIIElektronika,
1974
11. B.N. Barabanov, I.G. Blinov, and A.M. Dorodnov, High Energy Equipment: Cold Systems for Plasma Generation of Solid Conductors, Fizika I Khimiya Obrabotki
Materialov, 1, 1978, p 4446
12. S.A. Muboyadjan, Vacuum Plasma Unit of

MAP for Depositing Protective Coatings on
Machine Components, VIMI, NTD 6, 1989,
p 711
13. E.N. Kablov, Cast Blades of Gas Turbine
Engines, Moscow, MISIS, 2001, 632 p
14. E.B. Kachanov and Yu.A. Tamarin, Protective Coatings: Effective Path of Increase of
Reliability of the Blades, Aviation Materials
on the Eve of the 21st Century, Moscow,
VIAM, 1994, p 296304
15. S.A. Muboyadjan, S.A. Budinovsky, and
E.N. Kablov, Turbine Blade Coating Deposition by Ion-Plasma Method, Aircraft Industry, 34, 1997, p 6570
16. S.A. Budinovsky, E.N. Kablov, S.A. Muboyadjan, and Ya.A. Pomelov, High-Power
Vacuum Plasma Technique as a Promising
Means for Development of New Coatings
and Materials, Aviation Materials on the
Eve of 21st Century, Moscow, VIAM, 1994,
p 314325
17. L. Holland, Vacuum Deposition of Thin
Films, John Wiley, New York, 1956, 135 p
18. R. Bakish, Ed., Introduction to Electron
Beam Technology, John Wiley, New York,
1962, 395 p
19. J. Lafferti, Ed., Vacuum Arcs, Moscow,
Mir, 1982, 432 p
20. A.M. Dorodnov, Some Areas of Application of Plasma Accelerators in Engineering,
Physics and Application of Plasma Accelerators, Minsk, Nauka i Tekhnika, 1974, p
330365
21. A.M. Dorodnov, Technological Plasma Accelerators, J. Tech. Phys., 1981, Vol 51 (No.
3), p 504524
22. S.U. Kimblin, Erosion of Electrodes and
Ionization Processes in Applied Areas of
Vacuum Arcing and under Atmosphere
Pressure, Experimental Studies of Plasmatrons, M.F. Jukov, Ed., Novosibirsk, Nauka,
1977, p 226253
23. Yu.N. Andreev, T.M. Andronova, A.I.
Vovsi, et.al., Electric Arc Sputtering of Metals and Alloys under Vacuum, Riga, LatNIINTI, 1982, 45 p
24. V.M. Lunev, V.G. Padalka, and V.M. Khoroshikh, Research into Some Plasma Characteristics of Vacuum Metal Arc, J. Technical Phys., 1977, Vol 47 (No. 7), p 1491
1495
25. I.I. Aksenov et al., Research into Plasma of
Steady Vacuum Arc Discharge, Thermal
Physics of High Temperatures, 1983, Vol
21 (No. 4), p 646651
26. O.N. Mironov, To Determination of Forces

Affecting Stationary Hall Plasma Accelerators, J. Technical Phys., Vol 44 (No. 3),
1974, p 525535
27. S.D. Grishin, L.V. Leskov, and N.P. Kozlov, Plasma Accelerators, Moscow, Mashinostroenie, 1983, 243 p
28. S.A. Leontiev, V.G. Kuznetsov, A.I. Rybnicov, and I.V. Burov, Structure and Properties of Protective Coatings by Vacuum
Arc Deposition, Le Vide: Science-Technique et Application, N275, 1995, p 462
466
29. B.A. Movchan and I.S. Malachenko, Heat
Resistant Coatings Deposited in Vacuum,
Naukova Dumka, 1983, 232 p
30. B.A. Movchan, EB-PVD Technology in
Gas Turbine Industry: Present and Future,
JOM, Nov 1996, p 4045
31. V.G. Kuznetsov, C.A. Kukuchkin, and I.S.
Polipanov, State-of-Art and Prospects of
Developments in the Field of Vacuum Arc
Coating Deposition, preprint IMASh RAN,
Sankt-Petersburg, No. 105, 1994, p 714
32. V.G. Kuznetzov, Interaction between Accelerates Metal Plasma Flux and Surface of
Solid, Autoreferat of Dissertation Dr. Sc.,
Sankt-Petersburg, 1997, 38 p
33. V.G. Kuznetzov, V.P. Bulatov, and S.I.
Rybnikov, Unit for Coating Deposition under Vacuum. Patent 2155242, C23C14/16,
Bulletin of Inventions No. 24, 27.08.2000
34. D.H. Boon, T.E. Strengman, and L.W. Wilson, Some Effects of Structure and Composition on the Properties of Electron Beam
Deposited Coatings for Gas Turbine Superalloys, J. Vac. Sci. and Technol., Vol 11
(No. 4), 1974, p 641646
35. V.I. Lachkov, A.V. Demchishin, I.S. Vasileva, et al., Structure of Titanium Coatings
Deposited by Condensation of Non-Steady
Atom Beams, Problems of Special Electric
Metallurgy, Kiev, No. 9, 1978, p 6777
36. P.A. Pap, Research into Structure and
Properties and Design of Electron Beam
Technique for Deposition of Protective
Coatings on Gas Turbine Blades, Ph.D.
dissertation, scientific consultant B.A. Movchan, IES, 1982
37. A.A. Dalskiy and K.M. Pasukov, Effect of
Condition of Substrate Surface Layer of
GS6U Alloy on Adhesive Strength of
NiCrAlY Coating, Surface Layer and Reli-
ability of Machine and Instrument Components, MDNTP, Moscow, 1989, p 4244

38. B.A. Movchan and A.V. Demchishin, Research into Structure and Properties of
Thick Vacuum Condensate Layers of Titanium, Tungsten, Aluminum Oxide, Physics
of Metals and Metal Science, Vol 28 (No.

4), 1969, p 653660
39. D.F. Betteridge and R.G. Ubank, Quality
Control of High-Temperature Protective
Coatings, Mater. Sci. and Technol., Vol 2
(No. 3), 1986, p 232242

Y. Tamarin, p55-78
DOI: 10.1361/pctb2002p055

CHAPTER 4
Phase Composition and Structure of

Coatings on Superalloys
IN COATINGS obtained by aluminizing
nickel alloys in media containing aluminum and
its compounds, the following aluminides can be
formed: NiAl3, Ni2Al3, NiAl, and Ni3Al. Their
properties are indicated in Table 4.1 (Ref 1).
The phase composition of diffusion coatings
formed on superalloys at their saturation with
aluminum is determined by the activity of the
medium used for aluminizing, the aluminizing
temperature, the thermodynamics of aluminide
formation reactions, the kinetics of the diffusion
process, and the superalloy chemical and phase
compositions.
The initial structural constitution and phase
composition of overlay (condensation) coatings
are dependent on the chemical and phase compositions of the alloy evaporated, the temperature of the surface upon which it is condensed,
and the state of aggregation of the substance
transferred (vapor, plasma).

Diffusion Coatings
Judging by chemical and phase compositions
of the aluminides that are formed in the outer
zone of the coating (the latter constituting 60 to
80% of coating thickness), all aluminizing media can be divided conventionally into those of
high, moderate, and low activities (Table 4.2).
The main phases of the outer zone of the coatings, produced in high-activity media, are NiAl3
or Ni2Al3 aluminides. Concentration of aluminum in the outer zone of such coatings is above
40%. In low-activity media, the outer zone of
coatings is made up of NiAl compounds or of a
mixture of NiAlNi3Al compounds. In such
coatings, the outer zone contains less than 31%

Al. In moderate-activity media, the bulk of the
coating outer zone is generally made up of NiAl
compound containing 32 to 38% Al.
Diffusion Coatings Formed in High-Activity Media. High-activity media include powder
blends with high free-aluminum content, slurry
layers that contain powders of aluminum and its
alloys, and layers based on aluminum and its
alloys that are deposited by the electric arc (EA)
technique. High-activity media can be employed
to deposit diffusion coatings within a wide temperature range, beginning at 450 C.
Diffusion Coatings Formed at Temperatures
below 640 C. The EA-deposited coatings are
applied without special preheating of the blades
to be aluminized in EA units. That is why aluminum alloy layers are deposited on relatively
cool (300 to 500 C) surfaces. As the aluminum
alloy is being evaporated, a layer of that alloy is
being formed on these surfaces. The diffusion
coating is formed due to processes in the solid
phase during the course of subsequent annealing
at the temperature below its melting point (640
C).
In the temperature range below the melting
point of aluminum, the diffusion processes in
aluminum-nickel and aluminum-nickel alloy
systems can be thought of as the processes proceeding under rigorous boundary conditions,
when the aluminum concentration on the surface
of the formed diffusion layer is definite and
time-constant. To study diffusion processes,
thick aluminum layers (1 mm) were applied
to nickel specimens and to the specimens of
Ni20CrTi alloy (Ni base, 19 to 22% Cr, 0.15 to
0.35% Ti). The specimens were annealed under
vacuum at 470, 500, 600, and 615 C. Annealing
time was within the range of 2 to 74 h.
Table 4.1
Nickel-aluminum system aluminide properties (Ref 1)

Crystalline
structure
Aluminide
NiAl3 (e-phase), 57.97 wt% Al
NiAl (b-phase), 3622 wt% Al

at 900 oC
Orthorhombic
a 6.611 A
b 7.367 A
c 4.812 A
Rhombohedral
a 4.036 A
c 4.900 A
c/a 1.214
bcc
a 2.886 A
Ni3Al (c-phase),
12.614.4 wt% Al at 20 oC
fcc
a 3.589 A
Ni2Al3 (d-phase), 44.740 wt% Al

at 6001120 oC
Heat of
Density, g/m3, Hardness, formation,
(theoretical)
MPa
kJ/mol
Tmelt, C
Homogeneity
region
854
...
3.96
61007700
38.0
1132
44.740% Al at 6001120 oC
4.76
11,200
57.0
Violet
1638
3622% Al at 900 oC
5.35 6.50
56006200
59.2
1380
12.614.4% Al at 20 oC
7.29
46005600
39.0
Blue
at Al content
of 31%
and yellow
at Al content
of 31%
Pale yellow
Color
...
Note: bcc, body-centered cubic; fcc, face-centered cubic
After annealing the nickel-aluminum specimens under all the previously mentioned conditions, a violet-colored diffusion layer is observed at the interface of two metals (Fig. 4.1).
Some islets of lighter phase are found at the
diffusion layer/aluminum interface. The amount
of this phase increases with the increase in annealing temperatures. The diffusion layer
formed consists of Ni2Al3 compound (Table
4.3). The composition of the phase that forms
islets corresponds to that of NiAl3 compound
(59% Al). No other intermetallic compounds of
the nickel-aluminum system have been detected
in the diffusion layer.
On annealing the specimens of the Ni20CrTiAl system in the temperature range of 470 to 615
C, the diffusion layer is formed at the interface
of two metals. It consists of two zones of different-colored hues (Fig. 4.2). The crystalline
Table 4.2 Conventional classification of saturating media
Activity of
saturating media
Aluminides forming outer zone

of the coating
Aluminum concentration of the
outer zone, %
High
Moderate
Low
NiAl3, Ni2Al3; 40% Al

NiAl; 3238% Al
NiAl, NiAlNi3Al; 31% Al
structure of the zone adjacent to the aluminum

layer is that of NiAl3 compound, while that of
the zone adjacent to the alloy side is Ni2Al3. The
patterns of nickel and chromium distribution
through the thickness of the diffusion layer are
similar. They feature the decrease in nickel concentration at the transition from one zone to another that is accompanied by the proportional
decrease in chromium concentration.
Table 4.3 Phase and chemical compositions of diffusion layers formed at annealing of aluminumnickel and Al-Ni20CrTi specimens
Annealing
conditions
Specimen
470 oC, 74 h
Al-Ni
600 oC, 9 h
Al-Ni
Al-Ni20CrTi
Al-Ni20CrTi
470 C, 74 h
600 C, 9 h
Chemical
composition(a), wt%
Layer
zone
Crystalline
structure
of the zone
Hardness,
MPa
Ni
Cr
Al
Islets
Outer
Islets
Outer
Outer
Inner
Outer
Inner
NiAl3
Ni2Al3
NiAl3
Ni2Al3
NiAl3
Ni2Al3
NiAl3
Ni2Al3
...
9500
...
9500
8000
10,000
8000
10,000
41
5556
41
5556
30
40
2629
3742
...
...
...
...
6.8
9.3
6.06.7
8.39.6
59
4544
59
4544
63
50
6864
5448
(a) In a range, the first digit is the beginning of the zone (from aluminum side); the second one is its end.
Phase Composition and Structure of Coatings on Superalloys / 57
The increase in the thickness of the diffusion

layers formed at the aluminum/nickel interface,
as well as at the aluminum/NiCrTi alloy interface, fits the general pattern for all diffusion processes (Fig. 4.3a,b).
To describe phase composition of the coatings, it is expedient to use special diagrams on
the x-y coordinates, where the y-axis equals
phase-zone thickness in the coating (lm) and the
x-axis is saturation time (s1/2). The value equal
to zero on y-axis represents the initial superalloy
surface before its saturation with aluminum. The
diagrams clearly demonstrate the location of aluminides, the ratios of their thickness as well as
the changes in the sizes of the specimens during
aluminizing, the kinetics of each aluminide zone
growth, and the increase in the total coating
thickness. In the cases where the kinetics of
growth is determined by diffusion processes, the
diagram of phase transformations represents
straight lines dividing phase zones of the coating.
Figure 4.4 represents phase diagrams typical
of nickel and nickel-base superalloys at their sat-
Fig. 4.1
Diffusion layer between aluminum and nickel. Annealing at 600 C for 9 h. 365
Fig. 4.2
Diffusion layer between aluminum and Ni20CrTi alloy. Annealing at 600 C for 9 h. 350
uration from the aluminum layer deposited by

the EA method at the temperature below its
melting point. As mentioned previously, coating
formation proceeds due to diffusion of aluminum through the aluminide layer to the coating/
alloy interface. The kinetics of the increase in
coating thickness depend on aluminum diffusive
mobility in the aluminides formed. Alloying
with chromium retards aluminum diffusion; in
turn, it results in the formation of NiAl3 aluminide in the outer zone.
Phase compositions of the coatings identical
to those mentioned previously are also observed
on aluminizing from the mixture of 98% Al
powder and 2% NH4Cl at 500 to 600 C. At
aluminizing, a violet-colored coating is formed
on the nickel surface. In its metallographic structure, the coating is similar to the diffusion layer
formed at the annealing of nickel with a deposited aluminum layer. The aluminum content of
the coating and its crystalline structure (Table
4.4) evidence that the coating consists of a single
intermetallide zone of Ni2Al3 compound.
The coatings forming at low-temperature aluminizing Ni20CrTi alloys and Ni10CrWMoCo
superalloy (Fig. 4.5) have the same metallographic structure. It consists of two zones of different-colored hues. The outer zone is lighter
and the inner one is darker. No other phase precipitates are detected in them. The hardness
number of the outer zone for all the alloys is
7500 to 8000 MPa, while that of the inner zone
is 10,000 MPa. Crystallographic structure of the
outer zone corresponds to NiAl3 compound,
while that of the inner zone corresponds to
Ni2Al3 compound. Aluminum contents of the
outer and inner zones are 69 to 62% and 56 to
49% respectively (Table 4.4).
Nickel, chromium, and aluminum distribution
patterns in the coating zones on Ni20CrTi alloy
are the same as those of the layer formed at the
annealing of aluminum-Ni20CrTi specimens. In
the coating formed on Ni10CrWMoCo alloy,
nickel and all its alloying elements (chromium,
molybdenum, tungsten, cobalt) show the same
patterns of their distribution through the zones.
The proportional decrease in concentrations of
these elements is observed in the direction from
the base material toward the surface.
For low-temperature aluminizing, the
mixtures of aluminum-iron alloys with high aluminum content may be used. By saturation of
JS6K superalloy from the mixture containing
98% Al-Fe alloy (75% Al) and 2% NH4Cl at
600 C, the coatings have been formed. Their
structures and compositions are similar to those
of
the coatings on
Ni20CrTi and
Ni10CrWMoCo alloys discussed previously.
The aluminum content of the coating surface
zone on JS6K superalloy is 60%, while that of
the inner zone is 40 to 50%. The patterns of
Fig. 4.3
aluminum, chromium, molybdenum, and tungsten distribution are the same as at saturation
from aluminum-base mixtures. They feature the
stepwise decrease in aluminum concentration in
the direction toward the base material, accom-
Thickness of diffusion layers between (a) aluminum and nickel and between (b) aluminum and Ni20CrTi alloy vs. annealing
time. (A solid line is for the full layer; a dashed line is for its outer zone.)
80
80
Al
40
Thickness, m
Thickness, m
Al
-Ni2Al3
0
40
-NiAl3
0
-Ni2Al3
40
40
-(Ni20CrTi)
-(Ni)
80
80
0
16
25
16
25
(b)
(a)
Fig. 4.4
Aluminizing time, h
Aluminizing time, h
Diagrams of phase transformations during (a) nickel and (b) Ni20CrTi alloy aluminizing from deposited aluminum layer at
600 C
Table 4.4 Chemical compositions of diffusion coatings formed on nickel and Ni20CrTi and
Ni10CrWMoCo superalloys. Aluminizing in the mixture of 98% Al, and 2% NH4Cl at 580 oC for 4 h
Alloy
Ni
Ni20CrTi
Ni10CrWMoCo
Chemical composition(a), wt%
Layer zone
Crystalline structure
of the zone
Hardness, MPa
Ni
Cr
Mo
Co
Al
...
Outer
Inner
Outer
Inner
Ni2Al3
NiAl3
Ni2Al3
NiAl3
Ni2Al3
9500
8000
10,000
8000
10,000
5656
2527
3340
2126
3136
...
5.96.4
7.79.3
3.03.5
4.24.8
...
...
...
2.73.5
4.05.0
...
...
...
1.62.0
2.52.6
...
...
...
2.02.3
2.73.0
44
6966
5951
6962
5649
(a) The first digit is the beginning of the zone; the second one is its end.
panied by the stepwise increase in the alloying

element concentrations. Phase analysis allowed
the detection of NiAl3, Ni2Al3, and NiAl aluminides in the coating.
Thus, independently of the aluminizing technique chosen, the diffusion coatings applied in
the high-activity media at the temperatures below aluminum melting point are NiAl3 and/or
Ni2Al3 compound-base coatings. The presence
of either of the aluminides depends on the saturation process temperature as well as on the
alloy additives.
Research into the composition and structure
of NiAl3 and Ni2Al3 compound-base coatings
applied from high-activity media made it possible to describe the general mechanism of their
formation on nickel alloys (Ref 2, 3). As mentioned previously, the distribution patterns for
nickel, chromium, and other alloying elements
through the coating zones in diffusion layers between aluminum and Ni20CrTi alloy are the
same in the coatings on Ni20CrTi and
Ni10CrWMoCo alloys. Gradual decrease in
nickel concentration in the direction toward the
surface is accompanied by the decrease in the
concentrations of all alloying elements.
The formation of NiAl3 and Ni2Al3-base diffusion coating proceeds as follows. Because alu-
Diffusion coating on Ni10CrWMoCo superalloy.

Aluminizing in the mixture of 98% Al and 2%
NH4Cl at 580 C for 4 h. 300
Fig. 4.5
minum has high mobility in these compounds,

at aluminizing it diffuses to the depth equal to
the coating thickness. At the same time, the alloying elements (such as nickel, chromium,
tungsten, molybdenum, and cobalt) of the saturated alloy remain actually immobile, that is, as
if the saturated alloy is diluted with diffusing
aluminum. The increase in the coating thickness
takes place at the inner interfaces, in fact at the
outer zone/inner zone and inner zone/alloy interfaces.
Coatings formed in high-activity media at the
temperature below aluminum melting point
demonstrate high hardness and low ductility. In
addition, the melting point of NiAl3 compound
is 854 C. That is why blades and other components with these coatings should undergo
short-term annealing (for 1 to 2 h) at 950 to 1100
C before putting them into service.
Diffusion Coatings Formed at Temperatures
above 640 C. When the blades with layers of
aluminum alloys deposited on their surfaces are
annealed at temperatures above 640 C, the diffusion coatings are formed from the liquid
phase. In this case, their structure and phase
composition differ from that mentioned previously.
If the annealing temperature exceeds that of
melting of the deposited aluminum alloy, then
an active liquid phase is formed on the saturated
surface, in which aluminum content (mAl, g/m2)
is limited by the thickness (d) of the primary
layer deposited. In these circumstances, a diffusion coating forms a few minutes after the aluminum alloy heats to its melting point. Further
elevation of temperature and lengthening of annealing time bring about diffusion processes occurring between the coating formed and the superalloy. As a result, aluminum concentration is
lowered and coating thickness is increased.
The Al5SiY layer (aluminum-base, 5% Si,
1.5% Y) was deposited by the EA technique
onto the blades of JS6U superalloy. This was
followed by annealing at 700, 800, 900, and
1000 C for 1 to 4 h. Table 4.5 shows the analysis data of phase and chemical compositions
and the structure of the diffusion coatings
formed.
Initially, a layer is formed on the surface to
be saturated, its chemical and phase compositions being similar to those of the aluminum alloy evaporated. Further on, the changes in chemical and phase compositions that take place on
the saturated surface at annealing are fully de-
Table 4.5 Phase and chemical compositions of diffusion coatings on JS6U superalloy. Aluminizing by
EA method from alloy Al-5%Si-1.5%Y. Specific mass of alloy deposited, mAl 100 g/m2
Annealing
Without annealing
700 oC, 1 h
800 oC, 1 h
900 oC, 1 h
1000 oC, 2 h
Coating thickness, lm
40
45
50
55
65
Phase composition
Al, Si
Al, NiAl3, Ni2Al3
NiAl3, Ni2Al3
Ni2Al3
NiAl(a)
NiAl(b)
NiAl(c), carbides
Hardness, MPa
Al
Cr
Co
Si
680
10,800
10,800
8900
6300
...
6830
Base
70
48
40
26
26
13
...
2.5
4.5
5.0
7.0
3.7
13.4
...
3.0
5.0
6.0
...
...
...
...
3.0
5.0
6.0
...
...
17
5.0
3.5
2.5
2.0
1.4
0.1
0.5
1.5
1.0
0.7
0.6
0.1
0.1
0.1
(a), Outer zone of coating. (b) Middle zone of coating. (c) Intermediate zone of coating
Diffusion coating on JS6U superalloy. Electric arc

deposition of Al-5%Si-1.5%Y alloy and annealing at
700 C for 1 h. 1000
Fig. 4.6
pendent on the kinetics of diffusion processes

occurring in aluminides.
Annealing of specimens at 700 C results in
the formation of a coating on their surfaces, due
to dissolution of a thin layer of the superalloy in
the liquid Al5SiY alloy. The layer is structurally
dissimilar; the top part contains (NiAl3Al) eutectic while its inner layers are fully constituted
of NiAl3, with Ni2Al3 compound occurring at
the coating/alloy interface (Fig. 4.6).
Annealing at 800 C leads to the formation of
a Ni2Al3-base coating on the surface. Such an
annealing temperature makes chromium, silicon,
and tungsten spread evenly throughout the coating thickness. Elevating the annealing temperature to 900 C does not bring about any radical
changes in the structure of the coating formed;
its basis is still made up of Ni2Al3 compounds.
After annealing at 1000 C, the outer zone of
coating on the JS6U superalloy, as well as on
other superalloys, consists of two parts whose
basic component is NiAl. The first part lies on
the coating surface and includes phase precipi-
tates based on refractory elements and silicon.

The second part of the outer zone, bordering
the inner zone, is free of precipitates and contains refractory elements in solid solution. Such
an outer zone structure is formed due to the
changes of diffusion processes taking place
when the NiAl layer emerges in the coating.
The phenomenon is typical of aluminide coatings produced from aluminizing media of high
activity and subjected to diffusion annealing after saturation.
The inner zone of coating, also based on NiAl,
contains many carbide phases of M23C6 and
M6C type (where M represents Cr, W). After
being annealed at 1000 C, the structure and
composition of coatings formed with aluminumion energy in excess of 100 eV and 15 to 20 eV
are different (Fig. 4.7). The high energy of aluminum ions heats up the surface on which condensation (overlaying) occurs and simultaneously makes some part of aluminum
reevaporate. This leads to decreasing the mass
of aluminum deposited on the surface and to
forming NiAl3, Ni2Al3 aluminide coatings at deposition. When diffusion annealing is applied,
no liquid phase is formed on the surface, and, in
this case, a coating is formed due to gradual lowering of aluminum content in growing aluminides.
With annealing temperature elevated to 1100
to 1200 C and the annealing process taking 1
to 2 h, aluminum concentration in the coating
outer zone lowers to 16 to 18%, and Ni3Al compound emerges in it. The higher the temperature
of annealing, the lower the amount of phases
precipitating in the outer zone and based on
chromium, tungsten, and molybdenum, which
gradually dissolve in aluminides. Elevating the
annealing temperature to 1100 to 1200 C leads
to coagulation of carbide phases in the innerzone layer. The latter tends to form a columnar-
Fig. 4.7
Diffusion coating on JS6U superalloy. Electric arc deposition of Al-5%Si-1.5%Y alloy and annealing at 1000 C for 2 h: (a)
Aluminum-ion energy 100 eV. (b) Aluminum-ion energy 20 eV. 1000, reflected electrons
like structure. After annealing at 1200 C, the

inner zone is actually free of carbide phases.
Typical of the previously described operational procedure, the basic characteristics
composition, structure, thicknessof the resultant diffusion coatings are fully dependent on the
temperature and time of annealing as well as on
the specific mass of aluminum layer deposited
from plasma onto the surface (mAl). The specific
mass of the deposited layer is related with the
composition and thickness of the diffusion coating formed from it by the following relationship:
m
mAl
di Xi qi
i1
in g/m2, where di is the thickness of a certain (i)

zone of coating, Xi is aluminum mass fraction in
that zone, and qi is the density of that zone.
At certain stages of its forming in the course
of annealing, the diffusion coating is actually
constituted of zones of successively arranged aluminides: NiAl3-Ni2Al3 NiAl(32 to 38% Al)NiAl(20 to 26% Al)-Ni3Al. From aluminum
content in the coating aluminide zones, and from
density of the zones, the following approximation can be written:
mAl 2.3 dNiAl3 2.1 dNi2Al3 1.9 dNiAl(36%Al)
1.6 dNiAl(25%Al) 1.1 dNi3Al
Mass of deposited Al-5%Si-1.5%Y alloy vs. thickness of diffusion coating with the given phase composition theoretical relationship
Fig. 4.8
From this relationship, one may (judging by

the specific mass of aluminum layer deposited
on the specimen surfacemAl, g/m2) calculate
the thickness of a diffusion coating formed from
it and having definite phase composition. One

may also deduce how coating thickness varies
under certain phase transformations. The results
of such calculations are presented in Fig. 4.8.
They correlate with the experimental data obtained when the relationships between the mass
of deposited aluminum alloy and the thickness
of coatings and their compositions were investigated.

On annealing at 1000 C for 2 to 4 h, the
thickness (h) of the diffusion coating formed
bears the following relationship (derived experimentally) to the specific mass of aluminum alloy deposited (50 to 70 g/m2): h mAl 0.85.
Table 4.6 Phase and chemical compositions of outer zone of diffusion coatings on JS6K superalloy
formed at aluminizing from slurries of different compositions
Active component of slurry;
annealing conditions
Chemical composition(a), wt%
of the zone
Ni
Cr
Mo
Co
Ti
Al
Other elements(b)
NiAl
NiAl
NiAl
NiAl
NiAl
Base
Base
Base
Base
Base
4.6
4.6
4.7
5.6
5.9
2.5
2.6
2.6
1.1
0.5
1.8
1.7
1.8
1.7
0.8
3.8
3.6
3.8
4.0
4.1
1.2
1.2
1.4
1.0
1.4
25
25
25
22
21
...
3.00.3% Si
3.70.3% Si
1.00.4% Nb
1.00.4% Ta
100% Al; 950 oC, 4 h

(97.5% Al, 2.5% Si); 950 oC, 4 h
(92.5% Al, 7.5% Si); 950 oC, 4 h
(40% Al, 60% Nb); 1050 oC, 4 h
(30% Al, 70% Ta); 1050 oC, 4 h
(a) Average content of the zone. (b) The first digit is the beginning of the zone; the second one is its end.
External surface
of a coating
Phases
thickness
in a coating
Initial alloy boundary
NiAl
(36% Al)
NiAl
(24% Al)
Al Si Y
NiAl3
Ni3Al3
Ni3Al
Surface of aluminides
formation in a coating
Superalloy ( + )
I
600
II
700
III
800
IV
900
V
1000
1100
Temperature, C
Fig. 4.9
Diagram of phase transformations during annealing of superalloy with deposited Al-Si-Y alloy layer
Table 4.7 Characteristics of diffusion coatings formed on nickel. Aluminizing in the mixture of 98%
Al-Fe alloy (50% Al) and 2% NH4Cl at 800 to 1100 oC
Aluminizing
conditions
950 oC, 36 h
1000 oC, 8 h
1100 oC, 2 h
Ordinal number
of the zone
from the surface
Zone
thickness,
lm
1
2
3
1
2
3
1
2
3
105
35
4
72
19
5
90
24
8
Zone color
Aluminum
concentration(a), %
Hardness, MPa
Pale blue
Yellow
Pale yellow
Pale blue
Yellow
Pale yellow
Pale blue
Yellow
Pale yellow
3636
2822
15
36
24
15
36
24
15
84007800
5000
4000
8000
5000
4000
8000
5000
4000
(a) In a range, the first digit is the beginning of the zone; the second one is its end.
Aluminide;
its crystal
lattice parameter
NiAl; 2.86 A
NiAl; 2.86 A
Ni3Al; 3.56 A
NiAl; 2.86 A
NiAl; 2.86 A
Ni3Al; 3.56 A
NiAl; 2.86 A
NiAl; 2.86 A
Ni3Al; 3.56 A
The aluminum content of the outer zone of

such a coating is 22 to 24%.
The behavior of elements (silicon, yttrium,
etc.) introduced into the aluminum alloy for alloying of the coating formed depends on the
thermodynamic activity of these elements as related to the elements in the aluminized superalloy as well as on their solubility in the aluminides formed. At annealing temperatures up to
700 C, silicon is uniformly distributed throughout the (AlNiAl3) eutectic layer and then
throughout the NiAl3 layer. Its redistribution occurs at temperatures above 800 C, at the moment when phases are formed on the basis of
refractory elements. Silicon is included into
these phases, locating in the outer-zone surface
layer and in the inner zone of the coating. Yttrium content of the coating does not exceed its
solubility level in aluminides (0.1%).
The mechanism of diffusion-coating formation, when employing aluminum powder containing slurries as well as aluminum silicon
powder containing slurries, is similar to that
mentioned previously. Coatings are formed
through a liquid phase of aluminum alloy followed by diffusion processes evolving between
the aluminides and superalloy. Phase and chem-
Diffusion coating on nickel. Aluminizing in mixture of 98% Al-Fe alloy (50% Al) and 2% NH4Cl
at 950 C for 36 h. 300
Fig. 4.10
ical compositions of diffusion coatings, obtained

from slurry processes, are similar to those mentioned previously. The introduction of additional
elements (silicon, niobium, tantalum, etc.) into
the active component of the slurry makes it possible to produce diffusion coatings alloyed with
these elements (Table 4.6).
The process taking place upon formation of
diffusion coatings at the temperatures above the
melting point of the deposited aluminum alloy
is shown in Fig. 4.9. At the first stage
(AlNiAl3), eutectic formation takes place.
Then, NiAl3 aluminide-base zones form. The
subsequent process kinetics depend on aluminum diffusive mobility in the growing aluminides at the annealing temperature and on the
amount of aluminum deposited. At 800 C, the
Ni3Al zone transforms completely into Ni2Al3
within 1 h period. After it, a thin zone with
the structure of NiAl aluminide forms at a coating/superalloy interface. Under these conditions, all diffusion processes are retarded. At the
annealing temperature above 1000 C, Ni3Al
phase forms at the interface of the alloy and the
zone with the NiAl structure. Thus, diffusion
processes during annealing of the coating
formed in the high-activity media involve gradual changes of aluminide types in the following
succession: NiAl3Ni2Al3NiAlNi3Al.
When annealing temperature exceeds 800 C,
the outer and inner zones undergo precipitation
of refractory element-base phases. These elements (molybdenum, tungsten, chromium, etc.)
have low solubility in aluminides. This effect is
useful, because it imparts higher hot corrosion
resistance to diffusion coatings.
Diffusion Coatings Formed in Moderateand Low-Activity Media. When aluminizing
superalloys in moderately active media, diffusion coatings are formed. Such coatings are
based on NiAl and contain 32 to 38% Al. A
representative mixture of the kind that is broadly
used consists of 98% Al-Fe (50% Al) alloy and
2% NH4Cl.
After aluminizing nickel in the mixture of
98% Al-Fe alloy (50% Al) and 2% NH4Cl at
950, 1000, and 1100 C, the diffusion coatings
formed consist of three intermetallic zones (Table 4.7, Fig. 4.10). The surface zone is NiAl
compound with aluminum content of 36%. The
second zone also consists of NiAl compound
with aluminum content of 28 to 22%. Although
the crystalline structures and the lattice parameters of these zones are the same, there is an
interface between them, with a stepwise change
80
40
NiAl (36% Al)
0
NiAl (25% Al)
40
Ni
80
0
Ni3Al
4
9
16
Aluminizing time, h
External zone
NiAl (36% Al)
40
Internal zone
NiAl (25% Al)
Superalloy
40
Ni3Al + phases
based on
refractory elements
80
25
(a)
Fig. 4.11
Thickness, m
Thickness, m
80
4
9
16
Aluminizing time, h
25
(b)
Diagram of phase transformations on (a) nickel and (b) JS6U superalloy surfaces during aluminizing at 1000 C
in the aluminum content through it. Chemical

etching develops the third (deepest) zone. Both
aluminum content and crystalline structure of
this zone are typical of Ni3Al compound. Figure
4.11(a) shows a diagram of phase transformations developing on a nickel surface at its aluminizing in the mixture of 98% Al-Fe alloy
(50% Al) and 2% NH4Cl at 1000 C.
Within the full aluminizing temperature
range, it is observed that iron diffuses into the
coating to a depth of not more than 10 lm. Iron
content of the coating surface layer is below
1.0%.
The relative positions of Ni2Al3, NiAl, and
Ni3Al compounds in diffusion coatings on a
nickel surface corroborate the heterophase diffusion theory statement concerning phase positions in the phase diagram. Yet, the patterns of
aluminum and nickel distribution through each
coating zone differ substantially from those expected (Ref 4). The main distinction is that
within the region of NiAl compound homogeneity, two diffusion zones of different chemical
compositions form. Those are NiAl (36% Al)
and NiAl (28 to 22% Al). The zones demonstrate different colors and hardness numbers. A
concentration step is observed at their interface. Such an abnormality is due to crystalline
structure features of NiAl compound and some
specialties of diffusion in this compound (Ref
5).
Experimental data allow estimation of the
relative rate of nickel-atom diffusion toward the
specimen surface as well as that of aluminumatom diffusion from the surface inside the bulk
specimen. As calculations have shown, after alu-
minizing, almost all aluminum atoms are in intermetallic zones. The amount of aluminum that
has diffused through these zones and dissolved
in nickel is negligible. If the zone composition
and thickness are known, it is possible to calculate the amount of nickel and aluminum taking part in opposing diffusion through the sec-
Diffusion coating on JS-type superalloy. Aluminizing in mixture of 98% Al-Fe alloy (50% Al) and
2% NH4Cl at 1100 C for 8 h. 300
Fig. 4.12
the size of less than 1 lm are observed in this

zone. The inner zone consists of the mixture of
a few phases. There are no sharp boundaries between the inner and outer zones nor between the
inner zone and the base material.
The compound with the NiAl structure and
the aluminum content of 34 to 36% is the base
of the outer zone of the diffusion coating on superalloys after they aluminize. Some chromium-,
tungsten-, molybdenum-, and titanium-base
phases precipitate in the zone, but their fraction
is relatively little.
The inner-coating zone consists of the mixture of dispersed phases. Total nickel, chromium, tungsten, and cobalt content of the inner
zone is 84 to 87%. Phase analysis of the diffusion coating has revealed that its inner-zone ma-
ond zone to form the first and third zones.

Relative rates of nickel- and aluminum-atom
diffusion through the second zone are equal to
41 to 1, respectively. Thus, at aluminizing, the
directed flow of nickel atoms moves through
both the second and third zones toward the first
zone, while the opposite flow of aluminum atoms is less by a factor of 41 (Ref 4).
After aluminizing nickel alloys in the mixture
consisting of 98% Fe-Al alloy (50% Al) at 950
to 1100 C for 2 to 8 h, the diffusion coatings
on their surfaces show two structural zones different in their metallography (Fig. 4.12, Table
4.8). The outer zone that is 70 to 80% of the
coating thickness may have different-colored
hues, depending on aluminizing temperature.
Some isolated inclusions of a lighter phase with
Table 4.8 Characteristics of diffusion coatings formed on nickel alloys. Aluminizing in the mixture of
98% Al-Fe alloy (50% Al) and 2% NH4Cl at 950 oC for 4 h
Alloy
Ni20CrTi
Ni10CrWMoCo
JS6K
Coating zone
Ni
Cr
Mo
Co
Ti
Fe
Al
of the zone
Hardness, MPa
Outer
Inner
Outer
Inner
Outer
Inner
Base
Base
Base
Base
Base
Base
1.8
49
3.0
14
5.3
14.7
...
...
2.2
7.0
2.1
5.9
...
...
4.7
18
2.2
6.0
...
...
3.0
4.0
2.8
3.4
...
...
...
...
0.5
4.5
...
1.0
1.0
0.3
0.8
0.2
35
16
35
14
36
23
NiAl
...
NiAl
...
NiAl
...
8200
820
8400
8000
8200
9500
Table 4.9 Phase and chemical compositions of diffusion coating formed on JS6K superalloy in lowactivity powder mixtures
Saturating mixture and conditions
98% Al-Fe alloy (40% Al),
2% NH4Cl; 950 oC, 4 h
Powder mixture (5% Al, 45% Cr, 50% Al2O3);
vacuum, 1000 oC, 6 h
Table 4.10
Coating zone
Ni
Cr
Outer
Inner
Outer
Inner
Base
Base
Base
Base
1.2
15
5.0
10.0
Ti
Al
Phase composition
0.1 0.1 4.0 0.3

6.0
6.2 3.8 4.8
Was not analyzed
Was not analyzed
Mo
Co
24
10
24
15
NiAl
NiAl, Ni3Al, carbides
NiAl
NiAl, Ni3Al, carbides
Chemical and phase composition of Ni-Cr-Al alloys

Alloy
Ni22Cr11AlY
Ni14Cr3AlY
Ni18Cr5AlY
Ni16Cr9AlY
Ni21Cr9AlY
Ni18Cr16AlY
Ni19Cr12AlY
Ni8Cr18AlY
Ni8Cr15AlY
Ni7Cr12AlY
Ni7Cr10AlY
Ni20Cr6AlY
Ni28Cr11AlY
Ni5Cr25Al
Ni
Cr
Al
Phase composition at 20 oC
b, %
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
22.3
14.7
17.7
16.0
20.8
18.3
19.5
8.0
8.0
7.0
7.2
20.0
28.5
5.0
11.2
2.8
5.5
9.1
8.8
15.7
12.6
18.0
15.0
12.0
10.0
6.0
11.0
25.0
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
...
c, c, b,
c
c, c
c, c, b
c, c, b,
c, b,
c, c, b,
c, b
c, b
c, b
c
c, c
c, c, b,
b
50
...
...
15
20
80
60
95
70
15
...
...
45
100
trix consists of NiAl and Ni3Al compounds. The

precipitation of the phases on the base of refractory elements (chromium, molybdenum, tungsten, etc.), which alloy superalloys, has also
been detected.
Growth kinetics of diffusion coatings on
nickel alloys, within the accuracy of measurement, is the same as growth kinetics of the coatings on nickel. The changes in aluminide phase
size shown in Fig. 4.11(b) are brought about by
the processes taking place in the coatings on su-
Fig. 4.13
peralloys. The thickness of outer-zone diffusion

coating on superalloys coincides with the zone
of NiAl compound (36% Al). The inner zone
includes NiAl (24% Al) and Ni3Al aluminide
layers, in which refractory element-base phases
precipitate; the crystallographic structure and
morphology of such phases are dependent on alloy composition and aluminizing temperature.
The dependence of superalloy diffusion-coating thickness on the temperature and the time of
the aluminizing process is similar to the same
Microstructure of the alloys (a) Ni14Cr3AlY. (b) Ni16Cr9AlY. (c) Ni22Cr11AlY. (d) Ni7Cr12AlY. (e) Ni8Cr15AlY. (f)
Ni8Cr18AlY. 500
dependence for nickel. The coincidence should

be attributed to the fact that refractory elements
are difficult to solubilize in aluminides and actually do not affect the activation energy of aluminum and nickel diffusion in that compound.
Mixtures of moderate activity occupy intermediate position in a conventional gradation of
aluminizing media. This is why, under certain
conditions, the coatings obtained from such
mixtures may have structures and compositions
close to those of coatings obtained from highor low-activity mixtures. Thus, at 800 and 900
C, a thin layer of Ni2Al3 compound constituting
the substrate of the outer-zone coating is formed
on the coating surface, such coatings being deposited from highly active mixtures. On the
other hand, repetitive applications of a moderately active mixture, which is inevitable under
manufacturing conditions, lead to lowering aluminum concentration in the outer-zone coating
and forming coatings whose composition and
Thermal expansion coefficients vs. temperature.

(a) 1, Ni5Cr25Al alloy; 2, Ni7Cr12AlY alloy; 3,
Ni8Cr15AlY alloy. (b) 1, Ni22Cr11AlY alloy; 2, Ni18Cr5AlY alloy; 3, Ni16Cr9AlY alloy; 4, Ni14Cr3AlY alloy
Fig. 4.14
structure are typical of those produced from lowactivity mixtures.

Low-activity mixtures may be needed to obtain diffusion-aluminide coatings possessing
higher ductility and also when depositing coatings onto the insides of cooled blades. Chemical
compositions of the coatings formed on JS6K
superalloy in low-activity mixtures are presented in Table 4.9. The distinguishing feature
of the process is that the NiAl compound (constituting the substrate of the outer zone of coating) does not contain refractory elements. The
inner zone of the coating has a multiphase structure, comprised of (NiAl, Ni3Al) matrix and refractory-element based phases. The same structure and phase composition have a diffusion
coating forming under vacuum on the inside of
cooled turbine blades at 1000 C from a pouredin mixture of aluminum (5%), chromium (45%),
and Al2O3 (50%) powders.
The kinetics of forming a coating from lowactivity mixtures depend on the type of mixture
used, the ratio between the amount of aluminum
in the mixture and the quantity spent for coating
buildup, and finally on the kinetics of reaction
of halide or vapor formation. If the aluminizing
process is conducted in a vacuum using a powder mixture of aluminum, chromium, and aluminum oxide, the kinetics of the process slow
down with time, which is due to rapid mixture
depletion and a corresponding drop of aluminum
and chromium vapor pressures.
As revealed by the results of research into diffusion coatings applied at 950 to 1000 C in the
moderately active mixtures, the outer zone of the
coatings on nickel and nickel alloys consists of
NiAl compound with aluminum content of 34 to
36%. The inner zone of the coating on nickel
consists of NiAl and Ni3Al compounds going
one after another. Some phases on the base of
refractory elements (such as chromium, molybdenum, tungsten, etc.) precipitate in the innerzone matrix that consists of NiAl and Ni3Al
compounds in the coatings on superalloys. The
amount, fineness, and crystalline structure of
these phases depend on the alloy composition
and the aluminizing temperature conditions.
The distribution patterns of alloying elements
in the diffusion coatings formed on nickel alloys
in moderate- and low-activity mixtures in the
temperature range of 950 to 1100 C are basically different from those typical of the coatings
formed in highly active mixtures. Chromium,
tungsten, and molybdenum contents of the outer
zone are relatively low. At the same time, the
concentration of these elements in the inner zone

is higher than their concentration in the superalloy.
The mechanism of coating formation in the
moderately active media can be represented as
follows (Ref 2, 6). Almost all aluminum taken
up from the saturating medium is spent on the
formation of the outer zone of the NiAl compound-base coating. This compound is formed
due to aluminum flowing from the outer boundary and nickel atoms diffusing through the inner
zone from the alloy. Chromium-, molybdenum-,
and tungsten-atom diffusion rates are much less
than nickel-atom diffusion rate in the same direction. The coating inner zone is formed due to
partial depletion of deeper alloy layers of nickel
diffusing into the outer zone. Chromium, molybdenum, and tungsten are practically immobile under these conditions. Due to their low solubility in the growing aluminides, they form
their own phases. When coating outer-zone formation proceeds by this mechanism, it can take
place at either of two interfaces. In the case of
high-aluminum-diffusive mobility, it may be at
the interface between the inner and outer zones,
whereas in the cases where nickel-diffusive mobility in the aluminide of the outer zone is high,
Table 4.11
it may take place at the outer-coating boundary.

The first of the previously mentioned cases is
typical of the process of forming the NiAl compound-base outer-zone coating with aluminum
content beyond its stoichiometric concentration
(31.5%) in the media of moderate activity. The
second one occurs at forming the NiAl compound-base outer zone with aluminum content
below its stoichiometric concentration in the
media of low activity. When NiAl formation
takes place at the coating surface, low diffusive
mobility of refractory elements in it results in
considerable reduction of their concentration in
the outer zone. No chromium-, molybdenum-,
and tungsten-base phases are detected in this
zone.
This mechanism of forming diffusion coatings on nickel superalloys accounts for the experimental fact that at aluminizing under the
conditions discussed, the outer zones of the coatings on nickel and nickel superalloys demonstrate the same size, structure, and aluminum
contents.
Calculated specific-mass gain of the specimens of nickel alloy aluminized (mAl) at the expense of aluminum spent on coating formation
Chemical and phase composition of Ni-Co-Cr-Al alloys

No.
Alloy(a)
Ni
Co
Cr
Al
Qb(b), %
Qb(c), %
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Ni10Co15Cr8Al
Ni30Co15Cr8Al
Ni10Co30Cr8Al
Ni30Co30Cr8Al
Ni10Co15Cr14Al
Ni30Co15Cr14Al
Ni10Co30Cr14Al
Ni30Co30Cr14Al
Ni10Co15Cr8AlY
Ni30Co15Cr8AlY
Ni10Co30Cr8AlY
Ni30Co30Cr8AlY
Ni10Co15Cr14AlY
Ni30Co15Cr14AlY
Ni10Co30Cr14AlY
Ni30Co30Cr14AlY
Ni10Co22Cr11AlY
Ni30Co22Cr11AlY
Ni20Co15Cr11AlY
Ni20Co30Cr11AlY
Ni20Co22Cr8AlY
Ni20Co22Cr14AlY
Ni20Co22Cr11Al
Ni20Co22Cr11AlY
Ni20Co22Cr11AlY
Ni20Co22Cr11AlY
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
base
base
Base
11.3
27.5
10.1
29.4
10.5
27.5
10.6
29.7
11.2
30.2
11.1
28.8
10.4
29.5
10.5
29.9
10.1
29.8
19.6
19.7
19.7
19.8
20.3
19.7
19.6
22.5
15.5
14.4
28.5
29.6
15.1
15.2
30.0
29.7
15.2
14.7
29.3
29.5
14.6
15.9
28.8
29.7
22.2
22.2
14.6
29.3
22.2
22.7
21.7
22.7
22.5
22.5
8.9
9.3
7.8
7.8
13.5
13.6
14.1
14.1
8.2
8.1
8.0
8.1
13.6
13.8
13.8
14.1
10.6
10.9
11.2
11.1
8.3
13.7
11.1
11.1
11.3
11.2
0.04
0.05
0.11
0.10
0.07
0.08
0.06
0.08
0.04
0.45
0.20
0.33
0.04
0.25
0.32
0.46
0.30
0.23
0.35
0.33
0.40
0.24
0.04
0.28
0.19
0.39
c, c, b
c, c, b
c, c, b,
c, c, b, r
c, c, b,
b,
b,
b, r
c, c, b
c, c, b
c, c, b,
c, c, b, r
c, c, b,
b,
b,
b, r
c, c, b,
c, c, b, r
c, c, b
b,
c, c, b,
c, c, b,
c, c, b,
c, c, b,
c, c, b,
c, c, b,
36
33
60
68
71
78
78
91
...
...
...
...
...
...
...
...
63
76
53
72
35
79
60
60
60
60
38
33
42
48
87
79
76
68
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
...
(a) The alloys (except No. 26) include: 1.5% Hf, 1.0%W, 0.5% Nb, and 0.5% Si. (b) Phase quantity (Qb, vol%) in original state. (c) Phase quantity (Qb, vol%) after
annealing at 1200 oC for 200 h
Fig. 4.15
Microstructure of the alloys. (a) Ni10Co15Cr8AlY. (b) Ni30Co15Cr8AlY. (c) Ni10Co30Cr8AlY. (d) Ni30Co30Cr8AlY. (e)
Ni10Co22Cr11AlY. (f) Ni30Co22Cr11AlY. (g) Ni20Co15Cr11AlY. (h) Ni20Co30Cr11AlY. 335
Table 4.12 Chemical composition of structural

constituents in Ni-Co-Cr-Al alloys
b-phase(a)
c-c
Alloy
Al
Cr
Co
Al
Cr
Co
Ni10Co15Cr8Al
Ni30Co15Cr8Al
Ni10Co30Cr8Al
Ni30Co30Cr8Al
Ni10Co15Cr14Al
Ni30Co15Cr14Al
Ni10Co30Cr14Al
Ni30Co30Cr14Al
17
18
14
...
18
18
20
21
10
7.5
26
...
13
14
11
19
7.5
20
8.0
...
10
27
11
31
6.5
6.0
4.5
4.5
4.5
4.5
...
...
21
18
36
36
32
30
...
...
13
33
11
33
17
38
...
...
(a) The quoted average chemical composition includes -Cr: 70% Cr, 3% Al,
and 10 to 30% Co.
actually coincides with the experimental value

of specimen mass-specific change after aluminizing. Thus, when determined after aluminizing, the specific change of specimen mass interrelates two basic parameters of aluminide
coating: concentration of aluminum in a coating
and its thickness. The postaluminizing specific

change of specimen mass defines the aluminum
reserve in aluminide coating and is a vital characteristic applicable both to control of coatings
and to comparison of their protection properties.
Phase Composition and Structure of NiCr-Al, Ni-Co-Cr-Al, and Co-Ni-Cr-Al

Alloys and Overlay Coatings Made of
Them
After being deposited according to established technology, the overlay coatings of the
Ni-Cr-Al, Ni-Co-Cr-Al, and Co-Ni-Cr-Al systems have chemical and phase compositions
very close to those of the alloy evaporated.
The research into Ni-Cr-Al-, Ni-Co-Cr-Al-,
and Co-Ni-Cr-Al-system alloys was arranged to
determine the relationship between variations in
chemical composition and the resultant changes
of structure and properties of the alloys. The

range of composition the overlay coatings may
have after they have been deposited and subjected to various high-temperature tests is investigated.
Ni-Cr-Al Alloys. The alloy group under consideration and the coatings made of them consist
of the following main phases: c-Ni-base solid
solution, c-Ni3Al phase, b-NiAl phase, and Cr-base solid solution (Ref 7, 8). The number of
phases and temperature fields of their existence
depend on alloy chemical composition. Table
4.10 gives chemical composition and phase constitution of investigated alloys after the latter had
been annealed at 1100 C for 4 h; typical structures of alloys are shown in Fig. 4.13.
Phase transformations in Ni-Cr-Al alloys
were examined by dilatometric analysis (thermal
expansion coefficient values are given in Chapter 7, The Effect of Protective Coatings on the
Mechanical Properties of Superalloys). Specimens for this were made by extrusion from cast
blanks. Their microstructure was close to overlay-coating structure. Prior to testing, the specimens had been annealed at 1100 C for 2 h.
Alloys whose compositions fall within the phase
fields of c, c-(Ni7Cr12AlY), (bc)(Ni8Cr15AlY), and b-(Ni5Cr25Al) have no
phase transformations throughout the full range
of temperatures up to 1000 C (Fig. 4.14 a,b).
In the Ni18Cr6AlY alloy, whose composition
falls within a two-phase field (cc), marked
enlargement was noted at 700 to 1000 C, attributable to c-phase solution. During cooling,
the reversed transformation, cc, evolves
without hysteresis in the same interval of temperatures. The transformation cc is typical of
all the Ni-Cr-Al system alloys under consideration that have chromium content above 17%.
In contrast to cc transformation, the phase
reaction bcc that takes place at 1020
C involves considerable diffusion motion of
aluminum, nickel, and chromium. It proceeds
with time hysteresis, which depends on temperature, phase sizes, and forms. During cooling, the bcc process occurs with lessening of volume.
Thus, within the range of compositions under
consideration, a diversity of phase transformations and, hence, volumetric changes has been
noted, with all these processes taking place
within the interval of temperatures at which turbine blades run.
Ni-Co-Cr-Al Alloys. Experiments were
planned statistically to enable research into alloy
properties for the Ni-Co-Cr-Al-system coating

deposition. The compositions of the alloys produced are given in Table 4.11. In this plan of
experiments, cobalt (x1), chromium (x2), aluminum (x3), and yttrium (x4) are variables. A regression equation, obtained by statistical processing the experimental data, correlates the
property under study with the content of alloying elements (Ref 9).
Figure 4.15 shows typical structures of the alloys studied (Ref 10), while phase compositions
of the alloys determined from the data of x-ray
structural and metallographic analyses are given
in Table 4.11. The main phase constituents of
these alloys are c-Ni-base solid solution, b(Ni,Co)Al phase, c-(Ni,Co)3Al phase, and Cr-base solid solution. (For the sake of simplicity, these phases are subsequently written as
phases of the Ni-Cr-Al system.) In addition to
these listed phases, still another r-phase of CoCr
type evolves in high-chromium- and cobalt-content field. Chemical compositions of some alloy
phases can be found in Table 4.12.
It is the amount of NiAl phase in the coating
structure that is most essential for the coatingprotection capacity and physical-mechanical
properties. The amount was determined metallographically using the Quantimet instrument
for two conditions: as-cast and postannealing at
1200 C for 200 h. Regression equations, interrelating the amount of b-phase (Qb) with the
alloying of alloys, have been formulated while
statistically processing the results of metallographic measurements. These results are presented in the Appendix.
Aluminum is the main element controlling the
amount of b-phase in alloy and coating structures (Fig. 4.16). With aluminum concentration
rising from 8 to 14% and with chromium and
cobalt being minimum (15 and 10%, respectively), the amount of b-phase in cast alloys is
doubled, from 35 to 70%. Cobalt does not appreciably affect the amount of b-phase; chromium augments the phase content. In alloys containing minimum values of aluminum (8%) and
cobalt (10%) and with chromium concentration
rising from 15 to 30%, the amount of b-phase
increases from 40 to 60%.
In addition to b, c, and c-solid solution, the
structure of the studied alloys contains -Cr and
r-phases. As for -Cr in b-phase, it should be
attributed to its oversaturation with chromium
and to precipitation at cooling of -Cr particles
coherent with the matrix. The structure of the
alloys also has eutectic (b-Cr) when chromium and aluminum are present in their maximum contents (30 and 14%, respectively) in the
structure. The r-phase occurrences are typical
of alloys with maximum content of cobalt and
chromium (Ni30Co30Cr14AlY). Platelike rphase precipitation dramatically degrades the
ductility of alloys and, hence, the coatings
thereof.
Phase transformations, which occur in Ni-CoCr-Al alloys during heating and cooling, were
studied dilatometrically on sample pieces extruded from cast blanks and annealed at 1100 C
for 2 h. (TEC values are given in Chapter 6,
Turbine-Blade-Coating Protective Properties.) The volumetric changes taking place
when an alloy with minimum values of all alloying elements (Ni10Co15Cr8AlY) is heated in
a temperature range in excess of 600 C are due
to cc transformations (Fig. 4.17). Comparison of volumetric effects in the system under
consideration with those of the Ni-Cr-Al system
reveals that c-phase starts to dissolve at a lower
temperature and that the effect of such transformation is less. It is the evidence of less thermal
stability of c-phase when alloyed with cobalt.
Increasing cobalt content up to 30% does not
affect the phase transformations in progress.
In alloys with medium contents of all alloying
elements, phase transformations differ from that
previously mentioned. The difference consists
of the decrease in the total amount of (cc)phases from 65 to 40%, and, ensuing from that,
in a lesser volume effect of (cc) transformation. As heating continues, this transformation fully completes at 750 C. Subsequently,
the alloy is in a (bc)-double-phase field. In
the absence of cobalt, the same transformation
(cc) takes place at a temperature higher by
250 C, that is, at 1000 C. In alloys with high
contents of aluminum, cobalt, and chromium,
the volume effects caused by (cc) transformation are not great. Typical of these alloys is
uniform thermal expansion associated with thermal expansion of b-phase and -Cr solution.
Thus, alloying the Ni-Cr-Al system with cobalt reduces thermal stability of c-phase, decreases its quantity, and converts the Ni-Co-CrAl system into the state based on (bc). It is
this phase condition that makes coatings so
highly ductile when received in the system considered.
Co-Ni-Cr-Al Alloys. All alloys and coatings
based on the Co-Cr-Al system were thoroughly
studied earlier (Ref 11). When used for protec-
tion of nickel-base superalloys, the gradual rise

of nickel content in the coating has been noted.
Adding nickel to the Co-Cr-Al system can possibly reduce the processes of interaction between
the coating and the superalloy. Table 4.13 gives
chemical and phase compositions of the Co-NiCr-Al alloys produced. (Note that the main
phases in the Co-Ni-Cr-Al and Ni-Co-Cr-Al systems have similar notation.)
The main alloy phase constituents include cCo-base solid solution, b(CoAl)-phase, and
r(CoCr) phase. No low-temperature modification of cobalt solid solution with hexagonal
close-packed lattice was discovered in phase
composition. This is due to introducing nickel
into the Co-Cr-Al-system alloys, which repays
in stabilizing the high-temperature modification
of face-centered cubic solid solution. The bphase in the Co-Ni-Cr-Al system is similar to
the b-phase in nickel-base systems. It has bodycentered cubic lattice. No c-phase (one of the
most important phases in the Ni-Cr-Al system)
was discovered in the alloys studied. Microstructure of the alloys is illustrated in Fig. 4.18.
Low-aluminum alloys consist of solid solution with a small quantity of b-phase. Increased
contents of aluminum and chromium cause the
emergence of r-phase in the alloys. Alpha-chromium precipitates in b-phase of alloys containing 12% Al and 30% Cr.
Dilatometric investigations were conducted
on sample specimens cut out of as-cast blanks.
The specimens were annealed at 1000 C for 2
h. (TEC values are given in Chapter 6, TurbineBlade-Coating Protective Properties.) With regard to the thermal expansion coefficient, no appreciable alloying effects have been noted in the
temperature ranges of 20 to 500 C. In the 500
to 600 C range, the alloys experience magnetic
transformation (Fig. 4.19). At higher temperatures, b- and r-phases begin to dissolve. All of
the previously mentioned phase transformations
result in variations in the specimen coefficients
of linear expansion.

Overlay Coatings
The structure and phase composition of coatings produced by the electron beam (EB)-physical vapor deposition and EA techniques depend
on the chemical and phase composition of the
alloy evaporated and on a number of process
factors. The most important of these factors are:
Fig. 4.16
b-phase contents of (a) as-cast alloys and (b) those annealed at 1200 C vs. element content
Thermal expansion coefficients vs. temperature. (a) 1, Ni10Co15Cr8AlY alloy; 2, Ni30Co15Cr8AlY alloy; 3,
Ni10Co30Cr8AlY alloy; 4, Ni30Co30Cr8AlY alloy. (b) 1, Ni10Co15Cr14AlY alloy; 2, Ni30Co15Cr14AlY alloy; 3,
Ni10Co30Cr14AlY alloy; 4, Ni30Co30Cr14AlY alloy
Fig. 4.17
Table 4.13
Chemical and phase compositions of Co-Ni-Cr-Al alloys
No.
Alloy
Co
Ni
Cr
Al
Qb, %
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Co10Ni15Cr6AlY
Co30Ni15Cr6AlY
Co10Ni30Cr6AlY
Co30Ni30Cr6AlY
Co10Ni15Cr12AlY
Co30Ni15Cr12AlY
Co10Ni30Cr12AlY
Co30Ni30Cr12AlY
Co10Ni22Cr9AlY
Co30Ni22Cr9AlY
Co20Ni15Cr9AlY
Co20Ni30Cr9AlY
Co20Ni22Cr6AlY
Co20Ni22Cr12AlY
Co20Ni22Cr9AlY
Co20Ni22Cr9AlY
Co20Ni22Cr9AlY
Co22Cr9AlY
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
base
base
Base
Base
Base
Base
Base
Base
10.6
30.8
10.8
30.6
10.6
30.6
10.7
29.7
12.6
31.0
22.4
22.2
17.2
21.2
20.4
20.6
17.8
...
15.1
14.9
29.8
30.2
15.0
15.1
29.8
29.2
22.5
22.4
14.3
30.0
23.5
22.6
22.6
22.5
22.6
22.6
6.0
5.9
6.0
6.2
11.8
11.8
12.0
11.8
9.3
9.2
9.1
9.2
6.9
11.8
9.3
9.0
9.3
9.2
0.3
0.3
0.3
0.3
0.3
0.3
0.2
0.6
0.7
0.2
0.2
0.1
0.1
0.3
0.3
0.1
0.5
0.3
5
...
...
...
...
68
...
...
...
42
...
...
14
...
...
39
...
...
c, b
c, b
c, b
c, b
c, b
c, b, r
c, b, r
c, b, r
c, b, r
c, b, r
c, b
c, b, r
c, b, r
c, b, r
c, b, r
c, b, r
c, b, r
c, b, r
Microstructure of the alloys. (a) Co30Ni15Cr6AlY. (b) Co10Ni30Cr6AlY. (c) Co30Ni15Cr12AlY. (d) Co30Ni30Cr12AlY.
(e) Co10Ni22Cr9AlY. (f) Co30Ni22Cr9AlY. (g) Co20Ni15Cr9AlY. (h) Co20Ni30Cr9AlY. (i) Co20Ni22Cr9AlY. (j)
Co22Cr9AlY. 500
Fig. 4.18
Temperature of heating (the blades)

Stability of evaporation process
Conditions of blade revolution in vapor flow;
the radius of curvature and roughness of the
blade surface
Peening and heat treatment conditions of
coated blades
Thermal expansion coefficients vs.temperature (a)

1, Co10Ni15Cr6AlY alloy; 2, Co30Ni15Cr6AlY
alloy; 3, Co10Ni30Cr6AlY alloy; 4, Co30Ni30Cr6AlY alloy. (b)
1, Co10Ni15Cr12AlY alloy; 2, Co30Ni15Cr12AlY alloy; 3,
Co10Ni30Cr12AlY alloy
Fig. 4.19
Fig. 4.20
Any instability of the process involved in

coating deposition results in coatings whose
chemical and phase compositions vary and
whose properties are not stable.
One of the main parameters controlling the
coating structure is the temperature of a substrate on which the vapor is condensed. The diagram in Fig. 3.17 shows the relationship between coating structure and substrate
temperature. The diagram is plotted for pure
metals that have been condensed on nonrevolving substrates.
The most favorable structure of metallic coatings is formed in the zone above T2. In reality,
however, heating to the required temperature
above T2 can be complicated by a risk of overheating the superalloy structure. Figure 4.20
shows condensate structures that have been deposited by the EB technique on substrates having different temperatures (Ref 12). As a rule,
these are fine-grained structures oriented normally to the substrate surface. The orientation is
especially pronounced in coatings with highchromium content or in cobalt-base coatings.
Microstructure of condensate from Ni22Co22Cr12AlY alloy. Substrate temperatures of (a) 850 C, (b) 950 C, and (c)
1150 C. 400. Source: Ref 12
The increase in the substrate temperature improves structure stability but reduces its hardness.
The structure and phase composition of overlay coatings change notably in the course of
peening and heat treatment that follows the condensation process. Microball peening that brings
about mechanical hardening of the coating surface and subsequent annealing, causing recrystallization of the coating structure, strikes a balance between structure and phase composition
of coating. The orientation of structure with respect to the substrate disappears. Retaining the
Fig. 4.21
initial structural orientation in EB-produced

coatings unfavorably affects the coating thermal
fatigue.
Figure 4.21 shows typical microstructures of
overlay coatings obtained from Ni20Cr12AlY
and Ni20Co20Cr12AlY alloys, as well as those
from Ni8Co12Cr7Al alloy (4.5% Ta, 0.3% Re,
0.4% Hf), after a complete cycle of peening and
heat-treating operations. After depostion of the
coating from the Ni8Co12Cr7Al alloy by EA
technique, the diffusion coating of the
Al5%Si1.5%Y alloy was formed on its surface
by the technique just described. Although phase
Microstructure of overlay coatings. (a) Ni20Cr12AlY (EB), 200. (b) Ni20Co20Cr12AlY (EB), 300. (c)
Ni20Co20Cr12AlY (EA), 300. (d) Ni7Co12Cr17Al (Ni8Co12Cr7Al [EA] Al5% Si1.5%Y, [EA]). 480
compositions of the coating and feedstock alloy

principally coincide, there are still some differences caused by instability of the evaporation
process. In the original zone of condensation, the
phase composition was shifted to c-Cr field,
which can be attributed to increased chromiumcontent and lowered aluminum-content at the
first stage of feedstock evaporation.
The Ni-Cr-Al system is unique because at
1020 C, it has a point of bcc transformation. This is why the structure and phase
composition of coatings of this system depend
Fig. 4.22
on the annealing temperature and, subsequently,

on conditions of work. The results of x-ray
structural research are given in Fig. 4.22 and in
Table 4.14.
The coating obtained by the EB technique
from the Ni20Cr12AlY alloy is subject to a heat
treatment operation conducted at 1050 C for 2
to 4 h at the final stage of the whole processing
cycle. As research has shown, such heat treatment brings about the structure based on NiAl
(60 to 65%) phases and c-solid solution. Formation of such structure involves considerable
X-ray pattern of overlay coating surface from Ni20Cr12AlY alloy. 1, initial; 2, annealing 1050 C, 2 h; 3, 1050 C, 4 h;
4, 1100 C 2 h; 5, 900 C, 4 h
Table 4.14 Heat treatment effect on b-phase

quantity in Ni20Cr12AlY coating
Coating technique
applied
Electron beam
Electric arc
Annealing
condition
Quantity of
b-phase, %
After deposition
900 oC, 4 h
1050 oC, 2 h
1050 oC, 4 h
1100 oC, 2 h
After deposition
900 oC, 4 h
1050 oC, 2 h
1050 oC, 4 h
1100 oC, 2 h
55
40
60
65
70
(a)
45
50
60
60
(a) Cannot be determined by x-ray analysis
redistribution of aluminum and chromium between the previously mentioned phases. During
cooling, as the bcc transformation
point is passed through, the previously mentioned redistribution does not have enough time
to occur, and, as cooling goes on, Ni3Al phase
precipitates from c-solid solution in the temperature range of 1000 to 750 C. Thus, a nonequilibrium state b(cc) is fixed at room
temperature. With chromium in the coating exceeding 20%, coherent -Cr precipitates in the
NiAl phase.
The fixed structure is unstable, and it undergoes changes as blades and vanes are heated to
working temperatures. Because coating service
life covers thousands of hours, the coating phase
composition eventually comes to a state that is
equilibrium for a particular temperature. (The
changes caused by oxidation and the reactions
between coating and alloy are discussed in
Chapter 5, Phase and Structural Changes in
Coatings during High-Temperature Tests.)
At temperatures below 1000 C, c or
cb (for the top limit of aluminum concentration) comes out as an equilibrium phase
composition. bcc phase transformation is accompanied with a decrease of volume
and, hence, brings about growing tensile stresses
in the coating. In a coating annealed at 950 C
for 4 h, the NiAl phase lessened quantitatively
to 40%.
High-temperature state of bc is notable for
high ductility, whereas in c state, the ductility of the coating lowers. It follows, therefore,
that high-temperature heat treatments can be
used to good advantage to lengthen
bcc transformation in time, thereby
retaining the ductility. Such heat treatments augment the sizes of phases that interact and, hence,
decrease the rate of their interaction at temperatures below 1000 C.

After being deposited by the EA technique,
the Ni-Cr-Al coating is amorphous in terms of
x-ray picture; it has no reflections from phase
constituents of the coating. This is due to the
conditions of condensation on a cold substrate
(that is typical of the technique), 500 C, and
also owing to fineness of coating structure. After
annealing, the phase composition of EA coatings
is similar to that of EB coatings (Table 4.14). In
the structure of the coatings deposited by EA
technique, some banding can be detected. It
makes their structure similar to the structure of
the coatings deposited from powders by plasma
technique. The presence of a droplet phase that
behaves similar to melted powder particles is responsible for banding development.
The analysis of results from investigating
phase composition of alloys and coatings based
on Ni-Cr-Al, Ni-Co-Cr-Al, and Co-Ni-Cr-Al
systems revealed an important difference between these coatings and their diffusion counterparts. Phase composition of diffusion coatings remains stable when briefly heated in the
temperature range of 20 to 1100 C; however,
in the same range of temperatures, the alloys
used for overlay (condensed) coating deposition
and the coatings deposited show diverse phase
transformations followed by considerable volume effects.
REFERENCES
1. V.S. Sinelnikova, V.A. Podergin, and V.I.

Rechin, Aluminides, Kiev, Naukova Dumka, 1965, 240 p
2. G.W. Goward, D.H. Boon, and C.S. Giggens, Formation and Degradation Mechanism of Aluminide Coatings on NickelBase Superalloy, Trans. ASME, Vol 60,
1967, p 228241
3. A.G. Andreeva, V.A. Konstantinov, Y.A.
Tamarin, and V.V. Terekhova, Structure
and Composition of Aluminized Layer on
Nickel and Its Alloys, Met. Sci. Heat Treat.,
No. 8, 1967, p 2730
4. A. Konstantinov, V.V. Terekhova, and Y.A.
Tamarin, Nature of Aluminized Layers on
Nickel, Met. Prot., No. 2, 1970, p 213216
5. A. Teylor, X-Ray Metallography, Moscow,
Metallurgy, 1970, p 633
6. V.A. Konstantinov, Y.A. Tamarin, and
V.V. Terekhova, Structure and Composition Changes in Aluminized Layer on
Nickel and Its Alloys, Temperature Resis-
tant Protective Coatings, Moscow, Nauka,

1968, p 151157
7. A. Taylor and R.W. Floyd, The Constitution
of Nickel-Rich Alloys of Nickel-Chromium-Aluminum System, J. Inst. Met., Vol
81, 1952/1953, p 2532, 451464
8. S.M. Merchant and M.R. Natis, A Review:
Constitution of the Al-Cr-Ni System, Mater. Sci. Eng., Vol 66, 1984, p 4760
9. Y.M. Doljansky, G.B. Stroganov, and R.E.
Shalin, Optimization of Properties of Machine-Building Materials, Moscow, Voenizdat, 1980, 240 p
10. Y.A. Tamarin, V.I. Nikitin, and N.V. Zabrodina, Research into Structure and Physi-
cal-Chemical Characteristics of NiCoCrAl

System Alloys, Deposition and Application
of Protective Coatings, Leningrad, Nauka,
1987, p 6778
11. G.W. Goward, Coatings and Coating Processing for Gas Turbine Airfoils Operating
in a Marine Environment, Gas Turbine Materials in the Marine Environment Conf.,
J.W. Fairbanks and I. Machlin, Ed., Columbus, 1975, p 277296V
12. P.A. Pap, Research into Structure and
Properties and Design of Electron Beam
Technique for Deposition of Protective
Coatings on Gas Turbine Blades, Ph.D.
dissertation, scientific consultant B.A. Morchan, IES, 1982

Y. Tamarin, p97-117
DOI: 10.1361/pctb2002p097

CHAPTER 6
Turbine-Blade-Coating
Protective Properties
THE COATING should protect blade surfaces against damage caused by a gas flow consisting of a mixture of air and fuel combustion
products.
Different materials based on chromium, aluminides, a number of carbides, and other compounds have high resistance to gas corrosion
(oxidation and/or hot corrosion). Extensive
studies disclosed that aluminides and the alloys
based on them are the best materials for forming
protective coatings on nickel superalloys. This
is explained by two major characteristics of aluminides: their thermal stability on the surfaces
of superalloys strengthened by c-phase, and
their high resistance to oxidation.
The results of research into high-temperature
corrosion of aluminides and the protective coatings based on them are dealt with in numerous
publications, such as Ref 1.
Protective-Properties Evaluation
Methods
Test for Heat Resistance. Superalloy heat
resistance (resistance to oxidation) investigation
results depend on test methods. Alloy heat resistance is determined by exposing the samples
in a furnace in air to a constant or cycling temperature for a certain period. To describe the oxidation process, the following parameters were
used: sample specific weight loss q; and specific weight gain q; oxidation depth, h, for a
certain period, s:
q
m0 mk
in g/m2
S
mox m0
in g/m2
S
hox
q
in lm
q
where m0 is sample initial weight, mk is sample

weight after corrosion product removal, mox is
sample weight with all its oxidation products, S
is sample surface area, and q is sample material
density.
For a majority of the studied aluminide materials, specific weight change, q, on testing for
heat resistance shows its parabolic relationship
with oxidation time, s, at the specified temperature:
qn Kp s
where n is the parabola exponent (for the studied

spectrum of the materials n 2) and Kp is the
oxidation constant in g2 /cm4 s.
Testing is performed within the time allowed
for determining the oxidation general features.
Heat resistance test results depend on many
factors, such as furnace capacity and air circulation conditions, sample surface roughness, materials and dimensions of the crucibles in which
the samples are placed, sample heating and cooling rates, and the number of thermal cycles during testing.
The aircraft turbine blades work under heavy
thermal-cycling conditions. That is why the
evaluation of aluminide alloy and coating protective properties should be made both under
isothermal conditions and at the temperature cyclically changing according to the heating-holding-cooling cycle. Cyclic testing reveals oxide
cohesive strength on the sample surface. Under
cyclic testing conditions, oxidation speeds up
and test time reduces. Cyclic heat resistance testing results depend on the cycle temperature
range, DT, and time of holding the sample at the

cycle maximum temperature, Tmax.
Alloy heat resistance investigation methods
are based on the evaluation of oxidation under
the condition when the sample diameter is substantially bigger than the oxidation penetration
depth. This condition leads to steady-state oxidation when oxidation constant Kp and parabola
exponent n do not alter.
The approach to research into heat resistance
of coated alloys imposes limits on the test procedures and interpretation of the test results. In
this case, heat resistance test peculiarities are
caused by specific coating thickness and thus,
limited amounts of the elements in the coating,
which determine the oxidation general features.
During heat resistance testing, diffusion, together with oxidation, occurred between the protective coating and superalloy. This results in
decreasing the content of the elements that form
protective oxides. That is why the general features of coated sample oxidation processes, determined at the initial stage of testing, substantially change in time.
Heat resistance testing of the coated superalloys is carried out similar to that for uncoated
superalloys. The coated specimen heat resistance was determined by the previously mentioned characteristics, such as coating weight
change and oxidation penetration depth. In addition to those characteristics, a period of protective effect of the oxides formed on the coating
surface (oxide lifetime) or a period of protective
Diagram of coated bar specific mass change during

testing for heat resistance. 1, T1; 2, T2. T1 T2. sq,
oxide lifetime
Fig. 6.1
effect of the coating (coating protective lifetime)

can be determined.
Figure 6.1 shows the specific weight change
of the specimens with the coating (q) in the
form of a diagram versus the test time. At the
initial stage of oxidation, the specimen specific
weight change bears a parabolic relationship to
test time. At a certain moment, the coating composition substantially alters. This is due to the
consumption of the elements providing protective properties for oxidation and diffusion into
the superalloy. The coating composition change
results in changing its general oxidation behavior.
The transitional period is characterized by the
change in composition and structures of the oxides formed on the surface. On completion of
this period, sudden acceleration of the oxidation
processes occurs. It results in oxide spalling. Oxide spalling from the coating surface is not considered to be a criterion of the coating damage.
As testing goes on, new oxide is forming on its
surface. It can have protective properties, too.
However, oxide spalling is the evidence of a
transition to a new stage of oxidation, with a
higher rate that ultimately results in coating oxidation through its full thickness.
A period of oxide protective effect (oxide lifetime) is determined by time, sq, which, in turn,
determines the moment of the loss of protective
effect of the oxides formed on the coating surface. For the coating having a specific composition and thickness, sq depends on the test temperature and environment. Within the time
interval of up to sq, the coating oxidation behavior is determined by the same characteristics
that were used for the evaluation of superalloy
heat resistance.
Heat resistance can be determined metallographically by measuring the coating oxidation
penetration depth for the specified period of
time. The time required for oxidation penetration to the depth equal to the coating thickness
also determines its lifetime.
To speed up testing, the moment of the coating chemical and phase composition change can
be determined instead. This change results in deterioration of the protective oxides properties.
For overlay coatings, it may be the time until the
formation of the nickel-base c-solid solution
through the full coating thickness, with the aluminum and chromium content lower than is required for the formation of protective oxides.
For the diffusion coatings, this is the time for
complete disappearance of the NiAl phase in the
coating outer zone.
Turbine-Blade-Coating Protective Properties / 99
The initial coating thickness affects the heat

resistance test results. As a rule, coating lifetime
is directly proportional to its thickness (Ref 2).
To compare the heat resistance of the diffusion
coatings with different initial thickness values
and aluminum contents, it is convenient to introduce a parameter of aluminum reserve in
the coating.
Tests for Resistance to Hot Corrosion.
When studying aircraft gas-turbine engine turbine blades after their long service life, deposits
containing sodium, sulfur, vanadium, chlorine,
iron, and potassium, which substantially speed
up high-temperature oxidation, are detected on
the blade surfaces. These elements get into a gas
flow in the form of various compounds from the
environment as well as from the combustion of
fuel that contains small amounts of sulfur, vanadium compounds, and other elements.
The mechanism of hot corrosion caused by
the presence of Na2SO4, NaCl, V2O5, and other
aggressive compounds depends on the thermodynamic characteristics of these compounds as
well as temperature, velocity, and pressure of a
gas flow. The difference in the temperatures of
the gas flow and the blade surface contributes
much to corrosion development. The simulations of high-temperature corrosion caused by
sulfur-base compounds are covered in numerous
works (Ref 38).
Various methods can be used for hot corrosion testing. Dipping the specimens into crucibles with aggressive media is widely practiced
as the simplest method. The crucibles are held
at a certain temperature and time in the furnace.
Corrosion progress is determined by inspecting
the specimen appearance and the corrosion penetration depth. Under such testing conditions,
corrosion behavior depends substantially on the
depth the specimen is dipped into the aggressive
media and, as a rule, to the utmost extent, takes
place at the aggressive mixture/air interface.
This method lacks quantitative criteria for comparison of protective properties and shows poor
agreement of the test results with the data on
Table 6.1
(Ref 9)
Hot corrosion test conditions
Fuel at
service
Synthetic ash
composition (GZT)
Gas-turbine
fuel
66.2% Na2SO4; 1.8%

V2O5; 20.4% Fe2O3;
8.3% NiO; 3.3% CaO
Gas medium
composition at
testing
Air, flow rate of
20 L/min
actual coating properties under service conditions.

This drawback is rectified by applying special
synthetic ash that imitates carbon deposit
formed on blade surfaces. The synthetic ash is
deposited from aqueous or alcohol solutions
onto the surface of the specimens. It evenly covers the specimen surface, and its amount is determined by weighing. In the course of hightemperature testing, the synthetic ash can be
applied repeatedly to the specimen at the specified time intervals. The test results are analyzed
and presented as a function of the specimen specific weight change and corrosion penetration
depth versus time. It is done similarly to that
specified previously for oxidation.
The test conditions run on gas-flow test
benches (gas rigs), which provide the gas-flow
parameters similar to the aircraft-engine turbine
parameters and most closely approximate the actual operating conditions. Various combinations
of sodium salt, vanadium oxides, and more complex systems of compounds are used as aggressive mixtures. The aggressive environment composition should provide for the propagation of
corrosion according to the mechanism similar to
that existing under blade service conditions. The
drawback of this method is its high cost and environmental pollution. Shorter test periods, due
to additional aggressive components in the gas
flow, can change corrosion mechanism and
cause inaccuracy of test results.
The proper selection of the testing methods is
a crucial contribution to the successful selection
of the materials and coatings that feature high
resistance to hot corrosion. For this purpose, it
is advisable to employ the well-proven lab methods corroborated by the practice of blade testing
in engines.
To assess the resistance of different materials
and coatings to hot corrosion, which is presented
subsequently, the test method designed by the
authors of Ref 9 was used. Synthetic ash of gasturbine fuel (GZT) was applied to the specimen
surfaces (Table 6.1). Depending on the gas-turbine engine or the gas-turbine unit service conditions, the fuel in use, and the analysis of ash
deposit formed on the blades, the use of specially selected synthetic ash compositions is required. Gas-turbine fuel composition is chosen
on the basis of numerous studies of ash deposits
on turbine blades operating under different conditions (Ref 2).
The initial components of synthetic ash GZT
are thoroughly mixed and milled. Then, an
ethyl-alcohol-base slurry is made of this mixture. The slurry is applied uniformly to the specimen surface in the amount weighed out as 120
g/m2. The initial surface roughness of the specimens should be Ra 0.61.5 lm.
The samples with GZT ash applied to their
surfaces are placed in alundum crucibles, which,
in turn, are arranged in a closed container with
air forced through it at the rate of 20 L/min. In
specified intervals, the specimens are taken from
the container for inspection and weighing, and,
depending on test conditions, specific weight
loss or gain (g/m2) of the specimens is determined. During testing, slurry application can be
repeated more than once.
The tests are carried out at a constant temperature in the range of 700 to 900 C. Test time
depends on the corrosive activity of the synthetic
ash in use, the gas medium, the protective coating properties, as well as on test temperatures
and predicted coated-blade service lives.
After every test cycle, specimen surfaces are
examined by visual inspection and by weighing,
both before and after removal of corrosion products. One specimen from the test batch is randomly taken for metallographic examination.
In the case of testing coated specimens, a few
techniques for determining their lifetime are
available:
Corrosion products are not removed from the

specimen surface. After every stage of testing, specimens are weighed. A sudden change
in kinetics of specimen specific weight gain
or loss is considered to be the moment of
coating failure.
Corrosion products are periodically removed
from the specimen surfaces, and their mass
loss is determined. Sudden specific weight
change is evidence of the change in corrosion
mechanism and kinetics caused by coating
failure.
Scale is removed by methods that do not result in damage to the remaining coating. The
suitability of the scale removal method is veri-
fied by special check tests of the as-coated specimens. Corrosion products can be removed by
their reduction in metallic sodium with ammonia
bubbling through it (Ref 9). The degree of corrosion damage is determined from specimen
mass loss, q, and corrosion penetration depth,
hk.
The parameter chosen to characterize corrosion lifetime of the coating is the time interval
from the start of specimen testing to the moment
of coating damage, even though damage is in
one location on the surface. In all cases, the fact
of coating damage should be supported by the
data of metallographic examination.
Heat Resistance of Aluminides and

Alloys for Overlay Coatings
Heat Resistance of Aluminides. NiAl compounds or a combination of NiAl and Ni3Al
compounds are an outer-zone base for as-deposited and as-annealed aluminide diffusion coatings on nickel and superalloys. These very compounds determine the protective properties of
diffusion coatings at high temperatures.
To study heat resistance, the samples of
Ni2Al3, NiAl, and Ni3Al aluminides as well as
NiAl alloyed with chromium and silicon were
cast in vacuum induction furnaces. The compositions of the compounds are given in Table
6.2.
Heat resistance testing was carried out under
air atmosphere at 1000 C. During testing, specimen weight changes and the structure of oxides
formed on the specimens were examined. The
test results are shown in Fig. 6.2. Table 6.2 contains the x-ray diffraction analysis data of the
oxides after 500 h of testing.
Heat resistance of Ni2Al3 and NiAl compounds with different aluminum contents is actually the same. A protective, dense oxide layer
with -Al2O3 structure is formed on the surface.
Heat resistance of Ni3Al compound on whose
Table 6.2 Analysis of tested aluminides and crystal structure of oxides formed on their surface after
testing at 1000 C
Aluminide
Ni2Al3
NiAl
NiAl
Ni3Al
NiAl
NiAl
Oxidation time, h
Oxide structure
58% Ni, 41.6% Al

64% Ni, 36% Al
78% Ni, 22% Al
86% Ni, 14% Al
74% Ni, 23% Al, 3% Si
74% Ni, 23% Al, 3% Cr
75
500
500
500
500
500
-Al2O3
-Al2O3
-Al2O3
NiO, NiAl2O4, -Al2O3
-Al2O3
-Al2O3
surface a mixture of -Al2O3, NiAl2O4, and NiO

oxides was found is a bit lower than that of NiAl
compound.
Alloying NiAl with silicon has a beneficial
effect on heat resistance. During in-testing inspection of the NiAl specimens alloyed with silicon, a denser, less-spalling oxide is seen. However, the electron spectroscopy and x-ray
diffraction did not reveal structural differences
between the oxides formed on the NiAl specimens with and without silicon.
Alloying NiAl with such elements as zirconium, molybdenum, and tungsten, which, due to
their low solubility in this compound, form their
own phases of -Mo and -W types, results in
reducing its heat resistance. In this case, oxidation first penetrates the zones containing -Mo
and -W phases. It results in the penetration of
oxidation into the aluminide and the damaging
of the -Al2O3 protective oxide. Carbides and
other compounds of the previously mentioned
elements have a similar effect.
The mechanism of the NiAl oxidation is selective oxidation of aluminum at the Al2O3/NiAl
interface, due to the diffusion of oxygen to this
interface through the formed oxide layer. In this
case, the kinetics of oxidation is defined by diffusive mobility of oxygen within the Al2O3 lattice. However, no aluminide oxidation penetration depth is observed, because the oxide layer
is formed on the NiAl surface and does not penetrate the bulk compound. At a certain stage of
oxidation, the consumption of aluminum for
forming Al2O3 and simultaneous accumulation
Heat resistance of aluminides at 1000 C. 1, NiAl

(23% Al, 3% Si); 2, NiAl (24 to 26% Al); 3, NiAl
(23% Al, 3% Cr); 4, Ni3Al (14% Al)
Fig. 6.2
of nickel result in forming Ni3Al compound beneath the oxide layer. In this case, the mechanism of oxidation changes, because the formation of NiAl2O4 spinel and, hereafter, NiO oxide
results in oxidation penetration into the aluminide.
Heat resistance of Ni-Cr-Al alloys was investigated for the compositions simulating overlay coatings. Alloy compositions are given in
Tables 4.10 and 6.3. Testing was carried out isothermally and cyclically at 900 C. In all the
cases, sample specific weight gain (q, g/m2)
was measured.
During cyclic testing, the temperature varied
up and down the range of 20 to 900 C; the holding time at the maximum temperature was 1 h.
The total holding time in the vicinity of 900 C
was determined graphically by reference to the
temperature measurement chart. The longest isothermal test lasted 10,000 h; cyclic testing took
4500 h (4500 cycles).
All the Ni-Cr-Al-Y alloys under study have
high heat resistance at 900 C (Tables 6.4, 6.5).
During 4500 h of cyclic testing, no oxide scales
on any of the tested alloys of this system, except
for those containing hafnium, were damaged
(Table 6.6). This is explained by the formation
of -Al2O3Cr2O3 oxide scale on their surface.
Replacement of yttrium with hafnium impairs
the protective properties of the oxides formed.
Alloying Ni20Cr11Al alloy with 1% Hf results
in reducing the service life of the protective oxides to 5000 h under isothermal testing conditions and to 2000 h under cyclic testing conditions. The increase up to 2% Hf content of the
alloy reduces oxide protective life at isothermal
testing and increases it a little at cyclic testing.
Cyclic testing at 1200 C has revealed the unfavorable effect of hafnium and silicon on heat
resistance of Ni-Cr-Al alloys (Fig. 6.3). X-ray
diffraction of the oxides formed on the surface
of the alloys with hafnium revealed that HfO2
was observed to appear in addition to the crystal
structure similar to that of -Al2O3Cr2O3.
Heat Resistance of Ni-Co-Cr-Al Alloys. Alloy compositions are given in Tables 4.11 and
6.3. Heat resistance testing of this alloy system
was carried out at 900, 1100, and 1200 C. The
isothermal and cyclic test results at 900 C are
given in Tables 6.7, 6.8, and 6.9.
The Ni-Co-Cr-Al-system alloys alloyed with
yttrium, silicon, and hafnium have high heat resistance. The Ni20Co20Cr12Al alloy alloyed
with 2% Hf and 0.5% Si has an oxidation con-
stant less than that for the same alloy alloyed

with yttrium, and the oxide layer life under isothermal and cyclic testing conditions is more
than 10,000 and 5000 h, respectively. The same
Table 6.3
effect is produced when alloyed with 2% Si and

0.5% Hf. However, the life of the oxide layer on
this alloy is reduced a little, to about 5000 h
under isothermal testing conditions.
Chemical composition of alloys with hafnium and silicon additions

Alloy
Ni22Cr11Al2Hf
Ni22Cr11Al1Hf
Ni22Cr11Al1Si
Ni22Cr11Al2Si
Ni20Co22Cr11AlHfSi
Ni20Co22Cr11Al2HfSi
Ni20Co22Cr11AlHf2Si
Ni20Co22Cr11Al2Hf2Si
Ni20Co22Cr11AlY2HfSi
Table 6.4
Ni
Co
Cr
Al
Hf
Si
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
20.0
20.0
20.0
20.0
20.0
20.0
22.0
22.0
22.0
22.0
22.5
22.5
22.5
22.5
22.5
22.5
11.0
11.0
11.0
11.0
11.0
11.0
11.0
11.0
11.0
11.0
0.3
2.0
1.0
0.5
2.0
0.5
2.0
1.2
1.5
1.0
1.5
0.5
0.5
2.0
2.0
1.2
0.5
Heat resistance of Ni-Cr-Al alloys. Isothermal tests at 900 C

Weight gain (q), g/m2
Alloy
100 h
500 h
1000 h
2500 h
5000 h
10,000 h
1.4
1.6
3.0
2.5
3.4
1.8
2.9
2.1
2.8
3.8
3.5
4.1
3.6
4.6
2.8
3.6
4.6
4.5
4.9
4.5
5.9
3.5
4.3
5.2
4.8
5.4
4.5
6.5
4.9
5.8
6.5
6.4
5.9
3.9
6.9
5.9
5.1
7.3
7.1
7.0
1.6
1.4
Ni22Cr11AlY
Ni16Cr9AlY
Ni21Cr9AlY
Ni19Cr12AlY
Ni28Cr11AlY
Ni22Cr11Al2Hf
Ni22Cr11Al1Hf
Table 6.5 Heat resistance of Ni-Cr-Al alloys. Cyclic tests under 20 to 900 C (1 h),
20 C conditions
Alloy
Ni22Cr11AlY
Ni16Cr9AlY
Ni21Cr9AlY
Ni19Cr12AlY
Ni28Cr11AlY
Ni22Cr11Al2Hf
Ni22Cr11Al1Hf
Table 6.6
900 C
Alloy
Ni16Cr9AlY
Ni18Cr5AlY
Ni21Cr9AlY
Ni19Cr12AlY
Ni28Cr11AlY
Ni20Cr11Al1Hf
Ni20Cr11Al2Hf
100 h
500 h
1000 h
2500 h
4500 h
4.9
2.6
1.2
3.8
2.8
1.9
4.0
8.2
4.6
2.5
5.3
4.5
3.0
8.0
9.5
4.7
3.4
5.5
5.5
3.5
10.0
12.0
6.2
4.2
6.8
6.5
5.2
18.0
14.0
6.0
4.7
6.9
6.5
3.3
59.0
Life of Ni-Cr-Al alloys oxides at

Oxide life, h
Kp 1014,
Isothermal tests Cyclic tests g2/cm4 s
10,000
10,000
10,000
10,000
10,000
5000
2000
5000
5000
5000
5000
5000
2000
5000
3.6
5.6
5.8
2.2
6.6
9.6
5.6
In all cases, indicates that lifetime is longer than sample test time.
Heat resistance of alloys during cyclic testing in the

temperature range of 20 1200 C (1 h). 1,
Ni22Cr11AlY; 2, Ni22Cr11Al1Hf; 3, Ni22Cr11Al2Hf; 4,
Ni22Cr11Al1Si; 5, Ni22Cr11Al2Si
Fig. 6.3
Table 6.7
Heat resistance of Ni-Co-Cr-Al alloys. Isothermal tests at 900 C

Alloy
Ni20Co22Cr11AlY
Ni20Co22Cr11AlHfSi
Ni20Co22Cr11Al2HfSi
Ni20Co22Cr11AlHf2Si
Table 6.8
100 h
500 h
1000 h
5000 h
10,000 h
2.9
1.0
1.0
2.3
1.9
4.4
1.9
2.3
3.2
3.9
6.3
2.8
3.0
4.3
5.6
17.0
4.4
7.2
4.8
11.0
19.0
4.8
9.6
2.9
15.0
Heat resistance of Ni-Co-Cr-Al alloys. Tests under 20 to 900 C (1 h), 20 C conditions

Alloy
Ni20Co22Cr11AlY
Ni20Co22Cr11AlHfSi
Ni20Co22Cr11Al2HfSi
Ni20Co22Cr11AlHf2Si
100 h
500 h
1000 h
2500 h
4500 h
4.1
1.1
3.4
2.5
3.2
5.8
2.4
5.3
3.6
5.6
6.3
2.7
5.8
4.4
7.0
8.4
6.5
6.4
5.2
11.0
9.4
35.0
6.1
5.7
15.0
Comparison of oxidation constants of

Ni20Cr11AlY and Ni20Co20CrAlY alloys
shows that cobalt reduces heat resistance and
speeds up the oxide-layer formation and growth.
At the test temperatures of 1100 and 1200 C,
heat resistance of the alloys was determined
from specimen weight change, from the formed
oxide layer structures, and from the phase composition and structure of the subscale zones. The
results of the measurements of the specimen specific weight change during testing are presented
in Fig. 6.4, 6.5, and 6.6.
As a result of statistics-based processing of
specific weight measurements of the Ni-Co-CrAl alloys after testing at 1200 C for 100 h, the
regression equations were derived. They relate
specific weight gain, q, with the cobalt, chromium, aluminum, and yttrium contents of the
alloys. Common for the whole spectrum of the
investigated compounds is the fact that they
have high heat resistance at 1100 and 1200 C,
which hardly varies with the addition of alloying
elements. This fact is explained by the similar
Table 6.9
at 900 C
Life of Ni-Co-Cr-Al alloys oxides

Oxide life, h
Alloy
Ni20Co22Cr11AlY
Ni20Co22Cr11AlHfSi
Ni20Co22Cr11Al2HfSi
Ni20Co22Cr11AlHf2Si
Isothermal
tests
Cyclic
tests
Kp 1014,
g2/cm4 s
10,000
10,000
10,000
5000
10,000
5000
2000
5000
5000
5000
11.0
2.1
2.5
5.1
1.3
In all cases, indicates that lifetime is longer than sample test time.
crystal structure of the oxides, which, in all the

alloys, consists of -Al2O3Cr2O3 and (NiCo)
(CrAl)2O4 spinel. In the alloys with high cobalt
content, the spinel proportion increases. If the
cobalt content is about 30%, spinel, by its lattice
parameters, approaches CoAl2O4.
In all the alloys alloyed with hafnium, the formation of HfO2 is observed. The oxide weight
gain is observed in the alloys of a higher yttrium
content, due to their better adherence to the surface as well as to the extended zone of internal
oxidation.
More detailed information on the effect of alloying elements was obtained from the analysis
of b-(NiAl)-depleted outer-zone dimensions,
which are a measure of aluminum consumption
for protective oxide formation (Fig. 6.7). The
regression equation relating the depth of bphase-depleted subscale in NiCoCrAlY alloys
with their alloying is presented in the Appendix.
Aluminum has the main effect on the dimensions of the zone free from b-phase. This is explained by different aluminum contents of the
alloy outer zones. The calculated amounts of
aluminum in the 100 lm-thick outer zone of the
alloys, which contain 8, 11, and 14% Al, are 8,
30, and 55 g/m2, respectively. During calculations, it was assumed that the lower limit of
the aluminum content of the alloy that contains
b-phase is about 7%. The aluminum consumption rates for the formation of oxides for all the
alloys during testing are rather similar and are
10 to 15 g/m2. Thus, all the aluminum accumulated in the b-phase of the alloys with 8% Al
is consumed for the formation of the oxides,
while approximately only 14 is consumed in the
Fig. 6.4
Heat resistance of alloys under isothermal testing conditions at 1100 C. 1, Ni10Co15Cr8Al; 2, Ni30Co15Cr8Al; 3,
Ni10Co30Cr8Al; 4, Ni10Co15Cr8AlY; 5, Ni10Co15Cr14Al; 6, Ni30Co15Cr14Al
Heat resistance of alloys under isothermal testing

conditions at 1200 C. (a) 1, Ni10Co15Cr8Al; 2,
Ni30Co15Cr8Al; 3, Ni10Co30Cr8Al; 4, Ni30Co30Cr8Al; 5,
Ni10Co15Cr8AlY
and
(b)
5,
Ni10Co15Cr14Al; 6,
Ni10Co15Cr14AlY
Fig. 6.5
alloy with 14% Al. From the obtained results, it

follows that when comparing heat resistance
versus b-phase-free zone dimensions, the alloys
of high-aluminum contents have much higher
resistance than the alloys of lower-aluminum
content.
Addition of chromium to the studied alloys
results in decreasing the dimensions of the zone
depleted of b-phase. However, chromium has a
different effect on the alloys with varying aluminum contents. With 8% Al, and especially
with the lower-cobalt content, more chromium
reduces the zone depleted with b-phase and increases heat resistance, respectively. This effect
of chromium at 1200 C is explained by the fact
that chromium is active in oxidation, accompanied by the partial substitution of chromium in
Al2O3Cr2O3 oxide. With the high-aluminum
content, the effect of chromium on oxidation is
insignificant.
The effect of cobalt is most conspicuous in
the alloys with low-aluminum content. Increasing the cobalt content of the 8% Al alloy to 20
to 25% results in a slowdown of the oxidation
processes. The zone depleted of b-phase diminishes from 150 lm at 10% Co to 90 lm at 20%
Co. Further increase in the cobalt content results
in the gradual progress of oxidation processes
and a larger b-phase-depleted zone. The effect
of cobalt is associated with the changes in the
protective properties of (NiCo)(CrAl)2O4 spinel,
whose amount increases as the cobalt content
Heat resistance of the alloys under isothermal testing conditions at 1200 C. (a) 1, Ni20Co15Cr11AlY; 2, Ni10Co22Cr11AlY;
3, Ni30Co22Cr11AlY; 4, Ni20Co30Cr11AlY and (b) 1, Ni20Co22Cr14AlY; 2, Ni20Co22Cr11AlY; 3, Ni20Co22Cr8AlY; 4,
Ni22Co22Cr11AlY
Fig. 6.6
increases. At a certain stage, it transforms into

cobalt-base spinel of Co(CrAl)2O4, which has
worse protective properties. With the high-aluminum content, the effect of cobalt is insignificant. The effect of yttrium on the formation of
the b-phase-free zone is insignificant.
Cyclic heat resistance testing in the temperature range of 20 1200 C, with the holding
time of 1 h at the cycle maximum temperature
1200 C, confirmed the relationships generally
observed during the other types of heat resistance testing (Fig. 6.8). Hafnium and silicon in
the specified amounts (1.5% Hf, 0.5% Si)
increase cyclic heat resistance, especially if they
are used together with yttrium. Complete sub-
stitution of yttrium with hafnium and silicon in

the Ni-Co-Cr-Al system results in lower cyclic
heat resistance.
Heat Resistance of Co-Ni-Cr-Al Alloys. Alloy compositions are given in Table 4.13. The
results of research into heat resistance of Co-NiCr-Al-Y-system alloys at 900 C are presented
in Tables 6.10 and 6.11. The regression equations derived from statistics-based processing of
Heat resistance of the alloys during cyclic testing

under condition of 20 1200 C cycling (1 h). 1,
Ni20Co22Cr11AlY;
2,
Ni20Co22Cr11Al2Hf2Si;
3,
Ni20Co22Cr11AlY2HfSi;
4,
Ni20Co22Cr11Al2HfSi;
5,
Ni20Co22Cr11Al1Hf1Si;
6,
Ni20Co22Cr11AlHf2Si;
7,
Ni20Co22Cr11AlHfSi
Fig. 6.8
Effect of alloying on thickness of zone free from
nickel-aluminum phase in Ni-Co-Cr-Al alloys.
Tested at 1200 C for 200 h
Fig. 6.7
Table 6.10
ditions
Heat resistance of Co-Ni-Cr-Al alloys. Cyclic tests under 20 to 900 C (1 h), 20 C conWeight gain (q), g/m2
Alloy
100 h
500 h
1000 h
2500 h
5000 h
10,000 h
Co10Ni15Cr6AlY
Co10Ni30Cr6AlY
Co30Ni30Cr6AlY
Co10Ni30Cr12AlY
Co10Ni22Cr9AlY
Co20Ni30Cr9AlY
Co20Ni22Cr6AlY
Co20Ni22Cr9AlY
Co20Ni22Cr9AlY
Co20Ni22Cr9AlY
12
1.1
1.2
4.3
5.4
1.5
2.6
2.3
1.9
5.0
13
1.9
2.9
4.9
6.5
2.9
2.7
2.8
2.4
7.0
16
4
3.9
7.3
7.7
3.1
3.7
3.7
3.2
8.5
19
5.4
5.8
9
9.4
4.5
5.3
4.8
4.7
9.5
22
5.4
7.4
8.7
10
5.3
6.4
4.9
5.3
12
29
6.6
11
7.4
11
4.9
5.4
5.6
4.9
13
the test results are given in the Appendix. The

total test time was 8000 h.
All the alloys studied have high heat resistance. During testing, the oxide layers on the
alloys were not damaged, except for the alloy of
low-aluminum and -chromium contents. Increasing nickel, aluminum, and yttrium contents
of this system results in a slight rise in the alloy
oxidation rates. The oxidation rates for alloys
with the maximum content of the previously
mentioned elements increase 1.5 to 2 times.
Chromium exerts an opposite effect. Increasing its content from 10 to 30% decreases the
oxidation rate during cyclic heat resistance testing 5 times. The favorable effect occurs in the
chromium content range of 15 to 22.5%. These
general features of the alloying elements do not
depend on the test time. Despite alloying, the
oxides of all the tested alloys are highly stable.
Heat Resistance of Coated Superalloys

During heat resistance testing and other hightemperature tests, the following reactions beTable 6.11 Life of Co-Ni-Cr-Al alloys oxides. Cyclic testing under 20 to 900 C (1 h),
20 C conditions
Alloy
Co10Ni15Cr6AlY
Co10Ni30Cr6AlY
Co30Ni30Cr6AlY
Co10Ni30Cr12AlY
Co10Ni22Cr9AlY
Co20Ni30Cr9AlY
Co20Ni22Cr6AlY
Co20Ni22Cr9AlY
Co20Ni22Cr9AlY
Co20Ni22Cr9AlY
JS6K
Oxide life, h
Kp 1014(a),
g2/cm4 s
7000
8000
8000
8000
8000
8000
8000
8000
8000
8000
2000
71
4.4
4.2
15
16
2.7
3.8
3.8
2.8
20
(a) Designed for 1000 h of oxidation
tween the oxidizing environment, coating, and

superalloy take place:
Formation of a protective oxide on the coating surface

Diffusion of aluminum, chromium, and other
components of the coating into the oxide
scale
Diffusion of aluminum and elements added
for alloying the coating into the superalloy
Diffusion of the alloy components into the
coating
These processes result in the decrease in aluminum and chromium coating contents down to
the level when protective oxides no longer form
on the coating surface. Oxidation processes
speed up suddenly, and the protective effect of
the coating ends. For full and proper estimation
of coating protective properties during testing
for heat resistance, research into each of the previously mentioned processes is necessary.
Heat Resistance of Superalloys with Diffusion Coatings. The analysis of the results of
heat resistance testing of the superalloy samples
(e.g., Ni20CrTi, Ni10CrMoWCo, JS6K, JS6U,
and JS30) with various diffusion aluminide coatings shows that the distinctions in the kinetics
of their specific weight change (q, g/m2) are
insignificant. This is explained by the fact that
the oxides with the same or similar properties
are formed on the surface of the diffusion coatings.
X-ray diffraction of the diffusion coatings after 100 and 1000 h of testing at 950 C allowed
the detection of the oxides with -Al2O3 structure on their surfaces. The same crystalline
structure and protective properties of the oxides
that are formed on the surface of diffusion aluminide coatings deposited on different nickel alloys are the cause of their similar heat resistance.
During testing, the aluminide-coating oxidation
rate is considerably reduced; if, in the first 2 h

of holding at 950 C, it is approximately 0.3
g/m2 h, on a lapse of 100 h, it is about 0.025
g/m2 h. The sample specific weight change
after 100 h of testing corresponds to the formation of an Al2O3 oxide layer 2 to 4 lm thick.
The increase in protective oxide layer thickness at testing reduces layer adherence to the
coating surface and causes cracking and spalling. The oxide layer defects are rectified by aluminum diffusing from the coating. Aluminum
depletion of the coating outer zone at testing for
heat resistance results in altering the protective
oxide structure and speeding up oxidation. However, as long as aluminides remain in the coating
outer zone, including Ni3Al, the oxides provide
protection. They protect the coating and alloy
against accelerated oxidation.
Heat resistance testing of nickel superalloys
with aluminide diffusion coatings was carried
out by determining the sample weights during
isothermal oxidation in the range of 900 to 1000
C. The diffusion coating was deposited on the
samples by aluminizing in the mixture of 98%
Al-Fe alloy (50% Al) and 2% NH4Cl; initially
it was 40 to 50 lm thick, with 34 to 36% Al
Fig. 6.9
Parametric diagram of heat resistance of aluminized

coatings on superalloys at 950 C
content of the outer zone. During the analysis of

the test results for a period of more than 1000
h, the alteration of the oxidation mechanism related to the formation of NiAl2O4 spinel on the
coating and its partial spalling was observed.
The parametric heat resistance diagram
shown in Fig. 6.9 presents numerous results of
1000 h testing of the superalloys with diffusion
coatings. Plotting such diagrams is proposed
elsewhere (Ref 10). A parametric diagram allows integration of test results measured at different temperatures for different periods:
lgq
1
(lg k0 p)
n
where q is specific weight gain (g/m2), n and k0

are coefficients depending on materials properties and test medium, and p is the heat resistance
parameter.
The heat resistance parameter is calculated
from the equation:
p
0.218 Q
lg t
T
where Q is oxidation activation energy (in the

temperature range of 900 to 1000 C; Q
267,000 cal/mol); T is test temperature, in K;
and t is test time, in hours.
All the tested superalloys with aluminide diffusion coatings have high heat resistance practically independent of the alloy composition.
The test temperature rise to 1100 C results in
decreasing heat resistance of the diffusion coating. However, even at this temperature, the coating provides protection against oxidation for 200
to 400 h for the alloys of the JS-type (e.g., JS6K,
JS6U, and JS30).
The heat resistance test results for the JS6U
superalloy with different versions of diffusion
coatings deposited by the electric arc (EA)
method are presented in Fig. 6.10 and in Tables
6.12 and 6.13.
The oxidation constant of the coatings alloyed
with silicon and yttrium is approximately onethird of that of the diffusion coating formed in
the powder mixture with the composition of
98% Al-Fe alloy (50% Al) and 2% NH4Cl. If
the aluminum alloy that contains more than 10%
Si is used for saturation, oxide-layer spallation
occurs. High-silicon content reduces layer lifetime to 2000 to 3000 h, while the life of the
layers on the coatings with less silicon content
exceeds 10,000 h.
Alloying the aluminide coatings with silicon

and yttrium increases their heat resistance. High
heat resistance of aluminide coatings alloyed
with silicon is explained by the fact that silicon
Fig. 6.10
partially substitutes ions of aluminum and promotes a rise in the -Al2O3 oxide density in it.
Heat Resistance of Superalloys with Overlay Coatings. The characteristics of the coatings
Heat resistance of JS6U alloy with diffusion coatings at 900 C. (a) Isothermal test conditions, and (b) cyclic test conditions,
20 900 C (1 h). 1, Aluminized coating; 2, Al5Si1.5Y electric arc coating; 3, Al10Si0.7Y electric arc coating
Table 6.12 Oxidation constants of JS6U superalloy with diffusion aluminized coatings.
Isothermal tests
Oxidation
constant,
Kp 1014,
g2/cm4 s
Aluminizing technique
Alloy
900
C
950
C
1000
C
Aluminizing in mixture
of Al-Fe alloy (50%
Al), 2% NH4Cl
EA (Al5Si1.5Y)
EA (Al10Si0.7Y)
EA (Al15Si0.5Y)
Ni10CrMoWCo
JS6K
JS6U
JS6U
JS6U
JS6U
6.7
7.2
6.9
2.3
4.0
3.0
10.0
12.0
17.0
20.0
Table 6.13 Life of oxides on coatings on

JS6U superalloy. Cyclic testing under 20 to
900 C, 20 C conditions
Coating
Co20Cr11AlY
Co20Cr11AlY/ZrO2-15%Y2O3
Ni20Co20Cr11AlY/ZrO215%Y2O3
Fe12Ni20Cr12Al1SiY
Al
AlPd
Al10Si0.7Y
Al15Si0.5Y
Uncoated JS6U
Oxide
life, h
Kp 1014(a),
g2/cm4 s
4000
4000(b)
4000(b)
4.7
1.7
0.71
4000
1000
6000
4000
2500
2000
10.0
5.1
1.4
4.0
3.0
22.0
(a) Designed for 1000 h of oxidation. (b) Ceramic layer lifetime
studied are presented in Table 6.14. During testing in the temperature range up to 1000 C, an
oxide with the -Al2O3Cr2O3 structure is
formed on the surface of the superalloys with
overlay coatings in the same manner as on the
alloys with diffusion coatings. This explains
why the specific weight change of the superalloy
samples with overlay coatings is similar to the
specific weight change of the superalloy samples
with diffusion coatings (Table 6.13 and Fig.
6.11).
The use of the iron-base coatings for protection
of nickel-base superalloys cannot provide the desired heat resistance. The life of the protective
oxide layers on the Fe12Ni20Cr12Al1SiY coating at 900 C does not exceed 4000 h (4000 cycles). It is caused by instability of the coating
composition during testing.
The formation of a thin ceramic layer (15 lm
thick) on the coating surface by the electron
beam (EB) method allows the improvement of
coating heat resistance and retardation of its oxidation. After 400 h of cyclic testing, the ZrO215% Y2O3 ceramic spalls. However, the metallic
coating still preserves its protective properties.
At temperatures lower than 1000 C, interaction between nickel-base coatings and superalloys may be neglected. A temperature rise to
1100 C increases aluminum, chromium, and co-
Table 6.14
JS6U alloy
Characteristics of coatings on
Coating
Co20Cr11AlY
Ni20Co20Cr11AlY
Ni23Co28Cr9AlY
Ni28Cr6AlY
Fe12Ni20Cr12Al1.5SiY
Ni20Cr12AlY/Cr carbide
Co20Cr11AlY/ZrO215%Y2O3(a)
Ni20Co20Cr11AlY/ZrO215%Y2O3(a)
Al(b)
AlPd(b)
Al10Si0.7Y(b)
Al15Si0.5Y(b)
Ni8Co12Cr7Al(c)
Ni7Co12Cr17Al(c,d)
Deposition
method
Thickness of
layers, lm
EB
EB
EB
EB
EB
EB
EB
100 20
120 20
110 20
120 20
180 20
50 10(a)
100 20
EB
100 20
Aluminizing
Aluminizing
EA
EA
EA
EA
50
60
50
60
80
80
5
5
5
5
5
5
EB, electron beam; EA, electron arc. (a) Ceramic coating is 15 lm thick.
(b) Diffusion coatings. Al, aluminizing in powder mixture of 98% Al-Fe
(40% Al) alloy, 2% NH4Cl; AlPd, palladium layer (10 lm), aluminizing in powder mixture of 98% Al-Fe (40% Al) alloy, 2% NH4Cl. Chromium carbide sublayer 10 lm thick. (c) Coating additionally contains
4.5% Ta, 0.3% Re, and 0.4% Hf. (d) Two-layer (overlay diffusion)
coating (Fig. 4.21d)
balt diffusive mobility. The amounts of these

elements diffusing from the coating to the interaction zone become comparable with their
amounts spent on oxidation.
During heat resistance testing of the JS6U alloy with the Ni20Cr12AlY coating, which is
about 120 lm thick at 1100 C, the composition
and structure of oxide layers gradually change.
This change is caused by oxidation and diffusion
of chromium and aluminum into the alloy. At
the first oxidation stage, -Al2O3Cr2O3 is the
base of the oxide layer (Fig. 6.12). After a lapse
of 50 h, Ni(Al,Cr)2O4 spinel appears in the oxide
layer. The amount of the spinel increases gradually in the course of testing. The life of
Ni20Cr12AlY and Ni20Co20Cr12AlY coatings
about 100 lm thick at 1100 C is 600 to 800 h.
Long testing is required to reveal the differences in heat resistance of the coatings. Testing
duration should exceed coating lifetime, sq,
shown in Fig. 6.1. At temperatures less than
1000 C, the test time required to assess coating
lifetime is several thousand hours. Thus, shorttime heat resistance testing does not always permit one to define the coating protective properties. Under certain testing conditions, the similar
structures of the oxides formed and the kinetics
of their growth do not allow revealing differences in the coating protective properties, even
in the case of rather long test time.
To determine and compare the protective
properties of the coatings on nickel-base superalloys, a combination of different criteria should
be used. These may be the data on inspection of
the structures of the oxides, which are formed
on the surface and phase composition variations
during testing. The presence of spinels, the reduction of the aluminum content to 12%, and
the disappearance of NiAl phase from the outer
zone indicate that the diffusion coating protective properties are exhausted. For the overlay
coatings, the absence of NiAl phase does not
mean that the protective properties are exhausted. The coating ceases to be protective
when the aluminum content reduces to 4 to 6%
and the chromium content falls to 14 to 16%.
Resistance of Alloys for Overlay

Coatings to Hot Corrosion
The alloy resistance to hot corrosion was studied on the samples tested in the synthetic ash of
GZT (120 g/m2) at temperatures of 600, 750,
Fig. 6.11
Heat resistance of JS6U alloy with Ni20Cr12AlY overlay coating (100 lm thick) under isothermal test conditions
850, and 900 C for 200 h using the previously

mentioned technique. The corrosion rate was determined by measuring the loss of the sample
specific weight (q, g/m2) and the uniform corrosion depth (hhc, lm).
Hot Corrosion of Ni-Cr-Al Alloys. The results of testing Ni-Cr-Al alloys are presented in
Table 6.15 and Fig. 6.13. The corrosion damage
rate depends on the test temperature and the
chemical and phase compositions of the alloys.
From the results of testing in GZT ash, all the
Ni-Cr-Al alloys can be conventionally divided
into three groups: high, moderate, and low re-
Fig. 6.12
Kinetics of oxide structural change in overlay coatings on JS6U alloy at 1100 C
Table 6.15 Hot corrosion of Ni-Cr-Al alloys. Test time in GZT ash, 200 h
Weight loss (q), g/m2
Alloy
Ni22Cr11AlY
Ni16Cr9AlY
Ni21Cr9AlY
Ni18Cr16AlY
Ni19Cr12AlY
Ni28Cr11AlY
Ni8Cr18AlY
Ni7Cr12AlY
Ni22Cr11Al2Hf
Ni22Cr11Al1Hf
Ni22Cr11Al1Si
Ni22Cr11AlSi
600 C
750 C
850 C
900 C
2.3
0.6
3.8
1.9
2.7
2.1
3.6
3.6
13
14
52
23
32
53
1100
31
31
24
26
1500
6400
2000
280
860
150
1500
1600
130
910
910
1200
260
4900
180
140
100
2800
180
1300
Uniform corrosion penetration at testing in GZT

ash for 200 h. 1, Ni18Cr5AlY; 2, Ni28Cr11AlY; 3,
Ni18Cr16AlY; 4, Ni19Cr12AlY; 5, Ni22Cr11AlY; 6, Ni8Cr18AlY;
7, Ni7Cr12AlY; 8, Ni16Cr9AlY
Fig. 6.13
sistance to corrosion. The specified groups have

the following depths of corrosion attack: less
than 50 lm, 50 to 200 lm, and more than 200
lm, respectively.
In the temperature range of 600 to 750 C, all
the studied alloys, except for Ni7Cr12AlY, have
high resistance to hot corrosion (Ref 11). The
depth of corrosion attack does not exceed approximately 7 lm. The maximum depth of corrosion attack against the alloys was observed at
850 C. Classification of the alloys with high,
moderate, and low resistance to corrosion at this
temperature is based on their chemical and phase
compositions. The analysis of the corrosion test
results has revealed that, in the system in question, chromium has the maximum effect on hot
corrosion. The effect of chromium depends on
its amount and its distribution between certain
phases. This, in turn, is closely related with the
aluminum content of the alloy.
The effect of aluminum on resistance to hot
corrosion is exhibited mainly through the effect
on the phase composition of the alloys. The aluminum content governs the presence of heterogeneous or homogeneous structures based on the
phases having various chromium solubility. At
the temperature of 850 C, chromium solubility
in the main phases of the Ni-Cr-Al-system alloys
is about 30 to 35% in c-solid solution, 10 to 15%
in Ni3Al, and 10 to 13% in NiAl.
A c-solid-solution-based Ni20Cr6AlY alloy
that is homogeneous at 850 C and with a chro-
Uniform corrosion penetration at testing in GZT

ash for 200 h. 1, Ni22Cr11Al2Hf; 2,
Ni22Cr11Al1Si; 3, Ni22Cr11AlSi; 4, Ni22Cr11Al1Hf
Fig. 6.14
mium-aluminum content ratio of 3.3 has the

maximum resistance to hot corrosion. The
alloys
(e.g.,
Ni7Cr12AlY,
Ni3Al-base
Ni16Cr9AlY) were observed to have the minimal corrosion resistance. The NiAl-base alloys
(e.g., Ni8Cr18AlY) with the same chromium
content as Ni3Al-base alloys demonstrate higher
resistance to hot corrosion. The Ni28Cr11AlY
alloy with 28% Cr and a chromium-aluminum
content ratio of 2.7 has high resistance to hot
corrosion. Apart from phase heterogeneity of
this alloy, the previously mentioned chromium
content provides the alloy with high resistance
to hot corrosion within the entire temperature
range of 600 to 900 C.
The effect of hafnium on hot corrosion resistance of NiCrAl alloys depends on its amount
(Table 6.15, Fig. 6.14). The addition of 1% Hf
does not change corrosion resistance in the temperature range of 600 to 850 C. At the same
time, an expansion of the zone that is attacked
by corrosion is observed at 900 C. This expansion is caused by the shorter life of the formed
oxide layers that are noticed at heat resistance
testing of this alloy. Alloying the Ni22Cr11Al
alloy with 2% Hf transfers it to the group of
alloys with high resistance to corrosion in the
entire temperature range of 600 to 900 C. The
depth of uniform corrosion attack for this alloy
at 850 C does not exceed the depth of corrosion
attack for the alloy containing 28% Cr.
The effect of 0.5 to 1% Si on resistance of the
Ni22Cr11Al alloy to hot corrosion is insignificant. At 850 C, the addition of silicon to the
alloy reduces the corrosion penetration depth approximately 1.5 times compared with the
Ni22Cr11AlY alloy.
Hot Corrosion of Ni-Co-Cr-Al Alloys. Testing of Ni-Co-Cr-Al alloys for resistance to hot
corrosion was carried out at 850 and 900 C, that
is, under maximum corrosion-temperature conditions. The statistics-based processing of the results of the test run in GZT ash at 850 and 900
C allowed for deriving the regression equations
relating alloy addition contents to the sample
specific weight change. (The equations are given
in the Appendix.)
Similar to the Ni-Cr-Al system, the effect of
alloying elements on the resistance to hot corrosion is combined from two factors: the effect
on the phase composition of the alloys and, thus,
on their homogeneity and distribution of the elements between phases; and the direct effect of
the alloying element on corrosion characteristics
(Table 6.16, Fig. 6.15) (Ref 12, 13).
Chromium increases resistance of the lowaluminum content NiCoCrAlY alloys to hot corrosion. In this case, it is uniformly distributed
mainly in c-solid solution. In the alloys of medium-aluminum content (10 to 12%), the increase in chromium content does not result in
better resistance to corrosion. It is due to phase
heterogeneity of the structure that contains a
Table 6.16 Hot corrosion testing of Ni-CoCr-Al-Y alloys. Test time in GZT ash, 200 h
Alloy
Ni10Co15Cr8Al
Ni30Co15Cr8Al
Ni10Co30Cr8Al
Ni30Co30Cr8Al
Ni10Co15Cr14Al
Ni30Co15Cr14Al
Ni10Co30Cr14Al
Ni30Co30Cr14Al
Ni30Co15Cr8AlY
Ni10Co30Cr8AlY
Ni30Co30Cr8AlY
Ni10Co15Cr14AlY
Ni30Co15Cr14AlY
Ni10Co30Cr14AlY
Ni30Co30Cr14AlY
Ni10Co22Cr11AlY
Ni30Co22Cr11AlY
Ni20Co15Cr11AlY
Ni20Co30Cr11AlY
Ni20Co22Cr8AlY
Ni20Co22Cr14AlY
Ni20Co22Cr11Al
Ni20Co22Cr11AlY
Ni20Co22Cr11AlY
Fig. 6.15
C for 200 h
850 C
900 C
460
210
110
130
63
87
230
280
150
130
140
71
56
170
81
430
72
66
170
9
140
110
120
120
46
1300
68
86
58
42
74
42
33
3800
40
85
120
110
120
100
Effect of alloying on specific mass loss (q, g/m2)

of Ni-Co-Cr-Al alloys. Tested in GZT ash at 900
considerable amount of Ni3Al and NiAl-like

phases featuring limited solubility of chromium.
With high-aluminum content (13 to 14%), the
increase in chromium content to more than 15
to 20% reduces alloy resistance to hot corrosion
because of the formation of a great number of
-Cr-like phases and r-phase.
The effect of aluminum on the resistance of
the NiCoCrAlY alloys to hot corrosion is similar
to that described for the Ni-Cr-Al system. It exhibits itself in the change of the phase composition, the alloy structure, and the distribution of
alloying elements between phases. When
phases, which differ greatly by their composition, are present in the alloy structure, phase heterogeneity reduces the resistance to hot corrosion. There are two types of compositions that
have maximum resistance to corrosion. They
feature high- and low-aluminum contents.
Cobalt increases resistance to hot corrosion.
Unlike chromium, cobalt is uniformly distributed between the phases of the alloys and, due
to its positive effect, exhibits itself in the zones
with both low- and high-aluminum contents.
Cobalt reduces the amount of Ni3Al phase, thus
promoting more uniform distribution of chromium within the structure of the alloys. The effect of cobalt is extremely evident in the alloys
with 18 to 28% Co content. Further increase in
cobalt content promotes the formation of rphase, which bonds chromium with cobalt, creates structural heterogeneity, and reduces alloy
resistance to hot corrosion.
Yttrium in the content range of 0.04 to 0.08%
had no effect on the alloys tested for resistance
to hot corrosion. The effect of hafnium and silicon on the resistance of the NiCoCrAlY alloys
to hot corrosion is much weaker than for the
NiCrAl system (Table 6.17). This is due to the
high resistance of this system alloy to corrosion.
At the same time, alloying the alloys with hafnium and silicon causes the depth of corrosion
attack to be reduced a little.
Hot Corrosion of Co-Ni-Cr-Al Alloys. The
test results for hot corrosion resistance of the
Co-Ni-Cr-Al-system alloys are given in Table
6.18. The derived regression equations relating
sample specific weight loss to alloying are given
in the Appendix.
The chromium and aluminum contents and
their ratio have a crucial effect on this system as
well as on the previous one. To analyze the effect of alloying on resistance to hot corrosion,
the alloys were divided into three groups with
different ratios of chromium and aluminum con-
Table 6.17
Hot corrosion testing of Ni-Co-Cr-Al alloys. Test time in GZT ash, 200 h
Alloy
Ni20Co22Cr11AlY
Ni20Co22Cr11AlHfSi
Ni20Co22Cr11Al2HfSi
Ni20Co22Cr11AlHf2Si
Ni6Co22Cr11AlY
600 C
700 C
750 C
800 C
850 C
900 C
4.3
2.1
2.0
0.8
1.0
22.0
16.0
3.5
2.5
12.0
7.3
110
120
130
96
30
130
98
63
79
110
83
86
796
tents, CCr /CAl. The first group was characterized

by 3.4 to 4.9, the second by 2.0 to 2.5, and the
third by 1.3 to 2. Such a classification is explained by different compositions of the oxides
formed on the surface. If the first group is the
case, Cr2O3 oxides are formed mainly on the surface of the alloys; if the third group is the case,
Al2O3 oxides are formed; and if the second
group is the case, mixed oxides are formed.
Alloying with nickel is unfavorable for hot
corrosion resistance of the alloys, independent
of their chromium and aluminum contents. The
alloys with a CCr /CAl ratio of 3.4 to 4.9 are less
impaired by nickel than the others. Higher-chromium content results in better resistance to hot
corrosion. The effect of chromium is the strongest if the nickel and aluminum contents are
high.
In all the alloys studied, aluminum content
increase leads to a higher hot corrosion rate. This
Table 6.18 Hot corrosion testing of Co-NiCr-Al-Y-system alloys. Test time in GZT ash,
200 h
Alloy
700 C
800 C
850 C
900 C
0.34
0.55
1.5
1.3
1.3
2.7
0.4
0.29
1.0
1.0
0.6
1.1
0.81
0.55
0.82
1.2
0.53
1.3
4.7
5.3
6.0
25
9.2
11
7.5
3.2
9.1
14
3.1
19
24
2.5
2.0
22
3.1
3.1
47(a)
5.7(a)
32
15
41
33
21
17
25
5.3
26
55
15
50
33
41
31
30
23
27
45
18
72
27
29
35
333
86
28
7.7
36
149
278
57
170
37
26
31
29
31
Co10Ni15Cr6AlY
Co30Ni15Cr6AlY
Co10Ni30Cr6AlY
Co30Ni30Cr6AlY
Co10Ni15Cr12AlY
Co30Ni15Cr12AlY
Co10Ni30Cr12AlY
Co30Ni30Cr12AlY
Co10Ni22Cr9AlY
Co30Ni22Cr9AlY
Co20Ni15Cr9AlY
Co20Ni30Cr9AlY
Co20Ni22Cr6AlY
Co20Ni22Cr12AlY
Co20Ni22Cr9AlY
Co20Ni22Cr9AlY
Co20Ni22Cr9AlY
Co22Cr9AlY
Co22Cr11AlY
Fe22Cr11AlY
(a) Test temperature 750 C
effect of aluminum is explained by the moreevident phase heterogeneity of the alloy structure, the worse solubility of chromium, and the
presence of r-phase. The effect of aluminum on
resistance to hot corrosion is similar to that observed in nickel alloys.
Yttrium exerts a positive effect on resistance
to hot corrosion, if the alloys have high and medium ratios of chromium and aluminum contents. For the other alloys, the effect of yttrium
is detrimental. This is due to the formation of
yttrium oxides, which are easily damaged by
sulfur-base compounds.
Studies of the structure of the sample surface
zones after testing show that corrosion uniformly penetrates the alloys.
Table 6.18 contains the results of iron-base
alloy hot corrosion resistance tests for comparison. This system features high resistance to corrosion. However, when it is used for coatings,
there is a problem of interaction between the
coating and the superalloy.
Resistance of Coated Superalloys to

Hot Corrosion
Hot corrosion resistance tests of turbine-blade
coatings were described in numerous papers.
However, it is rather difficult to compare their
results because of the differences between the
testing procedures, the temperature conditions,
as well as some uncertainty about the compositions and thickness of the coatings tested.
The hot corrosion resistance test methods
used differ from the alloy sample test methods.
The difference is that for the coated specimens,
the time to coating damage is determined instead
of sample weight loss. In both cases, test results
are comparable. By knowing the kinetics of corrosion in the alloy, one can calculate the time
required for damage of the alloy top coat of a
certain thickness, which is equal to the coating
thickness.
During testing of coated specimens, a layer of

GZT ash, weighed out as 120 g/m2, is applied
to the surface. The specimens with the ash application are held in a furnace at the temperatures of 750, 800, 850, and 900 C. After a lapse
of the specified time period, the specimens are
taken from the furnace, weighed, and inspected.
The corrosion products are left untouched on the
specimen surfaces. After inspection, a layer of
ash is applied once again, and testing is resumed.
Until coating damage takes place, the specimen gains weight. As it does so, a dense film of
corrosion products is formed on its surface. Before the coating is damaged, the specimen
weight-gain rate increases. Coating damage is
accompanied by an abrupt specimen weight loss
due to spallation of corrosion products. From
this moment, the specimen starts losing weight,
and a thick, loose layer of corrosion products is
formed on its surface. Coating life is evaluated
by the time, up to the last holding, when the
Fig. 6.16
abrupt specimen weight loss occurs. If prior to

this, only weight gain had been observed, then
the moment when specimen weight loss occurs
may be considered as the moment of damage.
The choice of time to inspect the specimens is
dictated by the test temperature. When the temperature is raised, holding time to inspection is
shortened.
By way of example, Fig. 6.16 presents kinetic
curves of specific weight change at 800 and 900
C for the specimens with Co22Cr11AlY/ZrO215%Y2O3 coating deposited by the EB method.
The presented results are obtained on specimens 7 mm in diameter and 30 mm high. When
depositing overlay coatings on the specimens, a
problem of specimen fastening in the unit for
coating deposition arises. The areas where the
specimens are attached to a fixture are free from
the coating. This is why Co22Cr11AlY coating
was deposited on the specimens under study in
two steps. First, the coating was applied to one
Kinetics of hot corrosion of Co22Cr11AlY/ZrO2-15%Y203 coating tested in GZT ash. (a) 800 C, (b) 900 C
side of the specimen, then the specimen was fastened from the other side, and the coating was
applied again. In the middle, the specimen was
coated twice. This resulted in forming a zone
with a coating of double thickness. As-deposited
coating was peened with microballs, and specimens were annealed in vacuum at 1050 C for
2 h.
The ceramic coating does not allow overlapping and so should be deposited only once. As
a result, a small coating-free zone appeared on
the specimen. The presence of such a zone reduces the protective effect of the ceramic layer.
During testing, coating life depends on many
technological factors. These factors dictate the
similar composition of the evaporated alloy and
Table 6.19
the coating formed as well as the coating structure and the presence of various defects in it.
When continuously deposited, the coatings demonstrate chemical composition homogeneity
through the whole thickness and no directional
structure, porosity, and microdots. The resistance of these coatings to hot corrosion is no less
than that of the studied alloys. The presence of
different defects reduces coating lifetime. The
characteristics of the coatings studied are given
in Table 6.14.
The data on coating lifetime at hot corrosion
testing are given in Table 6.19 and in Fig. 6.17
and 6.18. Test results are affected by the coating
quality. This is why the table presents the data
for the best and the worst specimens.
Coating life. Hot corrosion tests in GZT ash

Coating life, h
Coating
750 C
800 C
850 C
900 C
Co22Cr11AlY
Ni20Co20Cr11AlY
Ni20Co20Cr11AlY/ZrO2-15%Y2O3
Ni23Co28Cr9AlY
Ni28Cr6AlY
Ni20Cr12AlY with chromium carbide sublayer
AlSiCrY
AlSiBY
AlPd
Ni8Co12Cr7Al
Ni7Co12Cr17Al (Ni8Co12Cr7Al/Al5Si1.5Y)
9000
9000
8500
9000
4000
4500
4500
4500
4500
3000
4500
3500
4500
1000
600
7000
4500
4500
1500
3200
3000
4000
1500
3000
700
2500
400
400
2500
3000
3000
2600
2000
800
900
500
1000
200
5500
200
150
100
1500
1500
Life of diffusion coatings on JS6U alloy on testing

in GZT ash (mAl 80 g/m2). 1, aluminized coating;
2, Al5Si1.5%Y electric arc coating; 3, Al14Si0.7Y electric arc
coating; 4, AlPd coating
Fig. 6.17
Coating life; tested in GZT ash. 1, Co22Cr11AlY;

3,
2,
Co22Cr11AlY/ZrO2-15%Y2O3;
Ni20Co20Cr11AlY; 4, Ni20Co20Cr11AlY/ZrO2-15Y2O3; 5,
Ni23Co28Cr9AlY; 6, Ni28Cr6AlY; 7, Ni20Cr12AlY with chromium carbide sublayer; 8, Fe12Ni20Cr12Al1.5SiY
Fig. 6.18
The common feature for all the coatings is an

accelerated hot corrosion at raising the test temperature from 750 to 900 C. At 900 C, only
the lifetime of iron-base coatings is higher than
that at lower temperatures.
Figure 6.17 presents the data on the corrosion
life of the coatings alloyed with silicon, yttrium,
and palladium and on the aluminized coatings
formed in the powder mixture of 98% of the
aluminum-iron alloy (50% Al) and 2% NH4Cl.
The results of testing were based on the assumption that the coatings had the same aluminum reserve of 80 g/m2.
Alloying diffusion coating with silicon makes
its life 3 to 4 times longer than that of the siliconfree coating. All modifications of the coatings
deposited by the EA method have longer life, in
comparison with conventional aluminized coatings.
The life of a diffusion coating, even if it is
alloyed with chromium, silicon, and boron, is
shorter than that of an overlay coating. The diffusion coating alloyed with palladium is distinguished by its protective properties. At 800 and
850 C, its properties are highly competitive
with those of cobalt-base overlay coatings,
which are 2 times as thick. At 900 C, protective
properties of the aluminum-palladium coating
Table 6.20 Life coefficient of coatings. Hot

corrosion tests in GZT ash
Life coefficient
Coating
Co22Cr11AlY
Ni20Co20Cr11AlY
Ni23Co28Cr9AlY
Ni28Cr6AlY
Ni20Cr12AlY/chromium carbide
AlPd
800 C
850 C
11.1
11.1
11.1
11.1
6.7
8.9
7.8
13.3
8.9
22.7
22.7
20.0
20.0
11.8
18.2
7.3
31.8
15.4
Table 6.21 Life of coatings on turbine

blades of JS6U superalloys operating at
800 C
Life, h
Coating
Co22Cr11AlY
Ni20Co20Cr11AlY
AlPd
Liquid fuel
Gas
20,000
20,000
20,000
20,000
25,000
70,000
70,000
70,000
70,000
80,000
55,000
worsen suddenly (Ref 14, 15). The effect of palladium is described in detail elsewhere (Ref 16).
At 750 C, all overlay coatings feature high
protective properties. A 900 h testing did not
reveal any distinctions between their properties.
At 800 C, most overlay coatings have lives of
more than 4500 h. The exception is the
Ni23Co28Cr9AlY coating. At this temperature,
none of the systems has an obvious advantage.
At 850 C, the differences in the coating lives
are more substantial. At this test temperature, the
cobalt-base coatings do not have advantages in
life. The Ni28Cr6AlY coating has longer life
than the widely used Co22Cr11AlY coating.
The ceramic coat applied to the metallic coating
protects it against damage and increases its life
1.5 times. This effect is achieved despite the previously mentioned defects, such as the presence
of ceramic-free zones, thin ceramic layer, and
imperfection of the ceramic deposition technique used. The effect caused by the ceramic
coating can be more substantial (Ref 1719).
The blade ceramic layer not only protects the
surface from aggressive compounds but also
changes the surface temperature, thus promoting
the reduction in condensation of aggressive
compounds from gas flow.
The cobalt-base coating advantages fully
manifest themselves at 900 C. Lives of these
coatings are 2 to 3 times longer than those of the
nickel- and nickel-cobalt-base coatings. Ironbase coatings have high resistance to hot corrosion. The Ni7Co12Cr17Al coatings alloyed
with tantalum, hafnium, and rhenium showed
good protective properties at this temperature.
Table 6.20 contains the calculated relative life
values of the coatings tested. The coefficient of
the coating lifetime is calculated as a ratio of
coating lifetime and lifetime of the uncoated alloy. The time-to-damage of the alloy to the
depth of 250 lm during its testing for hot corrosion was taken as an alloy lifetime characteristic. This corresponds to the specific weight loss
of q 0.2 g/cm2. This corrosion depth is considered to be allowable for stationary turbine
blades (Ref 2). Some modifications of the coating allow for prolonging sample lives 20 to 30
times at 800 and 850 C.
Table 6.21 contains the calculated data on the
life of JS6U alloy blades used in stationary turbines at 800 C. The calculations were made on
the basis of the studies of uncoated blade lives
and relative lifetime coefficients given in Table
6.20.
REFERENCES
1. High-Temperature Corrosion and Protection of Materials, Proc. International Symposiums on High-Temperature Corrosion
and Protection of Materials, R. Streiff, J.
Stringer, R.C. Krutenat, and M. Caillet, Ed.,
(Les Embiez, France), 1986, 1989, 1992,
1996, 2000
2. V.I. Nikitin, Corrosion and Protection of
Gas Turbine Blades, Leningrad, Machinostroenie, 1987, 272 p
3. E.L. Simoms, G.V. Browning, and H.A.
Liebhaisky, Sodium Sulfate in Gas Turbines, Corrosion, (No. 2), 1955, p 505514
4. J.A. Goebel, F.S. Petit, and G.W. Goward,
Mechanisms for the Hot Corrosion of
Nickel-Base Alloys, Metall. Trans., Vol 4
(No. 1), 1973, p 261278
5. R.F. Reising, Effects of Chromium, Molybdenum and Tungsten on Sodium Sulfate Induced High Temperature Corrosion of
Nickel, Corrosion, (No. 5), 1975, p 159
163
6. J. Stringer, Hot Corrosion of High Temperature Alloys, Annu. Rev. Mater. Sci.,
(No. 7), 1977, p 477509
7. M. Kawakami, K. Goto, and R.A. Rapp,
Accelerated Oxidation (Hot Corrosion) of
Alloys by Molten Salt, Iron Steel Inst. Jpn.,
(No. 9), 1980, p 646658
8. G.W. Goward, Low-Temperature Hot Corrosion in Gas Turbines, Trans. ASME, (No.
2), 1986, p 421424
9. V.I. Nikitin and I.P. Komossarova, Method
of Protective Coating Testing for Lifetime,
Proc. CKTI, (No. 176), 1980, p 6773
10. V.I. Nikitin, Calculation of Metal Heat Resistance, Moscow, Metallurgy, 1976, 208 p
11. V.I. Nikitin, Y.A. Tamarin, and N.V. Zabrodina, Corrosive and Mechanical Properties of Protective Coatings, Collected Volume: Aqueous-Chemical Conditions and
Corrosion of Power Equipment, No. 201,
Leningrad, CKTI, 1986
12. N.V. Zabrodina, Y.A. Tamarin, and V.I. Nikitin, High-Temperature Corrosion of Aluminide Alloys for Overlay Coatings, Col-
13.
14.
15.
16.
17.
18.
19.
lected Volume: Heat-Resistant Coatings,

St. Petersburg, Science, 1992, p 107110
Y.A. Tamarin, E.B. Kachanov, and N.V.
Zabrodina, Corrosion a Haute Temperature
des Alliages Base Aluminure Destines aux
Revetements Protecteurs, High Temperature Corrosion Advanced Materials and
Coatings, R. Streiff, J. Stringer, R.C. Krutenat, and M. Caillet, Ed., Elsevier, London,
1989, p 361365
A.I. Desipry, N.V. Zabrodina, and S.B.
Shutova, Heat-Resistant Aluminide Coatings Containing Platinum and Palladium,
Surface Layer, Accuracy and Performances
of Machine Parts, Moscow, MDNTP, 1989,
p 5557
Y.A. Tamarin, E.V. Lopatin, and N.V.
Klyueva, Research into Structural Characteristics of Diffusion Aluminide Coatings
Alloyed with Palladium, Surface Layer, Accuracy and Performances of Machine Parts,
Moscow, MAI, 1995
S. Alperine, P. Steinmetz, P. Josso, and A.
Constantini, High Temperature-Resistant
Palladium-Modified Aluminide Coatings
for Nickel-Base Superalloys, High Temperature Corrosion Advanced Materials
and Coatings, R. Streiff, J. Stringer, R.C.
Krutenat, and M. Caillet, Ed., Elsevier,
London, 1989, p 367372
B.A. Movchan, I.S. Malachenko, and K.Y.
Yakovchuk, Two- and Three-Layer Coatings Produced by Deposition in Vacuum for
Gas Turbine Blades Protection, Surf. Coat.
Technol., Vol 67, 1994, p 5563
B.A. Movchan, Y.D. Skliarov, A.I. Rybnikov, and I.S. Malachenko, Corrosion Lifetime of Two-Layer Overlay Co-Cr-Al-Y/
ZrO2 Coating in Gas Turbine Fuel Ash,
Adv. Spec. Electrometall., (No 3), 1988, p
5256
A.S. Osyka, A.I. Rybnikov, S.A. Leontiev,
N.V. Nikitin, and I.S. Malachenko, Experience with Metal/Ceramic Coating in Stationary Gas Turbine, Surface and Coatings
Technology, Elsevier, 1995, p 187192

Y. Tamarin, p119-160
DOI: 10.1361/pctb2002p119

CHAPTER 7
The Effect of Protective Coatings on

the Mechanical Properties of
Superalloys
COATINGS PROTECT the surface of turbine blades from damage caused by high-temperature corrosion and thus preserve the structural shape of blades and their mechanical
properties for the required time.
There are many papers in the technical literature that deal with experimental studies of heat
resistance, fatigue strength, and thermal fatigue
of coated superalloys. However, the connection
between the physical and mechanical properties
of protective coatings and their effect on the mechanical properties of turbine-blade superalloys
has not been studied sufficiently well. For thin
diffusion coatings, this can be explained by the
complexity of the task, both in reproducing the
correct composition and in testing the coatings
separately from the alloy. However, for these
coatings, information on their physical and mechanical properties is necessary for their proper
application and for accurate calculations of
coated-blades service lives.
The effect of coatings on mechanical properties of superalloys is a combination of several
factors:
The efficiency of revealing the effect of coatings on superalloy properties depends on the
testing methods employed. Cooled blades are essentially thin-walled constructions, and the effect of the coatings increases as the mismatch
between the coating thickness and the bladewall thickness rises. Thus, to correctly determine
the properties of blades protected with coatings,
the specimens preferable for mechanical testing
should have the ratio of the coating thickness
and the specimen cross section close to those of
the blades in use.
Mechanical and physical properties of NiAl
and Ni3Al aluminides, which form diffusion
coatings, are dealt with in numerous publications (Ref 1, 2). Such aluminide properties as
ultimate strength, yield strength, and elongation
values depend to a considerable extent on their
structure set by the process of test-specimen fabrication and their heat treatment.
Table 7.1 refers to the test results presented
in one of the earliest publications that dealt with
Table 7.1 Properties of Ni3Al and NiAl

intermetalide-base cast alloys at tensile tests
at 20 C
Coating deposition alters superalloy surface

stresses.
The coating can be damaged under action of
static or alternate stresses; the cracks nucleating in the coating may accelerate the superalloy damage process.
The mechanical properties of superalloys
may be affected by the coating deposition
thermal cycle.
The coating changes the surface layer microstructure of a superalloy, and this effect is
boosted as high-temperature exposure time
increases.
Aluminum
content,
wt%
13.3
14.0
25.0
28.0
30.0
31.5
Source: Ref 3
Ultimate
strength (rB),
MPa
Yield
strength (r0,2),
MPa
Elongation
(d), %
205
280
163
162
103
102
86
110
1.1
1.0
0
0
0
0
research into aluminide features (Ref 3). Mechanical properties of NiAl compound versus its
test temperatures are shown in Fig. 7.1 (Ref 3,
4). Despite different elongation values for the
NiAl samples reported by different authors, they
agreed that, in fact, elongation-temperature relation yielded quite the same patterns. At a certain temperature dependent on a NiAl compound
composition, it experiences ductile-brittle transition that is typical of all the materials with
body-centered cubic lattice.
Research into the high-temperature strength
of NiAl and Ni3Al compounds has revealed that
their high-temperature strength is not so high as
that of superalloys. It can be improved by alloying and by forming directional or single-crystal
structures (Ref 5). The previously mentioned
patterns of aluminide mechanical properties variations can be taken into consideration as a certain approximation in the research into the behavior of diffusion aluminide coatings on
superalloys.
Reproducing aluminide compositions identical to numerous diffusion coating modifications
is not a simple task. It is as difficult as their
testing, because a sample thickness is comparable with a coating thickness. That is why the
ductile-brittle transition temperature (DBTT) of
a diffusion coating is recommended as its service life criterion under static and cyclic loading
(Ref 6). The fact that aluminides experience
ductile-brittle transition is crucial for many diffusion coatings characteristics and their effect on
superalloy properties.
As an example, Fig. 7.2 demonstrates test results for diffusion coatings 50 lm thick (Ref 7,
8). The increase in aluminum content of the
coatings results in the increase in their ductilebrittle transition temperature. For coatings including 16 to 17% Al, the DBTT is below 0 C,
while for coatings including 36% Al, the DBTT
is 800 C. Ductile-brittle transition temperatures of the diffusion coatings on superalloys are
higher than those of the coatings on nickel.
Thermal Expansion Coefficients and

Elasticity Modulus of Coating Alloys
Mechanical properties of NiAl compound of stoichiometric composition at different temperatures. 1,
Ultimate strength; 2, yield strength; 3 and 4, elongation. Source:
Ref 3, 4
Fig. 7.1
Ductile-brittle transition temperatures for diffusion

coatings 50 lm thick with different aluminum
contents of their outer layer. 1, Coatings on JS-type superalloys;
2, coatings on nickel. Source: Ref 7, 8
Fig. 7.2
The mismatches between the thermal expansion coefficients (TEC) and the elasticity modulus of coatings and superalloys determine the
level of stresses that arise on the blade surfaces
after coating deposition, heat treatment, and under their service conditions. Thermal stresses on
the surface have an effect on all the mechanical
properties of coated superalloys.
For the Ni-Cr-Al and Ni-Co-Cr-Al systems,
TEC studies were carried out on samples made
by extruding cast billets. The Co-Ni-Cr-Al system samples are studied as-cast. To study TEC,
the 0.5 to 1.0 mm thick condensates deposited
by the electron beam (EB) method were also
used. Before dilatometric analysis, all the samples were subjected to vacuum annealing at 1100
C for 2 h followed by slow cooling down. The
structure and phase compositions of the extruded samples are close to those of overlay
coatings with the same chemical composition.
Although the cast alloys feature larger structural
constituent sizes, their phase composition is also
The Effect of Protective Coatings on the Mechanical Properties of Superalloys / 121
similar to that of the coatings of the same chemical composition.

Studies of TECs were carried out using a
quartz dilatometer of 1500 N (made by SincuRico) equipped with an induction transducer as
a sensor. Sample elongation and temperature recording were carried out with the interval of 2
C. In combination with high sensitivity of the
induction transducer, it allowed detection of
phase changes and transformations that cause
relative variations of the sample length as small
as 0.001%. The total error at measuring mean
TEC was (0.25 to 0.35) 106, deg1, assuming confidence coefficient (P) level of P 0.95.
However, in the case of measuring elongation of
flat samples made of condensates, the error was
as high as 5 to 10%.
Dynamic modulus of elasticity was determined using the Elasomat instrument and cylindrical cast samples that were 120 mm long and
8 mm in diameter. Flat samples made of condensate with the dimensions of 100 by 10 by 2
mm were also tested. Relative errors of measuring elasticity modulus for cast samples and flat
samples were 1% and 10%, respectively.
Ni-Cr-Al Alloys. The TECs for the Ni-Cr-Al
system are given in Table 7.2. (Regression equations derived from statistics-based processing of
the experimental data on TECs are given in the
Appendix.)
Thermal expansion coefficient values for the
alloy group under study are determined in each
alloy by the volume ratio of its basic phase components, such as c-solid solution, c-phase
Table 7.2
(Ni3Al), b-phase (NiAl), and -Cr. Each of these

phases has TEC monotonically rising as the temperature increases, whereupon its absolute value
is less for the phases with greater aluminum contents. By the mean TEC values in the temperature range of 20 to 1000 C, the phases of the
system under consideration are ranged as follows: c (19.8 106, K1), Ni3Al (16.7
106, K1), NiAl (16.3 106, K1), and Cr (9.6 106, K1).
The increase in the aluminum content of the
alloy and, hence, a greater number of aluminumrich phases (NiAl) cause reduction of the TEC
mean values in the range of 20 to 1000 C. More
chromium in the alloy produces an effect similar
to that of aluminum, because it leads to an -Cr
volume increase. The said general features of the
effect of aluminum and chromium on the TECs
of multiphase alloys of the Ni-Cr-Al system are
effective up to about 600 C; in a higher temperature range, phase transformations lead to
sudden TEC variations. On consideration of the
TEC effect on the coating-alloy system properties, it is not the absolute value of this characteristic that is really important but its relative
value while comparing it with superalloy TECs.
The Ni-Cr-Al-system alloys and superalloys are
much the same in their thermal coefficients.
When overlay coatings interact with superalloys while being high-temperature tested, cobalt
diffuses from the superalloys into the Ni-Cr-Alsystem coatings, and after about 100 h at temperatures higher than 1000 C, its content
reaches the average cobalt content of a super-
Thermal expansion coefficients, , of Ni-Cr-Al alloys

106, K1
Alloy(a)
Ni14Cr3AlY
Ni18Cr5AlY
Ni20Cr6AlY
Ni16Cr9AlY
Ni21Cr9AlY
Ni22Cr11AlY(b)
Ni7Cr12AlY
Ni19Cr12AlY
Ni8Cr15AlY
Ni18Cr16AlY
Ni8Cr18AlY
Ni5Cr25Al
Ni31Al
Ni19Cr10AlY(c)
CMSX-4(d)
20100
C
100200
C
200300
C
300400
C
400500
C
500600
C
600700
C
700800
C
800900
C
9001000
C
13.7
13.3
12.4
12.0
11.7
13.1
12.0
11.8
12.2
11.6
13.8
13.4
11.8
12.1
14.4
13.9
12.7
12.9
12.7
12.6
13.7
12.7
12.6
12.8
12.6
14.4
13.9
12.7
12.7
15.1
15.0
13.4
13.6
13.4
13.2
14.4
13.4
13.5
13.3
13.4
15.4
14.3
13.4
13.3
15.9
15.3
14.4
14.4
14.4
14.6
15.4
14.0
14.0
13.8
14.4
16.5
13.0
14.4
13.9
16.2
16.0
14.7
15.1
14.7
14.8
15.7
14.8
14.3
14.6
13.7
16.3
13.3
14.7
14.4
17.9
17.2
15.4
16.3
15.2
15.5
16.9
14.3
14.6
13.2
18.4
9.8
13.9
15.4
15.0
19.9
19.1
17.1
17.7
16.3
17.2
18.6
16.1
17.6
15.1
16.1
19.6
14.4
17.1
16.4
20.3
21.2
15.1
15.8
14.9
16.9
18.6
15.0
15.3
11.3
12.0
19.6
14.5
15.1
17.4
19.4
23.0
19.0
13.2
10.6
18.7
18.6
16.2
15.6
19.3
19.7
18.6
14.6
9.0
19.1
20.4
28.6
25.2
26.5
21.6
31.0
19.7
31.1
17.0
39.1
19.5
14.8
25.2
21.5
(a) Alloy chemical compositions are presented in Table 4.10. (b) In the range of 1000 to 1100 C, 28.0 106, K1; in the range of 1100 to 1200 C, 22.6
106, K1. (c) Condensate: nickel base, 18.5% Cr, 10.4% Al, 0.01% Y. (d) In the range of 1000 to 1100 C, 25.6 106, K1; in the range of 1100 to 1200
C, 31.4 106, K1
alloy. Therefore, when analyzing coating TEC,

especially if the coatings are intended for longterm service, it is worth using the TEC values
of the Ni-Co-Cr-Al system.
The results of elasticity modulus measurements for the Ni-Cr-Al alloys are given in Table
7.3. Information on statistics-based processing
of experimental data is presented in the Appendix.
The value of the alloy elasticity modulus is a
result of the combined elasticity modulus of the
phases that form it. The NiAl (Ni8Cr18AlY)base alloy features the lowest value of elasticity
modulus, wherein, with a temperature raising to
800 C, the modulus value does not actually
change. The maximum elasticity modulus is
demonstrated by the alloys whose phase composition corresponds to c-solid solution
(Ni15Cr3AlY) and c-phase (Ni7Cr12AlY) with
the low chromium content. The value of elasticity modulus for all the alloys (except for those
with the NiAl structure) monotonically decreases as the temperature rises. The absolute
value of elasticity modulus for the said group of
alloys is lower than that for nickel-base superalloys with an equiaxial structure.
Ni-Co-Cr-Al Alloys. To determine TECs of
the Ni-Co-Cr-Al alloy major phases, these
phases were extracted electrochemically and
their TECs were determined by high-temperature x-ray diffraction methods. As for the Ni-CrAl system, their mean TEC in the temperature
range of 20 to 1000 C decreases as the aluminum content of the phases increases: c (17.2
106, K1), Ni3Al (15.2 106, K1), and
NiAl (13.8 106, K1).
The experimental TEC data for the Ni-Co-CrAl-system alloys are given in Table 7.4 (Ref 9).
The statistics-based processing of the experi-
Table 7.3
mental TEC data for the Ni-Co-Cr-Al alloys was

conducted in different temperature ranges. In the
temperature range of 100 to 400 C, the influence of cobalt, chromium, and aluminum on
thermal effects of TEC variations was considered. In the range of 900 to 1000 C, the effects
of TEC variations during phase transformations
were discussed; in the range of 100 to 1000 C,
the effect of alloying on the mean TEC values
was determined. The regression equations are
given in the Appendix.
A cobalt-content increase from 10 to 30%
causes an increase of the maximum TEC level
for the Ni-Co-Cr-Al alloys in the temperature
range of 100 to 400 C (Fig. 7.3a). Alloying the
alloys with 8 to 14% Co reduces TEC. In this
case, with the chromium content raised from 15
to 30%, TEC values do not actually vary.
Aluminum, chromium, and cobalt exert a crucial effect on alloy TEC in the temperature range
of 900 to 1000 C (Fig. 7.3b). A cobalt content
increase from 10 to 30% results in the reduction
of the maximum TEC level. The increase in aluminum and chromium contents of the alloys
causes TEC reduction too, which is mainly associated with an increase in NiAl and -Cr volume fractions. If the aluminum, chromium, and
cobalt contents of the alloys are at their maximum levels, a region with comparatively low
TEC values of 20 106, K1, appears.
The TEC behavior in the temperature range of
900 to 1000 C is well correlated with variations
of the NiAl-Ni3Al ratio, depending on alloying.
The more NiAl is in the alloy, the lower the level
of TEC.
The results of the research into Ni-Co-Cr-Al
alloy elasticity modulus are given in Table 7.5.
Due to statistics-based processing of test results,
a regression equation is derived that links the
Elasticity modulus, E, of Ni-Cr-Al alloys

E 102, MPa
Alloy
20 C
200 C
300 C
400 C
500 C
600 C
700 C
800 C
Ni14Cr3AlY
Ni18Cr5AlY
Ni16Cr9AlY
Ni21Cr9AlY
Ni22Cr11AlY
Ni22Cr11AlY(a)
Ni7Cr12AlY
Ni19Cr12AlY
Ni8Cr15AlY
Ni18Cr16AlY
Ni8Cr18AlY
1784
1743
1665
1571
1696
1670
1806
1596
1424
1624
1079
1714
1681
1599
1517
1672
1749
1615
1370
1604
1065
1657
1614
1565
1477
1644
1696
1602
1346
1599
1092
1601
1564
1510
1438
1612
1656
1575
1329
1579
1091
1536
1500
1469
1393
1580
1609
1545
1315
1552
1085
1469
1431
1429
1352
1530
1545
1515
1283
1523
1072
1431
1361
1379
1305
1481
1488
1486
1249
1495
1062
1354
1304
1348
1270
1441
1440
1446
1462
1469
1047
(a) Condensate from Ni22Cr11AlY alloy
elasticity modulus value with alloying element

contents of the Ni-Co-Cr-Al alloys in the temperature range of 20 to 800 C. Calculation data
available from the equations for alloy elasticity
modulus at the temperatures of 20 and 800 C
are given in Fig. 7.4(a, b). Aluminum reduces
the value of elasticity modulus mainly due to a
higher volume of NiAl in the alloys. The cobalt
effect is much stronger, especially if the alumiTable 7.4
num content is high. The influence of both elements is explained by a rise in atom interaction
energy when the nickel-base c-solid solution is
alloyed with these elements.
Elastic modulus values determined for the vapor-deposited samples are close to those determined for cast alloys (Table 7.5). These similarities are caused by measurement errors in thin,
flat samples and by differences in the composi-
Thermal expansion coefficients, , of Ni-Co-Cr-Al alloys

106, K1
Alloy(a)
Ni10Co15Cr8Al
Ni30Co15Cr8Al
Ni10Co30Cr8Al
Ni30Co30Cr8Al
Ni10Co15Cr14Al
Ni30Co15Cr14Al
Ni10Co30Cr14Al
Ni30Co30Cr14Al
Ni10Co22Cr11Al
Ni30Co22Cr11Al
Ni20Co15Cr11Al
Ni20Co30Cr11Al
Ni20Co22Cr8Al
Ni20Co22Cr14Al
Ni20Co22Cr11Al
Ni20Co22Cr11AlY(b)
Ni20Co22Cr11AlY(c)
Ni6Co20Cr12AlY
Ni6Co10Cr13AlTaReHf(d)
100200
C
200300
C
300400
C
400500
C
500600
C
600700
C
700800
C
800900
C
9001000
C
13.2
13.6
13.0
13.3
13.3
13.4
13.0
13.3
13.2
12.1
13.7
13.4
14.1
13.9
13.8
13.5
13.0
13.2
13.2
13.9
14.4
13.8
13.6
14.2
14.1
13.5
13.6
13.6
14.2
14.3
14.3
14.5
14.1
14.5
14.0
15.3
13.6
14.2
15.0
15.3
14.6
14.6
14.6
15.5
14.1
15.0
15.2
15.7
15.5
15.2
16.3
15.4
15.3
15.1
15.3
15.2
14.7
15.5
16.0
15.0
15.4
16.7
17.1
14.8
15.4
15.8
16.7
16.6
16.4
16.7
16.1
16.8
15.7
15.3
15.8
14.2
16.7
17.8
15.7
16.6
16.7
17.5
16.2
15.7
16.0
18.3
17.7
16.4
17.4
17.2
16.8
16.9
16.7
16.0
13.6
19.9
20.7
18.1
19.0
19.1
19.3
17.1
17.3
18.0
20.8
20.9
20.1
21.8
18.3
21.3
20.0
16.7
18.0
17.5
21.5
22.8
21.8
19.7
21.9
21.0
18.8
18.1
20.5
21.3
23.7
20.4
25.2
21.6
25.0
22.8
16.7
20.5
19.2
25.3
25.7
25.2
19.3
23.7
21.0
20.6
17.8
25.3
21.3
25.4
21.3
27.7
23.7
24.7
22.9
20.0
25.3
22.1
29.5
28.3
29.9
20.1
28.4
21.5
21.6
19.8
31.1
23.5
28.5
25.3
27.7
25.9
25.1
22.8
20.0
31.1
27.8
(a) Alloys chemical compositions are presented in Table 4.11. (b) In the range of 1000 to 1100 C, 25.7 106, K1; in the range of 1100 to 1200 C,
23.8 106, K1. (c) Condensate from Ni20Co22Cr11AlY alloy. (d) Chemical composition: nickel base, 6.0% Co, 9.9% Cr, 12.9% Al, 3.9% Ta, 0.55% Re, 1.4%
Hf, and 0.4% Si
Fig. 7.3
Thermal expansion coefficients of Ni-Co-Cr-Al alloys (K1). (a) 100 to 400 C. (b) 900 to 1000 C
tions of the condensate samples and the alloys

used for their deposition.
Co-Ni-Cr-Al Alloys. The results of the research into Co-Ni-Cr-Al-system alloy TECs are
given in Table 7.6. Up to 500 C, the temperature elevation results in a monotonical TEC rise.
At the temperatures above 500 C, an abrupt
TEC change is observed due to magnetic transformation and dissolution of b- and r-phases.
The statistics-based processing of the experimental measurement results allows for drawing
the following conclusions. In the temperature
range up to 500 C, the increase in aluminum
and chromium contents reduces TEC from 16
106 to 13 106, K1. Alloying with nickel
produces the same effect. If the effect of alloying
elements within the whole temperature range
from 20 to 1000 C is considered, chromium has
the strongest effect on TEC. It reduces TEC.
Nickel slightly reduces TEC; the effect of aluminum is not so strong.
When the temperature rises to 800 C, the
elasticity modulus of Co-Ni-Cr-Al alloys monotonically decreases to 1262 102 to 1492
102, MPa (Table 7.7). The regression equation
linking elasticity modulus with alloying element
content and temperature is given in the Appendix. The effect of the alloying elements is similar
at both low and high temperatures.
Mechanical Properties of Coating

Alloys
The coatings deposited on turbine blade surfaces are affected by the same static and alterTable 7.5
nate stresses that have an impact on the blade

materials. The level of coating mechanical properties has an effect on the basic characteristics
of blades, such as their high-temperature
strength, endurance, and thermal fatigue. Determining the mechanical properties is a hard task.
Even fabrication of cast samples of the alloys
with high aluminum content is hampered by
their low ductility. It is also worth taking into
account that at high-temperature testing and using the blade coatings, their compositions and
properties vary continuously.
Strength and ductility of the alloys used for
coatings were investigated using the standard
procedures for cast and vapor-deposited samples. The cast samples 5 mm in diameter and the
flat samples 0.5 to 1.0 mm thick made by vapor
deposition were annealed before testing at 1050
C for 4 h.
Test results for Ni-Cr-Al-system alloys are
presented in Tables 7.8 and 7.9 and in Fig. 7.5.
The Ni20Cr6AlY alloy with (cc)-phase composition features moderate-strength values. Ductility of this alloy at temperatures up to 1000 C
does not exceed 5%. The Ni8Cr15AlY alloy
based on Ni3Al has higher strength. At temperatures above 800 C, its strength exceeds the
strength of all the alloys studied. Ductility of the
alloys with (cc)-phase composition depends
on the aluminum content and other alloying elements. The Ni3Al-base systems are used for fabrication of superalloys and can be considered
most promising for designing the coatings
highly resistant to thermal fatigue.
The Ni22Cr11AlY alloy is most widely used
as a coating on blades operating under conditions of high temperatures and thermal stresses.
Elasticity modulus, E, of Ni-Co-Cr-Al alloys

E 102, MPa
Alloy
20 C
200 C
300 C
400 C
500 C
600 C
700 C
800 C
Ni10Co15Cr8Al
Ni30Co15Cr8Al
Ni10Co30Cr8Al
Ni30Co30Cr8Al
Ni10Co15Cr14Al
Ni30Co15Cr14Al
Ni10Co30Cr14Al
Ni10Co22Cr11Al
Ni30Co22Cr11Al
Ni20Co15Cr11Al
Ni20Co22Cr8Al
Ni20Co22Cr11Al
Ni20Co22Cr11AlY(a)
1560
1699
1664
1858
1307
1550
1364
1486
1737
1566
1602
1633
1549
1499
1637
1602
1799
1308
1523
1341
1461
1688
1530
1548
1595
1514
1461
1586
1573
1747
1301
1501
1324
1440
1656
1500
1506
1563
1476
1461
1533
1538
1705
1286
1476
1306
1404
1625
1472
1472
1533
1444
1394
1482
1491
1662
1274
1446
1278
1373
1583
1435
1431
1496
1410
1358
1439
1449
1661
1255
1414
1259
1337
1541
1400
1379
1464
1365
1300
1357
1404
1556
1233
1373
1238
1299
1496
1359
1318
1422
1311
1237
1267
1322
1473
1173
1316
1193
1249
1426
1302
1263
1373
1248
(a) Condensate from Ni20Co22Cr11AlY alloy
Fig. 7.4
Table 7.6
Elasticity modulus of Ni-Co-Cr-Al alloys (MPa). (a) 20 C. (b) 800 C
Thermal expansion coefficients, , of Co-Ni-Cr-Al alloys

106, K1
Alloy(a)
Co10Ni15Cr6AlY
Co30Ni15Cr6AlY
Co10Ni30Cr6AlY
Co30Ni30Cr6AlY
Co10Ni15Cr12AlY
Co30Ni15Cr12AlY
Co10Ni30Cr12AlY
Co10Ni22Cr9AlY
Co30Ni22Cr9AlY
Co20Ni15Cr9AlY
Co20Ni30Cr9AlY
Co20Ni22Cr6AlY
Co20Ni22Cr12AlY
Co20Ni22Cr9AlY
Co22Cr9AlY
20100
C
100200
C
200300
C
300400
C
400500
C
500600
C
600700
C
700800
C
800900
C
9001000
C
14.3
12.8
13.7
13.3
13.9
13.8
12.3
13.8
13.4
14.2
12.8
13.8
13.0
13.1
15.2
13.6
14.2
14.2
14.9
14.7
12.9
14.3
14.0
14.9
13.5
14.6
13.6
13.8
13.3
16.1
14.6
14.9
15.1
15.7
15.7
13.5
14.8
14.7
15.6
14.0
15.3
14.5
14.4
13.7
17.1
15.5
15.5
16.1
16.9
16.7
14.2
15.3
15.7
16.5
14.8
16.1
15.3
15.2
14.1
17.8
16.5
16.2
16.7
17.5
17.4
14.8
16.6
16.2
17.1
15.8
16.8
16.5
16.1
14.5
21.4
17.3
19.2
19.3
19.9
18.8
15.3
17.6
17.5
19.4
17.0
19.3
17.4
18.0
14.8
35.5
18.0
20.6
21.7
27.6
20.3
15.8
15.6
22.5
26.6
18.8
22.4
19.0
19.9
14.8
23.1
19.0
19.2
25.2
27.7
21.4
16.5
14.8
22.8
21.0
15.6
19.3
17.3
18.6
14.7
20.2
20.3
18.7
18.1
21.3
20.3
19.2
16.8
20.1
18.9
16.8
18.5
17.0
15.2
20.5
22.0
18.2
15.6
15.4
20.3
21.7
21.8
18.5
18.1
23.3
20.6
16.1
(a) Alloy chemical compositions are presented in Table 4.13
Table 7.7
Elasticity modulus, E, of Co-Ni-Cr-Al alloys

E 102, MPa
Alloy
20 C
100 C
200 C
300 C
400 C
500 C
600 C
700 C
800 C
Co10Ni15Cr6AlY
Co30Ni15Cr6AlY
Co10Ni30Cr6AlY
Co30Ni30Cr6AlY
Co10Ni15Cr12AlY
Co30Ni15Cr12AlY
Co10Ni22Cr9AlY
Co30Ni22Cr9AlY
Co20Ni15Cr9AlY
Co20Ni30Cr9AlY
Co20Ni22Cr6AlY
Co20Ni22Cr12AlY
Co20Ni22Cr9AlY
2071
1893
1960
1820
1956
1826
2013
1820
1792
2037
1888
1778
1847
1956
1856
1923
1787
1911
1787
1982
1795
1759
2006
1843
1760
1799
1874
1803
1861
1726
1839
1763
1937
1749
1706
1970
1781
1728
1749
1809
1740
1792
1683
1779
1714
1888
1704
1657
1919
1717
1702
1700
1738
1655
1702
1603
1703
1674
1823
1649
1586
1866
1648
1662
1646
1662
1601
1618
1549
1641
1621
1764
1575
1540
1805
1576
1643
1600
1573
1549
1535
1469
1568
1554
1690
1516
1666
1739
1476
1578
1534
1492
1386
1447
1366
1458
1463
1625
1434
1363
1679
1372
1515
1426
1262
1266
1353
1284
1334
1308
1492
1345
1295
1492
1289
1438
1344
at 1100 C. Such a change in the properties of

this alloy is due to the development of phase
transformations. Intensive softening of the
Ni20Cr12AlY alloy starts at temperatures above
700 C, when c-phase is dissolving in the alloy.
Tables 7.10 and 7.11 and Fig. 7.5 present the
test results on the mechanical properties of the NiCo-Cr-Al alloys. At temperatures below 600 C,
Its properties are typical of all the materials of

the Ni(Co)-Cr-Al system used for depositing
overlay coatings. At temperatures under 800 C,
the alloy features high strength, which abruptly
diminishes as the temperature rises to 1100 C.
At 500 C, elongation of the Ni22Cr11AlY alloy
does not exceed 3%. However, at higher temperatures, elongation increases abruptly to 69%
Table 7.8
Ultimate strength, rB, of Ni-Cr-Al alloys

rB, MPa
Alloy
Ni22Cr11AlY
Ni22Cr6AlY
Ni20Cr6AlY(a)
Ni8Cr15AlY(a)
Ni5Cr25Al
500 C
600 C
700 C
800 C
900 C
975 C
1000 C
1100 C
728
316
550
530
650
300
672
406
390
500
250
396
341
190
216
306
400
200
358
150
68
236
65
120
34
157
50
(a) Samples are vapor deposited from alloy of the said composition
Table 7.9
Elongation, d, of Ni-Cr-Al alloys

d, %
Alloy
Ni22Cr11AlY
Ni22Cr6AlY
Ni20Cr6AlY(a)
Ni8Cr15AlY(a)
Ni5Cr25Al
500 C
600 C
700 C
800 C
900 C
975 C
1000 C
1100 C
1200 C
1
16
0
0
24
1.2
16
0
10
31
2.4
74
2.6
49
10
20
12
25
105
3.2
85
30
31
70
(a) Samples are vapor deposited from alloy of said composition
Fig. 7.5
(a) Ultimate strength and (b) elongation of alloys. 1, Ni20Co20Cr12AlY; 2, Ni20Cr12AlY; 3, Ni10Co20Cr12AlY; 4,
Co20Cr12AlY; 5, Ni20Cr7AlY; 6, NiAl (33% Al)
(bc) phase composition, which causes an

abrupt drop of its ultimate strength and an increase of its ductility as the temperature rises.
Cobalt-base alloys also feature monotonically
decreasing strength and increasing ductility as
the temperature rises from 500 to 1100 C (Tables 7.12 and 7.13 and Fig. 7.5). The strength of
these alloys depends slightly on their alloying.
It is related to the fact that the said system lacks
c-strengthening phase. The Appendix contains
the regression equations linking strength and
ductility of the Co-Ni-Cr-Al-Y-system alloys
with their alloying element contents.
The comparison of strength and ductility of
the cast and vapor-deposited samples of the
the Ni20Co20Cr12AlY and Ni10Co20Cr12AlY

alloys have high ultimate strength and relatively
low ductility (d 5%). As the temperature rises,
the alloy strength diminishes monotonically, and
at 1100 C, it is 30 to 40 MPa. Ductility of the
alloys under study rises monotonically, and at
temperatures above 1000 C, some alloys (e.g.,
Ni10Co22Cr11AlY) are in a superductile state;
their elongation exceeds 100%. Different alloy
behaviors, which manifest themselves in temperature-dependent variations of their strength
and ductility parameters, are caused by their
phase composition and alloying with cobalt. In
the temperature range higher than 600 C, the
Ni20Co20Cr12AlY alloy alloyed with 20% Co
shows no trace of c-strengthening phase. It has
Table 7.10
Yield strength (r0,2) and ultimate strength (rB) of Ni-Co-Cr-Al alloys

r0,2, MPa
Alloy
Ni10Co15Cr8AlY
Ni30Co15Cr8AlY
Ni10Co30Cr8AlY
Ni10Co22Cr11AlY
Ni10Co20Cr12AlY(a)
Ni20Co22Cr8AlY
Ni20Co22Cr11AlY
Ni20Co20Cr12AlY(a)
rB, MPa
800 C
900 C
1000 C
500 C
700 C
800 C
900 C
1000 C
1100 C
343
344
292
335
246
194
230
113
178
103
73
72
69
668
840
520
602
514
510
698
572
434
429
260
470
319
320
415
268
234
233
190
249
148
219
112
88
62
55
82
84
50
70
41
56
48
45
(a) Samples are vapor deposited from alloy of the said composition.
Table 7.11
Elongation (d) of Ni-Co-Cr-Al alloys

d, %
Alloy
500 C
700 C
800 C
900 C
1000 C
1100 C
0.5
4
3
28
4
5
27
14
20
39
25
7
7
32
23
3
41
65
70
14
15
68
84
130
42
68
130
34
37
66
15
Ni10Co15Cr8AlY
Ni30Co15Cr8AlY
Ni10Co30Cr8AlY
Ni10Co22Cr11AlY
Ni10Co20Cr12AlY (a)
Ni20Co22Cr8AlY
Ni20Co22Cr11AlY
Ni20Co20Cr12AlY (a)
Condensate from alloy of said composition
Table 7.12
Yield strength (r0,2) and ultimate strength (rB) of Co-Ni-Cr-Al alloys

r0,2, MPa
Alloy
Co10Ni22Cr9AlY
Co30Ni22Cr9AlY
Co20Ni15Cr9AlY
Co20Ni30Cr9AlY
Co20Ni22Cr6AlY
Co20Ni22Cr12AlY
Co20Ni22Cr9AlY
rB, MPa
500 C
700 C
900 C
500 C
700 C
900 C
1100 C
496
392
564
494
493
392
428
315
479
334
202
376
161
126
122
165
158
178
567
690
866
787
810
546
547
461
567
496
474
531
205
165
142
173
190
249
188
60
57
42
54
53
64
60
Ni20Co20Cr12AlY alloy (Fig. 7.6a,b) demonstrates that the principal trend of properties
change versus temperature is the same. At the
same time, the mechanical properties of the vapor-deposited sample are slightly worse than
those of the cast sample (50 and 84 MPa at 1000
C, respectively), but its ductility is higher. The
vapor-deposited and cast samples of other alloys
have the same relationship between their properties. Differences in the properties are explained by structural differences, such as more
dispersed phases in the vapor-deposited samples, different chemical compositions of vapordeposited samples and evaporated alloys, and
sample shapes and sizes.
To determine the moment when the alloys become ductile, the alloys of the Ni-Cr-Al and NiCo-Cr-Al systems were bend tested. Testing was
carried out by applying certain steady loading to
the central part of the sample and then measuring its strain. The test samples were fabricated
by the extrusion method and had a dispersed
structure similar to that of vapor-deposited samples and coatings.
The Ni20Co20Cr12AlY alloy experiences deformation even at room temperature. At 500 C,
the sample is quite ductile when bent (Fig. 7.7).
The cobalt-free alloy of the same chromium and
aluminum contents demonstrates noticeable
strains only at 600 C. The Ni10Co15Cr8AlY
and Ni30Co15Cr8AlY alloys differing only in
cobalt contents have different ductility levels.
The 30% Co alloy becomes ductile even at room
temperature, while the 10% Co alloy becomes
ductile at temperatures above 500 C. The increase in aluminum content of the alloy up to
14% results in overriding the cobalt effect. The
high aluminum content and, respectively, high
NiAl content (of 70 to 80%) raise the alloy plasticization temperature to 800 C.
The previously mentioned test results show
that at 500 to 1000 C, the strength levels of the
cast alloys used for coatings, their vapor-depos-
Table 7.13
Elongation, d, of Co-Ni-Cr-Al alloys

d, %
Alloy
Co10Ni22Cr9AlY
Co30Ni22Cr9AlY
Co20Ni15Cr9AlY
Co20Ni30Cr9AlY
Co20Ni22Cr6AlY
Co20Ni22Cr12AlY
Co20Ni22Cr9AlY
500 C
700 C
900 C
1100 C
11
12
9
10
13
15
19
29
25
28
49
27
72
66
53
37
50
74
37
61
72
56
57
63
ited samples, and the respective coatings are

considerably lower than those of modern superalloys. That is why the protective coating cannot
bear any static and alternate stresses generated
in turbine blades. This should be taken into account when calculating stresses and strains in
coated turbine blades.
Thermal Stresses in Superalloy Coatings

A study of the coating stressed state is one of
the basic elements used for development of coatings and their use. Thermal stress generation
during heating and cooling of coated parts
caused by a mismatch between the TECs of a
coating and a superalloy is typical for both diffusion and overlay coatings.
Thermal stress in a coating (rc) may be calculated from a ratio of coating and superalloy
TECs:
rc Ec
(c sa) (T T0)
1 lc
where Ec is coating modulus of elasticity; c and

sa are TECs of a coating and a superalloy in
the temperature range of T0 to T, respectively;
T0 is a temperature of zero thermal stress level
in a coating-superalloy system; T is a current
temperature; and lc is Poissons ratio of a coating.
Thermal stresses generated in the coating-superalloy system are crucial to the destruction of
the coating on turbine blade surfaces during
heating and cooling under transient conditions
typical of aircraft gas-turbine engines.
Experimental Studies of Thermal Stresses.
To study stressed coatings, two methods were
used:
Technique for measurement of strain caused

by coating chemical removal from a sample
surface
X-ray diffraction technique for assessing
coating stresses by measuring crystal lattice
strain (Ref 10, 11)
As mentioned in many publications, compressive stresses occur in diffusion aluminide
coatings at room temperature. Their generation
is caused by two factors:
When nickel superalloys undergo aluminizing, NiAl3, Ni2Al3, NiAl, and Ni3Al form,
their specific volumes (0.25, 0.21, 0.17, and
0.14 cm3 /g, respectively) being larger than

the specific volume of the superalloy (0.11 to
0.12 cm3 /g).
Thermal expansion coefficients of aluminides
are less than those of superalloys.
The compressive stress level in aluminide
coatings depends on the aluminizing technique
in use. The results of the studies pursued show
that on aluminizing a nickel superalloy of JStype from the slurry containing 100% Al at the
temperature of 700 C for 4 h, the residual com-
Fig. 7.6
(a) Ultimate strength and (b) elongation of samples made from alloy condensates. 1, Ni20Co20Cr12AlY; 2,
Ni10Co20Cr12AlY; 3, Ni20Cr7AlY; 4, Ni5Cr8AlY; 5, Ni20Cr12AlY
Ductility of alloys during bend testing. 1,

Ni10Co15Cr8Al;
2,
Ni30Co15Cr8Al;
3,
Ni22Cr11AlY; 7, Ni20Co22Cr11AlY
Fig. 7.7
pressive stress of r 200 MPa generates in

the coating surface layer at room temperature.
In the coating applied by aluminizing at 950 and
1200 C for 4 h, the compressive stresses are
160 and 140 MPa, respectively.
The stress level in diffusion coatings decreases with the temperature increase. The data
on thermal stresses in the diffusion coating measured by the x-ray method at elevated temperatures are presented in Fig. 7.8. The coating was
applied by aluminizing the JS6U superalloy in a
Thermal stresses in (1) diffusion coating on JS6U superalloy versus temperature. Thermal expansion coefficients of (2) NiAl and (3) superalloy. Source: A.I. Samoilov
and I.A. Ignatova
Fig. 7.8
mixture of 98% Al-Fe alloy (50% Al) and 2%

NH4Cl at 950 C for 4 h. The extreme points on
the stress-temperature curve are caused by different TECs of the superalloy and the coatingbase aluminide. At high-temperature testing, the
stress level of the diffusion coating depends on
its test time and temperature. However, on cooling from the test temperature, the compressive
stresses always rise in diffusion coatings.
When analyzing a stressed state of an overlay
coating, it is necessary to take into account the
following initial conditions. When the EB
method is used, the coating is deposited on a
substrate heated to 850 to 950 C, and evaporation heat generated during its condensation
also heats its thin surface layers. Despite the fact
that no special preheating equipment is used at
electric arc (EA) deposition, the substrate is
heated by plasma flow up to 500 to 600 C. Under these conditions, no stresses are generated at
the interface of the coating and the substrate on
which the coating is deposited from vapor or
plasma.
At the heat treatment temperatures of 1000 to
1050 C, the coatings are in a ductile or, in some
cases, superductile state (d 100%); their yield
stress is lower than 50 MPa. In addition, at
1000 C, recrystallization occurs in the coat-
ings of the Ni-Cr-Al and Ni-Co-Cr-Al systems.

On cooling from 1000 C, a mismatch between the TECs of the coating and the protected
alloy causes thermal stress generation. As the
temperature decreases, thermal stresses increase
and reach their maximum at room temperature.
The results of x-ray diffraction analysis of
stresses at different stages of the coating deposition process and their heat treatment and machining are shown in Fig. 7.9. The coatings 80
to 100 lm thick made of Ni20Cr12AlY, and
Ni20Co20Cr12AlY alloys were deposited by
the EB method. All the measurements were carried out at room temperatures; the JS6U superalloy was used as a substrate material.
The tensile stress level of (70 to 90) MPa
is typical of the coatings as-deposited by the EB
method on the substrate heated to 850 to 950 C.
On annealing at 1030 C for 2 h, tensile stresses
in the coatings increase. The tensile stress level
in the coating alloyed with cobalt is higher. On
annealing at 1030 C of the Ni20Cr12AlY coating applied by the EA method, its tensile stress
level is 90 MPa. This indicates that the level
of tensile stresses in the coatings depends little
on a deposition technique and is mainly determined by a mismatch between the TECs of the
coating and the superalloy.
Thermal stresses at 20 C in overlay coatings on JS6U superalloy after different technological treatment processes. 1, after
deposition (D); 2, after D annealing; 1030 C, 2 h (A); 3, after (DA) peening (P); 4, after (DAP) annealing:
700 C, 2 h; 5, after (DAP) annealing: 800 C, 2 h; 6, after (DAP) annealing: 900 C, 2 h; 7, after (DAP) annealing:
1000 C, 2 h; 8, after (DAP) annealing: 1030 C, 2 h
Fig. 7.9
The next step in the technological process for

the coatings deposited by the EB method is a
compacting treatment by microball peening,
which generates high compressive stresses,
whereupon their absolute value exceeds that for
the Ni20Cr12AlY coating. This is due to a
higher yield stress of this coating at room temperature.
Annealing the specimens peened with microballs at a temperature above 700 C completely
eliminates coating strain hardening. On annealing of the specimen peened with microballs, all
stresses are relaxed and tensile stresses regenerate during cooling. The threshold recrystallization temperature for the Ni20Cr12AlY coating
is 1000 C.
The processes of stress relaxation in the coating alloyed with cobalt run in another manner,
compared with the Ni20Cr12AlY coating. This
is because coating strain hardening and its subsequent heat treatment initiate phase transformations and recrystallization. This results in a
steadier state, compared to as-deposited and asannealed for the first-time coating. The increase
in thermal stresses of the Ni20Co20Cr12AlY
coating is caused by a greater mismatch between
its TEC and that of the JS6U superalloy, compared with the Ni20Cr12AlY coating and the
JS6U superalloy.
The
stressed
Ni20Cr12AlY
and
Ni20Co20Cr12AlY coatings were x-rayed in the
temperature range from room temperature to
900 C. The specimens tested were subjected to
a full treatment cycle, including annealing at
1030 C for 2 h, microball peening, and final
annealing at 1030 C for 2 h. The results of the
investigation are shown in Fig. 7.10. Stresses
Fig. 7.10
Thermal stresses in (1) Ni20Cr12AlY coating and

(2) Ni20Co20Cr12AlY coating on JS6U superalloy
were gradually relaxing in both coatings. For the

Ni20Cr12AlY coating, the coating-alloy system
shows no stresses at 800 C. Further temperature
increases to 900 C result in the generation of
low compressive stresses (20 MPa). When
cooling the alloy from the said temperature to
room temperature, stresses recover to their previous level.
In the Ni20Co20Cr12AlY coating, the temperature increase results first in the abrupt drop
of tensile stresses from 240 MPa at 20 C to 0
MPa at 650 C, then in the generation of low
compressive stresses (50 MPa); as the temperature approaches 900 C, their absolute value
decreases to 20 MPa. The compressive stress
generation in the Ni20Cr12AlY and
Ni20Co20Cr12AlY coatings is explained once
again by a mismatch between the TECs of the
coating and the alloy.
The results of the research into thermal
stresses using the technique of strain measuring
during chemical removal of coatings are shown
in Fig. 7.11. The stresses were measured in samples cut out of blades made of the JS6U superalloy and subjected to annealing at 1030 C for
2 h, compacting by microball peening, and further annealing at 1030 C for 2 h. Tensile
stresses at a distance of 20 lm from the surface
of the Ni20Cr12AlY coating are 100 to 180
MPa. With due account of possible errors introduced by both methods, these results are in good
agreement with the results of x-ray diffraction.
As the distance from the surface increases, the
stresses gradually decrease, and at a distance
equal to the coating thickness, the stresses approximate to a zero level. The study of the
stressed coating on the blades subjected to testing in an engine for 100 h demonstrates that
both the sign and the level of the stresses do not
actually change.
The experimental methods allow assessment
of the stressed state of the coated specimen surfaces. However, these methods are rather timeconsuming and may introduce considerable errors when the changes in chemical and phase
compositions take place during high-temperature testing. The method based on etching off
coatings allows studying the stressed state only
at their room temperature.
The use of calculation methods allows the
analysis of the coating physical and mechanical
properties effect on their thermal stress levels,
as well as the assessment of design features of
any type and composition of multilayer coatings
deposited on substrates of any type. The calcu-
Fig. 7.11
Distribution of thermal stresses at 20 C in Ni20Cr12AlY coating on JS6U superalloy blades
Calculated thermal stresses at 20 C in (1)

Ni20Co20Cr12AlY coating and (2) Ni20Cr12AlY
coating on JS6U superalloy specimens. hc, distance from a coating alloy interface into the coating; hs, distance into the alloy
Fig. 7.12
Fig. 7.13
Forces (P) acting in Ni22Cr12AlY coating on JS6U

superalloy versus coating thickness
lation methods are a step toward controlling

stresses in coatings and the respective characteristics of their thermal fatigue.
Coating-Thermal-Stresses Modeling. To
calculate the thermal stresses occurring in the
coating-superalloy system, a strained body
stressed-state numerical modeling method was
used (Ref 12). The specimen taken for model
calculation was 2 mm thick and 4 mm long; a
coating 0.1 mm thick was applied to one of its
sides. The length of 4 mm allows elimination of
any edge effect.
On the basis of the previously mentioned xray diffraction results, within tolerable limits,
the T0 temperature was taken to be 920 C. At
the T0 temperature, the stresses at the coating/
superalloy interface approximate to a zero level.
Thermal expansion coefficients and modulus of
elasticity determined from the investigation of
the coating alloys were used for calculations.
The calculation results for the stresses occurring in the Ni20Cr12AlY coating/JS6U superalloy and the Ni20Co20Cr12AlY coating/JS6U
superalloy systems at their cooling from 920 to
20 C are shown in Fig. 7.12. The calculated
stresses in the surface layer of the
Ni20Co20Cr12AlY coating are 392 MPa,
while in the surface layer of the Ni20Cr12AlY
coating they are 128 MPa. The coated specimen surface stress profile shows that the tensile
stresses peak is in the coating zone near the coating/alloy interface. In the vicinity of the interface, the stress sign reverses abruptly, and compressive stresses are observed in the alloy.
Beneath the Ni20Co20Cr12AlY coating, compressive stresses reach 8 MPa; beneath the
Ni20Cr12AlY coating they are 2 MPa. The
compressive stresses gradually decrease as the
distance from the coating/alloy interface increases.

The diagram showing the value of forces in
the coating-alloy system versus coating thickness is presented in Fig. 7.13. The force, P, in
the coating was calculated from the following
formula:
n
Pn
ri Fi
i1
Level of thermal stress ratio, r/r0, in (1, 3) coating

and in (2) substrate versus (1, 2) coating thermal
expansion coefficient variation, D, and (3) superalloy elasticity
modulus, DE
Fig. 7.14
Calculated
thermal
stresses
in
(1)
Ni20Co20Cr12AlY and (2) Ni20Cr12AlY coatings
on JS6F superalloy specimens. Orientations of (a) [001] and (b)
[111]
Fig. 7.15
where ri is stress in the sections into which the

coating is divided; Fi is the area affected by
stress ri; and n is the number of sections into
which the coating is divided.
The separation of the stresses and forces acting in the coating has a physical meaning. The
stresses in the coating have an effect on the coating material and may cause its failure. The
forces in the coating are applied to the coating/
alloy interface and may cause spalling of the
coating along this interface. The force in the
coating increases monotonically as the coating
becomes thicker. Thus, coating thickness is limited by the strength of the coating/coated alloy
interface (i.e., by its adhesive strength). The unsatisfactory adhesive strength level may cause
spontaneous spallation of a thick coating.
Spalling from the substrate surface may occur
without any stress applied.
Superalloy TECs vary slightly, depending on
their alloying. Coating TECs can vary within a
rather wide range. Coating TECs also vary during high-temperature testing due to the change
of the coating chemical and phase compositions.
The calculated thermal stresses in the coating
and the substrate versus differences of TEC (D)
and elasticity modulus (DE) for the coating and
superalloy are shown in Fig. 7.14. The
Ni20Cr12AlY coating/JS6U superalloy system
was taken as the basis for calculation. Coating
and substrate stresses vary linearly with D. As
a TEC mismatch between the coating and alloy
increases by 20%, the level of thermal stresses
in the coating increases 3.7 times. The D has a
weaker effect on the stresses in the substrate.
The effect of DE on the stresses in the coating
is not strong.
The calculation results for thermal stresses in
single-crystal specimens made of JS6F superalloy with grain orientations of [001] and [111]
are given in Fig. 7.15. The Ni20Cr12AlY and
Ni20Co20Cr12AlY coatings are deposited on
the specimens. In the case of the Ni20Cr12AlY
coating, the stress level in the substrate with the
orientation of [001] is 29 MPa, while in the
case of the Ni20Co20Cr12AlY coating it is
275 MPa. The stress level in the coatings on
the specimens with the orientation of [111] is
lower.
The main factor, which decides the value of
thermal stresses, is a mismatch between the coating and alloy TECs. If this mismatch is not great,
as for the Ni20Cr12AlY coating, the variation
of elasticity modulus is crucial to the value and
sign of thermal stresses. If the mismatch is high,
the use of the material with a high elasticity

modulus has a minor effect on the thermal stress
level, which decreased by only 5% for the
Ni20Co20Cr12AlY coating.
The development of multilayer coating systems is very promising and can improve coating
stability, for example, in the case of applying the
Ni3Al layer or thermodynamically stable carbides (NbC, HfC) to the coating/alloy interface.
By using multilayer coatings, it is possible to
control the level of thermal stresses in them. The
experimental determination of stresses in multilayer systems is a complicated process and, as
a rule, it is reduced to measuring total stresses.
The results of thermal stress calculations for different modifications of two-layer coatings are
shown in Fig. 7.16. Depending on the position
in the coating pattern, the layers of the same
chemical composition may have different stress
signs.
Effect of Coatings on High-Temperature

Strength of Superalloys
High-temperature strength of the protective
coating materials is substantially lower than that
of modern superalloys. That is why the coating
itself cannot bear the loads applied to the sample
during its high-temperature strength testing carried out at temperatures above 700 C. The
stresses arising in the coating at the moment of
Calculated thermal stresses at 20 C in two-layer

coatings on JS6F superalloy specimens [001]. 1,
Ni20Cr12AlY (95 lm)/Ni8Co20Cr12AlY (5 lm); 2,
Ni8Co20Cr12AlY (50 lm)/Ni20Cr12AlY (50 lm). hc, distance
from a coating alloy interface into the coating; hs , distance into
the alloy
Fig. 7.16
the sample loading, as a rule, exceed the coating

yield strength, so they are relieved and redistributed over the cross section of the superalloy. The
stresses in the tested sample should be calculated
taking into account the sample cross-sectional
area, while the coating thickness is negligible.
The nature of the protective coating effect on
high-temperature strength of nickel superalloys
is closely related with the peculiarities of their
failure in an oxidizing atmosphere at high temperatures. For a majority of the superalloys, oxidation at testing promotes the development of
surface cracks and contributes to their damaging
processes. Application of coatings that protect
the surface of such superalloys from oxidation
has a favorable effect on high-temperature
strength, especially in the case of long-term testing.
At the same time, the coating may have a negative effect on high-temperature strength of the
superalloy to be protected. This is due to the
formation of an interaction zone that reduces the
sample cross-sectional area, especially in the
case of precipitation of laminated, closely
packed phases (such as carbides, r-phase).
The scope of testing carried out to evaluate
the coating effect on high-temperature strength
is great enough. However, the authors of most
papers restricted themselves to comparison of
the time-to-failure of the coated and uncoated
samples. As a rule, in this case it is stated that
coating deposition does not reduce sample timeto-failure.
Creep-Rupture Strength of Superalloys
Tested in Media Provoking Hot Corrosion.
Testing for creep and stress-rupture properties
in a medium simulating fuel combustion products with elevated sulfur content substantially
accelerates the failure processes occurring in the
specimen surface layers and their destruction.
The testing procedure was as follows (Ref 13,
14). The gas turbine fuel (GZT) synthetic ash
was applied from an alcohol suspension to the
surface of the standard specimens intended for
high-temperature testing. The specimen was 6
mm in diameter, with the working part 30 mm
long. The ash layer was 120 g/m2, with the layer
application repeated every 500 h of testing. Air
supplied at a rate of 20 L/h was forced through
the furnace muffle where the specimen was held
for testing.
The results of the JS6U superalloy testing in
air and with a layer of ash on it at 800 and 900
C are shown in Fig. 7.17. The superalloy was
considerably damaged during testing where the
Fig. 7.17
Stress-rupture curves of JS6U superalloy in air (1, ) and in GZT ash (2, ). Ni20Co20Cr12AlY coating () and diffusion
coating (Al5Si1.5Y) () in GZT ash
GZT ash was used. It manifested itself in an intensive time-dependent drop of the superalloy
strength. In this case, the strength versus test
time is still the same, both under hot corrosion
conditions and in the air. It is an exponential
equation establishing a relationship of time-tofailure to stress (Ref 13):
s a rn
where s is time-to-failure, r is stress, and a and

n are constants depending on material properties, corrosive medium, and temperature.
At longer test time and higher test temperatures, the effect of the medium on the superalloy
is greater, as is shown in Fig. 7.17 (Ref 15).
After 100 h of testing at 800 C and 40 to 50
h at 900 C, creep-rupture strength of the JS6U
Table 7.14 Time-to-failure of JS6K superalloy in

air and in GZT ash at the same stress level. (Ref
13)
Test
Stress Time to failure
temperature, (r), at testing in air
C
MPa
(sA), h
800
850
900
800
850
900
410
297
179
293
190
112
1000
10,000
Time to failure at
testing in GZT ash
(sGZT), h
180
165
100
670
280
190
superalloy with the GZT ash starts to decrease.

If the initial life of the JS6U superalloy in air is
1000 h, the GZT ash on it will reduce it by 74%
at 800 C and by 91% at 900 C. The difference
in high-temperature creep-rupture strength levels of the specimens tested in air and in the corrosive media becomes more pronounced as their
testing time increases.
The same effect can be observed for the JS6K
superalloy (Table 7.14). At long-term testing,
time-to-failure decreases by a few orders of
magnitude. The higher the test temperature is
and the smaller the diameter of the specimen
tested, the greater the medium effect will be.
In addition to shortening superalloy specimen
lives under hot corrosion conditions, their initial,
secondary, and tertiary creep lives become
shorter, too; total ductility (specimen elongation) at failure and ductility at the initial and
secondary creep phases are reduced. The JS6U
superalloy behavior at testing is typical of all
nickel superalloys.
Two test-coated JS6U and JS6K superalloys
under corrosive conditions, two types of coatings were used: a diffusion coating made of the
Al5Si1.5Y alloy deposited by the EA method
and an overlay coating made of the
Ni20Co20Cr12AlY alloy deposited by the EB
method. The coating thickness values were 50
and 100 lm, respectively.
The test results on creep-rupture strength of
the coated alloys under hot corrosion conditions
are shown in Fig. 7.17. The Ni20Co20Cr12AlY

coating on the JS6U and JS6K superalloys guarantees their protection against hot corrosion as
well as ensures their long-term strength characteristics similar to those typical of testing in air.
The creep-rupture strength level of the coated
JS6U superalloy at 2000 h testing at 800 C is
400 MPa, while that for the uncoated specimen
is 240 MPa. At 900 C, the coating guarantees
the alloy protection for more than 1000 h. In this
case, the alloy high-temperature strength level is
220 MPa. At the same time, the uncoated alloy
has creep-rupture strength of less than 100 MPa.
Creep rates (V, %/h) of JS6U superalloy at 900 C

at (a) initial, (b) second, and (c) tertiary creep
stages. 1, Air, no coating; 2, GZT ash, Ni20Co20Cr12AlY coating; 3, GZT ash, diffusion coating (Al5Si1.5Y); 4, GZT ash, no
coating
Fig. 7.18
Diffusion coatings demonstrate worse protective properties. However, they reliably protect
the alloys against hot corrosion at 800 C for at
least 1000 h. The JS6U properties under these
test conditions are the same as at its testing in
air. At 900 C, the diffusion coating becomes
faulty after 200 h, and a loss of creep-rupture
strength is observed for the specimens with this
coating. However, it is higher than that of the
uncoated JS6U and JS6K superalloys.
The creep rate for the coated JS6U and JS6K
superalloys when tested in the GZT ash is the
same as at their testing in air (Fig. 7.18). In addition, the coated specimens do not lose their
ductility (Fig. 7.19), as compared with the uncoated specimens tested under the same conditions.
The study of the nature of coating failure during testing has not revealed any peculiarities or
distinctions from the failure of these coatings
without any load applied. The size of the
Ni20Co20Cr12AlY coating/JS6U superalloy interaction zone after 3000 h testing at 900 C does
not exceed 20 lm. The typical feature of the
Ni20Co20Cr12AlY coating detected during
testing is its high ductility. At testing, cracking
occurs beneath the coating that is strained above
the crack but is not faulty.
Creep-Rupture Strength of Superalloys
Tested in Air. High-temperature testing of superalloys with directional solidification (DS)
and single-crystal (SC) structures corroborates
the general features revealed of the protectivecoating effect on their high-temperature
strength. The results of testing for creep-rupture
strength of the directionally solidified JS26-DS
superalloy with the Ni20Co20Cr12AlY coating
are shown in Fig.7.20. At testing at 1100 C for
1000 h, creep-rupture strength values for the
coated and uncoated specimens are 55 and 45
MPa, respectively. The mismatch between
creep-rupture strength of the coated and uncoated alloys is much less than that obtained at
testing in the GZT ash medium and than could
have been expected from the uncoated alloy oxidation depth.
No evidence of the expected effect is likely
to result from the fact that under high-strain conditions typical of the specimens with directional
structures, the surface crack-growth rate exceeds
the specimen oxidation rate. Despite coating
ductility and protection of the alloy against oxidation, some cracks develop beneath the coating, too (Fig. 7.21). The cracks gradually propagate and cause damage to specimens.
The development of interaction processes between the coating and the alloy leads to a decrease of the specimen cross-sectional area. Dissolution of the c-strengthening phase in the
interaction zone beneath the coating reduces the
alloy strength. As a result, the working crosssectional area of the JS-type superalloy specimen 5 mm in diameter is reduced by 5 to 7%
after 500 h of testing. It is much less than in the
case of oxidation effect. Yet, for the blades less
than 2 mm thick and bearing high loads, the ef-
Relative elongation of JS6U superalloy at 900 C

versus stress at (a) initial, (b) second, and (c) tertiary creep stages. 1, Air, no coating; 2, GZT ash,
Ni20Co20Cr12AlY coating; 3, GZT ash, diffusion coating
(Al5Si1.5Y); 4, GZT ash, no coating
Fig. 7.19
fect of interaction between the coating and the

superalloy should be taken into account.
The structural investigations do not reveal differences in the interaction zone between coatings and the superalloys tested under no-load
conditions and with high stresses and strains applied. It is explained by the fact that the energy
contribution to the diffusion processes due to
high temperatures substantially exceeds the energy contribution of the stresses.
In all the nickel superalloy specimens tested
at long-term static loading, the initial failure
zone is located on the specimen surface, which
points to similar failure patterns based on crack
development and propagation causing destruction at the final creep stage. The coating does
not change the failure mechanism; it only neutralizes the effect of oxidation on it.
Mechanism of Coating Effect on Superalloy Creep-Rupture. Summarizing the experimental data on the coating effect on high-temperature strength of superalloys, the following
conclusion can be drawn. The plot of stress level
as a function of time-to-failure that is a straight
line has a point corresponding to a certain test
time of sc (Fig. 7.22). Starting from this moment, the uncoated specimens start losing their
creep-rupture strength due to oxidation effect,
surface softening, and reduction of their crosssectional areas. Starting from this moment,
creep-rupture strength of the specimens with
coatings that protect their surface against oxidation, softening, and respective reduction of
their cross-sectional areas is higher in comparison with that of the uncoated specimens.
Time of sc is a measure of an uncoated specimen test life. Testing for a longer period should
be carried out only on protected specimens. The
sc point position depends on heat resistance of
the alloy tested, test temperature, environment
aggressiveness, and the diameter of the specimen tested.
Time of sc for specimens 5 mm in diameter
and made of heat-resistant superalloys (such as
JS6K, JS6U, JS32, etc.) tested in air at 900 to
1000 C is more than 1000 h. That is why most
investigators do not reveal a positive effect of
coatings on creep and stress-rupture properties
of these alloys, because they reduced their test
time to 100 to 500 h. A more aggressive medium, elevated temperatures, smaller specimen
diameters, and reduced heat resistance shift the
sc point to the left. When testing the JS6U superalloy in the GZT synthetic ash medium at 800
Fig. 7.20
Stress-rupture curves of JS26-DS alloy in air at 1100 C. 1, D, No coating; 2, , Ni20Co20Cr12AlY coating
and 900 C, the sc time values are 100 and 500
h, respectively.
Effect of Coatings on Fatigue Strength

of Superalloys
One of the factors that may cause turbineblade failure is blade-material fatigue. Fatigue
cracks usually develop in the blade surface layer

in the zone of the greatest alternate stress effect.
Blade fatigue strength depends on the surface
condition and can vary substantially when a
coating with physical and mechanical properties
different from those of the superalloy is applied.
At testing for fatigue, it is necessary to take into
account the fact that the coating effect manifests
itself clearly during bend-testing when the maximum strain area is located on the surface. The
following factors have an effect on fatigue
strength of coated specimens:
Surface microstructure of JS26-DS superalloy sample with Ni20Co20Cr12AlY coating after testing
for stress-rupture strength. Applied stress (r) 150 MPa; time to
rupture (s) 22 h. 300
Fig. 7.21
Coating structure (defects, texture, grain size

and orientation, and phase composition)
Coating thickness and ratio of coating thickness and specimen cross-sectional area
Structure of the coating-alloy interaction zone
(presence of needlelike precipitate of closely
packed phases)
Sign and level of thermal stresses in the coating
Coating-material fatigue strength
Coating surface roughness
Fatigue Strength of Superalloys with Diffusion Coatings. Both thermal stresses generated in diffusion coatings and strength and ductility differ substantially from the respective
properties of overlay coatings. As a result, the
effect of these two groups of coatings on superalloy fatigue strength is different.

The results of extensive studies of fatigue
strength of superalloys with diffusion aluminide
coatings revealed the following main relationships.
In the temperature range of 20 to 250 C, fatigue strength of the specimens with diffusion
coatings is higher than that of unprotected specimens (Ref 16, 17). The increase in fatigue
strength depends on the coating chemical composition, the alloy properties, and the thickness
of the specimens tested. This effect of diffusion
coatings is explained by substantial compressive
stresses (150 to 300 MPa) in them. These
stresses compensate for the low strength and
zero ductility of coatings in the temperature
range in question.
Testing at high stresses when tensile stresses
arise in the surface layer may cause cracking
even at the first stages of testing. Such cracking
reduces total service lives of the specimens with
diffusion aluminide coatings compared to the
uncoated alloy in the range of 20 to 250 C.
At 900 to 950 C, no deterioration of fatigue
strength of the specimens and blade models
made of the JS6K and JS6U superalloys and protected with diffusion coatings has been detected
compared with unprotected specimens (Ref 18,
19).
Oxidizing protected and unprotected specimens before testing under the air atmosphere at
950 C for 1000 h reveals the positive effect of
the applied coatings. Depending on the test time,
preoxidation of the uncoated samples results in
the reduction of the JS6K superalloy sample fatigue lives by 10 to 17%. Fatigue lives of the
samples with the diffusion coatings also de-
crease, although this reduction is 5 to 12% and

also depends on the test time. Thus, fatigue lives
of the coated samples are higher than those of
uncoated samples by 6 to 9%. Similar results
were obtained on testing the blades.
Reduction of sample fatigue lives at longterm testing in the high-temperature range under
oxidizing atmospheres results from two processes: changes in superalloy structure and sample surface oxidation. For the uncoated sample,
its fatigue life drop is the consequence of more
intensive oxidation of its surface and softening
of its surface layer.
The same effect of diffusion coatings has been
revealed at testing in an aggressive medium of
fuel combustion products (Ref 20). Test results
for blade models with edge radii of 0.5 mm and
made of the Ni10CrWMoCo alloy are presented
in Fig. 7.23. The tests were carried out under
conditions of asymmetrical bending for the samples supported as a cantilever (cycle stress range
rm 200 MPa) in the T-1 aircraft fuel combustion products. The test temperature was 900
C, its cycle vibration frequency was 50 Hz, and
test time provided for 106 to 107 cycles. The diffusion coating was applied to the samples from
the powder mixture of 98% Al-Fe alloy (50%
Al) and 2% NH4Cl at 950 C. The coating thickness was 50 lm; its aluminum content was 34
to 36%.
In corrosive media, diffusion coatings on
blade models guarantee a fatigue-strength level
25 to 30% higher than that of the uncoated specimens. The JS6K and JS6U superalloys showed
the same results. Favorable outcome of fatiguestrength testing of blades with diffusion coatings
Fatigue strength of blade models made of JS6U

superalloy. Testing in combustion products of T-1
aircraft fuel at 900 C. 1, diffusion-coated samples; 2, uncoated
samples
Fig. 7.23
Fig. 7.22
Diagram illustrating the effect of coatings on

creep-rupture strength of superalloys
was also observed after their service in engines

(Ref 21).
Fatigue Strength of Superalloys with Overlay Coatings. Fatigue-strength testing of superalloys with different modifications of overlay
coatings in the temperature range of 20 to 1100
C (Ref 22, 23) has revealed the following relations.
At temperatures lower than 500 C, tensile
stresses arise in the surface layers of the specimens with overlay coatings. Their values depend
on the coating chemical composition. Ultimate
and yield strength values of overlay coatings are
rather high in this temperature range. They are
higher than the stresses applied to the specimens
during testing. Under these conditions, no increase in fatigue strength, which is typical of
diffusion coatings, is observed.
When testing round, standard specimens 8
mm in diameter with the Ni20Cr12AlY coating
60 lm thick under pure bending conditions at
20 C, no decrease in fatigue strength is detected
(Table 7.15). As a rule, a single crack that causes
failure (Fig. 7.24) appears on the specimen surface. The site of failure is located on the coating
surface. The same results were obtained on the
specimens made of the JS6F, JS26-DS, and
JS32-DS superalloys.
At 20 C, an insignificant rise in fatigue
strength is noted on round specimens made of
the VJL12U superalloy with the working part 3
mm in diameter and the Ni20Cr12AlY overlay
coating 100 lm thick (Fig. 7.25). It may result
from the fact that on testing the specimens 3 mm
in diameter, their stresses were calculated by ignoring coating thickness contribution to their
cross-sectional areas.
Fatigue-strength testing of finished blades at
20 C is a mandatory check operation in bladelot manufacturing at the plants. Heat treatment
of the blades and machining of their surfaces
have an important effect on the reference values
of blade fatigue strength (Table 7.16). Fatigue
life of the blade with a diffusion coating at the
Table 7.15 Fatigue strength, r1, of JS6U superalloy specimens 7 mm in diameter at 20 and
950 C
r1, MPa (N 2 107 cycles)
20 C
Coating
Uncoated
Ni20Cr12AlY
Ni20Co20Cr12AlY
190
190 (d 60 lm)
950 C
280
280 (d 100 lm)
300 (d 120 lm)
test stress of 180 MPa is more than N 2

107 cycles. When the Ni20Cr12AlY overlay
coating is applied instead of a diffusion coating,
reduction of fatigue life to N 6 106 cycles
is observed. Peening with microballs and generation of high-level compressive stresses did
not extend blade lives, although the site of failure shifted to their inner-cavity surface. The initial blade-fatigue life was obtained again after
reducing the temporary annealing temperature to
1000 C.
At 20 C, the defects of an overlay coating,
such as intercrystalline porosity, microdrops,
and so on that cause stress concentration, are the
sites of fatigue-crack nucleation. These blade
defects in the zones of high alternate stresses are
prohibitive.
At temperatures above 700 C, overlay coatings are free from stresses or show low compressive stresses. During fatigue-strength testing, the stress level on the specimen surface
considerably exceeds coating yield strength and
ultimate strength. The effect of combined
strains, depending on the superalloy mechanical
properties and test conditions, causes an accelerated accumulation of ductile strain followed
by a respective accumulation of defects. The
process is more intensive in the coating than in
the bulk superalloy. It results in the formation of
fatigue cracks and, as a rule, the cracks are not
longer than the coating thickness.
The results of fatigue strength testing of the
JS6U superalloy with the Ni20Co20Cr12AlY
and Ni20Cr12AlY coating at 950 C are given
in Table 7.15. Fatigue strength tested on the basis of 2 107 cycles for the coated samples was
the same as fatigue strength for the uncoated
samples. The lack of a wholesome effect of the
overlay coating is due to the fact that the rather
short test time of 120 h does not allow the
effect of oxidation that reduces fatigue strength
of uncoated samples to be noticed. Similar results were obtained for other superalloys. On
testing the coated samples 3 mm in diameter at
950 C, a slight increase in fatigue strength was
observed as compared with uncoated samples
(Fig. 7.25).
The salient feature of high-temperature testing of specimens with diffusion and overlay
coatings is the presence of numerous cracks on
their surfaces (Fig. 7.26). These cracks are stress
concentrators and cause formation of numerous
sites of fatigue failure beneath the coating. Further crack propagation depends on the failure
mechanism typical of the alloy tested at high
temperatures and on the specimen shape. Numerous surface sites merge and form a circular
crack, thereby reducing the specimen cross-sectional area. However, this reduction of the crosssectional area is insignificant for specimens 7
mm in diameter and does not affect their fatigue
strength. The main site causing failure of the
round specimens is located at a distance of 0.2
to 0.8 mm from the surface (Fig. 7.26c).
Studies of fatigue strength of superalloys with
EB thermal barrier coatings (TBC) (the coating
design is as follows: inner layer/intermediate
layer/outer
layer
of
Ni20Cr6AlY/
Ni20Cr12AlY/ZrO2-8%Y2O3, 35/45/60 lm
thick, respectively) were carried out for two
types of specimens. Round SC specimens 7 mm
in diameter and made of the JS32-DS superalloy
were tested under clear-bending conditions at
1100 C. Flat specimens with thickness varying
from 2 to 5 mm and made of the JS6U superalloy were bend-tested at 975 C. Bending vibrations were exhibited in a magnetostrictor at
a frequency of 3200 Hz.
The study of the nature of alloy round sample
failure has confirmed the general features stated
previously. The application of a TBC to a sample surface results (as has been observed for the
Ni20Cr12AlY coating) in nucleating numerous
sites of failure, followed by cracks propagating
into the coating and further into the alloy. However, the main site of nucleation of the crack
causing a sample failure is located beneath the
coating surface at a depth of 1 to 1.5 mm, and
it does not result from the coating effect. At
1100 C, fatigue strength (on the basis of N
2 107 cycles) was r1 150 MPa. The pres-
Fig. 7.24
ence of a ceramic layer changed neither the nature of sample failure nor fatigue strength values.
The situation changes on testing sheet (flat)
samples (Fig. 7.27). In this case, the surface site
of failure forms in the TBC in the maximum
stress zone in the corner of the flat sample. The
corner is a stress concentrator; it is the site from
where the fatigue crack causing failure propagates. In this case, fatigue strength after N 1
108 cycles diminishes by 30 MPa in comparison with unprotected samples, and the TBC
has the same effect as the Ni20Cr12AlY coating
lacking a ceramic layer.
Mechanism of Coating Effect on Superalloy Fatigue Strength. Fatigue-strength tests
showed that the effect of overlay coatings on this
superalloy property was different in the highand low-temperature ranges.
At room temperature, the level of coating
strength and respective fatigue strength approaches that of superalloys. That is why, despite tensile thermal stresses in the coating, sample and blade fatigue strength values do not
actually change when overlay coatings are deposited. On testing small-sized (3 mm) samples, deposition of a coating 100 lm thick,
which has high strength characteristics, can result in upgrading the level of fatigue strength.
On bend-testing in the low-temperature range,
coating defects act as stress concentrators. If the
coating is not free from defects, the sample and
blade fatigue strength values may decrease.
At temperatures above 800 C, overlay coating strength and respective fatigue strength are
substantially lower than superalloy fatigue
(a) Appearance and (b) fracture of JS6U superalloy specimen with Ni20Cr12AlY coating. Specimen tested at 20 C. r
200 MPa; N 1.8 107 cycles
Fatigue strength of VJL12U superalloy (3 mm

bars). 1 and 2, 20 C; 3, 700 C; 4 and 5, 900 C;
1 and 4, uncoated; 2, 3, and 5, Ni20Cr12AlY coating
Fig. 7.25
strength. The coating quickly accumulates defects, and numerous fatigue cracks, which are
effective stress concentrators, develop in it. In
addition, the cracks open the access of oxygen
to the alloy surface.
Further sample damage depends on the temperature, alloy heat resistance, sample shape,
and medium aggressiveness. The amount of oxygen passing through the cracks to the alloy surface is much less than if there is no coating at
all. The efficiency of the coating cracks as stress
concentrators, taking into account the high ductility of the coating, does not exceed the efficiency of stress concentrators formed due to oxidation. In addition, when testing round samples
under clear-bend conditions, a site of failure is
formed, as a rule, beneath the surface and inside
the sample. That is why on testing in the air at
900 to 1000 C for N 2 107 cycles (120

h), the values of fatigue strength for coated and
uncoated samples are the same. In the case of
testing in aggressive media, the coated samples
have higher fatigue strength in comparison with
uncoated samples.
The fatigue-strength testing procedure has
much in common with testing for creep-rupture
strength. A certain number of cycles, Nc, corresponds to the test time of sc (Fig. 7.22). On a
lapse of these cycles, surface damage caused by
oxidation and hot corrosion results in deterioration of uncoated sample fatigue strength. The
values of fatigue strength for coated samples are
higher than those for uncoated ones after the said
number of cycles.
At temperatures of 900 to 1000 C, heat resistance of the JS6U, JS6K, and VJL12U superalloys is rather high, and, in fact, their surfaces
remain free from oxidation products after the
said N 2 107 cycles (120 h). That is why
no effect of the coating on fatigue strength has
been noticed in the tests carried out. However,
when blades are used in turbines, the degree of
their surface damage substantially exceeds the
degree of sample surface damage at their testing
for fatigue strength. That is why fatigue strength
of the uncoated blades used in the engines is
always lower than that of the coated blades.
Effect of Coatings on Thermal Fatigue

of Superalloys
The blade designs featuring the most-effective cooling systems, the increase in temperature
gradients through the blade-wall thickness, and
the nonuniform temperature fields over the blade
surfaces have resulted in the growth of thermal
fatigue cracking events in the coatings. That is
why the study of coating effect on superalloy
resistance to failure during their heating and
Table 7.16
Fatigue strength of JS6U superalloy blades at 20 C
Stresses on
leading edge
Number of test cycles

before cracking
180 MPa
2 107
180 MPa
6 106
180 MPa
(8 15) 106
180 MPa
2 107
Location of cracks
Inner cavity at leading edge where stresses are
accumulated
Outer surface of loading edge where it
develops into shroud
Inner cavity at leading edge where stresses are
accumulated
Cracks are not detected
Type and characteristics of coating

and technological process
Diffusion coating, 40 lm thick, annealing:
1000 C, 4 h
Ni20Cr12AlY coating, 100 lm thick,
annealing: 1040 C, 5 h
Ni20Cr12AlY coating 100 lm thick,
annealing: 1040 C, 5 h; peening
Ni20Cr12AlY coating 100 lm thick,
annealing: 1000 C, 5 h
Fig. 7.26
(a) Appearance, (b) microstructure, and (c) fracture of JS6U superalloy specimen with Ni20Cr12AlY coating. Specimen
tested at 950 C. r 300 MPa, N 2.4 107 cycles
cooling cycles is an important point in the investigation of the properties of the coatings designed to protect turbine blades. Numerous publications deal with this problem.
A complicated dependence of thermal fatigue
on test rig parameters and on the structural design of the blades and test samples, as well as
the lack of a single-valued criterion for thermal
fatigue estimation, to some extent, make interpretation of numerous test results more complicated. Resistance of blades and samples to thermal fatigue depends on mechanical and
thermal-physical parameters of the materials,
such as strength, ductility, elasticity modulus,
thermal expansion coefficient, and thermal conductivity. The test conditions, which have a
great effect on thermal fatigue characteristics,
are as follows: rates of heating and cooling, cycle maximum and minimum temperatures, holding time at the maximum temperature, and static
loads applied.
During thermofatigue testing of the blades
and samples, their surfaces interact with gaseous
atmosphere. As a result of this interaction, the
top layers of the metal are depleted of alloying
elements. This reduces a fraction of strengthening phases in them and degrades their hightemperature strength. The changes in the chemical composition of the surface zone and the
respective variations of thermal expansion coefficients result in an additional increasing of
thermal stresses. They can reach a level considerably exceeding the surface zone strength.
Thermal fatigue cracking in the surface zone followed by crack propagation into the base material resulted in test sample failure.
The presence of heat-resistant coatings on
sample or blade surfaces, which protect the base
Fig. 7.27
Fatigue strength of JS6U superalloy (1) without any

coating and (2) with EB-TBC
material against oxidation and softening, is expected to have a favorable effect on resistance
to thermal fatigue. In this case, physical-mechanical and thermophysical characteristics of
the coating have a substantial effect on the test
results.
As a rule, rigidity of coatings (Ec h) is much
less than that of a blade wall or a sample (E
H):
Ec h
0.1
E H
where Ec and E are the coating and superalloy

elasticity modulus, respectively, and h and H are
the thickness of a coating and a blade wall or a
sample.
Under these conditions, even at slow cooling
or heating of the sample with the coating deposited at the temperature of T0, thermal stresses
arise in the coating. Under the conditions of
long-term cycling and accumulation of the strain
induced by thermal stresses, thermal fatigue
cracking may occur even if no additional
stresses are generated by impressed forces.
Thermal stresses may vary in the coating if a
temperature gradient in the coating/blade-wall
zone arises. It is always present in cooled blades
and substantially rises at turbine transient power
settings. If the sample or blade construction is
stressed or strained because of their nonuniform
temperature pattern along its full height and surface, coating thermal stresses are added to structural thermal stresses.
As a rule, thermal fatigue tests are carried out
on gas-dynamic test rigs at alternating heatingcooling cycles. Test results depend on specimen
shape, stiffness and rigidity, maximum and minimum cycle temperatures, rates of heating and
cooling, and gas parameters and composition.
Under these test conditions, thermal stress and
strain values in the specimens may be controlled
by specimen rigidity (e.g., solid or hollow specimens, etc.), rates of heating and cooling, and
maximum and minimum cycle temperatures.
Thermal fatigue is characterized by the number of cycles to appearance of the crack of the
specified size on the specimen surface. This
characteristic, to a considerable extent, is a qualitative one. It may be used for revealing the advantages of a certain blade construction or a certain superalloy or protective coating. However,
this characteristic cannot be used for blade service-life calculations. In addition, as a rule, gasdynamic test rigs do not simulate heat flows typ-
ical of the engines. That is why the results of the

rig tests cannot reveal the features of coating
behavior on the cooled blades at very high rates
of their heating and cooling that occur under
transition working conditions in modern engines.
Proceeding from extensive testing results for
specimens with different aluminide coatings, it
is noteworthy that aluminum content of the coating outer zone and the coating thickness are the
main factors that have an effect on the test results. The reduced aluminum content of the coating involving the respective reduction of the
NiAl compound ductile-brittle transition temperature improved thermal fatigue resistance of
the specimens and blades with diffusion aluminide coatings. The reduction of coating thickness
had the same effect.
Thermal fatigue resistance of JS6K superalloy

blades versus maximum cycle temperature. 1, Uncoated; 2, diffusion coating of aluminum, Al2O3, and NH4Cl mixture; 3, diffusion coating of aluminum, chromium, Al2O3, and
NH4Cl mixture. Source: Ref 24
Fig. 7.28
Fig. 7.29
Schematic diagram of test rig for thermal fatigue

tests
On setting up the program for testing specimens with diffusion coatings, it is necessary to
take into account some specific coating features,
such as low ductility of the coatings with the
aluminum content of more than 32% and gradual aluminum content reduction during testing,
followed by the increase in coating ductility.
Crack nucleation may take place in a diffusion
coating with high-aluminum content even after
the first few heating-cooling cycles at testing for
thermal fatigue under conditions of high thermal
stresses and respectively high strains. On running the same tests for the coated specimens
with low-aluminum content of the coating, their
cracking occurs much later than in uncoated
specimens.
Typical results of testing for thermal fatigue
of aluminized blades made of the JS6K superalloy on the gas-dynamic test rig are presented
in Fig. 7.28 (Ref 24, 25). The maximum gas
temperature available in the test rig is 1500 C,
the time of heating to the preset temperature is
8 s, and the cooling time is 17 s. The maximum test temperatures were 1000, 950, 900,
and 850 C. The minimum cycle temperature
was 300 C. After every 100 cycles, the specimens were examined visually for crack detection.
Under the conditions of the tests carried out,
the aluminized coatings increased thermal fatigue resistance. Under the conditions of 850
300 C cycling, aluminized blades withstood 4
times as many cycles to crack nucleation as uncoated blades. The increase in the maximum cycle temperature followed by the respective increase in thermal stress and strain values
resulted in less difference in thermal fatigue resistance of coated and uncoated blades.
The addition of salt to the gas flow under the
same test conditions gives rise to hot corrosion
and reduces blade service life considerably.
However, in all cases, the service lives of the
blades with aluminized coatings are longer than
those of uncoated blades.
Experimental Studies of Coated Superalloy
Thermal Fatigue. A large body of information
on comparison tests for specimens and blades
with different coatings is received using the gasdynamic test rig. Its layout is given in Fig. 7.29.
Knife-edged solid and hollow samples with
an edge radius of 1 mm (Fig. 7.30) are used for
testing. Eight samples are simultaneously fastened in the special fixtures of the test rig. Then,
they are moved in turn into a combustion chamber, where they are heated to 1000 or 1100 C,
and to compressed air supply systems, where

they are cooled to 200 C. The combustion
chamber uses aircraft fuel. The preset program
may change the cycle duration. Two options of
the cycle were used for testing:
sheat scool 30 s
sheat scool 30 s, holding time at the cycle
maximum temperature of 60 s
The second option was intended to determine

coating resistance to thermal fatigue at surface
Fig. 7.30
Table 7.17
No.
Sketch of specimens for thermal fatigue tests. R1,

radius equal to 1 mm
Test conditions
Temperatures
Time
1000 C 200 C
1000 C 200 C
1100 C 200 C
1200 C 200 C
1000 C 200 C
Time required for heating the

sample to the maximum
temperature is 30 s; cooling
time is 30 s
temperature is 30 s; holding
time at the maximum
time is 30 s
time is 60 s
time is 60 s
time is 120 s
exposure to severe conditions caused by oxidation.

Within the first seconds of heating hot gases
from the combustion chamber, heat the sample
central portion that is not large (20 to 30 mm).
It gives rise to a temperature gradient through
its full length and thickness. This gradient induces thermal stresses and strains, which peak
on the sample edge with a radius of 1 mm that
faces the flow of heating and cooling gases.
Strain accumulating in this place causes thermofatigue cracking of the sample. The number
of cycles to development of a crack 0.5 mm
long on the edge of the sample is its life criterion. Test conditions used are given in Table
7.17.
The test results are given in Table 7.18 and
shown in Fig. 7.31. When testing uncoated samples of the JS6U superalloy at the cycle maximum temperature of 1000 C, accelerated oxidation of their working edges and surfaces is
observed. Numerous cracks nucleate in these
zones. Protective coatings of all compositions
protect the superalloy surfaces against oxidation
and softening and, thus, increase the number of
cycles to thermal fatigue cracking. A narrower
range of strains arising on a hollow sample in
comparison with the solid one leads to a cyclic
life of 1.2 times as long.
The sample holding time at the cycle maximum temperature has a substantial effect on the
test results. The development of creep at a high
temperature and oxidation of coating surfaces
encourages accumulation of strains and facilitates a failure of both the coating and the superalloy. The samples tested at a holding time at the
maximum temperature of 60 s feature cyclic
lives 1.5 to 3 times shorter in comparison with
those of the samples tested without the hightemperature holding.
Under long-term cycling conditions (120 s),
the distinctions caused by sample physical-mechanical characteristics diminish. Cyclic life of
a JS6U superalloy sample with an overlay coating is 1.2 times longer than that of a sample
with a diffusion coating, while at testing under
short-term cycling conditions, the respective parameter is 2.2.
Thermal fatigue of the coated samples, which
are thermally stressed on the test rig under consideration, is closely related with both high-temperature properties of the alloys and sample fabrication techniques. The test results for the
samples made of the JS6U superalloy (equiaxial
solidification), the JS6F-DS superalloy (direc-
tional solidification), and the JS30-SC (single

crystal) with the diffusion aluminized coating
deposited from a mixture of 98% of the Al-Fe
(36% Al) alloy and 2% NH4Cl are presented in
Table 7.18. The aluminum content of the coating
outer zone is 22 to 24%; the coating is 30 to
50 lm thick. According to their resistance to
thermal fatigue, the alloys are ranged as follows:
JS6U (1), JS6F-DS (1.35), and JS30-SC (1.45).
Better ductility and strength of coatings as
well as a smaller mismatch between TECs of a
coating and a superalloy increase sample service
lives and reduce the distinction of thermal fatigue for superalloys with different structures.
The same alloys tested with the Ni20Cr12AlY/
Ni10Co20Cr12AlY coatings 100 lm thick deposited by EB technique have the same cyclic
lives: JS6U (1), JS6F-DS (1), and JS30-SC (1).
This is explained by the fact that long-term cycling gives rise to the accumulation of damages
not in a superalloy but in a coating, in the zone
of maximum oxidation and thermal strain effect.
On testing the samples made of the JS6U superalloy at long-term cycling (120 s) for their
resistance to thermal fatigue, the coatings are
ranged as follows: aluminized coating (1),
(Al5Si1.5Y)
diffusion
coating
(1.10),
Ni20Cr12AlY
coating
(1.20),
and
Ni20Co20Cr12AlY coating (1.30). The same
testing under short-term cycling conditions (60
s) gives the following ranges: aluminized coating (1), Ni20Cr12AlY coating (2.25), and
Ni20Cr12AlY/Ni10Co20Cr12AlY
coating
(2.40). Longer cyclic lives of the samples with
the Ni20Cr12AlY/Ni10Co20Cr12AlY coating
are due to their better surface roughness characteristics and fewer number of defects in the
coating structure when the two-stage surface
preparation technique is used, which is described in Chapter 3, Phase Composition of
Coatings on Superalloys.
The surface study of the samples tested on a
gas-dynamic test rig has shown that the nature
of their failures is similar in all cases; that is,
numerous thermal fatigue cracks develop on the
working edge at a distance of 0.6 to 1.0 mm
from one another, and the crack edges are open
and oxidized. Depending on coating ductility,
the cracks either propagate perpendicularly to
the surface or their abundant branching is observed in the coating.
Table 7.18 Number of cycles (N) to cracking for hollow samples at their testing for thermal fatigue
on 8UTS test rig
Superalloy
Coating
Deposition method
Thickness, lm
Test condition
NC
JS6U
JS6U(a)
JS6U
JS6U(a)
JS6U
JS6U
JS6U
JS6U
JS6U
JS6U
JS6U
JS6U
JS30-SC
JS30-SC
JS30-SC
JS6F-DS
JS6F-DS
JS6F-DS
VJL12U
VJL12U
VJL12U
VJL12U
JS26-DS(a)
JS30-SC(a)
JS30-SC(a)
JS30-SC(a)
JS32-DS(a)
JS32-DS(a)
Uncoated
Uncoated
Aluminized
Aluminized
Aluminized
Al5SiY
Ni20Cr12AlY
Ni20Cr12AlY
Ni20Cr12AlY
Ni20Co20Cr12AlY
Ni20Co20Cr12AlY
Uncoated
Aluminized
Uncoated
Aluminized
Al5SiY
Ni20Cr12AlY
Ni20Cr12AlY
Ni20Co20Cr12AlY
Ni20Cr12AlY/Al
Uncoated
Aluminized
TBC
Uncoated
TBC
Aluminizing
Aluminizing
Aluminizing
EA
EB
EB
EB
EB
EB
EB
Aluminizing
EB
Aluminizing
EB
EA
EB
EB
EB
EB/aluminizing
Aluminizing
EB
EB
40
40
40
40
100
50
100
20/80
100
100
40
40/60
40
40/70
40
50
100
100
100
50
180
180
1
2
1
2
2
2
1
2
2
1
2
5
1
1
1
1
1
1
2
2
2
5
4
3
3
3
3
3
1190
580
1610
830
1010
1080
3680
1160
1210
4720
1310
1330
2200
2420
4030
1900
2260
4000
700
740
860
870
620
1190
1300
1530
1320
1510
(a) Solid samples
The maximum range of thermal stresses and

strains in a coating at lab testing can be reproduced on rapid cooling of heated samples in cold
water. In this case, the range of strains in a coating induced by a temperature gradient between
the surface and the cross-sectional area of a sample wall increases. Such tests reveal the extreme
manifestation of thermal stresses that is a thermal shock. They allow the comparison of coatings by their ability to realize deformation
within the microzone of a coating. The typical
feature of thermal shock testing is a formation
of a network of small cracks in the coating. Their
length does not exceed coating thickness. Cracks
nucleate on cooling. At that period, the strains
exceeding coating ductility arise in surface layers. The results of thermal shock tests carried out
under 1050 C cold water conditions are
given in Table 7.19.
At such testing, an aluminized coating is fragmentized into a network of minor cracks after 8
to 10 cycles. The Ni20Cr12AlY coating is much
more stable. It shows a network of cracks after
40 to 50 cycles. The addition of a highly ductile
sublayer results in deceleration of crack growth

and in changing the mechanism of their propagation from brittle to ductile. The advantages of
the Ni20Cr12AlY coating tested under these
conditions are due to a better ratio of TECs of
an alloy and a coating and a respectively lower
level of thermal stresses.
Thermal fatigue tests of coated samples have
revealed that the coatings may both raise and
reduce superalloy thermal fatigue. When carrying out short-term tests in air, which did not result in unprotected sample surface damage by
oxidation, their lives could be longer than those
of the samples with protective coatings.
The detrimental effect of coatings is caused
by a wider range of strains on a coated alloy
surface due to additional thermal stresses in the
coating and lower mechanical properties of the
coating in comparison with those of the superalloy. Similar to testing for creep-rupture
strength, the situation changes at long-term
testing or at testing in aggressive media, which
prompt surface damage of unprotected samples.
Number of cycles to cracking caused by thermal fatigue of samples versus types of superalloys and coatings: JS6U
superalloy (1, 2, 3, 4); JS30-SC superalloy (5, 6, 7); JS6F-DS superalloy (8, 9, 10). 1, 5, 8, Uncoated superalloy; 2, 6, 9,
superalloy with aluminized coating; 3, superalloy with Ni20Cr12AlY coating; 4, 7, 10, superalloy with Ni20Cr12AlY/Ni10Co20Cr12AlY
coating
Fig. 7.31
Table 7.19
Alloy
JS6F-DS
JS6U-DS
JS6F-SC
JS6F-DS
JS6F-DS
JS6F-DS
Thermal shock testing of different-type coatings. Test conditions: 1050 C water

Coating
Thickness,
lm
Deposition method
Number of cycles before

cracking of coating
Coating effect factor,

K N/NAl
Aluminized
Aluminized
Aluminized
Ni20Cr12AlY
Ni20Co20Cr12AlY
40
40
40
100
100 m
40/60
Aluminizing
Aluminizing
Aluminizing
EB
EB
EB
8
10
10
50 60
30 40
50 70
56
34
57
Distinctions in service lives of the samples

with different coating compositions are caused
by a number of factors, such as the level of thermal stresses arising in the coating, its ductility,
and the strength and protective properties. The
efficiency of any coating type manifests itself,
to a greater extent, at high thermal stresses in a
coating, which are generated by high rates of
heating and cooling.
Coating Effect on Superalloy Low-Cycle
Fatigue. Low-cycle fatigue tests are an efficient
technique for evaluation of coating resistance to
cracking caused by high strains typical of thermal loading. Flat samples 1 mm thick of the
VJL12U superalloy with coatings of different
compositions were tested for low-cycle fatigue
by bending at 20 and 800 C. At 20 C, all types
of coatings tested have ductility less than 1%
and the maximum thermal stress level. At 800
C, all coatings are ductile and thermal stresses
are not high in them.
An aluminized coating deposited from a mixture of 98% Al-Fe alloy (36% Al) and 2%
NH4Cl
and
the
Ni20Cr12AlY
and
overlay
coatings 100 mm thick deposited by EB technique have been tested. The test results are presented in Fig. 7.32 and given in Table 7.20. At
20 C, all types of protective coatings reduce life
of the VJL12U superalloy. This feature is most
conspicuous in the samples with the
Ni20Cr12AlY coating and caused by defects
(microdots) in them, which are stress concentrators. A two-layer coating affects their lives to a
lesser extent. Unlike overlay coatings, diffusion
coatings are under compressive stresses at 20 C.
However, despite it, they also shorten lives of
the samples made of the VJL12U superalloy 5
to 7 times.
During testing (100 h) at 800 C, no failure
of the coating surface caused by oxidation occurs. Under these conditions, the coatings do not
exhibit their protective properties, and accumulation of strains in them due to low yield strength
enhances cracking and sample failure. Overlay
coatings having high strength and ductility margin at 800 C shorten sample lives to a lesser
extent in comparison with diffusion coatings
(Fig. 7.32). The lives of coated and uncoated
samples differ less and less and have a tendency
of becoming equal if cyclic testing lasts long
enough. The relationships derived for the
VJL12U superalloy can be extended to other superalloys.
It is worth determining coating material lowcycle fatigue separately from that of a superalloy
and comparing these results with coated sample
test results. A batch of samples was fabricated
from the Ni20Cr12AlY alloy condensate produced by EB evaporation technique. The samples were 0.6 to 0.8 mm thick. Low-cycle fatigue testing was carried out by tension at 950
C under two sets of test conditions:
On testing under the first set of test conditions, certain loads were cyclically applied to
the samples (4 cycles/min). Their levels neither varied nor depended on sample strains.
On testing under the second set of test conditions, samples were cyclically strained. The
strain range value, De, did not vary at testing.
Low-cycle fatigue of VJL12U superalloy at (a) 20

C and at (b) 800 C. 1, Uncoated; 2, aluminized
coating; 3, Ni20Cr12AlY coating
Fig. 7.32
The test results are presented in Fig. 7.33(a)

and (b) in comparison with the test results for
the samples made of the JS26-DS superalloy,
both uncoated and coated by the EA method
with the Ni20Cr12AlY alloy 100 lm thick.
As a result of testing under the first and second sets of test conditions, lives of coated samples are longer than those of uncoated ones. This
is due to the fact that coating ductility is very
high at 950 C. A coating material sample tested

under the first set of test conditions is shown in
Fig. 7.34. Its ductility exceeds 100%. In both
cases, the effect of the coating application is due
to the JS26-DS superalloy surface protection
against oxidation.
Under the first set of test conditions, when
high mechanical properties are most conspicuous, the coated superalloy life is much longer
than the coating material life. The conditions for
loading a coating on a superalloy sample are
fully dictated by the mechanical properties of a
superalloy. Under these conditions, the coating
accumulates failures much slower than a coating
material tested separately does.
Under the second set of test conditions, in the
certain range of strain range, a coating material
has a longer life. This is also due to high ductility of the coating material at 950 C. Failures
in the superalloy that features lower ductility are
accumulated much faster. Perhaps, coating failure occurs after cracking of the superalloy.
Coating Effect on Thermomechanical

Fatigue of Superalloys
Basic Principles and Methods. The most detailed information on coating properties under
conditions similar to blade service conditions
can be obtained from thermomechanical fatigue
(TMF) tests. For nickel-base superalloys used as
turbine blade materials, the main TMF test types
are load-adjusted thermomechanical fatigue test
(LA-TMF test) and strain-control thermomechanical fatigue test (SC-TMF test) (Ref 2527).
At LA-TMF testing, a constant stress range of
Dr const (gentle loading) is preset, while at
SC-TMF testing, a constant mechanical strain
Table 7.20 Experimental K and C in equation of De N k C characterizing low-cycle fatigue of

VJL12U superalloy with coatings of different types
Test temperature
20 C
800 C
Coating
Thickness, lm
Deposition method
Uncoated
Aluminum
Ni20Cr12AlY
Uncoated
Aluminum
Ni20Cr12AlY
40
100
100
40
100
Aluminizing
EB
EB
Aluminizing
EB
0.119
0.215
0.127
0.129
0.144
0.126
0.131
3.11
4.99
2.15
2.58
2.97
2.34
2.56
Results of low-cycle fatigue testing of JS26-DS superalloy and vapor-deposited samples under (a) the first and (b) the
second set of test conditions. 1, Uncoated superalloy; 2, superalloy with Ni20Cr12AlY coating; 3, samples made of
condensate alloy Ni20Cr12AlY
Fig. 7.33
range of De const (severe loading) is usually

held. At elastic deformation of a sample, both
types of testing are, in fact, similar and thus yield
similar outcome.
Each of the previously mentioned TMF tests
allows only approximate simulation of bladematerial deformation processes under service
conditions, because in a real blade, complicated
cyclic deformation takes place. This deformation is a combination of gentle loading caused
by centrifugal and gas loads and of severe loading caused by thermal stresses. It is considered
that severe cyclic loading that occurs at SC-TMF
test is most similar to the real deformation conditions of a blade airfoil material. However, this
statement is true under the conditions of elasticplastic cyclic deformation of a blade material
only when thermal stresses in a blade airfoil are
noticeably higher than the mechanical stresses
caused by centrifugal and gas loads. In addition,

the irreversible creep strains take place in the
blade airfoil under service conditions. This results in additional complication of a real blade
material deformation pattern.
Both of the previously mentioned test methods are used for experimental investigations of
TMF of overlay coatings deposited on turbine
blades. In the case of experimental research into
TMF of the coatings, the following factors typical of turbine blade coatings under their real
deformation and failure conditions should be
taken into consideration (Ref 28, 29):
Appearance of vapor-deposited Ni20Cr12AlY

sample after testing under the first set of test conditions. 1, Initial sample; 2, r 55 MPa, d 320%; 3, r 25
MPa, d 240%, N 2320 cycles; 4, r 15 MPa, d 500%,
N 5180 cycles
Fig. 7.34
Basically, thermal fatigue cracking starts in

the coating and then penetrates into the substrate. This is very typical for the SC blades
cast with the primary crystallographic orientation 001 because of the low elasticity
modulus of material and, therefore, low thermal strain ranges for thermal cyclic loading.
It is necessary to take into account the twoaxial stress state in the coating during analysis
of the TMF test results of coated bars.
Cracking of the coating on the cylindrical bar
during TMF testing can be both diametrical
and longitudinal.
The stress-strain state of the coating cannot
be directly measured by existing control
methods during TMF testing of coated bars.
For analysis of nonelastic straining of the
coating under these conditions, it is necessary
to use suitable constitutive models.
Strains in coatings are basically dependent on
the difference between TECs of the substrate
and the coating. These strains in the coatings
are usually non-linear (plastic) even when linear (elastic) straining of the substrate occurs.
Coating strain range values are, as a rule,
higher than corresponding strain range values
of the surface layer of the blade (or bar) during thermomechanical cyclic loading, especially for the SC and DS materials.
The number of cycles to the first crack initiation and their distribution on the outer surface of the SC blade airfoil cast with primary
crystallographic orientation 001 depend on
the secondary orientation of the crystallographic lattice in the airfoil cross section. This
is greatly dependent on the circumferential
thermal stresses, rs, in the blade, resulting in
longitudinal coating cracking, and the value
of circumferential elasticity modulus, Es, for
the SC superalloy. So, when TMF-cycle testing of SC-coated bar, it is necessary to know
the positioning of the axis of the secondary

orientations of 010 and 100 for each bar,
because distribution of the first crack on the
coated surface can depend on this axis in the
bar.
The following experimental method of TMF
investigation of the protective coatings for turbine blades is suggested.
Because elastic-plastic straining of the coating is possible even under elastic straining of the
substrate, it is suggested that TMF testing of the
coated bars be carried out under conditions of
elastic straining of the base material only. This
method is expedient for SC turbine blades with
primary crystallographic orientation 001, because there are no plastic strains in the outer
layers of such blades at operating conditions.
This is the result of the low elasticity modulus
and high yield strength of blade material for the
001 orientation. The exceptions to the rule are
small stress-concentration zones near perforation holes at the cooled blade walls where plastic
strains of the blade metal are possible.
In accordance with that mentioned previously, the LA-TMF test is recommended for
TMF testing of coated bars. The results of this
method are similar to SC-TMF testing under
conditions of elastic straining of the bar material
but are simple in practice. The main problem for
SC-TMF testing is the necessity for temperature
uniformity of the metal in the extensometer-controlled zone at stationary and nonstationary thermal conditions. This is not necessary for the LATMF test, because during the test, the loading of
Fig. 7.35
the bar is controlled. Moreover, nonuniformity

of temperature distribution along the central
work zone with a temperature maximum is preferable, because it helps to fix the location of
maximum failure in the cross section of the bar.
Using standard whole cylindrical bar with
constant square cross-sectional work zone is recommended for the LA-TMF test. The stress state
of the coating applied to the outer surface of the
bar is biaxial, although the stress state of the bar
itself during this test is uniaxial.
It is recommended to use a control system
with two independent channels: first, to control
the load, and second, to control the temperature
of the bar during LA-TMF testing. This control
system makes possible model changes in stress
and maximum temperature of the work zone of
the bar versus time with any phase shift among
these independently controlled parameters. The
possibility of stress asymmetry, Rr rmin /
rmax, varying for all types of LA-TMF tests
must also be limited. (rmin and rmax are correspondingly minimum and maximum values of
stress during cycle.)
The stress-strain state of the bar coatings is
calculated for the TMF test conditions by using
the developed constitutive model based on the
deformational theory of plasticity.
The calculational hysteresis loop for the overlay coating Ni20Co20Cr12AlY, obtained on the
basis of this constitutive model for the out-ofphase LA-TMF test of the coated bar cast of
MAR-M-002 superalloy, is shown in Fig.
7.35(a). The calculations are performed for out-
Hysteresis loops for (1) MAR-M-002 bar and (2) Ni20Co20Cr12AlY coating material of (a) whole cylindrical coated bar
and (b) when out-of-phase TMF testing
of-phase LA-TMF test conditions shown in Fig.

7.35(b) for a solid cylindrical specimen made of
the MAR-M-002 superalloy with the
For coating design, the mechanical properties
of cast specimens of the same chemical composition as the coating material were used in a
first approximation. It is worthwhile to perform
the calculations in question before testing, because it allows the proper choice of test conditions for coated specimens capable of realizing
the required range of elastic-plastic strains in the
coating. It is also worthwhile to perform calculations based on real measurements of the sample temperature and the load applied after testing, for estimation of the actual strain range in
the coating.
Calculations of strains in the coatings after
testing are also required for plotting thermal fatigue curves specialized for coatings. For this
purpose, sample thermal fatigue curves plotted
from LA-TMF test data are used.
Thermomechanical Test Results for
CMSX-4 Superalloy with Coatings. Three
modifications of the coating were deposited on
the test specimens (Ref 29, 30). The first modification of the coating 80 lm thick was deposited by EA method. After its deposition, the
coated specimens were annealed at 1080 C for
2 h. Then, they were shot peened with microballs and underwent another annealing under
vacuum at 1080 C for 2 h. The coating features
a dense, fine-grained structure typical of coatings deposited by EA method. The coating characteristics are presented in Table 7.21.
The second modification of the coating designated as Ni7Co12Cr17Al under study was deposited in two stages. At the first stage, a layer
of the Ni8Co12Cr7Al coating was deposited by
the previously mentioned technological process.
At the second stage, a layer of Al-5%Si-1.5%Y
alloy was deposited on the surface of
Ni8Co12Cr7Al layer by EA method at the rate
of 40 g/m2. A diffusion layer 50 lm thick was
Table 7.21
formed on the surface on annealing at 1080 C

for 2 h. The total coating thickness was 80 lm.
The average chemical compositions in the inner
and outer zones of the coating are given in Table
7.21. The outer zone of the Ni7Co12Cr17Al
coating consists of b-phase. Its structure is typical of the diffusion coatings deposited by EA
method (Fig. 4.21d). A multiphase zone is
formed under b-phase layer on diffusion annealing.
The Co32Ni21Cr8Al coating 100 to 120 lm
thick was deposited by the low-pressure plasma
spray (LPPS) method using the same technological process as for turbine blades.
Thermomechanical Fatigue Tests. Proceeding from the principles discussed previously, the
following test conditions were used: temperature
and load antiphase alteration, and asymmetric
cycle of load alteration without holding the specimen at the maximum cycle temperature (Fig.
7.36).
Fig. 7.36
Stress and maximum temperature vs. time distribution during out-of-phase LA-TMF testing of
coated bar
Characteristics of coatings tested

Coating
Ni8Co12Cr7Al
Ni7Co12Cr17Al
Co32Ni21Cr8AlY
Deposition method
Thickness, lm
Coating zone
Ni
Co
Cr
Al
Ta
Re
Hf
EA
EA
EA
LPPS
80
80
100120
Outer
Inner
Base
Base
Base
32
7.6
6.5
8.1
Base
12.4
7.5
16.9
21
6.5
17.0
6.8
8
4.5
2.0
4.5
0.3
0.2
1.1
0.4
0.5
Note: EA, electric arc; LPPS, low-pressure plasma spray
Fig. 7.37
Specimen for TMF tests (in mm)
Fig. 7.38
Temperature pattern in specimen working area
Table 7.22
No.
Specimen loading conditions
Temperature,
C
1
2
3
4
Fig. 7.39
100
1100
100
1100
100
1100
100
1100
Stress,
MPa
rtensile
rcompressive
rtensile
rcompressive
rtensile
rcompressive
rtensile
rcompressive
Stress range,
MPa
500
150
200
150
200
100
200
50
Dr 650
Dr 350
Dr 300
Dr 250
Thermal strain range distribution in the specimen

maximum temperature area
Whole cylindrical specimens of the CMSX-4

superalloy with diameters of 6.5 mm were used
to carry out TMF tests (Fig. 7.37). Specimens
were heated by conducting electric current with
current density of I 20 A/mm2. The temperature distribution pattern in the specimen working
zone is given in Fig. 7.38. Earlier research has
revealed that the previously mentioned heating
conditions do not cause changes in superalloy
mechanical properties (Ref 31). At the same
time, specimens were loaded using a hydraulic
loading device. The main parameters of specimen loading conditions are presented in Table
7.22.
The coating service life was set by the moment when cracking on the surface was detected.
Control over cracking was carried out by an optical method and by nondestructive liquid-penetrant testing. The resolving power of the technique in use was as follows: crack length 0.5
mm and depths of penetration of 0.05 mm.
To specify specimen-loading conditions during the testing cycle, strain was measured in the
zone of the maximum temperature. Marks made
of the wire with a diameter of 0.05 mm were
spot welded to the specimen surface. The distance between the marks was measured at the
minimum and maximum temperatures of the cycle using an optical measuring device (Fig.
7.39). Then, measurements were carried out at
heating under mechanical loading, and the total
strain was calculated in the central zone of the
specimen. The total strain range versus the applied stress range in the center of the specimen
is shown in Fig. 7.40.
For each coating modification, fatigue tests
under thermal cycling were carried out at three
levels of stress amplitudes in the specimen.
Based on the experimental data, a thermomechanical fatigue curve was plotted on the stress
Fig. 7.40
Strain range, De, of specimen versus stress range,

Dr, in the maximum temperature area
range (Dr, MPa)-service life (N, cycles) coordinates.

The terms used for test data description are as
follows: coating service life is the number of
cycles to crack(s) detection on the coated specimen surface, and CMSX-4 superalloy service
life is the number of cycles to crack(s) detection
on the uncoated specimen surface.
Service lives of uncoated specimens of the
CMSX-4 superalloy tested by the previously
mentioned method are of much interest, because
these data may be used for comparison with
coating service lives and the analysis of the effect of coatings on the alloy properties. When
cracking was detected on the specimen surface,
the test was not terminated. It was run to failure
of the specimen. Test results are given in Table
7.23 and Fig. 7.41.
Fractographic study of the specimens has
shown that they are typical of the failure at TMF
testing at narrow stress range. The nucleation
site of the crack is situated on the specimen surface (Fig. 7.42). The cracks grow into the specimen and cause its failure. During testing, most
of the cycles are spent on crack nucleation on
the specimen surface. After crack nucleation, the
service life of the cracked specimen is up to 8
Table 7.23
test results
Number of
cycles to
cracking
2850
6190
15,780
20,420
48,670
24,120
41,270
Fig. 7.41
Uncoated CMSX-4 alloy specimen
Number of cycles
to specimen failure
Stress range (Dr),

MPa
3630
6510
4840
16,950
21,820
25,480
44,860
650
650
650
450
350
300
300
250
Cyclic life of CMSX-4 superalloy. 1, Cracks on surface specimen; 2, failure of specimen
to 20% of the total service life of the specimen

under study. Measuring specimen dimensions
after their testing has revealed no elongation or
changes of their diameters. This is evidence of
the compliance of the tests with the chosen loading conditions in the range of superalloy elastic
deformation.
Test results for the specimens with the
Co32Ni21Cr8AlY coating are shown in Fig.
7.43. The number of cycles to crack appearance
on the specimen surface with the
Co32Ni21Cr8AlY coating is much less than for
an uncoated specimen. Crack appearance and
spacing in the surface network are similar to
those typical of thermal fatigue cracks, which
develop in the coatings on blade surfaces (Fig.
7.44). The cracks were detected along the full
length of the specimen. There were cracks in the
areas where the maximum temperatures did not
go over 1000 C. Crack nucleation is a consequence of the accumulation of plastic strain in
the coating as well as the formation of pores and
minor surface ruptures, which join together
gradually and form a thermomechanical fatigue
crack in the coating (Fig. 7.45a,b).
Crack propagation rate in the CMSX-4 superalloy is much less than that in the coating.
While cracks 80 to 100 lm long were detected
in the coating after 1000 cycles, a further 3100
cycles resulted in crack penetration of not more
than 100 to 120 lm into the alloy. Test results
for specimens with the Ni8Co12Cr7Al coating
are shown in Fig. 7.43.
The mechanism of crack formation and their
appearance and surface network are identical
with those typical of specimens with the
Co32Ni21Cr8AlY coating. The Ni8Co12Cr7Al
coating service life is 4 times as long as that of
the Co32Ni21Cr8AlY coating at the stress range
of Dr 300 MPa and 2 times as long at the
stress range of Dr 250 MPa.
Test results for specimens with the
Ni7Co8Cr17Al coating are shown in Fig. 7.43.
The mechanism of crack formation and their appearance and surface network are identical with
those typical of specimens with the
Co32Ni21Cr8AlY and Ni8Co12Cr7Al coatings.
The Ni7Co12Cr17Al coating service life is 13
times as long as that of the Co32Ni21Cr8AlY
coating at the stress range of Dr 300 MPa
and 8 times as long at the stress range of Dr
250 MPa.
Mechanism of Coating Effect on Thermomechanical Fatigue of Superalloys. A coating
with physical and mechanical properties identical to those of the protected superalloy can be
considered as an ideal coating, with high thermomechanical fatigue resistance. This coating
can hardly be developed, because a thin layer of
polycrystalline coating cannot have mechanical
properties typical of a modern SC superalloy.
However, the development of a coating with
characteristics that are as similar to those of the
uncoated superalloy as possible is an advantageous line in producing a coating with high thermal fatigue resistance.
The use of LA-TMF test methods with anti-
Fig. 7.42
phase change in temperatures and stresses allows the simulation of stress and strain conditions in the coatings on turbine blade surfaces.
Under the selected test conditions, the applied
loads contribute to the increase in compressive
stresses in coatings at high temperatures and to
the increase in their tensile stresses at low temperatures. The test technique in use allows the
distinctions in coating resistance to thermal fatigue crack nucleation and propagation to be revealed.
Fracture of specimen CMSX-4 superalloy, 3658 cycles, Dr 650 MPa. (a) Site of failure, 200; (b) crack development
area, 500
Table 7.24 Comparison of average lives of uncoated CMSX-4 specimens and CMSX-4 specimens with
different coatings
Number of cycles to cracking (%)
Stress range (Dr), MPa
650
350
300
250
CMSX-4
2850
20420
24120
41270
Co32Ni21Cr8AlY
Ni8Co12Cr7Al
Ni7Co12Cr17Al
368 (13%)
911 (4%)
2330 (6%)
2840 (14%)
4178 (17%)
6592 (16%)
10550 (52%)
12484 (52%)
18587 (45%)
(100%)
(100%)
(100%)
(100%)
Specimen fractography after thermal fatigue

testing has revealed that, in all cases, the crack
nucleation sites are located on the specimen surfaces and the fracture type is similar to that
caused by applying alternating loads. Cracks are
developed in the coatings after accumulation of
considerable strain. The tests have revealed a
noticeable distinction between the time to thermomechanical fatigue crack detection on the uncoated specimens and on the specimens with different coatings.
All test data are given in Table 7.24 for com-
Cyclic life of uncoated CMSX-4 superalloy (4) and

CMSX-4 with coatings. 1, Co32Ni21Cr8AlY coating; 2, Ni8Co12Cr7Al coating; 3, Ni7Co12Cr17Al coating
Fig. 7.43
Crack location on specimen surface with

Co32Ni21Cr8AlY coating (liquid penetrant testing), 900 cycles
Fig. 7.44
parison. The Co32Ni21Cr8AlY coating service

life is only 6 to 8% of uncoated CMSX-4 alloy
service life. Thus, if there are zones in turbine
blades with the stress range similar to that used
at testing, numerous thermal fatigue cracks will
develop in the Co32Ni21Cr8AlY coatings on
their surfaces. The development of these cracks
on the blades may be unexpected, because, as a
rule, coating service life is not taken into account
in blade service-life calculations.
Changes in coating physical-chemical properties can have a significant effect on their service
lives. The Ni8Co12Cr7Al coating service life is 2
to 4 times as long as that of Co32Ni21Cr8AlY
coating. The Ni7Co12Cr17Al coating has
also demonstrated high service life. The
Ni7Co12Cr17Al coating service life is 8 to 10
times as long as that of the Co32Ni21Cr8AlY
coating, and it is 45 to 52% of the CMSX-4 superalloy service life. This service life is primarily
due to close agreement of linear TECs of the coating and the CMSX-4 superalloy. However, the effect of technological factors in forming the outer
diffusion layer from a melted aluminum-base liquid phase cannot be excluded. The presence of the
outer diffusion layer has a desirable effect on the
coating service life, too. As is known, unlike the
applied coatings, diffusion coatings accumulate residual compressive stresses in the temperature
range below 900 C.
The chosen testing technique that uses the
specimens with diameters of 7 mm does not allow the estimation of the effect of crack formation in the coating on CMSX-4 specimen service lives. The comparison of the number of
cycles to failure for coated and uncoated specimens has revealed that the number of specimens
tested to failure and the test data are insufficient
for unbiased estimation of this effect.
The agreement of the test data from the coating service lives on the specimens with those on
the blades depends on blade service conditions
in the engine. In short-term cyclic tests of the
engines under high thermal loading conditions,
Fig. 7.45
Cracks in Co32Ni21Cr8AlY coating at (a) the initial stage of crack development and (b) the stage of its propagation into
the superalloy after 550 cycles. 500
there is a high probability that the data from the

coating service lives on the blades will show
close agreement with the respective results of
laboratory testing of the specimens. In the case
of long-term tests in the engines, these data will
depend to a considerable extent on such coating
characteristics as oxidation resistance and resistance to hot corrosion.
REFERENCES
1. M.I. Wood, The Mechanical Properties of

Coatings and Coated Systems, High Temperature Corrosion Advanced Materials
and Coatings, R. Streiff, J. Stringer, R. Krutenat, and M. Caillet, Ed., Elsevier Science
Publishers, 1989, p 633643
2. P.T. Kolomytzev, Gas Corrosion and

Strength of Nickel Alloys, Moscow, Metallurgy, 1984, 216 p
3. E.M. Grale, Research into NiAl and Ni3Al,
Mechanical Properties of Intermetallic
Compounds, J.H. Westbrook, Ed., John Wiley & Sons, New York, 1960, p 266299
4. A.G. Rozner and R.J. Wasilelevski, Tensile
Properties of NiAl and Ni3Al, J. Inst. Met.,
Vol 94 (No. 5), 1966, p 169175
5. V.P. Buntushkin and O.A. Bazyleva, Casting Ni3Al Intermetallic Compound-Base
Alloy for Single-Crystal Engine Turbine
Rotor Blades, Aviatsionnaya Prom-st., N 3
4, 1997, p 6165
6. G.W. Goward, Current Research on the Surface Protection of Superalloys for Gas Turbine Engines, J. Met., Vol 22 (No. 10),
1970, p 3139
7. P.T. Kolomytzev and N.V. Abraimov, Research into Temperature Effect on Coatings
Ductility, Phys. Chem. Mech. Mater., No.
2, 1982, p 104107
8. Y.S. Eliseev, N.V. Abraimov, and V.V.
Krymov, Chemical-Heat Treatment and
Protective Coatings in Aircraft EngineBuilding, Moscow, Vysshaya Shkola, 1999,
525 p
9. V.I. Nikitin, Y.A. Tamarin, and N.V. Zabrodina, Investigations of Structural and
Physical-Chemical Characteristics of NiCoCrAl System Alloys, Protective Coating
Application and Use, Leningrad, Nauka,
1987
10. A.I. Krivko, A.I. Epishin, and A.I. Samoilov, X-Ray Method of Research into
Stressed-Strained State of Composite Materials Based on Calculations and Experiments, Zavod. Lab., Vol 1 (No. 5), 1985, p
3438
11. A.I. Samoilov, I.A. Ignatova, A.M. Vorobiev, and V.S. Kozlova, Measurements of
Thermal Stresses in Ni-(Co)-Cr-Al-Y HeatResistant Coatings, Zavod. Lab., Vol 50
(No. 11), 1984, p 3639
12. E.G. Kuzovkov, Numerical Simulation of
Stressed State for Strained Bodies by Elementary Cells Method, Kiev, Institute of
Strength Problems, 1981, 37 p
13. V.I. Nikitin, Corrosion and Protection of
Gas Turbine Blades, Leningrad, Machinostroenie, 1987, 272 p
14. V.I. Nikitin and I.P. Komossarova, Method
of Protective Coating Testing for Lifetime,
Proc. CKTI, (No. 176), 1980, p 6773
15. V.I. Nikitin, Y.A. Tamarin, and N.V. Zabrodina, Corrosive and Mechanical Properties of Protective Coatings, AqueousChemical Conditions and Corrosion of
Power Equipment, No. 201, Leningrad,
CKTI, 1986, p 7281
16. J.F. Paskiet, D.H. Boon, and C.P. Sullivan,
Effect of Aluminide Coating on High-Cycle
Fatigue Behavior of Nickel-Base HighTemperature Alloy, J. Inst. Met., Vol 100
(No. 2), 1971, p 5862
17. M.M. Gell, G.P. Leverant, and C.H. Wells,
The Fatigue Strength of Nickel-Base Superalloys, ASTM STR, No. 64, 1970, p 113
153
18. I.I. Ischenko, V.I. Omelchenko, and B.N.
Sinaisky, Research into Effect of Heat-Resistant Coatings on Superalloy Fatigue
Strength, Strength Probl., No. 10, 1971, p
7679
19. I.I. Ischenko, V.I. Omelchenko, and B.N.
Sinaisky, Research into Effect of Heat
Treatment and Protective Coatings on JS6K
Superalloy Fatigue Strength, Strength
Probl., No. 8, 1976, p 1015
20. L.M. Akimov, Superalloys Fatigue Lives,
Moscow, Metallurgy, 1977, 220 p
21. P.T. Kolomytzev and P.G. Lebedev, Cyclic
and Long-Term Rupture Strength of Alloys
with Protective Coating, Strength Probl.,
No. 9, 1972, p 9295
22. L.B. Getsov, A.I. Rybnikov, and I.S. Malashenko, Fatigue Resistance of Coated Superalloys, Strength Probl., No. 5, 1990, p
5156
23. T.N. Rhys-Jones and T.P. Cunningham,
The Influence of Surface Coatings on the
Fatigue Behaviour of Aero Engine Materials Surface and Coatings Technology, Vol
42, 1990, p 1319
24. L.P. Lozinsky, B.I. Musienko, and A.A.
Ivanenko, Effect of Diffision Coatings on
Turbine Blade Damage at Thermal Cycling,
Vol 8, Kiev, Naukova Dumka, 1974, p 134
136
25. G.R. Halford, T.G. Meyer, R.S. Nelson, and
D.M. Nissley, Fatigue Life Prediction
Modeling for Turbine Hot Section Materials, NASA Technical Memorandum
100291, 33rd Intl. Gas Turbine and Aeroengine Congress and Exposition, (Amsterdam,
The Netherlands), ASME, 59 June 1988
26. E.G. Ellison, Thermomechanical Strain Cycling and Testing at Higher Temperature,
Measurement of High Temperature Mechanical Properties of Materials, M.S. Loveday and M.F. Day, Ed., HMSO, London,
1982, p 204218
27. D.A. Miller and R.H. Priest, Materials Response to Thermal Mechanical Strain Cycling, High Temperature Fatigue: Properties and Prediction, R.P. Skelton, Ed.,
Elsevier Applied Science, 1987, p 113175
28. Y.A. Nozhnitsky, R.A. Doulnev, and V.G.
Sundyrin, Damage Mechanisms for Thermomechanical Fatigue of Aircraft Engines
Materials, 81st Meeting of the AGARD
SMP, (Banff, Canada), 24 Oct 1995, and
published in CP-559
29. Y.A. Tamarin, V.G. Soundyrin, and N.G.

Bychkov, Thermo-Mechanical Fatigue
Tests of Coatings for Turbine Blades, High
Temperature Surface Engineering, J. Nicholls and D. Rickerby, Ed., IOM Communications Ltd., 2000, p 157169
30. Y.A. Tamarin and H.C. Low, Thermal Fatigue of Protective Coatings on CMSX-4
Superalloy, High Temperature Corrosion 5,
R. Streiff, I.J. Wright, R. Krutenat, M. Caillet, and A. Cailerie, Ed., Trans Tech Publication, 2001, p 647654
31. N.G. Bichkov, Test Rig for Thermo-Mechanical Fatigue Tests, Zavod. Lab., Vol 57
(No. 12), 1991, p 5758.

DOI: 10.1361/pctb2002p161

CHAPTER 8
Electron Beam Thermal Barrier

Coatings
DEVELOPMENT of thermal barrier coatings
(TBCs) applied to cooled blades is one of the
trends for improving gas turbines. Unlike aluminide protective coatings, ceramic coatings not
only protect blade surfaces from high-temperature oxidation and corrosion but also prevent
base material softening at high temperatures.
Thermal barrier coating application allows the
reduction of the blade temperature and the significant increase in its service life. Under both
steady-state and transient conditions, the application of TBCs can diminish temperature gradients over the blade surfaces as well as reduce
thermal stresses in them.
Between 1980 and 2000, a great number of
scientific papers and patents were published.
They covered the issues concerning development, deposition techniques, and use of TBCs.
A typical design of a TBC is presented in Fig.
8.1. The ceramic layer deposited directly on the
superalloy surface does not show the required
service life. Penetration of oxygen through the
ceramic layer to the superalloy surface results in
its quick oxidation and in spallation of the ceramic layer. That is why, as a rule, a TBC consists of at least two layers. An inner aluminide
heat-resistant bond coat may be formed by different techniques. It may be either a diffusion or
an overlay coating, depending on the requirements of its physical-mechanical properties and
protection targets. The requirements of bond
coat properties and protective coatings properties are much the same, yet the bond coat should
meet some special requirements. First of all, it
must be highly heat resistant; the oxides formed
on its surface should have high adhesion to both
the bond coat and the outer ceramic layer. When
choosing a bond coat composition, one should
pay special attention to its yttrium content as
well as to the contents of the other elements,

which guarantee high oxide adhesion to the surface and reactive element effect (Ref 1). It is of
special importance for bond coats deposited by
the electron beam (EB) technique, because their
yttrium contents depend on the yttrium content
of the liquid bath and vary within wide limits
(Ref 2). In this case, the required yttrium content
of 0.2 to 0.3% is guaranteed by different technological procedures, such as direct yttrium addition to the liquid bath. Under these conditions,
it is noteworthy that high yttrium contents of the
liquid bath cause slag formation on its surface,
thus resulting in occurrence of microdrops.
These microdrops on the bond coat surface may
provoke defects in the ceramic layer.
Fig. 8.1
Thermal barrier coating design
It should be taken into consideration that

TBCs are usually applied to the blades of hightemperature turbines. The blades of such turbines feature directionally solidified or singlecrystal structures, thin walls, and high cooling
efficiency. Under service conditions, high thermal stresses and strains arise in these blades, especially in their surface layers. That is why thermomechanical fatigue characteristics are as
important in choosing a bond coat composition
as its heat resistance. During thermal cycling,
the bond coat should not experience considerable plastic strain. For example, the effect of a
rippled blade surface (Fig. 2.10) always entails spallation of the ceramic layer.
The outer zirconium oxide/yttrium oxide
(ZrO2-Y2O3) system base ceramic layer can be
applied by two techniques: air plasma spraying
of powders (TBC-APS) or vapor condensation
at EB evaporation of ceramic pellets (TBC-EB).
For this system, ceramic coating service life
depends on Y2O3 content. The ZrO2-(6 to 9%)
Y2O3 compositions are usually applied, because
they have demonstrated maximum service lives
in the tests carried out (Ref 35). However, one
should bear in mind the fact that the coating service life depends not only on its chemical composition but also on its structure and adhesive
strength at the ceramic layer/bond coat interface, which depends on deposition technique.
For coatings deposited by different techniques,
the optimal chemical compositions may be other
than that stated previously. The ceramic layer
deposition technique determines such characteristics as ceramic layer structure and adhesive
strength, its corresponding service life, thermal
stresses in the ceramic layer, and its surface
roughness.
The main difficulty in designing TBCs for turbine blades lies in the combination of the ceramics on the blade surface and the superalloy that
they are made of. At heating-up/cooling-down
cycling, considerable difference between the ceramics and superalloy expansion coefficients
(D 5.0 106 1/C) causes the generation
of high thermal stresses in ceramics, which in
turn results in ceramic layer spalling from the
surface. To reduce thermal stresses, various
technological procedures are used. In the ceramic layer deposited by the APS technique,
special heat treatment is used to form a network
of microcracks that break the ceramics into isolated fragments (Ref 6). In the ceramic layer deposited by the EB technique, some specific columnar structure is formed that is readily
fragmentizing when tensile stresses arise in it

(Ref 7).
The point crucial to success in the development of TBCs lies in obtaining the required adhesive strength of the ceramic layer/heat-resistant bond coat, providing for holding of the
ceramics on the blade surface during all the
blade service life. As a rule, in aircraft engine
manufacturing, the technique of plasma deposition is used for nozzle vanes; in aircraft engine
turbine blades, the EB technique is considered
to be preferable. This is due to the fact that the
following properties can be rendered to the ceramic layer:
The specific columnar structure, with the

crystallites oriented perpendicular to the surface, forms in the ceramic vapor-deposited
layer. In the case of tensile stresses, the ceramic layer is readily fragmentizing, thus reducing ceramic tearing stress during thermal
cycling.
In the temperature range of 850 to 950 C,
which is below the blade heating temperature
at ceramic layer deposition, compressive
stresses arise in it. Their generation is due to
the different values of the ceramic and superalloy thermal expansion coefficients. These
stresses do not relax on subsequent process
annealing and under service conditions.
The adhesive strength of the ceramic layer is
controlled by physical-chemical reactions occurring between the ceramics and the metallic
bond coat. As-deposited ceramic layer adhesive strength is above 70 MPa.
The surface roughness of the ceramic coatings does not exceed 1.5 lm after their deposition.
This chapter deals with TBCs deposited by
the EB technique.
Ceramics for EB Evaporation

For EB evaporation, it is necessary to produce
ceramic pellets that are conveniently loaded into
a vaporizer and sufficiently strong to preserve
their shape during shipment and treatment with
EB. The composition of the pellets should be
similar and should provide for forming a coating
of the required chemical and phase compositions
when used for applying ceramic layers. Both
phase composition of the ceramic pellets and
their chemical homogeneity depend on chemical
and phase compositions of the initial ZrO2, the
Electron Beam Thermal Barrier Coatings / 163
conditions of its stabilization with Y2O3, the

production process for ceramic pellets, their annealing time and temperature, and their cooling
rate.
The process of TBC ceramic layer deposition
and the stability of its properties depend, to a
large extent, on the characteristics of the material evaporated. Zirconium oxide is used as a
base material to form the TBC ceramic layers.
The choice of this material is based on its low
thermal conductivity and high thermal expansion coefficient (TEC) of the ZrO2 cubic phase
( 11 106 K1) in the temperature range
of 20 to 1200 oC. The specific feature of ZrO2
is its polymorph transformations, causing significant volume changes (Fig. 8.2) (Ref 8).
When heating ZrO2, its low-temperature monoclinic (M) modification transforms into the tetragonal (T) and then into the cubic modification
(C): M T C.
M T transformation occurs martensitically
and features substantial hysteresis. When the T
phase transforms into the M one, the crystal lattice volume increases by 3 to 12%.
Making oxide additions prevents phase transformations in ZrO2. They form solid solutions
of fluorite crystalline structure with ZrO2. The
oxides stabilizing ZrO2 cubic modifications are
as follows: Y2O3, MgO, CaO, and Yb2O3 (Ref
Fig. 8.2
Ref 8
Low-yttrium region of ZrO2-Y2O3 phase diagram. M,

monoclinic; T, tetragonal; C, cubic phase. Source:
9). Among the previously mentioned oxides, the

most stable solid solutions are formed in the
ZrO2-Y2O3 system taken as a base of high-temperature-resistant ceramic coatings. The required addition contents are taken from the state
diagrams, although under actual conditions, it is
of great importance to consider the way of making additions, system homogeneity, and heat
treatment.
The quality of an overlay ceramic coating depends on the material homogeneity of the ceramic pellet being evaporated. This is caused by
the evaporation of a relatively shallow liquid
bath on the surface of the ceramics. Under other
equal conditions, mixture homogeneity depends
on the properties of the mixed components: the
finer they are, the less their difference in density
is, and the more their ranges of particle dimensions overlap, the more homogeneous mixture
can be produced.
A few techniques for ZrO2-Y2O3 solid-solution synthesis are available. Each of them has
both advantages and disadvantages (Ref 10):
Mixing ZrO2 and Y2O3 powders, followed by

mixture heat treatment at 1600 to 1800 C and
sintering in electric-arc furnaces
Mechanical mixing of hydroxides or zirconium and yttrium salts, followed by their heat
treatment at 1000 to 1200 C
Simultaneous deposition of homogeneous hydroxide mixtures or other unsolvable zirconium and yttrium compounds from salt solutions
If ZrO2 and Y2O3 powders are used and the
initial mixture is prepared by their mechanical
mixing, the phase composition of the pellet fabricated depends on a mixing process quality and
an annealing temperature. Zirconium oxide full
stabilization does not occur even after adding
20% Y2O3 to it (Table 8.1). Some Y2O3 remains
in the free state after annealing. As a rule, different densities of the initial components and the
size and shape of their particles do not guarantee
the uniform distribution of both oxides within
the pellets on mixing. In the mechanically mixed
pellets, one can detect the zones of different
chemical compositions (Table 8.2).
The pellets made of ZrO2-Y2O3 powders using the hydroxide method are found to be much
more homogeneous in their chemical and phase
compositions (Ref 11). Zirconium and yttrium
hydroxides are deposited simultaneously from
chloride solutions at pH 8 to 9, using aqueous
Table 8.1 Phase composition of ZrO2-Y2O3 ceramic pellets made of mechanical mixture of ZrO2 and
Y2O3 powders(a)
Phase content(b), vol%
Annealing at 1350 oC(c), 10 h
Annealing at 1700 oC(c), 10 h
Chemical composition
Y2O3(d)
Y2O3(d)
ZrO2
ZrO2-2Y2O3
ZrO2-4Y2O3
ZrO2-6Y2O3
ZrO2-8Y2O3
ZrO2-10Y2O3
ZrO2-15Y2O3
ZrO2-20Y2O3
ZrO2-25Y2O3
ZrO2-30Y2O3
ZrO2-40Y2O3
79
72
60
42
20
16
15
15
5
5
5
21
28
40
58
60
43
45
42
40
40
40
0
0
0
0
20
41
40
43
55
55
55
Not detected
Not detected
Detected
Detected
Detected
Detected
Detected
Detected
Detected
Detected
Detected
75
75
50
40
10
5
0
0
0
0
0
25
25
50
60
60
40
40
20
15
10
0
0
0
0
0
30
55
60
80
85
90
100
Not detected
Not detected
Not detected
Not detected
Not detected
Not detected
Not detected
Not detected
Not detected
Detected
Detected
(a) Size of mixed powders, 2060 mm. (b) Cooled in furnace. (c) M, monoclinic; T, tetragonal; C, cubic. (d) Free Y2O3
solution of ammonia in accordance with the reaction:

ZrOCl2 YCl3 NH4OH Zr(OH)4
Y(OH)3 NH4Cl
After simultaneous deposition of the components, the deposit is isolated by filtration and
annealed at 1000 C. Some characteristics of
commercial ZrO2-Y2O3 powders produced by
the hydroxide method and the phase compositions of the powders after annealing under various conditions are presented in Table 8.3. After
agglomeration, the powders feature a particle
size of 30 to 80 lm. Unlike the powders produced by mechanical mixing, the powders produced by simultaneous deposition, as well as the
pellets made of them, have more homogeneous
phase compositions.
To fabricate ceramic pellets from ZrO2-Y2O3
powders for their further evaporation, different
techniques can be used, for example, the semidry molding technique (Ref 12). The main feaTable 8.2 Changes in phase composition along
side face of ZrO2-8%Y2O3 ceramic pellet with dimensions of 50 50 50 mm made of mechanical mixture of ZrO2 and Y2O3 powders
Phase constituents(a), vol%
Surface zone
End
Center
End
21
12
20
10
15
60
88
60
65
75
19
...
20
25
10
(a) M, monoclinic; T, tetragonal; C, cubic
tures of the ceramic pellet fabrication technique

are as follows: the ZrO2-Y2O3 powders produced by the hydroxide process are screened
and, if necessary, mixed with certain oxide additions in the special equipment designed for this
purpose. The prepared powders are moistened
with a 4% solution of polyvinyl alcohol, and,
after becoming granulated, they are kept in a
closed vessel for 24 h. After being conventionally molded in special dies to form pellets of the
specified size, they are annealed in a furnace at
1350 oC for 10 h. The furnace heating-cooling
rates should be below 100 C/h.
On heat treatment, noticeable pellet shrinkage
may occur. It depends on the granulometric
compositions of the powders in use as well as
their molding and firing conditions. The results
of research into the effect of dimension ratio on
cylindrical pellet densities before and after firing
are presented in Table 8.4. The pellets for evaporation were produced using the semidry technique.
Firing zirconium-dioxide-base semiproduct
pellets for EB evaporation does not bring about
glassy phase formation, and the process requires
a long time and high temperatures. The firing
process may be divided into several stages: preheating, with a preset rate of temperature rise;
firing at a preset temperature for a preset period;
and cooling down. At the preheating stage, the
rate of temperature rise is chosen depending on
the parameters of the process under way, such
as rates of moisture and plasticizer removal and
product hardening. The rate of temperature rise
and drop is controlled, to avoid thermal stresses
causing cracking and straining. The maximum
firing temperature and holding time depend
mainly on oxide compositions and their particle
Table 8.3 Chemical and phase compositions (vol%) of ceramic powders fabricated by simultaneous
deposition from zirconium and yttrium hydroxides
Annealed(a)
at 1350 C, 10 h
No annealing(a)
Annealed(a)
at 1700 C, 10 h
Grade
Specific
surface, cm2/g
Average grain
diameter, mm
ZrO2
Y2O3
M, %
T, %
C, %
M, %
T, %
C, %
M, %
T, %
C, %
YSZ-3
YSZ-8
YSZ-12
YSZ-15
1996
2007
2265
...
5.83
5.30
4.73
...
Base
Base
Base
Base
34
810
1112
1516
40
13
0
0
69
74
20
0
0
13
80
100
...
13
0
0
...
74
22
0
...
13
78
100
...
6
0
0
...
74
15
0
...
20
85
100
Table 8.4
Effect of molding and annealing conditions on ZrO2-8%Y2O3 pellet density

Molding
force, kg
Cylinder height before/

after annealing, mm
Density before/
after annealing, g/cm3
Porosity
after annealing, %
YSZ-8
10,000
YSZ-8
20,000
22/18
45/38
25/21
43/36
2.12/3.59
2.07/3.29
2.53/4.18
2.45/4.05
36
42
25
28
Composition(a)
(a) Compositions are additionally doped with 5% NiO.
sizes. Firing temperature is chosen with allowance for completion of the sintering process for
powder particles within acceptable time intervals.
At firing, the access of air to the pellets is
limited. Under conditions of air deficit, polyvinyl alcohol decomposes, forming more simple
compounds and releasing elemental carbon that
disperses in the pellet in the form of the finest
coke particles. The carbon content of as-produced pellets may be as high as 0.02 to 0.05%.
Its presence can affect the processes of ceramics
evaporation.
The stability of ZrO2-Y2O3 initial powder
properties and the process of its manufacturing
can guarantee stability of as-produced pellet dimensions. The properties of the finished products depend not only on chemical compositions
Table 8.5 ZrO2-8%Y2O3 ceramic pellet phase

composition
Phase composition(a), vol%
Lot No.
1
2
3
4
5
6
7
30
30
40
40
40
20
10
30
30
30
30
30
25
25
40
40
30
30
30
55
65
of oxide powders but also on their phase compositions, grain sizes, and manufacturing processes. That is why an important component of
the quality-control system is a procedure of initial powder control on delivery from the suppliers. On compacting semiproduct pellets, their
visual inspection is carried out to detect outer
pellet defects. Their density and geometry
should also be controlled. After firing, the pellet
quality is controlled by visual inspection and
measurements of their shrinkage and density.
Their phase and chemical compositions are also
analyzed. Phase composition of the pellets depends on the chemical compositions of their initial powder components, the firing temperatures,
and the cooling rates. Data on phase compositions for several commercial batches of ZrO28%Y2O3-5%NiO pellets are presented in Table
8.5. Firing conditions were 1350 C for 5 h. The
presence of monoclinic phase is caused by slow
cooling of the pellets in the furnace.
Main Features of TBC Deposition

Technique
There are basic differences between the deposition techniques for metallic and ceramic
overlay coatings. Deposition of metallic coatings is not a finishing procedure in forming their
structure. Subsequent mechanical, heat, or
chemical treatments can cause noticeable
changes in their structures, and, in some cases,

they may change chemical composition of the
coating as well. These treatments can favor the
healing of defects formed in the case of violation
of the coating deposition technique parameters.
As for ceramic coatings, their deposition technique is the only factor that sets a coating structure pattern and its adhesive strength on the bond
coat surface. These characteristics cannot be
changed by subsequent treatments. That is why
the technique of ceramics deposition should feature consistency. Malfunction within a technological cycle results in spoilage. The only way
to reclaim this spoilage is to remove the deposited ceramic layer and repeat its deposition. This
is not an easy task for finished blades arriving
for TBC deposition.
Thermal barrier coating deposition technique
includes the following main stages: bond coat
deposition, preparation of a bond coat surface,
ceramic layer deposition, and coating inspection. The technological operations of a TBC
bond coat deposition are similar to the technological operations of heat resistant coating deposition discussed previously. The peculiarities
of the ceramic layer deposition technique are
mainly concerned with the preparation of the
heat resistant bond coat surface for ceramics deposition as well as with ceramics evaporation
and condensation conditions.
The main characteristics governing a TBC
service life are ceramic layer structure and adhesive strength at the ceramic layer/metallic
bond coat interface. The ceramic layer structure
depends on ceramics evaporation and condensation conditions, while adhesive strength at the
ceramics/bond coat interface is affected by different process parameters. It depends on such
factors as methods of bond coat surface preparation, blade heating method and temperature,
conditions of its revolution past evaporator and
heating systems, ceramics evaporation rates, and
the quality of vacuum in the unit. Insufficient
initial adhesive strength at the ceramic layer/
bond coat interface may be caused by various
optimal parameter violations of the designed
technique for ceramic layer deposition, especially at the early stage of ceramics vapor condensation.
Bond Coat Surface Preparation. The quality of bond coat surface preparation is crucial
for physical-chemical processes of bond coat/
ceramic layer interaction, adhesive strength at
this interface, its stability at high temperatures,
and ultimately, for a TBC service life. There are
two basic approaches to bond coat preparation

for ceramic layer deposition:
Clearing oxides from the bond coat surface

and finishing it to the surface roughness of
0.32 to 0.80 lm
Preoxidizing the bond coat surface to form
Al2O3 on it
The first finishing variant is a conventional
process that is used for the surfaces to be coated.
As a rule, it includes any abrasive machining of
the surface, such as vibrogrinding, vibropolishing, and so on. For bond coats deposited by EB
technique, a compacting procedure of microball
peening is required. Bond coat surface hardening that occurs at vibroprocesses and microball
peening facilitates further activation and development of diffusion processes between the metallic bond coat and the ceramic layer.
When selecting the machining technique and
tolerable surface roughness of the bond coat surface, it is necessary to take into account the requirement of the maximum heat resistance. The
surfaces with low roughness feature the best heat
resistance. It is advantageous that the surface
roughness (R) of the bond coat surface prepared
for ceramics deposition should not exceed 0.80
lm. Sharp microprojections on the surface become readily oxidized and cause the spallation
of the ceramic layer. As experience shows, bond
coat surface roughness affects coating texture.
X-ray diffraction analysis of crystallite texture
has shown that, on condensing the ceramic layer
on the surface with R 0.65 lm, the grainboundary angle for the crystallites oriented in
[011] direction is not more than 1 to 2. The
ceramic layer deposited on the surface with the
surface roughness of 1.25 lm has columnar
structure, with a grain-boundary angle of 10. It
contains crystallites oriented in the directions of
[311] and [111]. This structure is more susceptible to erosion.
Special attention should be paid to the surface
quality. In view of earlier publications (Ref 13),
glass particles left on the surface after peening
with glass microballs may form phases, causing
premature spallation of the ceramic layer. Similar results are reported (Ref 14) that throw light
on the presence of silicon on the surfaces processed with the abrasive material, and the presence of iron on the surfaces peened with steel
microballs.
To prepare a bond coat surface that is clean
of oxides, the following technique can be used.
Immediately before the ceramic layer condensation onset in the EB unit, a thin metallic layer
(1 to 5 lm thick), with the composition identical
to that of a bond coat or close to it, is deposited
for 1 to 2 min. A thin metallic layer with a juvenile surface free from oxides and dirt facilitates the formation of a strong diffusion bond
between the ceramics and the metal.
The second variant of bond coat surface preparation includes its oxidation before ceramic
layer deposition (Ref 15). Bond coat preoxidizing should guarantee the formation of a thin Al2O3 layer (1 to 2 lm thick) on its surface. It
should ensure high adhesive strength of a bond
coat/ceramics interface and its long service life
at elevated temperatures. Preoxidizing may be
achieved by different methods. It may be carried
out either in special units or in an EB unit before
ceramics evaporation. The crucial feature of the
process is a need for preoxidizing under the conditions, which guarantees the formation of
Al2O3 with the most perfect crystal structure
available over the whole blade surface. To meet
this requirement, the following factors should be
controlled: thorough cleaning of foreign particles, such as SiO2, from the surface before the
process; the specified aluminum content on the
surface and high homogeneity of bond coat surface composition; and a special pattern of supplying the surface with oxygen.
It has been established that one can find a variant of surface preparation procedure combining
the two methods just discussed (Ref 16). A tablet
made of a powder mixture, such as 13%Al7%Al2O3-80%(ZrO2-7%Y2O3), or of mixtures
of more complex compositions (Ref 17) is
placed on the top of the ZrO2-Y2O3 pellet to be
evaporated.
On EB heating, the processes of melting and
continuous evaporation, first of the tablet and
then of the ceramic pellet itself, take place. Because the partial vapor pressures of the previously mentioned mixture components are different, selective evaporation of the molten tablet
elements occurs. Aluminum is the first to evaporate. It is followed by Al2O3 and finally by
ZrO2. The corresponding composition and structure gradients through the coating thickness accompany the deposition of such a vapor flow on
the substrate. It is obvious that variation of tablet
composition and thickness may result in modification of the composition, structure, and thickness of the coating deposited.
At the metal substrate temperature of 850 to
1000 C, which is higher than the aluminum
melting temperature, condensation proceeds according to the vapor-liquid-solid phase pattern.

Melting of the thin bond coat surface layer took
place at condensation onset. Then, (nickel, cobalt) aluminum intermetallide-base structure
was forming. Those deposition conditions made
it possible to achieve high adhesion at the coating-substrate interface. It allows less stringent
requirements to the initial bond coat surface
roughness. The previously mentioned features
of Al-Al2O3-ZrO2 multicomponent mixture
evaporation and condensation open up possibilities of bond coat and ceramic layer deposition
within one technological cycle, eliminating intermediate heat treatment and mechanical finishing of the bond coat.
The main evidence of the efficiency of the
previously mentioned bond coat surface preparation technique is ceramic layer service lifetime
on a blade surface. Comparative analysis of a
large body of test results for TBCs after various
bond coat surface preparation processes on the
basis of this characteristic is a difficult task, because a service-life characteristic depends on
many other factors. On designing a bond coat
preparation technique, one should first take into
account the requirements of a TBC service life.
In addition, all the procedures of service preparation should be reproducible for each blade in
a big lot. As a rule, any complication of a technological process makes it less stable and reproducible. The choice of the surface preparation
technique also depends on bond coat composition and its deposition technique. Mechanical
finishing of a bond coat deposited by diffusion
techniques (such as aluminizing, platinum-aluminum application, and etc.), which make it
relatively brittle, should not entail surface flaking and cracking. For bond coats deposited by
low-pressure plasma spray (LPPS) or electric arc
technique, mechanical finishing that guarantees
the required surface roughness is a mandatory
procedure. For bond coats deposited by EB technique, the mandatory procedure is surface compacting. It is also noteworthy that heat treatment
used traditionally at different stages of bond
coating preparation affects its surface composition. Heat treatment in vacuum at temperatures
above 1000 C results in reducing chromium
content on the surface, because of its evaporation, and involves the increase in aluminum content (Table 8.6) (Ref 14). This effect can also be
purposely used for improving surface homogeneity before its preoxidizing.
Table 8.6
Effect of technological factors on element contents of bond coat surface

Element contents of the surface, wt%
Bond coat treatment

As-deposited
Annealing at 1050 oC for 2 h
Finishing with steel microballs
Annealing at 1050 oC for 2 h
Abrasive finishing
Ni
Co
Cr
Al
48.2
58.3
58.4
60.9
60.4
19.4
14.2
14.7
13.2
13.1
20.5
8.5
9.1
6.7
6.7
11.2
17.2
15.5
17.2
16.8
0.08
0.11
0.09
0.11
0.09
Source: Ref 14
Pellet Evaporation and Ceramic Layer Formation. There are two variants of the technique
for ceramics evaporation and condensation. In
the first variant of the technological process,
some oxygen is supplied into the evaporator
zone that is enough for compensation of its loss
in ZrO2-Y2O3 dissociation. It results in forming
a white ZrO2-Y2O3 ceramic layer of stoichiometric composition. As a rule, this process is
used for ceramic layer deposition on the preoxidized bond coat.
In the second variant of the technique, the effects of partial reduction of the ZrO2-Y2O3 ceramics, which takes place at its evaporation, are
used for developing diffusion processes of interaction between the condensing ceramic layer
and the metal bond coat. It guarantees high adhesive strength of the ceramics/bond coat bonding and long service life of the ceramic layer.
For this variant of technique, the bond coat surface should be free from oxides. The results of
the application of this variant of the technique
are discussed subsequently.
Ceramics evaporation conditions, vapor density, and its condensation conditions are set by
the evaporator construction used in the EB unit.
The EB coaters UE described in Chapter 3,
Technological Processes for Deposition of Protective Coatings to Turbine Blades, is equipped
with flat beam guns for material evaporation.
These EB guns form a beam with a cross section
Table 8.7
ciation
Products and degree of ZrO2 dissoTemperature, K
Vapor phase composition, vol%
2950
3500
4000
O
O2
Zr
ZrO
ZrO2
Degree of dissociation,
36.60
0.33
0.02
36.95
26.10
0.59
38.42
0.82
0.06
40.00
20.70
0.65
39.67
1.40
0.13
42.00
16.80
0.72
Source: Ref 18
of 100 mm2 on the surface of the evaporated

material. Evaporation occurs on scanning the
surface of the ceramic pellet 70 mm in diameter
with the EB. Its power is within the range of 17
to 25 kW (the voltage is 19 kV; the current is
0.8 to 1.3 A). In this case, the ceramic surface
layer melts and forms the liquid bath, with a
depth of 2 to 5 mm. Under these conditions, the
evaporation rate of ZrO2-8%Y2O3 ceramics is 1
to 2 kg/h, and the rate of condensation on rotating blades is in the range of 0.6 to 2.0 lm/
min.
Zirconium oxide dissociation takes place during ceramic pellet evaporation (Table 8.7). Oxygen that is present in dissociation products is
partially removed from the evaporator zone under vacuum. The degree of ceramics reduction
depends on a large number of technological parameters, such as a bath temperature, an evaporation rate, and vacuum in the unit. An evidence of the partial reduction and the increase
in the number of anionic vacancies is a change
in the ceramics surface color. After cooling, the
color of the solidified bath surface and the layer
5 to 8 mm thick under it is black. It is typical of
zirconia with a lack of oxygen.
Reduction effects may be enhanced by the use
of ceramics condensates doped with NiO that
feature much less thermodynamic potential of
formation (Ref 18, 19). During heating and
evaporation of this system, the reduction of NiO
to metallic nickel takes place. Nickel atoms facilitate metallic bonding between the ceramic
layer and the metallic bond coat and give high
adhesive strength to their interface.
Annealing in air of the ceramic layer solidified after evaporation results in its mass gain,
due to oxygen uptake. Kinetics of mass change
on annealing is the same for both the bath and
the condensates (Fig. 8.3). After 30 min of annealing at 700 C, ceramics gain 70% of the oxygen lost. Its full oxidation takes place in 2 h.
Total loss of oxygen by the ceramics during
evaporation, calculated from mass gain, was 3.0
to 3.5%. At oxidizing annealing, no phase composition change of ceramics has been detected.
As the oxygen content increases to its initial
level, the ceramics whiten.
The phase composition of the black ceramics
differs from the phase composition of the ceramic pellet used for evaporation, as mentioned
previously. The initial phase composition depends on the ZrO2-8%Y2O3 powder used and
the procedure of pellets manufacturing. After
annealing at 1350 oC and retarded furnace cooling, the ceramic pellet contains 30% M
phase, 30% T phase, and 40% C phase.
The phase composition of the black bath after
evaporation includes only cubical and tetragonal
phases (Fig. 8.4). The research into chemical
compositions of the ZrO2-8%Y2O3 bath and the
Fig. 8.3
Kinetics of ZrO2-8%Y2O3 black ceramics oxidation

at 700 and 1000 C
Fig. 8.4
X-ray patterns of (1) ZrO2-8%Y2O3 ceramic pellet

and of (2) bath surface after evaporation
ceramics condensated from it has shown that

both oxides evaporate cooperatively. The difference between the Y2O3 contents of the condensate and the bath is not more than 0.5%.
The condensed ceramic layer also shows different hues. As microstructure examination has
shown, a change in a color of the ceramic layer
does not involve a change in its phase composition (Table 8.8). The degree of ceramics reduction and its color also depend on the substrate surface preheating temperature and its
position relative to the evaporator. As published
data have shown (Ref 20), the degree of ceramic
layer blackening increases with the increase in
the EB power and the evaporation rate. The increase in the angle of vapor condensation has
the same effect (Table 8.9). Temperature conditions of blade heating at ceramic layer deposition affect its properties to a larger extent than
at metallic coating deposition. While texture and
porosity in the metallic coatings deposited by
EB technique result in degradation of their mechanical and protective properties, the presence
of porosity and columnar texture that is perpendicular to the substrate surface facilitates ceramics fragmentation on generation of thermal
stresses and thus ensures its high thermal stability. On the contrary, dense ceramics provoke
generation of high thermal stresses, entailing ceramic layer spallation.
Ceramic layer structure depends on the temperature of the surface on which vapor condenses and on blade rotation conditions in a vapor flow. According to the scheme presented in
Ref 21 columnar structure of ZrO2 layers forms
on substrate heating within a temperature range
of 800 to 1350 C. The blade heating temperature used on ceramic layer deposition is conventionally in the range of 850 to 950 C. However,
measured substrate temperatures can be substantially different from the temperature of the thin
surface layer, where vapor condensation and
evaporation heat release take place. This real
temperature depends on the conditions of blade
surface heating, condensation rates, the parameters of blade rotation/revolution around the
evaporators, the space between the blades and
the evaporator, ceramic layer thermal conductivity, and optical characteristics of the ceramic
layer.
The structure and phase composition of the
condensed ceramics experiences considerable
changes, depending on the substrate heating
temperature. X-ray diffraction analysis has
shown that the decrease of T-phase content
Table 8.8
Structural features of ZrO2-8%Y2O3 ceramic layers of different colors
Lattice parameters, A
Ceramics color
a (C)
a (T)
c (T)
c (T)/a (T)
Phase composition
Light gray
Gray
Dark gray
Brown
5.115
5.115
5.119
5.115
5.136
5.141
5.045
5.159
5.163
5.163
5.169
5.156
1.005
1.004
1.024
0.9994
TC
TC
TC
TC
Note: Cubic lattice was determined by reference to line (620); tetragonal lattice was determined by reference to lines (004) and (206).
and the increase of the C-phase content with

the temperature increase are general trends (Fig.
8.5). At the same time, the texture gets less pronounced, and at substrate temperatures above
1050 oC, it is no longer observed. These effects
are likely to result from variation of temperature
gradients at the vapor/condensation surface interface.
The use of electron beams for surface heating
and temperature control promotes intensive heat
Table 8.9 Effect of angle of vapor incidence on
substrate on degree of ceramics blackening
Angle of substrate
with vapor flow
direction, degrees
90
7085
4570
045
Condensed
ceramics color
Black with luster
Black without luster
Gradual change in color from
black to white
White
Note: The substrate temperature is 550 oC; the beam power is 13 kW.
Source: Ref 20
X-ray patterns of EB-deposited ZrO2-8%Y2O3 ceramics on substrates heated up to different temperatures.

(1) 400 C. (2) 600 C. (3) 800 C. (4) 1000 C
Fig. 8.5
release in the thin ceramic layer. That is why

such a heating technique is a method of surface
treatment that governs composition, structure,
and properties of a thin interlayer between the
ceramics and the bond coat (Ref 19). Calculations have shown that at the surface heated with
EB having the power of 0.1 kW/cm2, the surface
temperature increases by 300 to 500 C during
the first 10 to 20 s of ceramics condensation and
a ceramic layer 0.5 to 1.2 lm thick forms. An
increase in the temperature at the onset of ceramics condensation results in intensification of
the diffusion process between the ceramic layer
and the metallic bond coat. The analysis of element contents in the ceramic layer adjacent to
the bond coat (Table 8.10) has revealed that the
bond coat elements diffuse into the ceramic
layer to the depth of 10 lm. This results in the
formation of a special interlayer between zirconia and the metallic bond coat that guarantees
high strength of bonding between the layers with
different physical-chemical properties. The formation of the interaction zone at the ceramicsmetal interface is the feature of the previously
mentioned variant of EB technique. It ensures
an adhesive strength of more than 70 MPa at the
ceramic layer/bond coat interface.
If the working parameters of blades heated
with an EB do not change during the period of
ceramic layer formation, it results in formation
of thick, dense zones between the ceramics and
the metallic bond coat. These layers contain an
M phase that causes stress rise and ceramic
layer spalling immediately after their deposition.
The fragment of the blade with partial spallation
of the ceramic layer on check annealing at 700
o
C is shown in Fig. 8.6. The zone where ceramics spallation is detected has a lustrous black
color. It is located on the blade airfoil pressure
surface on a thin wall, where surface overheating at EB heating is possible. The research has
shown that thick layers (5 lm) of dense, black
ceramics formed on that surface area on condensation. It contained bond coat elements,
Table 8.10
Element contents in ceramic surface layer adjacent at ceramics/bond coat interface

Distance from surface of bond coat Ni20Cr12AlY, lm
Element
Zr, %
Y, %
Ni, %
O, %
59.1
5.7
3.4
30.9
54.9
5.1
7.2
32.6
49.9
4.5
13.9
31.0
41.5
3.0
20.8
33.1
40.2
3.9
23.6
30.7
Source: Ref 22
Fig. 8.6
(a) Appearance of blade with spalled ceramic coating after check annealing. (b) Structure of interaction zone between
ceramics and Ni20Cr12AlY bond coat. Reflected electrons, 800
mainly nickel. On structure analysis, the monoclinic phase that was not detected on the other
blade surface areas was found in the phase composition of the dense zone (Table 8.11). The
presence of the M phase is probably due to
high nickel content that degrades the degree of
Table 8.11 Properties of dense zone at ceramics/bond coat interface and tendency to ceramics
spalling
Failure after
annealing
at 750 oC, 2 h
Not detected
Detected
Phase composition
Zone
thickness, lm
Outer layer
Dense zone
2
8
C15T
C50T
CT
CM
ZrO2 stabilization. At the same time, dense interlayers 3 lm thick improve adhesive
strength of the ceramic layer. They contribute to
its service life. To prevent the formation of thick
interlayers, it is advantageous to reduce the heating electron gun power at the starting moment
of ceramic condensation.
The average ceramics condensation rate is a
convenient characteristic of the technological
process (Ref 23). The increase in a condensation
rate results in a corresponding thickness increase
of the dense ceramics zone that forms on the
bond coat surface at the starting moment of condensation. It also makes ceramic layers less porous and much harder. An average crystallite diameter increases from 1.5 to 2.0 lm at a rate of
1 lm/min to 3 to 4 lm at a rate of 2 lm/

min (Fig. 8.7a, b). Structural changes in the ceramic layer at different condensation rates have
an effect on its service life. Cyclic testing is carried out, using the samples 7 mm in diameter,
under the following conditions: heating to 1100
C for 4 min, holding at this temperature for 20
min, cooling to room temperature for 6 min. The
ceramic layer deposited at the condensation rate
of 1.2 to 2.0 lm/min has shown the longest service life (Fig. 8.7c).
Ceramic layer structure is affected by the
composition of the ceramics evaporated. Surface
structures of the ZrO2-8%Y2O3 and ZrO28%Y2O3-5%NiO ceramic layers are shown in
Fig. 8.8. Both layers are formed under the same
working conditions of evaporator guns, heating
guns, and at the same specified substrate temperatures. The ZrO2-8%Y2O3-5%NiO crystal-
Effect of condensation rate on characteristics of ceramic layer. Thickness of dense zone. (a) Thickness.
(b) Porosity and hardness. (c) Number of cycles to spallation.
Source: Ref 23
Fig. 8.7
lites have diameters that are less than those of

the nickel-free layers by a factor of 2 to 3. The
use of compositions with higher Y2O3 contents
has a similar effect. The crystallites of the ZrO220%Y2O3 ceramic layer formed under the same
conditions have diameters 2 to 3 times as large
as the diameters of the ZrO2-8%Y2O3 ceramic
layer crystallites. These effects are caused by the
ceramics composition effect on its melting temperature and by condensation conditions on the
bond coat surface.
Blade surface temperatures and condensation
rates affect the thermal stress level in the ceramic layer. This depends on the temperature of
the surface on which condensation occurs and
on the difference between TECs in the temperature range from condensation temperature to
room temperature. Stresses in blade airfoil pressure and suction surface layers depend on the
actual heating conditions on these surfaces (Table 8.12).
The need for maintaining the parameters of
condensation and ceramic layer crystallite
growth at a steady level requires a certain layout
of relative positions of the blades, the vapor generator, and the EB guns for blade heating. As is
shown in Ref 24, substrate rotation speeds have
the same effect as temperatures. This behavior
is caused by the effect of rotation on the time of
growing crystal presence in the zones with different vapor density. The higher the temperature
and rotation rate, the larger the diameter of an
individual crystallite of the condensing ceramics
(Fig. 8.9). Using different rotation speeds, structural characteristics of the ceramic layer can be
governed.
From the experience of ceramic layer deposition and taking into consideration intricate
blade profiles and a need for simultaneous coating deposition on several blades, the best results
can be achieved by combining blade revolution
around the evaporator and rotation about their
axes. An illustration of blade arrangement and
their revolution/rotation is given in Fig. 8.10.
The fixture in use revolves in the vapor flow
with the speed of 12 rpm. At each fixture revolution the blades additionally revolve once
around the fixture axis. The choice of blade rotation conditions depends also on the requirements to the ceramic layer thickness and its
spread over the blade surface.
A ceramic layer fracture pattern at the temperature of 850 to 950 C is shown in Fig. 8.11.
The ceramic structure features the pronounced
texture of growth perpendicular to the surface.
Fig. 8.8
Ceramic layer surfaces. (a) ZrO2-8%Y2O3. (b) ZrO2-8%Y2O3-5%NiO. 5000
Table 8.12 Thermal stresses in ZrO2-8%Y2O3 ceramic layer on turbine blades. Ceramic layer is deposited on blades heated to 850 oC
Suction surface
Vapor deposition rate
1 lm/min
2 lm/min
4 lm/min
Pressure surface
Phase composition
Texture
Stress, MPa
Phase composition
Texture
Stress, MPa
FT
FT
F
[100]
[100][111]
[100]
30
130
87
FT
F
T
[100]
[100]
[100]
76
113
152
Fig. 8.9
Schematic representation of substrate temperature and rotational speed effect on columnar microstructure evolution of EB
ceramic layer. Tm, melting temperature. Source: Ref 24
Some individual ceramics crystals are preferably

oriented in a [100] direction. Their diameters are
in the range of 0.6 to 1.2 lm. They do not vary
much along the full crystal lengths.
Ceramic layer crystallites should have high
cohesive strength and withstand an attack of a
high-temperature gas flow. That is why the ceramic evaporation process feature is a requirement to its continuity. Unlike metallic bond coat
deposition, in which no process interruption is
Scheme of cassette for blade fixation and revolution. (1) Cassette. (2) Blades. (3) Ceramics pellets.
(4) EB gun of evaporator. (5) EB gun for blade heating
Fig. 8.10
harmful for the coating quality, any interruption

of ceramic layer deposition forms an additional
boundary in the ceramics. The strength of this
boundary is much lower than the crystallite
strength. Thus, under these conditions, the ceramic layer will never meet the requirements of
its properties. In the case of any pause in ceramics evaporation, all the lot of blades being coated
are rejected and sent to ceramic layer removal
procedure, followed by its redeposition.
When ceramic layer deposition is carried to
its completion, the blades are removed from the
unit and passed to heat treatment. This includes
two steps: annealing in a vacuum furnace at
1000 to 1050 C for 2 h, followed by annealing
in air at 750 C for 4 to 6 h. The first annealing
improves the properties of TBCs formed by the
technique that excludes bond coat preoxidizing.
Ceramic layer oxidation in air at 750 C corrects
its composition toward the stoichiometric composition. Its color changes from dark gray to
white. Two-step annealing does not change ceramic layer structure and phase composition.
Check operations in the TBC quality control
include a visual inspection to guarantee that its
surface is free from ceramics droplets; measurements of ceramic layer thickness in the specified
blade zones; and a bend test of a flat check sam-
Fig. 8.11
Fractures of ZrO2-8%Y2O3 ceramic layers. (a) 850. (b) 3000. Substrate temperature, 850 to 950 C; rotational speed,
12 rpm
ple on the radius of 3 mm to assess its adhesive

strength. On its bending to the angle of 90, ceramic layer spallation is prohibitive (Fig. 8.12).
Some cracking of the ceramic layer is allowed.
Then, the check samples 10 mm in diameter are
subjected to oxidizing annealing at 1100 C. No
ceramic layer spallation should occur within the

check annealing period.
Thermophysical Properties of
Condensed Ceramics
Thermal barrier coating application efficiency
depends on ceramic layer thermal conductivity,
which determines the cooled blade temperature
drop and corresponding increase in its service
life. To measure thermal conductivity of a TBC
ceramic layer, a laser flash method is used (Ref
2526). The method is based on irradiating the
surface of a flat sample surface with an energy
pulse, followed by recording a temperature rise
on its backside (Fig. 8.13). Thermal diffusivity
and heat capacity are determined experimentally
using the pulse method. Then, thermal conductivity (k) is calculated from these characteristics:
k a q Cp
Samples after bend testing: pale sample with

Ni20Cr12AlY bond coat, dark sample with ZrO28%Y2O3 /Ni20Cr12AlY (140/80 lm)
Fig. 8.12
in W/m K, where a is thermal diffusivity

(cm2 /s); q is density (g/cm3); Cp is heat capacity
(J/g K).
Measuring each thermophysical characteristic
is an independent task. The most developed
method is that of thermal diffusivity calculation,
because the main formula for thermal diffusivity
includes only one experimentally measured parameter. It is a period for the temperature to
reach half of its maximum level:

a 0.1388
d2
s1/2
in cm2 /s, where d is sample thickness, and s1 / 2

is the time required for the temperature of the
sample backside to reach the level equal to onehalf of the maximum temperature. Coefficient
0.1388 corresponds to an ideal case when the
following conditions are met: instantaneous and
uniform heat pulse, heat pulse absorption in a
thin surface layer, and no heat losses. For experimental thermal diffusivity determination,
one should know neither absolute temperatures
nor parameters of a heat flow affecting a sample.
Measuring heat capacity by the flash technique,
especially for coated samples, is a much more
complicated task. Analysis of thermal diffusivity
and thermal conductivity of ceramic coatings are
discussed elsewhere (Ref 2830).
For thermophysical studies of ceramics condensates, the TC-3000H unit manufactured by
the Sinku-Riko Company was used. A ruby laser
with a wave length of 6.943 lm was used as an
energy source, and as a temperature pickup on
the backside of the sample, either a thermocouple (Pt-PtRo) or an infrared sensor was used
(Ref 31). The tested sample is essentially a flat
disc 10 mm in diameter and 0.8 to 2 mm thick.
When thermal diffusivity is studied in this
unit, two types of experimental errors are possible. The first type of errors results from some
Fig. 8.13
lack of information on the parameter values used

in the design formulas. They are due to the available accuracy of sample thickness and time of
s1 / 2 measurements, exactness of the maximum
temperature rise assessment, and of catching the
moment of the sample irradiation start. These
errors are covered in detail in Ref 32 and 33. On
the basis of the results reported in the literature,
one can deduce that, with the modern data collection systems used, the contribution of this
type error does not exceed 0.5%.
The second type of errors is due to the difference between the experimental conditions and
assumptions in the mathematical model used for
calculating thermal diffusivity and heat capacity.
These errors are related to the finite pulse duration and its spatial inhomogeneity, to heat
losses (due to irradiation, mainly), and to violation of pulse absorption conditions in the thin
surface layer. These errors may be avoided by
using certain corrections (Ref 3436). For the
TC-3000H unit, pulse duration and spatial inhomogeneity errors determined according to the
Sinku-Riko Company recommendations are unessential (less than 1%).
Heat losses in the experiment result in a quick
temperature rise to its maximum and then a
sharply defined smooth temperature decrease.
The main cause that gives rise to measurement
errors is radiation heat exchange, whose effect
rises simultaneously with a temperature rise.
The errors caused by radiation may account for
30%. To meet the requirements of pulse absorp-
Scheme of apparatus for measuring thermophysical properties by flash method. Q, heat flow; DT, change in temperature
of sample backside due to its frontside irradiation; s1 / 2, time to temperature rise to DT/2
tion in the thin surface layer, the ceramic samples, which are partially transparent, were coated
with a thin layer (10 to 12 lm) of the NiAl intermetallic compound (20% Al). This layer ensured steady surface optical parameters of the
samples as well.
To avoid the errors of heat capacity measuring, which are possible in the flash method, the
HT-1000 calorimeter, manufactured by the Setaram Company, was used to determine heat capacity of the ceramic powders and condensates
within the temperature range of 20 to 800 oC.
The maximum relative errors of test measurements for thermophysical properties account for
7% for heat capacity and 4% for thermal diffusivity.
To study thermophysical properties of the
condensates, three systems have been chosen
(Ref 22). These are ZrO2-8%Y2O3; ZrO28%Y2O3-5%NiO, and ZrO2-20%CeO2. A NiO
dopant makes a ceramic layer optically opaque.
The ceramics with a CeO2 dopant may be considered as promising for service in media containing sulfur compounds. The initial ZrO28%Y2O3 powders were produced by the
hydroxide method. The ZrO2-20%CeO2 composition was produced by mechanical powder
mixing, followed by annealing at 1350 C for 10
h. The ceramic pellets shaped as cylinders 50
mm high and 70 mm in diameter, produced by
the previously mentioned method, were used for
EB evaporation.
Ceramics condensates were deposited on the
substrates of the JS6U superalloy in the ES 1/
3/60 (Leybold) unit. The substrate-measured
temperatures were 800 to 900 C; the condensation rate was 8 to 10 lm/min; and the substrates did not rotate in the vapor flow. After
deposition, the ceramic layers were separated
from the substrates and annealed in air at 700 C
for 2 h. The characteristics of the condensates
deposited are presented in Table 8.13.
Heat capacity versus temperature, plotted using the experimental data measured in the HT1000 calorimeter, is presented in Fig. 8.14. The
Table 8.13
ZrO2-8%Y2O3 and ZrO2-8%Y2O3-5%NiO powders and condensates deposited from them were
used for measurements. Before measurements,
the powders were annealed at 1350 C for 10 h.
The main trend for all the compositions is a
gradual heat capacity increase with a temperature rise. At 800 C, heat capacity is 0.6 J/g
K. As far as heat capacity is concerned, condensate behavior does not differ from that of the
initial powders (Fig. 8.14b). All experimental
heat capacity measurements were made in the
temperature range of 20 to 800 C. To assess
heat capacity values at temperatures above 800
C, extrapolation of the experimental results was
done. It was based on the effect of slight gradual
increase in zirconia heat capacity with the temperature rise reported in Ref 37.
The results of thermal diffusivity measurements and calculation results for ceramics condensate thermal conductivity are given in Table
8.14. It also includes experimental data on heat
capacity and thermal diffusivity. The results of
heat capacity measurements made on the powders were used for calculations, because, in this
case, one can avoid errors of measuring sample
characteristics inevitable at measuring heat capacity by the flash method. Heat losses caused
by radiation effects in the pulse method of thermal diffusivity measuring were calculated using
the method described elsewhere (Ref 33). Correction coefficients for the ZrO2-8%Y2O35%NiO composition are given in Table 8.15.
The values of correction coefficients decrease
with the temperature rise. They do not go above
13%. Thermal conductivity values measured for
the electron beam-physical vapor deposition
(EB-PVD) condensed ZrO2-8%Y2O3 ceramics
are similar to thermal conductivity of the ZrO210%Y2O3 single crystal at room temperature,
which is 3 W/m K (Ref 9).
High thermal conductivity values typical of
the ceramic layers deposited by vapor condensation are attributed to their specific structure.
Ceramic layer crystallites are oriented perpendicularly to a condensation surface. Insignificant
Materials for investigations
Specimen
ZrO2-8%Y2O3
ZrO2-8%Y2O3-5%NiO
ZrO2-20%CeO2
ZrO2-8%Y2O3
Thickness, lm
Method of manufacturing
Density, g/cm3
Color
1000
815
900
1100
EB
EB
EB
APS
5.58
5.42
5.66
5.00
White
Dark gray
Pale pink
White
EB, electron beam; EB-PVD, electron beam-physical vapor deposition; APS, air plasma spray
Fig. 8.14
Table 8.14
Ceramics heat capacity. (a) Powders. 1, ZrO2-8%Y2O3; 2, ZrO2-15%Y2O3; 3, ZrO2-8%Y2O3-5%NiO. (b) Condensates. 1,
ZrO2-8%Y2O3; 2, ZrO2-8%Y2O3-5%NiO
Thermophysical properties of ceramic specimens

Temperature, oC
Specimen
Characteristics
200
400
600
800
1000
1200
ZrO2-8%Y2O3, EB
Cp, J/g K
a 102, cm2/s
a, W/m K
Cp, J/g K
a 102, cm2/s
k, W/m K
Cp, J/g K
a 102, cm2/s
k, W/m K
Cp, J/g K
a 102, cm2/s
k, W/m K
0.40
1.18
2.63
0.56
0.70
2.12
0.52
1.05
3.09
0.40
0.42
0.84
0.54
1.07
3.22
0.52
0.61
1.72
0.58
1.05
3.45
0.54
0.40
1.08
0.55
1.05
3.22
0.56
0.60
1.82
0.58
0.95
3.12
0.55
0.39
1.07
0.60
1.04
3.48
0.61
0.65
2.15
0.58
0.95
3.12
0.60
0.40
1.20
0.62
1.03
3.56
0.66
0.63
2.25
0.6
0.88
2.99
0.62
0.41
1.27
0.70
1.04
4.06
0.67
0.62
2.25
0.6
0.9
3.06
0.70
0.44
1.54
ZrO2-8%Y2O3-5%NiO, EB
ZrO2-20%CeO2, EB
ZrO2-8%Y2O3, APS
Note: Corrections for heat irradiation losses excluded. Source: Ref 22
porosity (10 to 15%) brought about by substrate

rotation in the vapor flow is oriented in the same
direction. This porosity cannot ensure heat resistance similar to that demonstrated by the ceramic layers formed by APS. In addition, the
increase in the oriented porosity fraction is detrimental for a ceramic layer service life, because
of easier oxygen penetration to the bond coat
surface through the ceramic layer.
Research into thermal diffusivity of solid nonmetal materials has shown that this characteristic
depends on the presence of point defects in the
crystalline lattice of the material and on its strucTable 8.15 Correction factors for irradiation
losses, K a/ao
Temperature (T), oC
700
840
930
1170
K
0.949
0.924
0.920
0.872
tural characteristics (Ref 38). As mentioned previously, structural characteristics of condensed

ceramic layers depend on many technological
factors of their formation. Peculiarities of technological processes and corresponding structural
difference between ceramics layers deposited
under different conditions have more significant
effect on their thermal diffusivity than the errors
due to salient features of measuring techniques.
This is likely to be an explanation of the spread
in thermal diffusivity and thermal conductivity
data for TBC-EB reported by different authors.
In this case, structural difference between the
ZrO2-8%Y2O3 and ZrO2-8%Y2O3-5%NiO condensates was characterized by the nontextured
condensate with a NiO dopant and predominant
T phase in its phase composition. In the ZrO28%Y2O3-5%NiO ceramics crystallites, some
precipitation was detected; it may be NiO precipitate. In addition, NiO has an effect on the
ceramic layer structure, thus reducing crystallite
size. It is likely that NiO effect on thermal con-
ductivity is due to both additional vacancies in

the ZrO2 lattice and structural peculiarities of the
ZrO2-8%Y2O3-5%NiO ceramic layer caused by
its deposition technique in use. Thermophysical
properties of a TBC, including a ceramic layer
with a NiO content of 5%, were studied independently in a few laboratories. According to
their data, thermal conductivity of the 80 to 200
lm thick ZrO2-8%Y2O3-5%NiO ceramic layers
was 0.8 to 1.2 W/m K in the temperature
range of room temperature to 1200 C.
To reduce thermal diffusivity of EB-deposited
ceramic layers, some artificially introduced
point defects may be used (Ref 22, 38). For this
purpose, point defects of a crystal lattice can be
used. They include foreign atoms in the lattice
nodes instead of inherent ones (substitutional atoms), vacancies, interstitial atoms, or combinations of these defects. Phonon scattering by the
boundaries may also be considered as an effective way of thermal conductivity reduction. For
example, if a polycrystalline material with the
size of 1 mm and grain size of 1 lm is taken
instead of a single crystal with the size of 1 mm,
its thermal conductivity at low temperatures is
1/1000 of single-crystal thermal conductivity
(Ref 38). Thus, the technique of ceramic layer
deposition may have an impact on its thermal
conductivity.
On doping ZrO2 with cations exhibiting a valence other than zirconium, additional vacancies
form to keep a crystalline cell electrically neutral. They increase heat resistance of the crystal
lattice (Fig. 8.15). As experiments have shown,
NiO dopants make it possible to reduce thermal
Fig. 8.15
conductivity of the electron beam-physical vapor deposition ZrO2-8%Y2O3 ceramic layers.

Doping the ZrO2 crystal lattice with ions having
larger or smaller ion radii (such as ytterbium or
erbium) produces the same effect.
The results of thermal conductivity studies for
different ceramic coatings formed by the EBPVD technique are presented in Ref 39. On the
basis of the studies (Ref 40, 41), a two-zone
model of a ceramic layer is suggested. Thermal
conductivity of a dense, inner ceramic zone that
forms at the starting moment of condensation is
much lower than thermal conductivity of an
outer zone (Fig. 8.16). This effect is attributed
to the presence of numerous boundaries in the
dense zone. Therefore, for reducing thermal
conductivity of EB-TBC ceramic layers, it is advantageous to form thin layers 0.2 to 2.0 lm
thick in the crystallites. Their boundaries ensure
effective phonon scattering. Multilayer structure
may be formed by plasma discharge to vary the
density of the ceramics during deposition. According to the research, the efficiency of thermal
conductivity reduction by means of multilayer
structure may be as high as 37 to 45%.
The investigation also included the analysis
of doping the ZrO2-8%Y2O3 ceramics with erbium, neodymium, gadolinium, ytterbium, and
NiO. These elements reduce ceramic layer transparency and facilitate formation of point defects
in the ZrO2 crystal lattice. The efficiency of thermal conductivity reduction with these dopants
may be as high as 52%. At room temperature,
thermal conductivity values for ceramic layers
250 lm thick, with neodymium and ytterbium
(a) Pattern for ZrO2 lattice cell and (b) ZrO2 lattice cell with Ni2 cation
dopant contents of 4 mol% each, are 0.86 and

0.89 W/m K, respectively. The results of thermal conductivity measurements made on the ceramic layers deposited by the technique used in
the research led to the conclusion that the efficiency of the NiO dopant is not as high as it was
earlier supposed.
For efficient temperature control of surfaces
with deposited ceramic layers, their emissivity
factor is also required. The integral emissivity
factor of the condensed ceramic layers was determined in the unit, using a black body model.
All measurements were made in a quiescent air
atmosphere under isothermal conditions (Ref
22). Measurement error did not go above 5%.
Test results are presented in Table 8.16.
Some insignificant difference in emissivity
factors for the Ni20Cr12AlY coating variants
with different surface preparation is due to the
fact that the coating contains up to 60% NiAl.
For intermetallides, the effect of surface roughness on the emission is insignificant. The ZrO2Y2O3 ceramic layer deposition changes optical
characteristics of the surface. For ceramic layers
10 lm thick or thinner, the increase in a test
temperature involves slight decrease of emissivity factor. This results from partial transparency
of the ceramic layer in the infrared spectrum
range and from the substrate effect. As ceramics
layer thickness increases to 100 lm, emissivity
factor also increases.
On calculating TBC thermal conductivity,
knowledge of thermophysical characteristics of
metal bond coat materials is required. These are
presented in Table 8.17. Measurements were
made on condensates formed by EB evaporation
of the alloys in question.
Thermal expansion coefficients were determined using the technique described in Chapter
7, The Effect of Protective Coatings on the Mechanical Properties of Superalloys. Measurements were made on the samples produced by
compacting corresponding powders. After compaction, the samples were fired at 1700 C for
10 h. Test results are presented in Table 8.18.
Thermal expansion coefficient measurements
for the ZrO2-8%Y2O3 ceramics in the temperature range of 400 to 800 C cover the T M
phase transformation. The results of TEC measurements made on ceramics condensates are
shown in Fig. 8.17. Oxidation of the initial black
ceramics during the first heating results in its
volume increase. On repeated measurement,
TEC values get stable and do not vary much.
TBC Ceramic Layer Durability

The moment of ceramic layer spallation from
a cooled blade surface may be considered as a
TBC failure, because after spallation, its main
duty of protecting blade walls against high-temperature attack is no longer fulfilled. Thermal
barrier coating ceramic layer cracking and formation of different-sized fragments are evidence
of its response to blade surface stresses and
strains but not evidence of the lack of TBC protective properties.
Main Causes of TBC Ceramic Layer Damage. A TBC ceramic layer failure on its deposition by an optimized and reproducible technique may be caused by the following different
factors:
Scheme of an EB-TBC and corresponding variations of thermal conductivity of a two-layer model.

Source: Ref 39
Fig. 8.16
Ceramic layer/metallic bond coat boundary

oxidation, formation, and growth of thermally
grown oxide (TGO) on this boundary
Changes in ceramic layer phase composition;
recrystallization and sintering
Hot corrosion of the ceramic layer
Erosion of the ceramic layer
Deformation of a bond coat and a blade under
the action of thermal stresses and centrifugal
force
Each of these processes can occur separately and

in combination with other processes.
According to a large body of test data, the
weak point of TBCs is a boundary between the
ceramic layer and the heat-resistant bond coat
(Ref 42, 43). During high-temperature testing or
under service conditions, oxygen from the oxidizing environment (air, combustion products)
penetrates to the ceramic-metal boundary. Its
penetration to a bond coat surface occurs in two
Table 8.16
Coating emissivity
Temperature, oC
Coating
Preparation of bond coat surface
Ni20Cr12AlY
Polishing
Abrasive blasting
Polishing
ZrO2-11%Y2O3
Thickness, lm
20
200
400
600
800
100
100
2
5
10
2
5
10
100
0.23
0.27
0.54
0.77
0.79
0.79
0.86
0.81
0.80
0.25
0.30
0.50
0.62
0.65
0.66
0.70
0.71
0.80
0.27
0.32
0.45
0.48
0.50
0.52
0.60
0.57
0.79
0.29
0.37
0.43
0.46
0.47
0.49
0.60
0.52
0.77
0.35
0.46
0.46
0.47
...
0.52
0.60
0.56
0.76
Abrasive blasting
Source: Ref 22
Table 8.17
Thermophysical properties of JS32 superalloy and condensed alloys for coatings

Density
(q), g/cm3
Temperature
(T ), oC
Heat capacity
(Cp), J/g deg
Thermal diffusivity
(a 102), cm2/s
Thermal conductivity
(k), W/m K
JS32
8.60
Ni20Co20Cr12AlY
7.05
Ni20Cr12AlY
7.60
Co20Cr12AlY
7.10
Ni8Co20Cr12AlY
7.40
Ni20Cr6AlY
7.30
20
400
600
800
1000
1200
20
600
800
1000
1200
20
400
600
800
1000
1200
20
600
800
1000
1200
20
600
800
1000
1200
20
400
600
800
1000
1200
0.40
0.44
0.49
0.55
0.64
0.78
0.50
0.57
0.65
0.71
0.93
0.38
0.40
0.42
0.45
0.53
0.65
0.50
0.62
0.67
0.73
0.90
0.44
0.48
0.54
0.66
0.76
0.39
0.43
0.46
0.50
0.58
0.73
2.05
3.00
3.55
4.00
4.00
4.00
2.90
5.40
5.50
5.55
5.80
5.40
5.80
6.00
5.90
5.60
4.90
3.20
4.58
4.93
5.10
5.60
3.98
4.79
5.38
5.20
5.22
3.30
4.70
5.40
6.10
6.50
7.30
7.05
11.35
14.96
18.92
22.02
16.83
10.22
21.70
25.20
27.78
38.03
15.60
17.63
19.15
20.18
22.56
24.21
11.36
20.16
23.45
26.43
35.78
12.96
17.01
21.50
25.40
29.36
9.40
14.75
18.13
22.27
27.52
38.90
Trademark
Courtesy of Dr. S.V. Zherzdev
Table 8.18
Thermal expansion coefficients for ceramics sintered at 1700 oC for 10 h

Coefficient of thermal expansion ( 106), K1
Ceramics
ZrO2-8%Y2O3
ZrO2-15%Y2O3
ZrO2-7%Y2O3-5%NiO
ZrO2-7%Y2O3-15%NiO
Courtesy of Dr. S.V. Zherzdev
20400
400600
600800
8001000
10001200
11.90
8.50
11.00
11.20
20.30
9.00
11.10
11.30
2.30
10.30
11.10
11.30
8.00
10.30
11.40
11.30
7.00
10.50
10.50
10.50
ways: by gas transfer through open porosity of

the ceramic layer and by oxygen ion diffusion
through anion vacancies in the ZrO2 lattice. Depending on the temperature and ceramic layer
state, the contribution of these mechanisms may
be different. It is noteworthy that oxygen demonstrates very high diffusive mobility in the
ZrO2 lattice (a few orders of magnitude higher
than the Al2O3 lattice) (Ref 44). It may set limits
on the use of ZrO2 in TBCs at temperatures
above 1100 C.
Oxygen penetration through the ceramic layer
results in bond coat oxidation. Oxides form on
its surface. Their composition and structure depend on the amount of oxygen reaching the surface and on the bond coat composition. As a
rule, the oxides forming on the bond coat surface
Thermal expansion coefficient of ZrO2-8%Y2O3

ceramics condensates. (1) Black (no annealing).
(2) After annealing in air at 1000 C for 2 h. , along crystallites;
, across crystallites
Fig. 8.17
Service lives of nickel superalloy samples 7 mm

in diameter with TBC. 1, ZrO2-8%Y2O3 /
Ni23Co20Cr11/5Al0.26Y; 2, TBC-EB (references); 3, TBC-APS
(references). Source: Ref 50
Fig. 8.18
have Al2O3 structure, but some spinel may also

form there. Oxide formation and growth on the
bond coat surface was independent of the surface preparation technique in use. It did not matter whether it was preoxidized or if ceramics
condensed on the surface free from oxides. Formation and growth of oxides on the ceramics/
bond coat interface result in additional stress
generation, degradation of ceramic layer adhesion, and ultimately, in its spallation. Ceramics
spallation takes place even if no stresses are generated by outside forces in a sample after hightemperature testing.
Ceramic layer spallation is caused by tearing
stresses arising on the convex surfaces of samples and blades (Fig. 9.16). These stresses are
induced by compressive stresses, which arise in
the ceramic layer on sample cooling from the
temperature equal to or below its condensation
temperature. Compressive stress rise is caused
by the fact that the TEC of ceramics is much
less than the TEC of the superalloy. The stresses
of ceramic layer tearing reach their maximum at
cooling to room temperature. Bond coat surface
oxidation and tearing stress of ceramic layer are
the main factors causing TBC damage on lab
testing and under service conditions. The higher
the temperature, the quicker oxidation and ceramic layer spallation occur.
To assess a TBC service life, it is advantageous to use a criterion of TGO layer thickness
that involves ceramic layer spallation (Ref 45)
and a criterion of adhesive strength of the ceramic layer/bond coat interface (Ref 46). The
latter criterion assumes that the adhesive
strength of the ceramic layer/bond coat interface
should be higher than the stress of the ceramic
layer tearing from the blade surface over the
specified period of blade service (Chapter 9,
Some Principles of Strength Designing for Turbine-Blade Protective Coatings).
Ceramic layer service life is affected by inhomogeneity of ceramics phase composition,
ceramics crystallite recrystallization processes,
and sintering at temperatures above 1000 C. On
evaporating and vapor depositing ZrO2-8%Y2O3
ceramics, the ceramic layer is fixed in its hightemperature metastable state. Under certain conditions in the temperature range of 500 to 1200
o
C, the high-temperature C and T phases
dissociate in the ceramic layer, and the M
phase may appear. The T M phase transformation occurs martensitically, with the volume
expanding by 3 to 12%. This results in higher
compressive stress in the ceramic layer and in
Table 8.19
Characteristics and service lifetime of samples tested
Alloy
Bond coat
thickness, lm
Ceramics
thickness, lm
Ceramics condensation
rate, lm/min
Ceramics
density, g/cm3
Ceramics
hardness, MPa
Service
lifetime, h
IN625
Hast X
IN625
Hast X
IN625
IN625
Hast X
Hast X
100
130
110
105
140
150
120
120
125
130
165
170
190
95
150
185
2.9
2.9
3.56
4.25
3.28
2.38
3.75
4.40
4.72
4.69
4.73
...
4.32
...
...
...
5085
5260
4870
6530
4484
5300
...
5772
2120
831
1047
832
497
945
1023
1722
Source: Ref 50
the corresponding increase of the stress, causing

its tearing from the bond coat.
Ceramics sintering at temperatures above
1000 C results in gradual ceramic layer densification, in shrinkage effects, and also in gen-
Fig. 8.19
Thermally grown oxide (-Al2O3) thickness versus

test time at 1135 C. Source: Ref 50
Fig. 8.20
Effect of TBC ceramic layer thickness on its service

life (sample diameter of 7 mm). Source: Ref 51
eration of additional tearing stresses of the ceramic layer on cooling. Sintering effects depend
on ceramic layer composition (Ref 47). The addition of NiO to the ceramics composition enhances sintering effects. The development of the
ceramics recrystallization processes in TBC-EB
ceramic layers at high temperatures also results
in reducing the strength of some crystallites and
in worsening their resistance to damage by erosion on gas flow attack.
The ceramic layer of a TBC does not actually
show any ductility. That is why any plastic deformation causes cracks in the ceramic layer, facilitating oxygen access to the bond coat surface.
Turbine blades with TBCs including ductile
Service lives of samples 10 mm in diameter with

ZrO2-8%Y2O3 /Ni20Co20Cr11/6Al0.25Y (thickness of 130/120 lm). 1 to 9, different sample lots
Fig. 8.21
Table 8.20
Coating
TBC-4
RT31/RT33
Coating characteristics
Bond coat composition
Ceramic layer
Ceramics thickness, lm
Ni14%Cr12%AlY(a)
Co38%Ni18%Cr8.9%AlY
ZrO2-7%Y2O3
ZrO2-7%Y2O3
120
100
(a) Yttrium content 0.1%. Source: Ref 52
Ni20Co20Cr12AlY bond coat have shown bond

coat deformation at temperatures above 1100
o
C. In these areas, the ceramic layer was easily
damaged by gas flow.
Ceramic layers of ZrO2-8%Y2O3 deposited by
the EB technique feature resistance to erosion
that is 3 to 7 times higher than that of ceramic
layers deposited by the APS technique (Ref 48).
This type of TBC damage is mainly met with on
turbine blades of stationary units. The ZrO28%Y2O3 ceramics may be damaged by hot corrosion on exposure to fuel combustion products
containing sulfur compounds and some other aggressive agents. Mechanisms of these processes
are discussed elsewhere (Ref 49).
Lab Tests of TBC Service Life. The main
body of publications on the research into TBC
durability deals with different high-temperature
tests of samples and blades under conditions
similar to those of testing for isothermal and cyclic heat resistance. The peculiarity of TBC testing is that one should detect the moment of ceramic layer spallation from their surfaces instead
of measuring sample mass gain/loss. Testing cylindrical samples with different diameters is the
most advantageous method for TBC durability
studies, optimization of their compositions, design, and deposition technique. This is due to
Service lives of TBCs on cylindrical specimens of

Hastelloy X samples 6 and 10 mm in diameter. Nf,
cycles to spallation of 20% sample surface area. Source: Ref 52
Fig. 8.22
the fact that on cooling the samples from the test

temperature, compressive stresses occur in the
ceramic layer and induce stresses of its tearing
from the cylindrical sample with the radius of R.
The magnitude of tearing stress, rncr, is directly
proportional to the ceramic coating thickness,
hcr, and inversely proportional to the test sample
radius, R.
rncr
hcr
rscr
R
where rscr is ceramic coating compressive stress.

As an approximation, the stress in the ceramic
layer may be calculated from the following
equation:
rscr
Ecr (T Tcr) (T To)

1 lcr
where T is temperature, To is conventional initial

temperature (temperature of ceramic layer deposition), Ecr is the ceramic layer modulus of
elasticity, and Tcr and T are, correspondingly,
TECs of the ceramics and substrate materials in
the temperature range of T to To, and lcr is Poissons ratio for the ceramic layer.
Ceramic layer spalling from a cylindrical
sample surface occurs at the moment when its
tearing stress is equal to or higher than the adhesive strength at a ceramic layer/bond coat interface; that is, rncr racr.
TBC service life depends on ceramic layer
composition and thickness. After the choice and
optimization of these characteristics for certain
turbine blades, it depends on ceramic layer deposition technique. At high-temperature cycling
testing, ceramic layer service life shortens with
the increase in maximum cycle temperature and
rates of heating and cooling of the samples
tested. By now, a large body of TBC tests for
service life evaluation have been carried out.
Unfortunately, comparative analysis of these test
results and their use for practical purposes is a
difficult task, because of the lack of information
on their deposition techniques and salient features of their test conditions.
Examples of testing samples and blades with

TBCs deposited using EB units and techniques
designed by Paton Electric Welding Institute
(Ukraine) are given subsequently. Thermal barrier coatings (including their bond coats) were
deposited in EB units of the UE type. Their bond
coat compositions were chosen by taking into
account the blade service conditions in an engine
or a stationary unit. After bond coat deposition,
the samples underwent microball peening and
recrystallization annealing in vacuum at 1000 to
1050 C for 2 to 4 h. Before deposition of the
ceramic layer, the bond coat surface underwent
hydroabrasive processing (with fused corundum
with a grain size of 30 to 60 lm). Then, the
samples were rinsed in an ultrasonic bath. Bond
coat surface roughness before ceramics deposition was Ra 0.6 to 0.8 lm.
Bond coat surface preoxidation was not used.
Table 8.21
The samples were arranged in special fixtures

and loaded into the unit for ceramic layer deposition, where they were EB heated to a temperature of 900 to 930 C. On reaching this temperature, evaporation of a previously prepared
ceramic bath started. The rotation speed of the
fixture with the samples in the vapor flow was
12 rpm. After ceramic layer deposition, the
samples were unloaded from the unit and annealed under vacuum at 1080 C for 2 h. The
ZrO2-8%Y2O3 powders from the same supplier
were used for ceramic layer deposition, and the
pellets for evaporation were made by the same
technique.
The results of high-temperature TBC cycling
testing are presented in Ref 50. In that case, the
metallic bond coat was deposited sequentially
from two alloys in a one-step process. First, an
inner layer 30 to 40 lm thick was deposited
Service lives of samples 10 mm in diameter

Service life, cycles
Coating
High
Mean
Low
Thermally grown oxide thickness, lm
Growth parameter (Kp), g2/cm4 s
TBC-4
RT31/RT33
1080
710
460
260
96
19
5.5
7.5
3.9 1013
8.4 1013
Source: Ref 52
Fig. 8.23
Structure of surface injured by spallation. (a) Ceramic layer side. Dark surface, Al2O3; pale surface, ZrO2-8%Y2O3. (b)
Bond coat side. Pale surface, bond coat; dark surface, Al2O3. 910
Structure of TBC bond coat on CMSX-4 superalloy. 1000. Element contents in points 1 to 5 are
given in Table 8.22
Fig. 8.24
Corrosion lives of EB-TBC (1, 2, 3) and

Co22Cr12AlY coating (4, 5, 6) in ash of gas-turbine fuel at different temperatures. 1 and 4, 750 C; 2 and 5, 800
C; 3 and 6, 850 C. Source: Ref 51
Fig. 8.25
from the Ni20Co20Cr5AlY alloy, and then, an

outer layer 60 to 80 lm thick was deposited
from the Ni20Co20Cr12AlY alloy. Yttrium content of the bond coat was controlled, and it was
within the range of 0.2 to 0.26%. The layer with
low aluminum content was intended to improve
bond coat ductility. Deposition of the ceramic
layer involved the formation of a dense zone 2
to 3 lm thick and enriched with nickel, chromium, and cobalt due to ceramic layer and bond
coat interaction. The samples were tested in air
in a 1710BL unit under the cycling conditions
of 50 1135 C, a sample cooling rate of 8 to
10 min, and a holding time of 1 h at the maximum temperature. The information on the samples tested and their test results are presented in
Table 8.19. Service life characteristic was taken
as a period (number of cycles) to the onset of
ceramic layer spalling.
Considerable spread in the experimental results is attributed to the problems of technological parameter controlling on TBC deposition.
The spread range is a characteristic of technological process reproducibility. The best results
are typical of the achievable level of TBC properties on optimizing all the TBC deposition process parameters. It is noteworthy that each product type should be manufactured using its
optimal technological process, because even insignificant variations of a sample or a blade
shape or mass entail some changes in the conditions of its heating and vapor flow condensation. The test results for the best samples of the
previously mentioned work, together with the
results known from references, are given for
comparative analysis in Fig. 8.18.
Ceramic layer spalling from the samples at
their testing is caused by TGO (Al2O3) that
forms and grows on the bond coat surface (Fig.
8.19). In the authors opinion, the optimal TBC
characteristics for the said deposition technique,
bond coat, and ceramics compositions are as follows: bond coat thickness of 110 to 135 lm,
ceramic layer thickness of 135 to 140 lm, and
condensation rate of 4 lm/min. In the case of
thicker or thinner TBCs, their service lives
shorten.
This effect of ceramic layer thickness on its
service life under different test conditions has
already been mentioned in Ref 51. The TBC-EB
service life on JS6U-type superalloy versus ceramic layer thickness is presented in Fig. 8.20.
The bond coat layer 90 to 120 lm thick was
deposited from the Co22Cr12Al0.4Y alloy; the
ceramic layer was deposited from the ZrO2-
8%Y2O3 pellet. Cycling conditions were as follows: heating to 1100 C for 4 min, holding at
this temperature for 20 min, and cooling to 40
to 50 C for 6 min. The TBC service life is determined as a period to ceramic layer spalling
from 50% of its surface. Maximum service life
was demonstrated by the TBC including the ceramic layer 100 to 140 lm thick. General decrease of the number of cycles to ceramic layer
spalling in comparison with the previously mentioned data is caused by yttrium content of the
bond coat (0.1%Y).
Test results for the samples 10 mm in diameter with TBCs deposited by the same technique are presented in Ref 2. The TBC bond
coat was deposited sequentially from two alloys in a one-step process. First, an inner layer
was deposited from the Ni20Co20Cr5AlY alloy, and then, an outer layer was deposited from
the Ni20Co20Cr12AlY alloy. Yttrium content
of the bond coat was 0.2 to 0.26%. The outer
ZrO2-8%Y2O3 ceramic layer was 120 to 130
Table 8.22
lm thick. The samples were tested under the

cycling conditions of 50 1100 C, heating
for 10 min, and cooling for 6 min. The test results are presented in Fig. 8.21. As compared
to the earlier tests, it may be concluded that
both reducing the temperature and the number
of cycles (30 cycles) and enlarging the sample diameter had an effect on TBC durability.
Ceramics spalling from the best samples occurred after 2500 h.
An investigation compared two variants of
coatings deposited by different techniques (Ref
52). The previously mentioned technique was
used for TBC-4 coating deposition. Some minor
distinctions in its deposition technique from the
previously mentioned were concerned with the
deposition of a thin layer (1.5 to 2 lm thick) on
the bond coat surface before ceramics evaporation onset. The thin layer composition was identical to the bond coat composition. The RT31/
RT33 coating was used for comparison.
The Hastelloy X samples 10 and 6 mm in di-
Composition of bond coat on CMSX-4 superalloy after testing at 1100 C for 120 h
Analyzed zone (Fig. 8.24)

1
2
3
4
5
Fig. 8.26
Al
Ti
Cr
Co
Ni
Mo
Ta
Re
2.5
6.6
9.6
6.5
17.4
0.4
0.6
0.5
0.2
0.2
12.4
10.4
6.7
19.9
8.0
4.5
5.9
2.6
4.2
2.3
29.3
64.1
74.1
65.0
71.7
1.1
0.3
0.1
0.2
0
2.8
6.0
4.8
1.1
0.6
24.7
4.7
1.6
1.2
0.2
20.5
1.4
0.1
0.5
0
Appearance of samples after adhesive strength testing at 1050 C for (a) 200 h and (b) 500 h
ameter were tested. The test conditions were as

follows: 20 1100 C, heating for 5 min, holding at the maximum temperature for 55 min, and
cooling for 15 min. Ceramic layer damage was
reported at 20% spalling from the surface. The
characteristics of the coatings and the test results
are presented in Table 8.20 and in Fig. 8.22.
In all cases, ceramics spalling was caused by
TGO layer formation between the ceramic layer
and the bond coat. Independently of the initial
state of a bond coat surface, TGO is forming on
its surface under high-temperature test conditions. The thickness and the growth parameter
of this layer differ for the coatings deposited by
different techniques (Table 8.21). Ceramic layer
spalling took place along the ceramics/TGO
boundary for the samples with long service lives
and along the bond coat/TGO boundary for the
samples with short service lives. Short service
lives of the samples with the TBC-4 coating may
be caused by low yttrium content (0.1%Y).
Similar ceramics spalling along the TGO/
bond coat interface was detected on testing the
samples of CMSX-4 alloy at 1100 C. The ZrO28%Y2O3 ceramic layer spalled after 250 cycles
(cycle of 60 min) of 20 1100 C, which was
much worse than other test results. The largest
part of the spalling surface area from the ceramics inside consisted of Al2O3 (Fig. 8.23a), while
from the Ni20Cr12AlY (Y 0.1%) bond coat
side, spalling took place along its surface. No
oxides were detected on that surface (Fig.
8.23b). The changes in the bond coat are the
same as the changes in the coating without the
ceramic layer, as discussed previously (Fig.
8.24). A great deal of aluminum from a bond
coat is spent on TGO layer formation (Table
8.22).
In addition to high-temperature areas of ap-
Fig. 8.27
Variation of adhesive strength of TBC ceramic

layer versus test conditions
plication, TBCs may also be advantageous for

improving service lives of blades operating in
media containing sulfur compounds. In this
case, a wholesome effect of the ceramics may
be attributed to both blade surface temperature
increase and less condensation of aggressive
compounds on it as well as to bond coat surface
protection from aggressive compound attack.
The test results for a few coating modifications
with outer ceramic layers are presented in Fig.
8.25 (Ref 51), together with test results for the
coating deposited from the Co22Cr12AlY alloy
for comparison.
In this case, the procedures of bond coat surface preparation and coating deposition are the
same as that mentioned previously. The bond
coat was vapor deposited sequentially from two
alloys, first from Co24Cr5Al0.4Y, then from
Co24Cr12Al0.4Y. Bond coat thickness was 90
to 120 lm. The ZrO2-8%Y2O3 ceramic layer
thickness varied from 30 to 160 lm. A synthetic
ash of gas-turbine fuel was used in the tests. Its
composition was as follows: Na2SO4, 66.2%;
Fe2O3, 20.4%; V2O5, 1.8%; NiO, 80.3%; and
CaO, 3.3%. The ash was applied to the sample
surfaces as 120 g/m2 every 250 h. The samples
were tested at 750, 800, and 850 C to the moment of corrosive damage of its coating (both
the ceramic layer and the bond coat). Ceramic
layer spalling on testing for 20,000 h (at 750 C)
and for 7500 h (at 800 C) was not reported. In
all cases, service lives of the coatings with ceramic layers were longer than service lives of
the coatings without ceramic layers.
Testing TBC Ceramic Layer for Adhesive
Strength. For TBC service life prediction, one
may use the adhesive strength characteristic of
the ceramic layer/bond coat interface (Ref 53).
To do so, strength variations at this interface
should be determined as ra f(T,s), depending
on test time and temperature.
Testing for adhesive strength may be carried
out by different methods used for evaluation of
this property (Ref 54). A commonly used glue
method is suitable. To study ceramic layer adhesion, a TBC was deposited on a flat surface of
a cylindrical sample 20 mm in diameter. The
coating consisted of the Ni20Cr12AlY bond
coat 80 lm thick and the ZrO2-8%Y2O3 ceramic
layer 120 lm thick. Sample heating temperature
at ceramics deposition was 850 C. The ceramic
layer condensation rate was 10 lm/min.
The samples and their TBCs were held under
the furnace air atmosphere at temperatures of
1050, 1100, and 1150 C for periods of up to
Table 8.23 Blade test results on gas-dynamic test rig. Test cycling conditions: 2001100 oC, heating
time is 30 s, cooling time is 30 s
Blade lot
Number of cycles at testing
Injury description
1265
650
805
805
Ceramics started to spall after 1100 cycles

Insignificant ceramics spallation along a leading edge
Insignificant ceramics spallation along a leading edge
Undamaged ceramics
A
B
C
D
Note: Ceramic layer thickness 120 to 160 lm
500 h. After this, the ceramics surface of one

sample glued to the respective metallic surface
of the other sample with the same geometrical
characteristics. Then, the glued sample was
tested for breaking at room temperature. The
samples, after testing for adhesive strength, are
shown in Fig. 8.26. In their initial state, the samples after ceramic layer deposition always break
down along the glued joint, which has a strength
of 70 MPa. On testing after holding at high
temperatures, the samples break down partially
Fig. 8.28
Turbine blades with EB-TBC after testing in aircraft

engine for 1200 h
along the ceramics/bond coat interface and partially along the glue joint (Fig. 8.26b). The
higher the test temperature and the longer the
test time, the larger the breakage area along the
ceramics/bond coat interface. The adhesive
strength data presented in Fig. 8.27 are a little
lower than their real level, because calculations
were carried out on the base of test sample crosssectional areas without due account of the surfaces where ceramic layer breakage occurred
along the glue. Ceramic layer phase composition
did not change after holding at high temperatures. The oxides with -Al2O3 structure and
spinels were detected on the inside ceramics surface adjacent to the ceramics/bond coat interface. The general trend in adhesive strength behavior of the ceramic layer is its drop as test time
and temperature increase.
TBC Testing on Gas-Dynamic Test Rigs
and Engines. Testing on gas-dynamic test rigs
makes interpretation and comparison of different test data even more difficult. In addition to
tearing stresses generated during slow cooling,
thermal stresses are generated on the test rig, due
to nonuniformity of the heat pattern along the
height and across the cross section of the samples and blades. Thermal stresses and strains result in ceramics fragmentation and facilitate oxygen access to the bond coat surface. The main
cause of ceramic layer spalling is bond coat oxidation and the corresponding decrease in adhesive strength of the ceramic layer.
Test results on gas-dynamic test rigs for TBCEB with a bond coat 60 to 80 lm thick deposited
from the Ni20Cr12AlY alloy and the ZrO28%Y2O3 ceramic layer 120 to 140 lm thick are
given in Table 7.18 (Chapter 7, The Effect of
Protective Coatings on the Mechanical Properties of Superalloys). Thermal cycling is 1100
200 C, time required for heating the sample
up to the maximum temperature is 60 s, and
cooling time is 60 s. Ceramics spalling was detected on the edge with the radius of 1 mm after
1500 cycles (25 h).
Similar results were obtained on testing turbine blades on this gas-dynamic test rig. In Table
8.23, test results are presented for a few turbine
blade lots of the MAR-M-002 alloy on cycling
that included heating the blades in the flow of
combustion products to 1100 C and cooling
them to 200 oC with compressed air. Heating
and cooling periods were the same (30 s). The
examination of the coatings after their testing
has revealed the TGO layer formed at the ceramic layer/bond coat interface. The thickness
of this layer increases with the increase in the
number of cycles. For the best blades, ceramic
layer spalling started after 1100 cycles.
Maximum information on life and efficiency
of TBCs is obtained through testing in engines.
However, this has specific features connected
with the peculiarities of blade design and testing
conditions. To determine the general features of
TBC behavior, the results of testing of different
engines should be systematized.
Tests of aircraft and power-turbine blades
with TBCs are dealt with in Ref 5559. Blade
examination after long-term testing in engines
reveals that the main TBC damage types are as
follows: ceramic layer spalling because of bond
coat oxidation in the high-temperature zones, ceramic layer spalling at the punched holes because of bond coat oxidation from the inside perforation space, erosion wear of the layer on the
blade shroud due to unfavorable ceramics structure on these surfaces, and cracking in the zones
where high thermal stresses arise. Phase transformations in the bond coat are similar to those
in the coatings without ceramics, provided that
less aluminum content is spent on oxidation.
The TBCs (consisting of the Ni20Cr12AlY
bond coat 80 to 120 lm thick and the ZrO28%Y2O3 or ZrO2-8%Y2O3-5%NiO ceramic
layer 120 to 160 lm thick) were tested in different aircraft engines. Thermal barrier coating
lifetime is closely allied with testing conditions,
such as maximum temperature, time, and number of transient situations. A blade, after 1200
h of testing in an aircraft engine, is shown in
Fig. 8.28. Under extreme conditions, the temperature on some blade surface areas is as high
as 1120 to 1150 C. Within the said time, ceramic layer damage was detected only on a
shroud surface.
Under testing conditions equivalent to
10,000 h of engine operation, TBCs did not
show any damage on the coated blades of MARM-200 alloy. The same coating was used. Examination revealed that it was undamaged even
on the surfaces where maximum temperatures

were as high as 1100 to 1150 C. As in all hightemperature tests, the TGO layer formed and
grew on the ceramics/bond coat interface.
All the durability tests of samples and blades
demonstrate that the vapor-deposited ceramic
layer has some unique properties that allow it to
hold on turbine blade surfaces for a long time.
Its columnar structure compensates thermal
stresses and strains arising at the operating temperatures in the cooled blade walls and on their
surfaces. High adhesive strength of the ceramics/bond coat interface keeps ceramics on the
bond coat surface on blade cooling.
The problems that set limits to a TBC service
life are caused primarily by the high-oxygenpermeable ZrO2 lattice. To a large extent, a TBC
ceramic layer service life depends on its deposition technique. Optimization of the technological cycle of TBC deposition, with reference to
technological equipment available and the TBC
composition and design chosen, is the crucial
consideration in producing TBCs with long service lives.
REFERENCES
1. J. Stringer, The Reactive Element Effect in

High-Temperature Corrosion, High Temperature Corrosion 2, R. Streiff, J. Stringer,
R. Krutenat, and M. Caillet, Ed., Elsevier
Science Publishers, 1989, p 129137
2. I.S. Malashenko, N.P. Vashchilo, V.A. Belotserkovsky, and K.Y. Yakovchuk, Effect
of Yttrium on Functional Characteristics of
Vacuum Condensates and Protective Coatings MCrAlY and MCrAlY/ZrO2-8%Y2O3
at Thermocyclic Loading, Adv. Spec. Electrometall., Vol 1, 1997, p 2433
3. R.A. Miller, R.G. Garlick, and J.L. Smialek,
Phase Distribution in Plasma-Sprayed Zirconia-Yttria, Am. Ceram. Soc. Bull., Vol 63
(No. 12), 1983, p 13551358
4. R.P. Ingel, D. Lewis, B.A. Bender, and
R.W. Rice, Physical, Microstructural and
Thermomechanical Properties of ZrO2 Single Crystals, Advances in Ceramics, Vol 12,
American Ceramics Society, Inc., 1984, p
408414
5. S. Stecura, Optimization of the Ni-Cr-Al-Y/
ZrO2-Y2O3 Thermal Barrier System, Adv.
Ceram. Mater., Vol 1, 1986, p 6876
6. D.L. Ruckle and D.S. Duvall, QuenchCracked Ceramic Thermal Barrier Coatings,
U.S. Patent 4457948, 3 Aug 1984
7. T.E. Strangman, Columnar Grain Thermal

Barrier Coatings, U.S. Patent 4321311, 23
March 1982
8. H.G. Scott, Phase Relationship in the Zirconia-Yttria System, J. Mater. Sci., Vol 10
(No. 9), 1975, p 15271535
9. D.S. Rutman and Y.S. Toropov, High-Temperature Zirconium Dioxide-Base Materials, Moscow, Metallurgy, 1985, 137 p
10. V.S. Bakunov, V.L. Balkovich, and A.S.
Vlasov, Ceramics of Highly Refractory Oxides, Moscow, Metallurgy, 1977, 304 p
11. I.D. Abramson, Ceramics for Aviation,
Oborongiz, Moscow, 1993, p 240
12. A. Karklit, A. Larine, et al., Producing of
High-Temperature Strength Materials by
Semi-Dry Molding, Moscow, Metallurgy,
1982, p 368
13. K. Fritscher, M. Schmucker, C. Leyens, and
U. Schulz, TEM Investigation on the Adhesion of YPSZ EB-PVD TBCs, High Temperature Corrosion 4, R. Streiff, J. Stringer,
R. Krutenat, M. Caillet, and R. Rapp, Ed.,
Trans Tech Publication, 1997, p 965970
14. I.S. Malashenko, G.C. Marijnissen, V.A.
Belotserkovsky, A.N. Shelkovoj, and V.V.
Orlenko, Microstructural Peculiarities of
Metal/Ceramic Condensation Coatings Under Thermocyclic Loading Conditions, Adv.
Spec. Electrometall., Vol 2, 1997, p 3243
15. T.E. Strangman and P.A. Solfest, Ceramic
Barrier Coating with Alumina Interlayer,
U.S. Patent 5015502, 14 May 1991
16. B.A. Movchan, Y.E. Rudoj, and I.S. Malashenko, Method of Producing on a Substrate of Protective Coats with Chemical
Composition and Structure Gradient Across
the Thickness and with Top Ceramic Layer,
Ukraine Patent 17473A, 6 May 1997
17. B.A. Movchan and Y.E. Rudoj, Gradient
Heat-Resistant Coatings, Produced by
Vacuum Electron-Beam Deposition of Vapour Phase, Adv. Spec. Electrometall., Vol
2, 1997, p 2531
18. I.S. Kulikov, Thermodynamics of Oxides,
Moscow, Metallurgy, 1986, 384 p
19. Y.A. Tamarin and S.V. Zherzdev, Technological Aspects of Forming TBC-EB Ceramic Layers, High Temperature Corrosion
5, R. Streiff, I.J. Wright, R. Krutenat, M.
Caillet, and A. Cailerie, Ed., Trans Tech
Publication, 2001, p 587593
20. A.V. Demchishin, Structure and Properties
of Thick Vacuum Condensate of Metallic
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
and Nonmetallic Materials, Ph.D. dissertation, Kiev, 1980

B.A. Movchan and A.V. Demchishin, Research into Structure and Properties of
Thick Vacuum Condensate Layers of Titanium, Tungsten, Aluminum Oxide, Physics
of Metals and Metals Science, Vol 28 (No.
4), 1969, p. 653660
Y.A. Tamarin, E.B. Kachanov, and S.V.
Zherzdev, Thermophysical Properties of
Ceramic Layers in TBC, High Temperature
Corrosion 4, R. Streiff, J. Stringer, R. Krutenat, M. Caillet, and R.A. Rapp, Ed., Trans
Tech Publication, 1997, p 949955
B.A. Movchan and I.S. Malashenko, Twoand Three-Layer Vacuum-Deposited Coatings for Gas Turbine Blades Protection, Automat. Weld., Vol 2, 1994, p 3038
U. Schulz, K. Fritscher, H.-J. RatzerScheibe, et al., Thermocyclic Behavior of
Microstructurally Modified EB-PVD Thermal Barrier Coatings, High Temperature
Corrosion 4, R. Streiff, J. Stringer, R. Krutenat, M. Caillet, and R. Rapp, Ed., Trans
Tech Publication, 1997, p 957964
W.J. Parker, R.J. Jenkins, C.P. Bulter, and
G.L. Abbott, Flash Method of Determining
Thermal Diffusivity, Heat Capacity and
Thermal Conductivity, J. Appl. Phys., Vol
32 (No. 9), 1961, p 16791684
R.D. Cowan, Pulse Method of Measuring
Thermal Diffusivity at High Temperature,
J. Appl. Phys., Vol 34 (No. 4), 1963, p 926
927
R.F. Blumer and R.E. Taylor, Measurement
by the Flash Method of Thermal Diffusivity
in Two-Layer Composite Sample, High
Temp.High Press., Vol 6, 1974, p 491
497
L. Pawlowski, D. Lombard, A. Mahlia, et
al., Thermal Diffusivity of Arc Plasma
Sprayed Zirconia Coatings, High Temp.
High Press., Vol 16, 1984, p 347359
P. Morell and R. Taylor, Thermal Diffusivity of Thermal Barrier Coatings of ZrO2 Stabilized with Y2O3, High Temp.High
Press., Vol 17, 1985, p 7988
AGARDS-R-823, NATO workshop on
Thermal Barrier Coatings, (Aalsborg, Denmark), 1998
A. Maesono, Measurement of Thermal
Constants by Laser Flash Method, SinkuRiko Co. LTD, April, 1983
J.A. Cape and G.W. Lehman, Temperature
and Finite-Pulse Time Effect in the Flash
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
Method for Measuring Thermal Diffusivity,

J. Appl. Phys., Vol 34 (No. 7), 1963, p
19091913
R.C. Hekman, Finite-Pulse Time and Heat
Loss Effects in Pulse Thermal Diffusivity
Measurements, J. Appl. Phys., Vol 44 (No.
4), 1973, p 14551460
L.M. Clark and R.E. Taylor, Radiation Loss
in the Flash Method for Thermal Diffusivity, J. Appl. Phys., Vol 46 (No. 2), 1975, p
714719
R.F. Blumer and R.E. Taylor, Measurement
by the Flash Method of Thermal Diffusivity
in Two-Layer Composite Samples, High
Temp.High Press., Vol 6, 1974, p 491
497
K.B. Larson., K. Koyama, Correction for
Finite-Pulse Time Effects in Very Thin
Samples Using the Flash Method of Measuring Thermal Diffusivity, J. Appl. Phys.,
Vol 38 (No. 3), 1967, p 465473
R.E. Krzhinhanovsky and Z.J. Shtern, NonMetallic Material Thermophysical Properties, Leningrad, Energy, 1973, 336 p
R. Berman, Thermal Conduction in Solids,
Claredon Press, Oxford, 1976, 283 p
J.R. Nicholls, K.J. Lawson, A. Johnston,
and D.S. Rickerby, Low Thermal Conductivity EB-PVD Thermal Barrier Coatings,
High Temperature Corrosion 5, R. Streiff,
I.J. Wright, R. Krutenat, M. Caillet, and A.
Cailerie, Ed., Trans Tech Publication, 2001,
p 595606
J.R. Nicholls, K.J. Lawson, D.S. Rickerby,
and P. Morrel, Advanced Processing of
TBCs for Reduced Thermal Conductivity,
Paper 6, AGARDS-R-823, NATO workshop on Thermal Barrier Coatings (Aalsborg, Denmark), 1998
K.J. Lawson, J.R. Nicholls, and D.S. Rickerby, The Effect of Coating Thickness on
the Thermal Conductivity of CVD and PVD
Coatings, Fourth International Conf. on
Advances in Surface Engineering (Newcastle, U.K.), 1996
S. Suhr, T.E. Fitchell, and R.J. Keller, Microstructure and Durability of Zirconia
Thermal Barrier Coatings, Advances in Ceramics, Vol 12, American Ceramics Society, Inc., 1984, p 503517
L. Lelait, S. Alperine, and R. Mevrel, Alumina Scale Growth at Zirconia-MCrAlY Interface: A Microstructural Study, J. Mater.
Sci., Vol 27 (No. 1), 1992, p 512
44. O. Kubaschewski and B.E. Hopkins, Oxidation of Metals and Alloys, Butterworths,
London, 1962, 424 p
45. R.A. Miller, Oxidation-Based Model of
Thermal-Barrier Coating Life, J. Am. Ceram. Soc., Vol 67 (No. 8), 1984, p 517521
46. Y.A. Tamarin, V.G. Sundyrin, and V.Y.
Kanaev, Stress Calculation of the Ceramic
Thermal Barrier Coatings for the Cooled
Turbine Blades, Mechanics of Creep Brittle
Materials-2, A.C.F. Cocks and A.R.S. Ponter, Ed., Elsevier, 1991, p 308312
47. D. Zhu and R.A. Miller, Sintering and
Creep Behavior of Plasma-Sprayed Zirconia and Hafnia-Based Thermal Barrier
Coatings,
NASA/TM-1998208406/
REV1, Aug 1998
48. J.R. Nicholls, Y. Jasler, and D.S. Rickerby,
Erosion and Foreign Object Damage of
Thermal Barrier Coatings, High Temperature Corrosion 4, R. Streiff, J. Stringer, R.
Krutenat, M. Caillet, and R.A. Rapp, Ed.,
Trans Tech Publication, 1997, p 935948
49. S. Alperine, Molten Salt Induced High
Temperature Degradation of Thermal Barrier Coatings, Preprint AGARD, High
Temperature Surface Interaction, Conference Proceedings N461
50. I.S. Malashenko, G.C. Marijnissen, V.A.
Belotserkovsky, A.N. Shelkovoj, and V.V.
Orlenko, Life of Condensed Coatings
NiCoCrAlY/ZrO2-8%Y2O3 at Thermocyclic Loading, J. Adv. Spec. Electrometall., Vol 1, 1997, p 3447
51. B.A. Movchan, I.S. Malashenko, K.Y. Yakovchuk, A.I. Rybnikov, and A.A. Tchizik,
Two- and Three-Layer Coatings Deposited
Under Vacuum for Gas Turbine Blades Protection, J. Surf. Coatings Technol., Vol 67,
1994, p 5563
52. Y. Jaslier and S. Alperine, EB-TBC Coatings: A Comparative Evaluation of Competing Deposition Technologies, Preprint
AGARD, 85th Meeting of the Structures
and Materials Panel, 1516 Oct 1997, p 81 to 8-10
53. Y.A. Tamarin, E.B. Kachanov, and S.V.
Zherzdev, Properties of Ceramic Coatings
for Turbine Blades Deposited by ElectronBeam Technique, Probl. Mach. Build. Reliab., Vol 1, 1994, p 7480
54. P.R. Chalker, S.J. Bull, and D.S. Rickerby,
A Review of the Methods for the Evaluation
of Coating-Substrate Adhesion, Mater. Sci.

Eng. A., Vol 140, 1991, p 583592
55. A. Bennett, Rolls-Royce Experience with
Thermal Barrier Coating, Proc. Br. Ceram.
Soc., Vol 34, 1984, p 207217
56. F.C. Toriz and A. Bennett, A Philosophy for
TBCs and Its Collaboration by 10,000 h
Service Experience on RB-211 Nozzle
Guide Vane, J. Surf. Coatings Technol., Vol
32, 1987, p 359375
57. T.N. Rhys-Jones and F.C. Toriz, Thermal
Barrier Coating for Turbine Application in
Aero Engines, High Temp. Technol., Vol 7

(No. 2), 1989, p 73
58. F.C. Toriz, A.B. Thakker, and S.K. Gupta,
Flight Service Evaluation of Thermal Barrier Coating by Physical Vapor Deposition
at 5200 h, J. Surf. Coatings Technol., Vol
39/40, 1989, p 161172
59. I.S. Malashenko, V.V. Orlenko, S.V. Domoroslov, et al., Overlay Coatings on Turbine Blades for Gascompressor Units, J.
Adv. Spec. Electrometall., Vol 2, 1990, p
5770

CHAPTER 9
Some Principles of Strength Designing

for Turbine-Blade Protective Coatings*
WHEN USING different types of protective
coatings for turbine blades, specific problems
emerge that are related to the need for strength
and lifetime ensurance of both a blade and a
coating itself under the action of thermal and
mechanical loads brought about by turbine service conditions. Basic principles of the design
approach to aircraft-engine turbine blades are
discussed elsewhere (Ref 14). In calculating
the coated-blade service life, the following considerations should be taken into account:
The application of a coating has an effect on

the main mechanical properties of the blade
material. Coatings that have minimum effect
on the substrate properties are considered to
be preferable.
A coating gives rise to additional centrifugal
load for a blade. This can result in partial loss
of the benefits of coating application. This
problem is most critical for thermal barrier
coatings (TBC).
Because of distinctions between the thermal
expansion coefficients as well as the mechanical and thermal-physical properties of the
coating and substrate materials, additional
thermal stresses can arise in the coatings, initiating cracks in the coating or resulting in
spalling of its fragments from the blade surfaces. As a result, with further running, there
is a risk of coating crack propagation into the
blade material, causing its failure on the
whole.
In the elevated working temperature range
(1100 oC), the reduction of the blade working cross-sectional area can take place in its
hot zone, because of diffusion interaction between the coating and the superalloy.
When designing a turbine blade, the effect of
the coating must be taken into account. Blademass gain resulting from coating application is
allowed for by taking into account its contribution to the centrifugal load. Then, stresses and
strains in the blade can be calculated using different model approximations, which are currently used for calculations of turbine blade
stressed-strained state, neglecting the contribution of coating stiffness. In the calculations, due
regard must be given to the effect of the coating
on the mechanical properties of the blade material, involving the test results on the coated
specimens. When applying a TBC, there is a
need to take into account its influence on the
blade temperature field.
Calculation of Stresses and Strains in

Coatings
For all heat resistant coatings (such as diffusion coatings and overlay coatings) whose thickness is significantly less (by a factor of 10 to 20)
than the thickness of cooled turbine-blade walls,
one can assume that the strains in the coating are
defined completely by the strains in the blade;
that is:
eZc eZ; eSc eS
(Eq 1)
where eZc and eSc are strains in the coating, and

eZ and eS are strains in the blade surface layer
*This chapter was written in cooperation with Dr. V.G. Sundyrin.

adjacent to the coating corresponding to longitudinal direction z and transverse direction along
the surface line S in the blade cross section (Fig.
9.1).
The strains eZ and eS in the blade surface layer
are determined in advance as a result of the calculation of its stress-strain state with the use of
one-, two-, or three-dimensional models, which
are usually applied to blades. At this stage, the
coating is taken into account only as an additional mass that increases the centrifugal load on
a blade. With the use of a one-dimensional
model (beam model), it is taken that eS 0.
The strain characteristics of (eZc, eSc, eZ, and
eS in equalities (Eq 1) are, correspondingly, the
sums of strains generated by effective stresses
and thermal strains in the coating and the substrate:
eZc efZc eTc
eSc efSc eTc
eTc Tc (T T0)
eT T (T T0)
where T is the blade surface temperature, T0 is

the conventional initial temperature, and Tc and
T are, correspondingly, thermal expansion coefficients of the coating and substrate materials
in the temperature range of T to T0. The temperature characterized by zero thermal stress
level in the coating and the blade is used as the
initial temperature of T0. As a rule, it is a temperature of coating deposition or that of hightemperature annealing of coated blades.
In addition to the previously mentioned
strains, the strain enc also occurs in the coating
in the direction on the normal, n, to the blade
surface (Fig. 9.1), which also must be considered in the calculations. The coating strain enc
can be defined from the expression:
(Eq 2a)
enc
and
eZ efZ eT
eS efS eT
(Eq 3)
(Eq 2b)
where efZc, efSc, efZ, and efS are strains being directly caused by the stresses involved, and eTc
and eT are, correspondingly, the coating and
substrate thermal strains.
Thermal strains (eTc , eT) are calculated from
the corresponding thermal expansion coefficients of the coating and superalloy materials:
l*c
1 l*c
(eZ eS)
1 l*c
1 l*c
TC (T T0)
(Eq 4)
In Eq 4, l*c is Poissons ratio for the coating

material, in case of its elastic-plastic straining.
The plastic strains in the coating can be taken
into account as a result of a number of elastic
strain calculations using the method of successive approximations discussed elsewhere (Ref
2). With the use of ordinary strain plasticity theory, l*c in Eq 4 is of the following form:
l*
c
vc (1 lc) (1 2lc)
2vc (1 lc) (1 2lc)
(Eq 5)
where lc is Poissons ratio for the coating in case

of its elastic straining, and vc is the coefficient
of plasticity:
vc
3Ec
1
2(1 lc) EcS
(Eq 6)
In Eq 6, Ec is the elasticity modulus of the coating:

ESc
Fig. 9.1
Stresses in the coating and blade wall
ric
eic
(Eq 7)
the so-called secant modulus for the coating

material, where ric and eic are the stress and
strain intensities in the coating, respectively.
When considering Eq 1 in Chapter 8, the
strain intensity in the coating is defined by the
expression:

Some Principles of Strength Designing for Turbine-Blade Protective Coatings / 197
eic
3
(ezc esc)2 (ezc enc)2 (esc enc)2
(Eq 8)
In this expression, only the coating strain enc is

the unknown term. For determining enc and,
consequently, eic, it is best to use the iterative
method, which is often called the method of
variable elasticity parameters presented elsewhere (Ref 2).
For cyclic loading, the strain-intensity range
in the coating can be estimated from the following formula:
Deic |sign(emax
1c )
max
min
min
eic
sign(e1c
) eic
|
(Eq 9)
min
In this formula, emax
and eic
are, correspondic
ingly, maximum and minimum values of the
strain intensity in the coating during cycle of
min
loading; sign (emax
1c ) and sign (e1c ) are signs of
the first principal strain e1c for, correspondingly,
min
maximum (emax
ic ) and minimum (eic ) values of
strain intensity in the coating.
Stresses in the coating in the case of its elastic-plastic straining can be estimated, if necessary, with the use of the formulae:
1 l*
E*
c
rZc g
(eZ l*e
c S)
rSc
E*
c Tc (T T0)
1 l*c
(Eq 10)
3Ec
2(1 lc) vc 1 2lc
1
Ec h
1
EH
(Eq 11)
(Eq 13)
where N0 is the centrifugal load in the given

blade cross section, neglecting the coating mass,
and b is the coefficient of the coating-mass effect on the centrifugal load in the blade:
b
where E*c is the reduced modulus of elasticity in

the strain theory of plasticity, which is also calculated in the process of successive approximations described elsewhere (Ref 2).
(Eq 12)
As a rule, protective coatings 2 5%

are quite typical. For plastic straining of the
coating, the value of this error is significantly
less.
Value Correction for Centrifugal Load in a
Blade. Coating application results in an increase
in the value of centrifugal load in a blade.
The value of the total centrifugal load in the
blade with the consideration of coating mass in
a certain specified blade cross section can be estimated with the use of the approximate formula:
Here
E*
c
Ec h
100%
EH
N bN0
E*
c
g
(eS l*e
c Z)
1 l*c 2
E* Tc (T T0)
c
1 l*c
is the coefficient that allows one to take into account tensile stiffness of coating Ec h, where h
and H are, correspondingly, the coating and
blade-wall thickness values.
In the case of elastic straining of the coating,
it should be taken that E*c Ec and l*c lc
for Eq 9 and 10. In this case, the calculation is
performed once without recourse to the process
of successive approximations.
With the use of the previously described
model, the stresses and elastic-plastic strains in
a protective coating can be calculated for any
blade surface points under transient and steadystate conditions. It allows the determination of
the elastic-plastic strain range in the coating under cyclic thermomechanical blade loading.
The error of the calculations performed for the
given model using elastic strains is:
qc h
qH
Ec h
1
EH
1
(*)
(Eq 14)
where qc and q are, correspondingly, densities

of coating and blade materials.
The parameter values from the midheight
blade portions are used for calculations in this
formula.
Equations 13 and 14 are obtained from the
consideration of the coated blade wall as the sys-

tem of two bars having different tensile stiffness

values, E H and Ec h, in the case of their joint
elastic straining, when the strains of these bars
are the same (Fig. 9.2).
If b 1, then according to Eq 13, the stresses
generated in the coated blade by centrifugal load
are less than those in the uncoated blade. This
is the case for hard coatings.
If b 0, then the blade material is being additionally loaded by extra centrifugal load associated with the coating mass. This is the case
for soft heat resistant metallic coatings at their
plastic deformation.
If b 1, then the total extra centrifugal load
generated by the coating mass is carried by the
coating itself. This is the case for neutral coatings, typical of heat resistant metallic coatings
at their elastic deformation and also, to a certain
extent, of the ceramic layers of TBCs if the network of developed cracks has not appeared in
these layers.
Actually, for the condition of b 1 to be met,
wherein there is no undesirable extra blade loading, it is required, as follows from Eq 14, to
fulfill the following condition:
qc
Ec
q
E
(Eq 15)
The condition (Eq 15) is not met for the case of

developed plastic deformation of a corrosion-resistant coating when Ec Esc E (qc q),
as well as for a TBC with a crack network in the
ceramic layer when one can take Ec 0.
In such a case:
b1
qc h
1
qH
It means that the coating contributes to bladematerial loading with the extra load.
Fig. 9.2
Centrifugal force in a coated blade
Presented dependencies (Eq 13 and 14) may

be used when correcting the value of centrifugal
load in the blade, in view of actual coating elastic-plastic properties and thickness in calculating
stresses and strains in the blade.
Coating Life Estimation. One of the most
common types of failures of turbine-blade protective coatings is the initiation of individual
cracks or a network of multiple, closely spaced
cracks (fragmentation) in the coating. As a rule,
the mechanical coating failures occur after longterm operation under thermal cyclic loading.
However, in some cases, cracking in the coating
may be detected even after short-term cycling,
such as 10 to 100 cycles of start-maximumshut conditions. It happens by choosing an unsuitable coating because of its physical-mechanical properties. The following criteria may be
used for estimation of coating serviceability under the conditions of thermal cycling (Ref 5).
Criterion of Ultimate Strain. High-temperature, corrosion-resistant aluminide coatings are
characterized by a steep change in ductility
within the blade working temperature range of
20 to 1100 oC. The ductile-brittle transition temperature (DBTT) range, in which a steep change
in ductility and strength properties is taking
place, depends on the coating structure and its
chemical composition. If, within the reducedcoating ductility temperature range, considerable thermal stresses occur in the coating due to
the difference between the thermal expansion
coefficients of the coating and substrate materials, coating cracking may result after a small
number of loading cycles (10 to 100 cycles).
In this case, the failure mechanism is of a static
(quasi-static) nature. Here, as the criterion for
coating serviceability, the following condition
may be used:
eic dc
(Eq 16)
where dc is coating ductility and eic is the strain

intensity in the coating. Equation 16 should be
checked under transient cooling conditions at
the blade surface temperatures, which are less
than the DBTT of the coating material. In this
case, the coating ductility is minimum, and thermal stresses in the coating caused by the difference between the thermal expansion coefficients
of the coating and blade materials may be at their
maximum levels. In this case, the residual temperature stresses in condensated coatings usually
are found to be tensile, and in diffused coatings
they are compressive. Under the service condi-

tions in the blade temperature range of 1000 to

1100 oC, the temperature stresses in the heat resistant coatings are actually at the zero level.
As an example, Fig. 9.3 and 9.4 illustrate the
plots of strain intensity eic (curves 1) and ductility dc (curves 2) versus temperature for two
overlay coatings of Ni20Cr5AlY (Fig. 9.3) and
Ni20Cr12AlY (Fig. 9.4) applied to the blade
cast from the nickel-base directionally solidified
superalloy of MAR-M-200 type. The final annealing temperature, once the coatings are applied, is T0 1050 oC.
It follows from the presented plot that the
Ni20Cr12AlY coating is more preferable by the
criterion of ultimate strain (Eq 16) when compared with the Ni20Cr5AlY coating, despite the
fact that the Ni20Cr5AlY coating is more ductile. Actually, curves 1 and 2 in Fig. 9.3 are almost the same at T 180 oC. It evidences that
condition (Eq 16) for the Ni20Cr5AlY coating
is not met; that is, there is a certain probability
for crack nucleation in the coating after a small
number of loading cycles.
On the contrary, for the Ni20Cr12AlY coating, curves 1 and 2 in Fig. 9.4 are located far
from each other, and they do not intersect over
the entire temperature range; that is, condition
(Eq 16) is met and a considerable margin is left.
The main reason, because of which the
Ni20Cr5AlY coating is found to be susceptible
to cracking despite its rather high ductility, is an
unfavorable combination of the coating and
blade material thermal expansion coefficients,
resulting in enhanced thermal stresses and
strains in the coating. Thus, despite its high ductility, the Ni20Cr5AlY coating can provoke
cracking on the blade surface rather than hinder
it.
Criterion of Thermomechanical Fatigue.
With large cyclic running time, a network of microcracks (fragmentation) can occur in the turbine blade coatings, which are caused by thermomechanical fatigue (TMF) by nature. The
criterion for the coating cyclic life, when applied
to such a failure, can be presented as follows
(Ref 6):
N NRC
(1) Strain intensity, eic, and (2) ductility, dc, of

Ni20Cr5AlY coating on a blade uniformly heated to
temperature, T
Fig. 9.3
(Eq 17)
where N is the number of loading cycles and NRC

is the number of cycles to crack detection under
the given cyclic loading conditions.
For experimental determination of the coating
TMF characteristic (NRC), the approach can be
used that is based on the determination of the
number of loading cycles for the coated specimens to the first crack detection in the coating
at their testing for thermomechanical fatigue
(see Chapter 7, The Effect of Protective Coatings on the Mechanical Properties of Superalloys). As a result, an experimental TMF curve
for a coated specimen can be obtained:
De De(NRc)
(1) Strain intensity, eic, and (2) ductility, dc, of

Ni20Cr12AlY coating on a blade uniformly heated
to temperature, T
Fig. 9.4
where De is the strain range of a specimen at

cycling loading.
In the case of thermomechanical cyclic loading of the coated specimen, the strain-intensity
range in the coating, where a plane stress state
is observed, can be approximately presented as

a sum of the mechanical component, DeM

ic ,
caused by the external mechanical load applied
to the specimen, and the thermal component,
DeTic, caused by the difference between the coating and blade material thermal expansion coefficients under cyclic heating and cooling of the
specimen:
T
Deic DeM
ic Deic
(Eq 18)
Deic 2DT DT
where
DT T Tc and DT Tmax Tmin.
As a result, the total range of elastic-plastic

strains in the coating under its nonisothermal
loading can be expressed as:
If it is taken that l lc l*c 1/2 (where

l is Poissons ratio for specimen material) for
the developed plastic specimen and coating
straining conditions, then, from Eq 1 to 9, it is
obtained:
For the specimen and coating elastic straining

(if l lc 0.3 is used):
DeM
ic De
Deic 0.87 De 1.24 DT DT
(Eq 19)
DeTic 2|DT(Tmax)(Tmax T0)

DT(Tmin)(Tmin T0)|
where Tmin and Tmax are the minimum and maximum cycle temperatures:
DT(Tmax) T(Tmax) Tc(Tmax)
DT(Tmin) T(Tmin) Tc(Tmin)
If one takes, approximately, for the Tmin

Tmax temperature range that:
DT(Tmin) DT(Tmax) T Tc const
then Eq 18 is of more simple form:
Thermomechanical fatigue (LA-TMF test) for JS6U

superalloy (T 100 } 950 oC, t 1.5 min). 1,
uncoated; 2, Ni10Cr17AlY coating (80 lm); 3, Ni10Cr17AlY
coating to detection of the first crack 0.5 mm long; 4, calculated
relationship Deic Deic(NRC)
Fig. 9.5
Deic De 2DT DT
Thus, if the experimental TMF curve of De

De(NRC) for the coated specimen is available,
then, with the use of dependencies (Eq 18 and
19), the TMF curve of Deic Deic(NRC) may
be plotted for the coating strain-intensity ranges,
which might be directly applied in combination
with the coating stressed-strained model discussed earlier to predict its life.
Figure 9.5 presents the results of strain control
thermomechanical fatigue tests (when De
const) of standard cylindrical specimens 6.5 mm
in diameter with the Ni10Cr17AlY coating. The
specimens are cast from the JS6U superalloy and
have equiaxial structures. Coating application
reduces the lives of the specimens at standard
tests in air by a factor of 1.5 to 2, depending on
strain-range value, De. The figure also illustrates
the relationship for the Ni10Cr17AlY coating
that is plotted on the basis of measurements of
the number of cycles to detection of the first
crack 0.5 mm long in the coating. The period to
detection of such a crack appearing in the coating is 2 to 3 times shorter than the time to coated
specimen fracture.
The calculated plot of the Ni10Cr17AlY coating service life against its strain-intensity range,
Deic Deic(NRC), is obtained with the use of
Eq 18 and 19, based on the experimental relationship of De De(NRC). It can be used to
predict the thermomechanical life of the given
coating applied to a blade, if, when defining a
stressed-strained state of the coating, the calculation method discussed previously is used.
For approximate estimations of protective
coating cyclic lives, when there is a lack of experimental data for coatings, the empirical dependencies used for corrosion-resistant alloys
may be applied. In particular, with this purpose,

Mansons formula (Ref 7) may be used as applied to a coating as follows:

De
3.5 rBc
0.12
NRc
Ec
1n
0.6
1
1 wc
N0.6
Rc .
(Eq 20)
where rBc rBc(T) is coating material ultimate

strength:
wc
dc
1 dc
is the parameter of coating ductility (transverse

cross-sectional area reduction when ruptured).
Stress intensity (ric) in coatings at minimum (1, 3, 5)

and maximum (2, 4, 6) gas temperatures. 1 and 2,
Ni20Co20Cr12AlY coating; 3 and 4, Ni20Cr12AlY coating; 5 and
6, Ni8Co20Cr12AlY coating
Use of Mansons formula enables one to estimate only approximately the number of cycles
to coating cracking. At the same time, this formula is found to be rather convenient for comparison of the calculated assessments of coating
cyclic durability at the preliminary stage of selection of the coatings, which would be best
matched to the given blade by the TMF criterion
(Eq 17).
As an example, Fig. 9.6 to 9.8 show the results of calculated estimates of the stressedstrained state of different condensated coatings
deposited on turbine blades, and Fig. 9.9 shows
the assessed values of their cyclic lives performed from Mansons formula (Ref 7). The
blade was cast of the nickel-base directionally
Fig. 9.6
Strain intensity (eic) in coatings at minimum (1, 3, 5)

and maximum (2, 4, 6) gas temperatures. 1 and 2,
Ni20Co20Cr12AlY coating; 3 and 4, Ni20Cr12AlY coating; 5 and
6, Ni8Co20Cr12AlY coating
Fig. 9.7
Strain-intensity range (Deic) in coating at thermocyclic loading of blade. 1, Ni20Co20Cr12AlY coating; 2, Ni20Cr12AlY coating; 3, Ni8Co20Cr12AlY coating
Fig. 9.8
The calculated number of cycles to coating failure

(NRc) at thermocyclic loading of blade. 1,
Ni20Co20Cr12AlY coating; 2, Ni20Cr12AlY coating; 3,
Ni8Co20Cr12AlY coating
Fig. 9.9

solidified superalloy of MAR-M-200. The calculation is performed for mean blade cross-sectional area for two stationary regimes at minimum and maximum gas temperatures. The blade
surface temperature under these conditions is
shown in Fig. 9.10. The coating strain-intensity
range was defined approximately as the difference of the coating strain intensities at maximum
and minimum gas temperatures from Eq 9.
Stresses and strains in the blade itself were calculated based on the generalized plane-strainstate model and the finite-element method
(FEM), with due regard to anisotropy of the mechanical properties of the directionally solidified
blade material.
The minimum calculated cyclic lives of the
coatings correspond to points 27 to 50 on the
outer blade surface. For the blade under con-
Fig. 9.10
Temperature of blade surface at minimum (Tmin)

and maximum (Tmax) gas temperatures
Table 9.1 Intensities of strains (eic) and stresses

(ric) on trailing edge of turbine blade made of
MAR-M-002 alloy with Ni20Cr12AlY coating
within start-maximum-shut cycle
Temperature, oC
13
216
322
401
726
915
1009
893
785
630
392
248
127
13
eic 103, %
ric, MPa
0.979
1.048
1.054
1.098
1.106
0.341
0.386
0.638
0.840
0.960
0.977
1.013
1.031
0.979
163.5
165.9
162.1
164.9
150.5
43.0
46.1
81.3
112.3
134.5
147.1
159.1
167.1
163.5
sideration, the best coatings, from the cycliclife standpoint, are Ni20Cr12AlY and
Ni8Co20Cr12AlY, ones whose minimum calculated cyclic-life values are similar: NRcmin
3100 cycles for the Ni20Cr12AlY coating
and NRcmin 3800 cycles for the
The test results for blades with the
Ni20Co20Cr12AlY and Ni20Cr12AlY coatings
run in the engine have confirmed the calculationbased conclusions. For the blade under consideration, preference should be given to the
Ni20Cr12AlY
coating
over
the
Ni20Co20Cr12AlY coating, by the coating
TMF criterion considerations. After the first 80
h of engine tests, many cracks (fragmentation)
have appeared on the Ni20Co20Cr12AlYcoated blades, while no cracks have been detected on the Ni20Cr12AlY-coated blades
within 380 running hours.
As follows from Fig. 9.9, the coating servicelife characteristics in the different points of the
blade surface differ greatly because of different
stresses, strains, and temperatures in these
points. That is why coating service lives may be
calculated only for the most loaded points on the
blade surface. The calculation results of stress
and strain intensities arising in the
Ni20Cr12AlY coating on the blade of MAR-M002 alloy in the most loaded point on the trailing
edge are presented as an example in Table 9.1
(Ref 8). The calculations are based on the data
of temperature and elastic strains arising at different moments of the start-maximum-shut cycle
of turbine operation.
For this blade point, the calculations of
stressed-strained state have been carried out, together with estimation of service lives for different coating variants. For calculations, the data
on coating-material properties discussed in previous chapters were used. The thickness of all
the coating variants was 100 lm.
The calculated number of cycles to coating
failure, NRc, was determined from Mansons formula, Eq 20. Strain-intensity range in the coating was determined from Eq 9. For calculations
of NRc values, the coating mechanical properties
corresponding to the maximum blade surface
temperature within the loading cycle were used.
The calculation results are presented in Table
9.2.
As follows from Table 9.2, variations in the
coating chemical composition open up possibilities of considerable changes in its service life
under thermomechanical loading conditions.

The minimum calculated cyclic life of NRc

1800 cycles is typical of the Ni20Co20Cr12AlY
coating, while the maximum cyclic life of NRc
141,000 cycles is demonstrated by the
Ni7Cr12AlY coating. Low calculated value of
NRc for the Ni20Co20Cr12AlY coating, despite
its high ductility, is due to considerable difference between thermal expansion coefficients of
this coating and the MAR-M-002 alloy used as
a blade material. Among the large number of
coatings studied, it is the Ni20Co20Cr12AlY
coating whose thermal expansion coefficient differs to the greatest extent from the thermal expansion coefficient of the blade material in the
temperature range of 20 to 800 oC, which is
characterized by the highest stresses and strains
in the coating.
Long cyclic life (NRc) of the Ni7Cr12AlY
coating is due to quite similar values of the coating and blade material thermal expansion coefficients, especially in the temperature range of
low coating ductility (20 to 600 oC), as well as
to high yield and ultimate strength values of the
coating over its full-service temperature range.
As a result, the Ni7Cr12AlY coating deposited
on the blade is capable of withstanding a considerable number of cycles without cracking, despite its relatively low ductility. Under service
conditions, only elastic strains arise in this coating.
It is noteworthy, however, that the choice of
coating should be guided by the required combination of coating properties. First of all, the
protective properties should be considered. The
Ni7Cr12AlY coating, as has been mentioned in
Chapter 6, Turbine Blade Coating Protective
Properties, has relatively low resistance to hot
corrosion. In choosing it for application in constructions exposed to hot corrosion attack, its
special alloying with such elements as tantalum
and hafnium is required.
Thermal Barrier Coatings

The main causes of a TBC failure are the thermal stresses, which arise at the ceramic layer/
Table 9.2 Calculated number of cycles to crack
appearance, NRc, in protective coatings on trailing edge
Coating
Ni20Co20Cr12AlY
Ni20Cr12AlY
Ni7Cr12AlY
Ni8Cr18AlY
Ni5Cr25Al
NRc, number of cycles

1800
12,500
141,000
76,500
15,500
bond-coat interface, because of the difference

between the thermal expansion coefficients of
the ceramics and the superalloy. Both these
stresses and oxidation at the ceramic/bond-coat
interface result in ceramic layer spallation.
To estimate stresses and strains in different
(ceramic and bond coat) TBC layers, the calculation model described previously for heat resistant metallic coatings can be used. In this case,
the calculation for each TBC layer with the use
of the previously mentioned model is performed
separately (independently of the other layers).
This results in determination of the rZc and rSc
stresses and the eZc and eSc strain components
averaged over the thickness for each TBC layer.
When calculating for the ceramic layer, the temperature averaged over its thickness should be
used. It can differ by 30 to 50 oC or more from
the blade-metal temperature.
The calculation model described previously is
applicable to TBCs, if the error () for this
model determined by Eq 12 for elastic stresses
does not exceed some reasonable limit, for example, 10 to 15%. If the TBC ceramic-layer
stiffness is found to be comparable with the
blade-wall stiffness (which is possible only for
thick coatings, when ceramic-layer thickness is
0.3 to 0.4 mm), then one has to use more complicated calculation models. In particular, the
two-dimensional generalized plane-strain-state
models based on FEM application are effective
for this case. Such models allow natural consideration of the combined effect of the coating and
blade-wall stiffness characteristics, especially if
specific finite elements are involved, for example, finite elements in the form of curvilinear
strips intended for the calculation of stresses and
strains in thin-surface blade layers (Ref 9).
Thermal Barrier Coating Life Estimation.
Criterion of TBC Ceramic-Layer Cracking.
Spalling of ceramic-layer fragments from a
blade surface is the main type of TBC mechanical failure. The occurrence of a network of microcracks in the ceramic layer (fragmentation)
in some blade-surface zones is considered to be
the initial stage of TBC failure. At the same
time, ceramic cracking does not yet mean its
failure. Safe operation for a long time period can
be possible for blades having such cracks
formed in the TBC ceramic layer usually within
the first hours or tens of hours of turbine operation.
The condition for TBC cracking can be presented as follows (Ref 9):

rmax rcr
(Eq 21)
where rmax is the maximum tensile stress in the

center of the fragment of cracked ceramic layer/
metallic bond-coat interface (point B in Fig. 9.11
and 9.12), and rcr rcr(T) is the cohesive ultimate strength of the ceramic layer (minimum
tension stress bringing about instant ceramic
cracking at the temperature of T). The subscript
cr symbol stands for a TBC ceramic layer.
Figure 9.12 illustrates the plot of the maximum relative tensile thermal stress in the center
of the ceramic-layer fragment at the moment of
its cracking versus the relative size, d/hcr, of
these fragments:
rmax
At the moment when the ceramic layer is being cracked, the stress (rmax) is as high as the
cohesive ultimate strength (rcr), that is
rcr rmax f (d/hcr)
Ecr (T Tcr) (T T0)
1 lcr
(Eq 22)
Using this formula and the plot given in Fig.
rmax (1 l)
f (d/hcr)
Ecr (T Tcr) (T T0)
where d is fragment spacing and hcr is thickness

of a TBC ceramic layer.
This dependence is obtained from calculations using the two-dimensional FEM for elastic
thermal stresses in the three-layer plate, whose
top (ceramic) layer has a regular network of
cracks with a fragment size designated as d. The
calculation is performed for a two-layer TBC of
ZrO2-8%Y2O3/Ni20Cr12AlY (100/80 lm) deposited by electron beam technique.
Fig. 9.11
Fig. 9.12
Maximal relative stress (r*max) versus relative size

(d/hcr) of ceramic layer fragment
Thermal stresses in cracked ceramic layer of TBC. (a) Finite-element grid. (b) Thermal stress isolines (MPa). (c) Thermal
stress variations along blade-wall thickness in the center of the cracked ceramic-layer fragment

9.12, one can determine the value of the ceramic-layer cohesive ultimate strength, rcr, for
the specified temperature conditions, if the mean
size of the cracked fragments of the ceramic
layer and the ceramic temperature, T, at its
cracking are defined.
Obviously, cracking of the ceramic layer is
possible if tensile thermal stresses are generated
in it. It can take place only when the coating
temperature, T, is higher than the annealing temperature, T0, (as for ceramic layers Tcr T),
and in cases where considerable creep strains are
accumulated in the blade-metal surface layer and
in the metallic bond coat, which are capable of
giving rise to tensile stresses in the ceramic
layer.
If the rcr values for different service conditions are known, then, with the use of Eq 22, the
plot in Fig. 9.12, and Eq 21, it is possible to
Fig. 9.13
Model of bond-coat surface oxidation
Fig. 9.14
Maximal principal stress (r1) in ZrO2-8%Y2O3 ceramics in the vicinity of Al2O3 ridge (MPa)
estimate by calculations the probability of TBC

cracking and even to define the mean size of the
ceramic layer fragments that appear on cracking.
Criterion of TBC Ceramic-Layer Spallation.
Thermal barrier coating failure results from the
decrease in adhesive strength at the ceramic/
bond-coat interface during long-term service at
high temperatures, causing ceramic-layer spalling (Ref 10). This effect is due to the formation
of the oxide of Al2O3 type at the ceramic/bondcoat interface. Aluminum oxide not only reduces
adhesive strength but also changes the stressed
state at the ceramic/metallic bond-coat interface,
because the properties of this oxide are different
from those of the ceramic and bond-coat layers.
To simulate the damaged ceramic layer, it may
be thought that Al2O3 formation does not occur
uniformly over the ceramic/bond-coat interface,
but by the formation of the specific ridges intruding into the ceramic layer (Fig. 9.13). The
dimensions of these ridges and their spacing depend on ZrO2 Y2O3 ceramic-layer crystal sizes,
ceramic porosity, blade-surface roughness, and
so on. However, the most intensive growth of
these ridges is observed at the base of fragmentation cracks in the ceramic layer.
Formation of Al2O3 occurs at a high temperature, and while cooling, additional internal thermal stresses are generated in the ceramic layer
near the ridges, which are caused by the difference between the values of the thermal expansion coefficients for the TBC ceramic layer and
Al2O3 oxide.
As an example, Fig. 9.14 illustrates the results
of calculations done using a FEM for the twodimensional problem of joint straining of two
semiinfinite bodies of ZrO2 and Al2O3 materials
uniformly cooled to the same temperature of 20
o
C and joining over the contact surface. At the
interface of these materials, in the ZrO28%Y2O3 body, there is a ridge of Al2O3 oxide in
the form of a wedge 1 lm high and 2 lm wide
at its base. The calculation used the following
initial data: T0 1000 oC; for ZrO2-8%Y2O3
E 10 GPa, T 9.3 106 1/oC; lcr
0.3; for Al2O3 E 40 GPa; 8.0 106
1/oC; l 0.3.
As is seen from the offered model, extra concentration of thermal stresses in the ceramic
layer near the Al2O3 oxide ridge facilitates both
ceramic-layer spalling from the substrate in the
vicinity of point A at the ridge base, and the
opening of a crack formed earlier in the ceramics
or a new crack nucleating between the adjacent
columnar crystallites of ZrO2-8%Y2O3 in the vicinity of point B at the ridge top (Fig. 9.14).

As an integral characteristic describing the

failure mechanics of the TBC ceramic layer, the
adhesive strength value racr racr(T, s) defined
experimentally can be used. Here, racr is the
minimum stress of ceramic-layer separation
from the substrate at room temperature after
holding the specimen in a furnace at the temperature of T for a period of time, s. In order to
take into account the effect of thermocyclic
loading on TBC adhesive strength, the experiments on determination of the racr characteristic
can be performed with simulation of blade-service thermocycles by means of periodic specimen cooling to room temperature followed by
heating to the test temperature.
The criterion of TBC adhesive strength can
be presented as:
s sRcr
where sRcr is time of operation at high temperature before spalling of TBC ceramic layer from
the blade surface after cooling to room temperature, and s is time of blade operation at high
temperature.
Time to ceramic-layer spalling, sRcr, can be
defined from the experimental relation of ceramic-layer adhesive strength to its test time and
temperature. They are shown in Fig. 9.15. The
TBC adhesive strength curves can cross the time
axis. Ceramic spalling caused by formation of
Al2O3 oxide at the ceramic/bond-coat interface
can also take place at zero stress level in the
ceramic layer at low blade temperatures.
The stresses rncr, which facilitate TBC ceramic-layer separation from the blade surface
(Fig. 9.16), may be approximately defined by the
formula:
rncr
Fig. 9.15
Fig. 9.16
(Eq 23)
hcr
rscr
R
(Eq 24)
Adhesion strength curve
(a) Stresses of spalling (rncr) and (b) rule of signs for

these stresses
where R is a radius of curvature of the outer

surface of the blade airfoil in its cross-sectional
area. The values of rncr stresses are the greatest
in the blade leading-edge zone, where the value
of radius of curvature, R, is minimum.
Ceramic layers of TBC are capable of withstanding high compressive thermal stresses
caused by the difference between the thermal
expansion coefficients of the ceramics and the
blade material. However, with the occurrence of
even minor tensile thermal stresses in the ceramic layer, this layer experiences immediate
cracking, resulting in tensile stress relaxation.
Thus, if the calculation performed gives the tensile stresses rZcr or rScr, in the ceramic layer,
these stresses should be taken as zero.
In the temperature range below the temperature of ceramic-layer deposition on the blade
surface, only compressive stresses, rScr, in convex surface layers of the blade (Fig. 9.16), such
as its suction face and leading edge, can give
rise to spalling of ceramic-layer fragments. In
concave blade-surface layers (such as its pressure face), these compressive stresses press the
ceramic layer to the blade.

The following rule of signs for the stresses of

layer separation rncr determined by Eq 24 can
be used. Under the compressive stresses, rncr, in
the ceramic layer, the stresses of separation, rScr,
in convex blade-surface layers are positive, and
those in concave blade-surface layers are negative.
If the coating is affected by the positive stress,
rncr, (which separates the coating from the blade
surface), time to coating failure (spalling of ceramic layer), sRcr, can be found from the adhesive strength curves racr racr(T, s) at the specified stress of separation, rncr (point A, Fig.
9.15). If the coating is affected by the negative
stress, rncr, (which presses the coating to the
blade), then, in this case, one can assume that
rncr 0 and define the time to ceramic spalling,
sRcr, as the point where the appropriate adhesive
strength racr(T, s) curve crosses the axis s (point
B, Fig. 9.15). If no effect of rncr stress on the
ceramic layer is observed, that is, rncr 0, then
sRcr is also defined by point B in Fig. 9.15.
Stresses in the TBC ceramic layer vary with
time, mainly due to the accumulation of creep
strains in the blade. This process is most intensive in the leading- and trailing-edge zones,
where the blade material temperature is maximal. The residual thermal stresses in the ceramic
layer also vary with the blade running time.
As an example, the results of durability calculations for a TBC ceramic layer on a turbine
Fig. 9.17
blade are presented subsequently. The CMSX-4

single-crystal alloy was used as a blade material.
The ceramic layer was deposited by electron
beam technique. The TBC consists of two layers: Ni20Cr12AlY bond layer (80 lm) and
ZrO2-8%Y2O3 ceramic layer (100 lm).
The steady-state temperature field at the blade
leading edge is shown in Fig. 9.17 (a) and (b).
The stressed-strained state calculation in the
blade under stationary thermal conditions is carried out with the use of a generalized plane (twodimensional) strain-state model and the FEM.
Triangular finite elements with a cubic approximation of form functions for displacements
were used. The calculation is carried out with
account of anisotropy of elastic, plastic, and
creep properties of the single-crystal alloy. The
calculations for the coating bond coat are carried
out using the models described in this chapter.
The calculation results are shown in Fig. 9.18,
9.19, and 9.20.
Maximum thermal stresses in TBC ceramic
layer arise on cooling the blade to room temperature (20 oC) immediately after coating deposition and subsequent high-temperature annealing of the coated blade (T0 950 oC). In
this case, compressive residual stresses of rres
rZcr rScr 400 MPa arise in the TBC
ceramic layer at its interface with the metallic
bond coat, and tensile residual stresses of rres
560 MPa arise in the bond coat. In this case,
(a) Stationary thermal field and (b) temperature distribution over blade leading edge

Fig. 9.18
Actual stresses (rScr) in ceramic layer of TBC on

the blade leading edge
Fig. 9.19
Residual stresses (rScr) in ceramic layer of TBC on

the blade leading edge
Adhesion strength, racr, (full lines) and stress of

spallation, rncr, (dotted line) versus time, s, in
blade with TBC after cooling to room temperature
Fig. 9.20
the stresses separating the ceramic layer from

the blade surface are maximum and account for
rncr 50 MPa.
At blade heating, the internal elastic compressive stresses in the TBC ceramic layer decrease; under the steady-state operating conditions, they become tensile in the blade
leading-edge zone and reach the maximum values of rScr rZcr 70 75 MPa in points
6 and 7 of the blade surface at the initial time
moment of s 0 (Fig. 9.18, dotted lines).
When taking, for example, the racr 15 MPa
value as the TBC ceramic layer cohesive
strength for the blade leading-edge thermal conditions, then the indicated TBC surface zone
should be covered with a network of microcracks. Their spacing is 100 lm in accordance
with the plot given in Fig. 9.15. Under these
conditions, the actual tensile stresses, rscr and
rzcr, in the ceramic layer on the blade leading
edge (points 4 to 9, see Fig. 9.17 and 9.18) relax.
In this case, the plots for actual stress, rscr, in
Fig. 9.18 between points 4 and 9 are full straight
lines, rscr 0 parallel to the abscissa.
At long-term running, creep strains are accumulated in the blade body, resulting in the decrease in residual stresses arising in the TBC
ceramic layer at blade cooling. In the example
being considered, on reaching running time of s
300 h, these residual stresses in the leadingedge zone in points 6 and 7 are reduced to rScr
rres 200 MPa (Fig. 9.19).
With decreasing residual stresses, rScr, the
stresses of rncr determined by Eq 24, which
separate the ceramic layer from the blade, are
also being reduced. The plot of ultimate adhesive strength racr racr(T, s) variation versus
running time, s, under the service conditions in
the most hazardous point, 6, of the highest
blade-metal temperature of T 1120 oC is presented in Fig. 9.20, together with the plot of the
stress of separation, rncr, in this point at cooling
the blade to room temperature (dotted line).
This plot shows that the rate of reducing the
stress of separation, rncr, in point 6 versus running time is less than the rate of reducing the
adhesive strength, racr, in this point. The curves
racr(s) and rncr(s) in Fig. 9.20 intersect at the
time moment of s 265 h.
Heat resistant bond-coat life may be estimated
using the approaches discussed in this chapter.
It is noteworthy that the calculation results are
approximate. For more reliable calculation of estimate of time to TBC spalling, one should have
experimental data on modulus of elasticy, Ecr,

for the ceramic layer at its compression. The

value has a noticeable effect on the level of calculated stresses of separation, rncr, and on the
calculated service life of the TBC ceramic layer
in the leading-edge zone of the blade. The experimental studies of the dependence of adhesive strength of the TBC ceramic layer on test
time and temperature are also required.
6.
REFERENCES
1. Y.A. Nozhnitsky, R.A. Doulnev, and V.G.

Sundyrin, Damage Mechanisms for Thermomechanical Fatigue of Aircraft Engines
Material, 81st Meeting of the AGARD SMP
(Banff, Canada), 24 Oct 1995, and published in CP-559
2. I.A. Birger and B.F. Chorr, Thermal Fatigue of Engine Details, Moscow, Mashinostroene, 1975, 455 p
3. I.A. Birger and N.I. Koterov, Ed., Strength
Calculations for Aircraft Gas Turbine Engines, Moscow, Mashinostroene, 1984,
208c
4. V.K. Kostege, V.A. Khalturin, and V.G.
Sundyrin, Mathematical Models of Gas
Turbine Engines and Their Components,
AGARD-LS-198, Advisory Group for
Aerospace Research and Development, Dec
1994 (France), North Atlantic Treaty Organization
5. Y.A. Tamarin, E.B. Kachanov, and V.G.
Sundyrin, Gas Corrosion and Thermal Fatigue of Protective Coatings for Turbine
7.
8.
9.
10.
Blades, High Temperature Corrosion and

Protection, H. Guan, W. Wu, J. Shen, and
T. Li, Ed., Liaoning Science and Technology Publishing House, China, 1991, p 161
166
Y.A. Tamarin, V.G. Sundyrin, and N.G.
Bychkov, Thermo-Mechanical Fatigue
Tests of Coatings for Turbine Blades, High
Temperature Surface Engineering, J. Nicholls and D. Rickerby, Ed., IOM Communications Ltd., 2000, p 587593
S.S. Manson, Thermal Stress and Low-Cycle Fatigue, New York, 1964, p 187213
Y.A. Tamarin, H.C. Low, Thermal Fatigue
of Protective Coatings on CMSX-4 Superalloy, High Temperature Corrosion and
Protection of Materials5, R. Streiff, I.G.
Wright, R.C. Krutenat et al., Ed., Trans
Tech Publication, Switzerland, 2001, p
647654
V.G. Sundyrin, Y.A. Tamarin, and V.J.
Kanaev, Stress Calculation of the Ceramic
Thermal Barrier Coatings for the Cooled
Turbine Blades, Mechanics of Creep Brittle
Materials 2, A.C.F. Cocks and A.R.S.
Poner, Ed., Elsevier App. Science, 1991, p
308313
Y.A. Tamarin and V.G. Sundyrin, Adhesive
Strength and Life of TBC-EB Ceramic
Layer, Proc. Intern. Symposium on High
Temperature Corrosion Protection (Abstracts Booklet), SEFRACOR, Universite
de Provence, EPRI, Les Embier, France,
May 1996

Publication title
Product code
06738G

Fax 1-440-338-4634
Mail


www.ameritech.co.uk
Neutrino Inc.
Telephone: 81 (0) 424 84 5550
ASM International.

DOI: 10.1361/pctb2002p211

APPENDIX
Regression Equation
FOR SIMULATION of the relationship and
plotting models for the properties of the alloys
for the coatings depending on their chemical
compositions, the methods of regression analysis were used (Ref 1). They included methods
of experiment planning and an application package of Stark (statistical analysis regression
correlation).
The models for a Ni-Co-Cr-Al system were
plotted using the data on the properties of the
alloys, whose chemical compositions corresponded to the two-power plan matrix of B4 (Table A.1). To plot models for Co-Ni-Cr-Al and
Ni-Cr-Al systems, the passive statistic data on
Table A.1
the properties of the alloys, whose chemical

compositions are given in Tables A.2 and A.3,
were used.
Standardly normalized magnitudes of variables (x1 xn) are given in Tables A.1, A.2,
and A.3. They are correspondingly used in calculated regressions:
xi
xie xia
Dxi
where xi is normalized magnitude of variable i;

xie is actual magnitude of variable;
Ni-Co-Cr-Al alloys
Chemical
composition, wt%
Normalized number
of variable
No.
Alloy
Ni
Co
Cr
Al
x1
x2
x3
x4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Ni10Co15Cr8Al
Ni30Co15Cr8Al
Ni10Co30Cr8Al
Ni30Co30Cr8Al
Ni10Co15Cr14Al
Ni30Co15Cr14Al
Ni10Co30Cr14Al
Ni30Co30Cr14Al
Ni10Co15Cr8AlY
Ni30Co15Cr8AlY
Ni10Co30Cr8AlY
Ni30Co30Cr8AlY
Ni10Co15Cr14AlY
Ni30Co15Cr14AlY
Ni10Co30Cr14AlY
Ni30Co30Cr14AlY
Ni10Co22Cr11AlY
Ni30Co22Cr11AlY
Ni20Co15Cr11AlY
Ni20Co30Cr11AlY
Ni20Co22Cr8AlY
Ni20Co22Cr14AlY
Ni20Co22Cr11Al
Ni20Co22Cr11AlY
Ni20Co22Cr11AlY
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
10
30
10
30
10
30
10
30
10
30
10
30
10
30
10
30
10
30
20
20
20
20
20
20
20
15
15
30
30
15
15
30
30
15
15
30
30
15
15
30
30
22.5
22.5
15
30
22.5
22.5
22.5
22.5
22.5
8
8
8
8
14
14
14
14
8
8
8
8
14
14
14
14
11
11
11
11
8
14
11
11
11
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.3
0.3
0.3
0.3
0.3
0.3
0.1
0.5
0.3
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
1
1
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
1
1
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
1
1
0
The alloys include 1.5% Hf, 1.0% W, 0.5% Nb, and 0.5% Si.
Table A.2
Co-Ni-Cr-Al alloys
Normalized number of variable
No.
Alloy
Co
Ni
Cr
Al
x1
x2
x3
x4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Co10Ni15Cr6AlY
Co30Ni15Cr6AlY
Co10Ni30Cr6AlY
Co30Ni30Cr6AlY
Co10Ni15Cr12AlY
Co30Ni15Cr12AlY
Co10Ni30Cr12AlY
Co30Ni30Cr12AlY
Co10Ni22Cr9AlY
Co30Ni22Cr9AlY
Co20Ni15Cr9AlY
Co20Ni30Cr9AlY
Co20Ni22Cr6AlY
Co20Ni22Cr12AlY
Co20Ni22Cr9AlY
Co20Ni22Cr9AlY
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
10
30
10
30
10
30
10
30
10
30
20
20
20
20
20
20
15
15
30
30
15
15
30
30
22.5
22.5
15
30
22.5
22.5
22.5
22.5
6
6
6
6
12
12
12
12
9
9
9
9
6
12
9
9
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.6
0.6
0.2
0.2
0.1
0.1
0.5
0.3
0.3
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
1
1
1
1
1
1
1
1
0
0
1
1
0
0
0
0
1
1
1
1
1
1
1
1
0
0
0
0
1
1
0
0
0.2
0.2
0.2
0.2
0.2
0.2
0.2
1
1
0.6
0.6
1
1
0.2
0.2
1
xia is average magnitude of variable in the

range under consideration; and
Dxi is range of varying the variable during
experiment.
Processing of experimental results makes it
possible to obtain mathematical models connecting system properties and alloying element
contents (regression equations) in the form of
square polynomials, which may be presented in
general as a series:
n
y bon
bi xi i1j
bij xi xj
i1
n
bii x i2 . . . .
S2R
(yci yei)2
i1
where yei is actual value (average experimental

magnitude of the property) in i-point;
yci is magnitude of the property under study
calculated from the model equations (response
value) in this point; and
n is number of experiments in the plan matrix.
In addition to the above, the following criteria
were used for selection and statistic analysis of
regression equations.
Multiple correlation coefficient (Rf) corrected on the basis of degree of freedom:
i1
The procedure of increment inclusion of the

terms in the model was used to select the best
regression because it allowed maximal reduction
of the residual sum of squares:
Table A.3
Ni-Cr-Al alloys
Chemical
composition,
wt%
Normalized
number of
variable
No.
Alloy
Ni
Cr
Al
x1
x2
1
2
3
4
5
6
7
8
9
10
11
12
Ni22Cr11AlY
Ni14Cr3AlY
Ni18Cr5AlY
Ni16Cr9AlY
Ni21Cr9AlY
Ni18Cr16AlY
Ni19Cr12AlY
Ni8Cr18AlY
Ni8Cr15AlY
Ni7Cr12AlY
Ni7Cr10AlY
Ni20Cr6AlY
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
Base
22.3
14.7
17.7
16.0
20.8
18.3
19.5
8.0
8.0
7.0
7.2
20.0
11.2
2.8
5.5
9.1
8.8
15.7
12.6
18.0
15.0
12.0
10.0
6.0
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.97
0
0.38
0.16
0.78
0.45
0.61
0.87
0.87
1
0.97
0.67
0.07
0.97
0.71
0.18
0.18
0.73
0.21
1
0.60
0.21
0.05
0.57
Rf
1 (1 R2) (n 1)
n k
where R is assessment of multiple correlation

coefficient for the effects of the factors included
in the model at j-increment;
k is number of independent assessments of coefficients in the model of j-increment calculation; and
n is number of experiments included in the
plan matrix.
Accumulated cancelled ratio (ACR) of variances of the given increment of model selection to the sum of variances cancelled at
the increments from the first to the current
one. The regressions, for which the variance of
nonadequacy (S2na) seemed to be minimal, multiple correlation coefficient was close to 1, and
the current increment variance cancelled was
comparable with an experiment error, were chosen as the best ones.
Appendix / 213
Ni-Co-Cr-Al alloys
Variability of values
Variable
Co
Cr
Al
Y
T, oC
Symbol
min
max
X1
X2
X3
X4
X5
10.10
14.40
7.80
0.04
30.20
30.00
14.10
0.45
Liquidus temperature
Regression equation:
Tliq 1366.6 9.8 X2 65.6 X3 42.7 X12
14.6 X1 X2 X3, (oC)
2
S na 239.8; (DSj2) 0.94; Rf 0.97
Solidus temperature
Tsol 1275.2 6.6 X3 10.7 X12 18.4
X32 21.3 X2 X3 12.8 X1 X2 X3, (oC)
2
S na 93.7; (DSj2) 0.88; Rf 0.94
Crystallization interval
DT 90.9 10.1X2 60.7X3 52.0X12
16.0X32 3.6X1X3 27.2X2 X3, (oC)
2
S na 73.0; (DSj2) 0.98; Rf 0.99
NiAl (b-phase) quantity
S2na
0.62X12 0.20X22 0.14X32

0.11X1X2X3, (106, K1)
0.01; (DSj2) 0.93; Rf 0.96
Thermal expansion coefficient, range 100

500 oC
15.0 0.14 X1 0.22 X2 0.54 X12
0.20X22 0.08X2X3 0.09 X1 X2 X3,
(106, K1)
2
S na 0.03; (DSj2) 0.88; Rf 0.93
600 oC
15.5 0.22 X1 0.31 X2 0.47 X12
0.25 X22, (106, K1)
2
S na 0.03; (DSj2) 0.89; Rf 0.94
700 oC
16.3 0.27 X1 0.40 X2 0.17 X3
0.64 X12 0.31 X22 0.08 X1 X3,
(106, K1)
2
S na 0.06; (DSj2) 0.89; Rf 0.94
800 oC
Qb 60.3 3.8X1 13.0X2 20.8X3
9.5X12 6.6X32 2.5X1X3 11.2X2X3,
(vol%)
S2na 14.5; (DSj2) 0.97; Rf 0.98
17.4 0.19 X1 0.50 X2 0.28 X3
0.88 X12 0.43 X22, (106, K1)
2
S na 0.09; (DSj2) 0.89; Rf 0.94
NiAl (b-phase) quantity, annealing 1200 oC,

200 h

900 oC
Qb 63.6 3.9X2 25.0X3 7.3X12
8.6X22 7.6X32 3.5X1X3 11.3X2X3,
(vol%)
S2na 19.4; S (DSj2) 0.96; Rf 0.98
18.3 0.64 X2 0.48 X3 0.99 X12
0.59 X22 0.17 X1X2
0.20 X1 X3 0.31 X1 X2 X3, (106, K1)
S2na 0.07; (DSj2) 0.95; Rf 0.97

300 oC

1000 oC
14.01 0.15 X2 0.06 X3 0.72 X12
0.22 X32 0.08 X1 X3 0.16 X1 X2
X3, (106, K1)
2
S na 0.01; (DSj2) 0.92; Rf 0.96
Y 19.3 0.31X1 0.80X2 0.65X3
0.96X12 0.66X22 0.19X1X2
0.21X1X3 0.54X1 X2 X3, (106, K1)
S2na 0.03; (DSj2) 0.98; Rf 0.99

400 oC
Elasticity modulus, 20800 oC
14.5 0.10X1 0.17X2 0.07X3
E 1515 118.7X1 107.2X2 41.7X3
134.6X5 85.6X22 100.6X32
37.2 X52 50.6 X1 X2 69.7 X1 X3

20.6 X1 X5 2.8 X2 X3 7.4 X2 X5
52.5X3X5 25.9X1 X2X3 14.0X1X2X5
12.0X2X3X5 (100, MPa)
S2na 220; (DSj2) 0.99; Rf 0.99
Hot corrosion, 850 oC, 200 h
q 200.6 22.1X2 223.3X12 75.3X22
81.5X32 77.5X42 123.2X1X2
45.4 X1 X3 130.9 X2 X3 77.7 X2 X4
38.1 X3 X4 66.2 X1 X2 X3
87.8 X1 X2 X4 167.5 X1 X2 X3 X4 (g/m2)
S2na 653; (DSj2) 0.98; Rf 0.99
o
Hot corrosion, 900 C, 200 h

log q 1.868 0.224 X12 0.189 X22
0.303 X32 0.089 X1 X3, (g/m2)
2
S na 0.015; (DSj2) 0.70; Rf 0.84
Heat resistance, isothermal test at 1200 oC,
100 h
q 30.1 6.56X1 2.83X2 18.74X3
20.74X4 4.50X12 0.07X22
24.50 X42 9.52X1X2 12.38X1X3
5.91X1X4 35.16X2X3 3.86X2X4
14.29X3 X4 13.32 X1 X2 X3 13.14 X1
X2 X4 22.29 X1 X3 X4
25.92 X2 X3 X4 29.31X1 X2 X3 X4, (g/
m2)
S2na 28; (DSj2) 0.96; Rf 0.97
o
Heat resistance, isothermal test at 1200 C,

300 h
q 62.6 13.60X1 4.69X2 37.37X3
37.80X4 8.98X12 28.71X22
29.42X32 34.84X42 11.05X1X2
2.39X1X3 22.28X1X4 59.72X2X3
6.57X2X4 10.21X1X2X3 21.88X1X2X4
47.49X1X3X4 25.68X2X3X4
58.37X1X2X3X4, (g/m2)
S2na 20; (DSj2) 0.99; Rf 0.99
b-phase-free zone after heat resistance testing
at 1200 oC, 200 h
Hb 69 9.2X1 14.1X2 47.5X3
65.5X12 28.9X22 46.2X32 4.8X42
7.1X1 X2 16.3 X1 X4 7.2 X2 X3 11.7

X1 X2 X3 13.4 X2 X3 X4
28.3X1X2X3X4, (lm)
S2na 59; (DSj2) 0.99; Rf 0.99
Co-Ni-Cr-Al alloys
Variable
Symbol
Variability
of values
min
max
Ni
Cr
Al
X1
X2
X3
10.60
14.30
5.90
31.00
30.20
12.00
Y
T, oC
X4
X5
0.1
0.7

400 oC
14.5 0.21 X1 0.68 X2 0.32 X3
0.21 X1 X2 0.14 X1 X3
0.56X2X3 0.26X1X2X3, (106, K1)
S2na 0.04; (DSj2) 0.95; Rf 0.97
500 oC
14.8 0.22X1 0.76X2 0.36X3
0.16X1X2 0.62X2X3 0.32X1X2X3,
(106, K1)
2
S na 0.03; (DSj2) 0.96; Rf 0.97
600 oC
15.4 0.22X1 0.78X2 0.42X3
0.31X1X2 0.18X1X3 0.62X2X3
0.27X1X2X3, (106, K1)
S2na 0.03; (DSj2) 0.97; Rf 0.98
700 oC
16.0 0.27X1 1.01X2 0.37X3
0.69X22 1.05X1X2 0.68X1X3
0.29X2X3, (106, K1)
S2na 0.28; (DSj2) 0.93; Rf 0.96
800 oC
16.0 0.96X2 0.67X12 0.83X22
1.4X1X2 0.75X1X3 0.35X2X3,
(106, K1)
2
S na 0.52; (DSj2) 0.88; Rf 0.94
Appendix / 215

900 oC
16.1 0.99X2 0.84X12 0.73X22
1.24X1X2 0.64X1X3 0.32X2X3,
(106, K1)
S2na 0.51; (DSj2) 0.87; Rf 0.93
1000 oC
16.5 0.93X2 0.52X12 0.31X22
0.62X32 1.01X1X2 0.62X1X3,
(106, K1)
2
S na 0.31; (DSj2) 0.89; Rf 0.94
E 1626 99.9X1 98.4X2 103.9X3
254.1X5 113.4X12 109.2X22
48.4X32 67.8X52 70.9X1X2
48.3X1X3 22.3X1X5 125.1X2X3
31.1X2 X5 55.4 X3 X5 59.9 X1 X2X 3
11.9 X1 X2 X5 24.0X2 X3 X5, (100,
MPa)
S2na 1639; (DSj2) 0.96; Rf 0.98
Ultimate strength, 7001100 oC
rb 16.7 1.67X1 22.93X5 2.52X22
2.55X32 9.90X52 1.41X3X5, (10,
MPa)
S2na 7.52; (DSj2) 0.98; Rf 0.99
Yield strength, 700900 oC
r0.2 24.3 2.50X1 4.29X2 0.24X3
12.36X5 1.46X12 5.97X22
16.81X32 1.53X2X5 36.36X1X2X5
33.31X1X3X5, (10, MPa)
S2na 0.90; (DSj2) 0.99; Rf 0.99
Relative elongation, 7001100 oC
d 55.1 12.40X1 5.93X2 16.89X5
10.05X12 10.43X52 6.74X1X5
46.30X1X2X5 56.02X1X2X3X5, (%)
S2na 47.15; (DSj2) 0.93; Rf 0.96
q 0.57 0.13 X1 0.21 X2 0.29 X12
0.42 X42 0.28 X1X2
0.8424501 X2 X3 0.56 X1 X2 X3 X4, (g/

m2)
S2na 0.08; (DSj2) 0.85; Rf 0.92
q 9.56 2.34X1 2.93X2 1.94X1X2
2.93X1X3 4.20X2X3
3.36X1X2X3, (g/m2)
S2na 18.85; (DSj2) 0.80; Rf 0.89
q 31.11 8.73X1 17.24X2 7.96X3
8.38X12 2.09X22 13.91X1X2
15.54X1X3 11.73X2X3 13.17X1X2 X3,
(g/m2)
2
S na 15.93; (DSj2) 0.95; Rf 0.97
q 36.43 3.88X1 8.31X2 6.99X3
20.36X22 6.73X32 5.36X1X2
16.29X1X3 7.05X1X2X3, (g/m2)
S2na 85.33; (DSj2) 0.87; Rf 0.93
Heat resistance, cyclic test 20 } 900 oC,
1000 h
q 8.48 (11.01 X2 6.56 X12 3.78 X22
0.79 X32 5.40 X42 14.62 X2 X4,
(g/m2)
2
S na 0.06; (DSj2) 0.99; Rf 0.99
5000 h
q 11.88 9.22X2 1.82X4 2.72X22
4.06X42 6.06X2X4
3.59X1X2X3X4 , (g/m2)
S2na 0.65; (DSj2) 0.99; Rf 0.99
8000 h
q 12.05 23.48 X4 16.08 X42 18.52
X1 X2 11.40 X1 X2 X3
29.50X1X2X4 8.85X2X3X4, (g/m2)
S2na 0.59; (DSj2) 0.98; Rf 0.99
S2na 0.31; (DSj2) 0.74; Rf 0.86
Ni-Cr-Al alloys
Variable
Cr
Al
T, oC
Symbol
X1
X2
X3
Variability of
values
min
max
7.00
2.80
22.5
18.00

300 oC
13.4 0.66 X1 1.14 X2, (106, K1)
S2na 0.12; (DSj2) 0.77; Rf 0.87
400 oC
13.84 0.63 X1 1.08 X2, (x106, K1)
S2na 0.13; (DSj2) 0.73; Rf 0.85
500 oC
13.8 0.63 X1 1.08 X2, (106, K1)
S2na 0.13; (DSj2) 0.73; Rf 0.85
600 oC
14.3 0.59 X1 1.02 X2 0.49 X22,
(106, K1)
S2na 0.26; (DSj2) 0.68; Rf 0.82
700 oC
14.5 0.70 X1 1.45 X2 0.59 X12
0.53 X22, (106, K1)

800 oC
14.7 1.01 X1 2.12 X2 0.85 X12,
(106, K1)
S2na 0.59; (DSj2) 0.67; Rf 0.82
900 oC
15.0 1.06 X1 2.17 X2 0.90 X12,
(106, K1)
2
S na 0.83; (DSj2) 0.60; Rf 0.77
1000 oC
16.5 1.25 X2 0.63 X1 X2, (106,
K1)
2
S na 1.00; (DSj2) 0.46; Rf 0.68
E 1557 87.5X1 183.1X2 122.8X3
92.2X12 187.5X22 22.2X32
414.5X1X2 114.3X2X3 66.8X1X2X3,
(100, MPa)
S2na 2754; (DSj2) 0.94; Rf 0.97
REFERENCE
1. Y.M. Doljansky, G.B. Stroganov, and R.E.

Shalin, Optimization of Machine-Building
Material Properties with a Computer, Moscow, Voenizdat, 1979, 240 p

Index
A
Accumulated cancelled ratio (ACR), of variances
of the given increment of model selection to
the sum of variances cancelled at the
increments from the first to the current
one ................................................ 212
ACR. See Accumulated cancelled ratio.
Activity coefficients of each component ..........39
Adhesion, of overlay coatings, testing of ..........52
Adhesive strength
of ceramic layer/bond coat ...................... 162
of ceramic layer of thermal barrier
coatings ....... 187(F), 188189(F), 190, 206
of thermal barrier coatings ....................2122
of thermal barrier coatings, criterion of .. 206, 207
Aging, with aluminizing ..............................29
Aircraft turbine blades, superalloy chemical
compositions ............................. 10, 11(T)
Air plasma spraying of powders
(TBC-APS) ..................................... 162
Alitizing .................................................18
Alkaline metal salts ................................... 6
Alloying factor (K) ...............................1819
Alpha-aluminum oxide ................. 83, 101, 189
of diffusion coating/superalloys system surfaces,
heat resistance testing results .............. 106
formation on superalloy surface ............. 81, 83
Alpha-aluminum oxide chromium oxide,
formation with overlay coatings on JS6U
superalloy ................................... 109(T)
Alpha-chromium phase
of nickel-chromium-aluminum system .. 87, 88(F)
in nickel-cobalt-chromium-aluminum alloys ....70
Alpha-molybdenum phase, of NiAl ............. 101
Alpha-tungsten phase, of NiAl ................... 101
Alumina/chromia coatings, to protect superalloys
from oxidation during 10,000 hours ...1112
Alumina-type oxides, formation on ceramic-metal
interface with thermal barrier coatings ........22
Aluminides
formed, by aluminizing nickel alloys ............55
heat resistance testing ............... 100106(F,T)
Aluminizing ............................................. 8
with aging ............................................29

aluminum content decrease with reused mixture
applications ....................................30
annealing of powders to increase stability under
H2 atmosphere .................................29
causes .................................................26
circulation technique for diffusion coating
application ............................. 3132(F)
cleaning for preparation before ...................29
halide activator amount affecting
thickness .................................... 26(F)
nickel alloys, aluminides formed .................55
outcomes .............................................26
powder method of application ....................30
in powder mixtures with activators ...............25
and silicification ...................................... 1
temperature range with halide activators ........25
time considerations .................................29
Aluminum
amount the mixture feeds to the saturated surface
in aluminizing .................................26
bonded ................................................. 1
coefficient of transfer in flux, electric arc
evaporation .................................40(T)
consumption variation with temperature
level ............................... 1617, 18(F)
content effect of Ni-Co-Cr-Al alloys on thermal
expansion coefficients ................ 122, 123
content effect on heat resistance of Co-Ni-Cr-Al
alloys .......................................... 106
content effect on hot corrosion and phase
composition of Ni-Cr-Al alloys ........... 111
content effect on hot corrosion resistance of
Co-Ni-Cr-Al alloys ................... 112, 113
Ni-Co-Cr-Al alloys ......................... 112
content effect on protective nature of overlay
coatings on superalloys .................... 109
content effect on thermal expansion coefficient of
Co-Ni-Cr-Al alloys ......................... 124
content effect on thermal expansion coefficient of
Ni-Cr-Al alloys .............................. 121
content effect on thickness of beta-phase-depleted
outer zone ....................103104, 105(F)

218 / Index
Aluminum (continued)
content in diffusion coatings affecting ductilebrittle transition temperature .............. 120
content in NiAl compounds and thermal
fatigue ......................................... 145
degree of ionization .............................40(T)
free ..................................................... 1
as impurity of jet fuels .............................. 5
mass-spectrometric analysis data for material
evaporation .................................40(T)
Aluminum-boron alloys, slurry technique for
depositing diffusion coatings ...................35
Aluminum chlorides, formation of .................25
Aluminum-chromium alloys
circulation technique for depositing diffusion
coatings .........................................32
as powder mixtures used in aluminizing ........27
slurry technique for depositing diffusion
coatings .........................................35
Aluminum halides ..................................... 1
Aluminum iodides, use in circulation technique for
diffusion coating application ...................32
Aluminum-iron alloys, as powder mixtures used
in aluminizing ....................................27
Aluminum mass fraction in a definite zone of
coating (Xi) ..................................1718
Aluminum oxide
as corrosion product penetrating the coating ....13
formation at ceramic/bond-coat
interface ................................ 205, 206
formation due to interaction of overlay coating
and superalloy .................................96
Aluminum powder, content effect on coatings
deposited by slurry technique ..................35
Aluminum reserve
calculation of value ............................1718
as criterion for coating life expectancy ..... 17, 19
definition ........................................ 18, 69
of diffusion aluminide coatings ......... 1516, 38
with gamma double-prime phase compound
coatings .........................................81
and high-temperature testing ............. 8485(F)
parameter introduced in coating ..................99
significance ...........................................69
Aluminum-silicon alloys
as base system for development of electric arc
coating compositions .........................36
circulation technique for depositing diffusion
coatings .........................................32
slurry technique for depositing diffusion
coatings ..................................... 35(F)
Aluminum-silicon-yttrium alloy, as diffusion
layer, phase transformations during annealing
of superalloys ............................ 62(F), 63
Aluminum-to-surface transfer .....................31
Alundum crucibles ................................. 100
Ammonium chloride, and activator distribution
for use ..............................................29
Amylacetate .................................. 33, 34, 35
Angle of vapor incidence, and ceramics
blackening ............................. 169, 170(T)
Annealing ................................... 130(F), 131

after electric arc technique of applying diffusion
aluminide coatings ............................37
powder mixture of diffusion aluminide
coatings .........................................30
Annealing temperature, elevation effect on
aluminum concentration in diffusion aluminide
coatings ............................................60
Approximate life expectancy of a new
coating ........................................1415
Arc plasma spraying .................................. 8
B
Barium, as impurity of jet fuels ...................... 5
Beam model .......................................... 196
Bending stresses ........................................ 8
Beta-phase, in cobalt-nickel-chromium-aluminum
alloys ...............................................71
Beta-phase-free zone after heat resistance testing
at 1200C, 200h, regression equation for
calculation of ................................... 214
Beta-phase nickel aluminide (NiAl) ........... 2, 55
alloying with molybdenum heat resistance
effect .......................................... 101
alloying with silicon, heat resistance effect ... 101
alloying with tungsten, heat resistance
effect .......................................... 101
alloying with zirconium, heat resistance
effect .......................................... 101
aluminum content ............................. 119(T)
chromium solubility in Ni-Cr-Al system
alloys .......................................... 111
color ...............................................56(T)
crystalline structure .............................56(T)
density ............................................56(T)
in diffusion coating formed in high-activity
media ........... 59, 60(T), 61(F), 62(F,T), 63
in diffusion coating formed in moderate and lowactivity media ......... 6369(F,T), 70, 72(F)
disappearance in coating outer zone .............98
elongation ...................................... 119(T)
elongation vs. temperature ................... 120(F)
hardness ..........................................56(T)
heat of formation ................................56(T)
heat resistance at high temperatures101(F)
high-temperature strength ........................ 120
homogeneity region .............................56(T)
melting temperature .............................56(T)
of Ni-Cr-Al system ........................ 87, 88(F)
oxidation mechanism ............................. 101
oxidation time ................................. 100(T)
oxide structure ................................. 100(T)
phase transformations in Al5SiY diffusion
coatings after high-temperature
tests ..................................84(F), 85(F)
in phase transformations of overlay coatings on
superalloys ....... 89(F), 90(F), 91(F), 92(T),
93(F), 95(T)

Index / 219
quantity, annealing 1200C, 200h, regression

equation for calculation of ................. 213
quantity, regression equation for calculation
of .............................................. 213
specific volume ..............................128129
two-way diffusion in diffusion coatings .... 80(F),
81(T), 82(F), 83(T), 84(F), 85(F)
ultimate strength ............................... 119(T)
ultimate strength vs. temperature ........... 120(F)
yield strength .................................. 119(T)
yield strength vs. temperature ............... 120(F)
x-ray diffraction analysis ..................... 100(T)
Black ceramics
oxidation kinetics ............................. 169(F)
phase composition of ......................... 169(F)
Blade fatigue limit, thermal fatigue of overlay
coatings ............................................. 9
Blade leading-edge zone ........................... 206
Blade pressure surfaces, preparation for overlay
coatings ............................................45
Blade suction surfaces, preparation for overlay
coatings ............................................45
Blade-wall thickness value ........................ 197
Blasting, as preparation of inner cavity for
aluminizing ........................................31
Bond coats ............................................ 161
effect of technological factors on element
contents of surface ................ 167, 168(T)
oxidation ............................................ 182
plastic strain consideration ....................... 162
preoxidizing ........................................ 167
protecting blade from oxidation ..................21
surface preparation ................166167, 168(T)
surface roughness ................................. 166
Boron, as alloying element for diffusion
coating ........................................... 116
Burners .................................................. 8
Burnouts ................................................. 7
C
Calcium, as impurity of jet fuels ..................... 5
Carbide phase precipitation
Cr7C3 in diffusion coatings on Ni20CrTi
alloy ....................................8081(T)
Cr23C6 in diffusion coatings on Ni20CrTi
alloy ....................................8081(T)
of diffusion coatings .......... 60(T), 65(T), 83(T),
84(F), 85(F)
in interaction mechanism between overlay
coatings and superalloys .....................93
on nickel superalloy Ni20CrTi ....................81
of overlay coatings in high-temperature tests ..87,
88(F), 89(F), 90(F), 92(T), 93(F),
94(F), 95(T)
Carbides, effect on NiAl with alloying
elements ......................................... 101
Carbon ................................................... 6
Carbon dioxide ......................................... 6

Carbon hydroxide (CH4) ............................. 6
Carbon monoxide ...................................... 6
Cathode material transfer coefficient (l) ........40
Cathode spots ..........................................39
temperature of ...................................40(T)
vapor pressure ...................................40(T)
Centrifugal force, in coated blade ................ 198
Centrifugal load in a blade, value correction .. 197
Centrifugal load in a given blade cross
section ........................................... 197
Centrifugal loading, blade-mass gain from
coating application ............................. 195
Ceramic coating compressive stress ............ 184
Ceramic coatings. See also Ceramic condensates;
Thermal barrier coatings.
effects on blade-life expectancy ..................15
requirements .......................................... 8
Ceramic layer blackening ......................... 169
Ceramic layer modulus of elasticity ............ 184
Ceramic layer spalling. See also Spalling ...... 162
Ceramics
for electron-beam evaporation ..... 162165(F,T)
spalling on turbine blades with thermal barrier
coatings ................................ 2022(F)
Ceramics condensates, specific types
ZrO2, phase composition ..................... 164(T)
ZrO2-20%CeO2
color .......................................... 177(T)
density ....................................... 177(T)
EB, thermophysical properties ........... 178(T)
manufacturing method ..................... 177(T)
thickness ..................................... 177(T)
ZrO2-7%Y2O3-5%NiO, thermal expansion
coefficients ............................... 181(T)
ZrO2-7%Y2O3-15%NiO, thermal expansion
coefficients ............................... 181(T)
ZrO2-8%Y2O3 ................................. 164(T)
APS, thermophysical properties ......... 178(T)
as ceramic layer of thermal barrier coating,
service life .................186187(F), 188
as ceramic layer tested for adhesive
strength ................ 187(F), 188189(F)
as ceramic layer tested on gas-dynamic test
rigs ................................... 189, 190
color .......................................... 177(T)
density ....................................... 177(T)
heat capacity versus temperature ........ 178(F)
maximum principal stress ................. 205(F)
phase composition ......................... 164(T)
powder, heat capacity versus
temperature ............................ 178(F)
spalling ...................................... 185(F)
thermal expansion coefficient ............... 180,
181(T), 182(F)
thickness ..................................... 177(T)
ZrO2-8%Y2O3-5%NiO
ceramic layer tested on gas-dynamic test
rigs ......................................... 190

220 / Index
Ceramics Condensates (continued)

color .......................................... 177(T)
density ....................................... 177(T)
heat capacity versus temperature ........ 178(F)
temperature ............................ 178(F)
thickness ..................................... 177(T)
ZrO2-11%Y2O3, emissivity, coating ........ 181(T)
ZrO2-15%Y2O3 ................................ 164(T)
phase composition ......................... 164(T)
temperature ............................ 178(F)
thermal expansion coefficients ........... 181(T)
Chemical compositions
diffusion aluminide coatings ............. 10, 11(T)
of overlay coatings ..............................10(T)
superalloys for aircraft turbine blades ... 10, 11(T)
Chlorine, in deposits detected on gas-turbine
engine blade surfaces ............................99
Chromium
addition effect on beta-phase depleted outer
zone ........................................... 104
addition effect on saturating mixtures for
aluminizing ................................ 27, 29
as alloying element for diffusion coating ...... 116
as alloying element, retarding of aluminum
diffusion ........................................57
alloying with aluminide to retard diffusion
processes .......................................81
evaporation .................................40(T)
consumption variation with temperature
level ............................... 1617, 18(F)
content effect of Ni-Co-Cr-Al alloys on thermal
expansion coefficients ................ 122, 123
alloys .......................................... 106
Co-Ni-Cr-Al alloys ................... 112, 113
Ni-Cr-Al alloys .............................. 111
Ni-Co-Cr-Al alloys ...................111112
content effect on protective nature of overlay
coatings on superalloys .................... 109
content effect on thermal expansion coefficients
of Co-Ni-Cr-Al alloys ...................... 124
of Ni-Cr-Al alloys ........................... 121
content effect on thickness of interaction zone of
overlay coating and superalloy ..... 93, 95(F)
in diffusion aluminide coatings .................... 1
diffusion in overlay coatings ......................88
diffusion into interaction zone with cobalt content
addition in overlay coatings .................91
diffusion with interaction mechanism of
overlay coating and superalloy at high
temperatures ..................89, 91(F), 92(T)
Chromium-base carbides, stable phases formed

from interaction mechanism ....................94
Chromium-base phases ..............................82
Chromium mass fraction (XCr) .....................19
Chromium oxide, as corrosion product penetrating
the coating .........................................13
Chromizing-aluminizing method ................... 1
Circulation technique, for applying diffusion
coatings ................................... 3132(F)
Circumferential elasticity modulus ............. 151
Circumferential thermal stresses ................ 151
Cleaning, before aluminizing ........................29
Coastal regions ......................................... 9
Coated-blade service life, consideration affecting
calculations ...................................... 195
Coating cyclic life expectancy ......................20
Coating ductility .................................... 198
Coating ductility parameter ...................... 201
Coating emissivity ........................ 180, 181(T)
Coating lifetime. See Coating protective lifetime.
Coating-mass effect on blade centrifugal load,
coefficient of .................................... 197
Coating protective lifetime ..................... 98, 99
related to aluminum reserve .......................19
techniques for determination .................... 100
Coatings
efficiency showing coating effect on superalloy
properties ..................................... 119
factors affecting mechanical properties of
superalloys ................................... 119
multilayer systems ............................ 134(F)
Coating serviceability criterion .................. 198
Coating service life. See also Service
life. ......................................... 155, 162
definition .............................................14
and thermomechanical fatigue testing .......... 157
Coatings, specific types
aluminized
life of diffusion coatings on JS6U
superalloys ............................ 115(F)
low-cycle fatigue ................. 149(F), 150(T)
thermal shock testing of coating/superalloy
systems ................................ 148(T)
thermofatigue testing of coating/superalloy
systems ...................... 147(T), 148(F)
aluminum
oxide life on JS6U superalloy, cyclic
testing .................................. 108(T)
thickness of coatings on JS6U
superalloy ............................. 109(T)
aluminum palladium (AlPd)
coating life (hot corrosion test in GZT
ash) ..................................... 115(T)
coating life on JS6U superalloys ........ 116(T)
life coefficient (hot corrosion test on GZT
ash) ..................................... 116(T)
life of diffusion coatings on JS6U
superalloy ............................. 115(F)
testing .................................. 108(T)

Index / 221

superalloy ............................. 109(T)
AlSiBY, coating life (hot corrosion test in
GZT ash) ................................. 115(T)
AlSiCrY, coating life (hot corrosion test in
GZT ash) ................................. 115(T)
Al5SiY
phase transformations diagram ............. 84(F)
systems ...................... 147(T), 148(F)
Al5Si1.5Y
creep-rupture strength ..... 134, 135(F), 136(F)
electric arc coating, life of diffusion coatings
on JS6U superalloy .................. 115(F)
elongation of JS6U superalloy ..... 136, 137(F)
heat resistance of JS6U superalloy with
diffusion coating ..................... 108(F)
Al10Si0.7Y
heat resistance of JS6U superalloy with
diffusion coating ..................... 108(F)
testing .................................. 108(T)
superalloy ............................. 109(T)
Al14Si0.7Y electric arc coating, life of diffusion
coatings on JS6U superalloy .......... 115(F)
Al15Si0.5Y
testing .................................. 108(T)
superalloy ............................. 109(T)
Coatings, specific types, Cobalt-chromiumaluminum alloys
Co20Cr11AlY
testing .................................. 108(T)
superalloy ............................. 109(T)
testing .................................. 108(T)
superalloy ............................. 109(T)
Co20Cr12AlY
chemical composition .......................10(T)
density ....................................... 181(T)
elongation vs. temperature ................ 126(F)
heat capacity ................................ 181(T)
temperature .................................. 181(T)
thermal conductivity ....................... 181(T)
thermal diffusivity ......................... 181(T)
ultimate strength vs. temperature ........ 126(F)
Co22Cr9AlY
microstructure ................................ 73(F)
phase composition ...........................72(T)
thermal expansion coefficients, various
temperatures .......................... 125(T)
weight loss (hot corrosion test in GZT
ash) ..................................... 113(T)
Co22Cr11AlY
coating life ...................................... 116
coating life, hot corrosion test in GZT

ash .................................... 115(F,T)
life coefficient (hot corrosion test in GZT
ash) ..................................... 116(T)
ash) ..................................... 113(T)
ash) ................................... 115(F,T)
ash) ..................................... 116(T)
weight change (hot corrosion test in GZT
ash) ..................................... 114(F)
Co22Cr12Al0.4Y, bond coat service
life .............................183(F), 186187
Co22Cr12AlY, corrosion lives after sulfur attack
from ash ............................ 186(F), 188
Co23Cr12AlY, chemical composition .......10(T)
Co24Cr5Al0.4Y, for bond coat, service
life ................................... 186(F), 188
Co24Cr12Al0.4Y, for bond coat, service
life ................................... 186(F), 188
Co25Cr4AlY, chemical composition .........10(T)
Co28Cr5AlY, chemical compositions of cathode
and deposited coating ............... 40, 41(T)
Cobalt-chromium-iron alloys
Co28Cr10FeY, chemical composition .......10(T)
Cobalt-nickel-chromium-aluminum alloys
Co8Ni23Cr12AlY, chemical composition ...10(T)
Co10Ni15Cr6AlY
chemical composition .............72(T), 212(T)
elasticity modulus values, various
temperatures .......................... 125(T)
normalized number of variable ........... 212(T)
oxide life .................................... 106(T)
temperatures .......................... 125(T)
thermal expansion coefficients vs.
temperature .............................. 74(F)
weight gain (heat resistance cyclic
test) ..................................... 106(T)
ash) ..................................... 113(T)
Co10Ni15Cr12AlY
temperatures .......................... 125(T)
temperatures .......................... 125(T)
temperature .............................. 74(F)
ash) ..................................... 113(T)

222 / Index
Coatings, (continued)
Co10Ni22Cr9AlY
temperatures .......................... 125(T)
elongation, various temperatures ........ 128(T)
oxide life .................................... 106(T)
temperatures .......................... 125(T)
ultimate strength, various
temperatures .......................... 127(T)
test) ..................................... 106(T)
ash) ..................................... 113(T)
yield strength, various temperatures ..... 127(T)
Co10Ni30Cr6AlY
temperatures .......................... 125(T)
oxide life .................................... 106(T)
temperatures .......................... 125(T)
temperature .............................. 74(F)
test) ..................................... 106(T)
ash) ..................................... 113(T)
Co10Ni30Cr12AlY
oxide life .................................... 106(T)
temperatures .......................... 125(T)
temperature .............................. 74(F)
test) ..................................... 106(T)
ash) ..................................... 113(T)
Co20Ni15Cr9AlY
temperatures .......................... 125(T)
temperatures .......................... 125(T)
temperatures .......................... 127(T)
ash) ..................................... 113(T)

Co20Ni22Cr6AlY
temperatures .......................... 125(T)
oxide life .................................... 106(T)
temperatures .......................... 125(T)
temperatures .......................... 127(T)
test) ..................................... 106(T)
ash) ..................................... 113(T)
Co20Ni22Cr9AlY
temperatures .......................... 125(T)
oxide life .................................... 106(T)
temperatures .......................... 125(T)
temperatures .......................... 127(T)
test) ..................................... 106(T)
ash) ..................................... 113(T)
Co20Ni22Cr12AlY
temperatures .......................... 125(T)
temperatures .......................... 125(T)
temperatures .......................... 127(T)
ash) ..................................... 113(T)
Co20Ni30Cr9AlY
temperatures .......................... 125(T)
oxide life .................................... 106(T)
temperatures .......................... 125(T)

Index / 223

temperatures .......................... 127(T)
test) ..................................... 106(T)
ash) ..................................... 113(T)
Co22Ni23Cr12AlY, chemical
composition ................................10(T)
Co25Ni20Cr12AlY, cyclic testing and
fragmentation ......................... 20, 21(F)
Co30Ni15Cr6AlY
temperatures .......................... 125(T)
temperatures .......................... 125(T)
temperature .............................. 74(F)
ash) ..................................... 113(T)
Co30Ni15Cr12AlY
temperatures .......................... 125(T)
temperatures .......................... 125(T)
temperature .............................. 74(F)
ash) ..................................... 113(T)
Co30Ni22Cr9AlY
temperatures .......................... 125(T)
temperatures .......................... 125(T)
temperatures .......................... 127(T)
ash) ..................................... 113(T)
Co30Ni30Cr6AlY
temperatures .......................... 125(T)
oxide life .................................... 106(T)
temperatures .......................... 125(T)

temperature .............................. 74(F)
test) ..................................... 106(T)
ash) ..................................... 113(T)
Co30Ni30Cr12AlY
ash) ..................................... 113(T)
Co32Ni21Cr8AlY
chemical composition and other
characteristics ......................... 153(T)
coatings service life ............................ 157
cyclic life of coatings/superalloy CMSX-4
system ................ 155, 157(F,T), 158(F)
EB-TBC, thermal barrier coating testing of
MAR-M-002 superalloy on gas-dynamic test
rigs and engines ................... 189(F), 190
Iron-chromium-aluminum alloys
Fe22Cr11AlY, weight loss (hot corrosion test
in GZT ash) .............................. 113(T)
Fe25Cr5AlY, chemical composition .........10(T)
Coatings, specific types, Iron-nickel-chromiumaluminum alloys
Fe12 Ni20Cr12Al1SiY
oxide life ........................................ 109
testing .................................. 108(T)
coating life, hot corrosion test in
GZT ash .............................115(F,T)
life coefficient (hot corrosion test in
GZT ash) .............................. 116(T)
superalloy ............................. 109(T)
Fe22Ni24Cr5AlY, chemical composition ...10(T)
Coatings, specific types, Nickel-aluminum alloys
Ni31Al, thermal expansion coefficients, various
temperatures ............................. 121(T)
NiAl (33% Al)
Nickel-chromium-aluminum alloys
Ni5Cr8AlY
elongation ................................... 129(F)
ultimate strength ................127128, 129(F)
Ni5Cr25Al
calculated number of cycles to crack
appearance ............................ 203(T)
elongation ................................... 126(T)
temperatures .......................... 121(T)
temperature ......................... 67(F), 70

224 / Index
Coatings (continued)
ultimate strength ............................ 126(T)
Ni7Cr10AlY
Ni7Cr12AlY
appearance ............................ 203(T)
corrosion depth in GZT ash .............. 110(F)
elasticity modulus .............................. 122
temperatures .......................... 122(T)
hot corrosion resistance ....................... 111
maximum cyclic life ........................... 203
temperatures .......................... 121(T)
temperature ......................... 67(F), 70
weight loss due to hot corrosion ......... 110(T)
Ni8Cr15AlY
temperatures .......................... 122(T)
elongation ................................... 126(T)
temperature ......................... 67(F), 70
ultimate strength ...................... 124, 126(T)
Ni8Cr18AlY
appearance ............................ 203(T)
elasticity modulus .............................. 122
temperatures .......................... 122(T)
temperatures .......................... 121(T)
Ni10Cr17AlY
service life vs. strain-intensity range ........ 200
thermomechanical fatigue tests .......... 200(F)
Ni14Cr3AlY
crystallization intervals, 49(T)
temperatures .......................... 122(T)
liquidus-solidus temperatures ..............49(T)

temperatures .......................... 121(T)
temperature ......................... 67(F), 70
Ni15Cr3AlY, elasticity modulus ................ 122
Ni16Cr9AlY
crystallization intervals ......................49(T)
temperatures .......................... 122(T)
oxide life at 900C ......................... 102(T)
temperatures .......................... 121(T)
temperature ......................... 67(F), 70
weight gain (cyclic heat resistance
tested) .................................. 102(T)
weight gain (isothermal heat resistance
tested) .................................. 102(T)
Ni16Cr11AlY, phase transformations, interaction
mechanism with superalloy at high
temperatures .....8788(F), 89, 90(F), 91(F)
Ni18Cr5AlY
temperatures .......................... 122(T)
oxide life at 900C ......................... 102(T)
temperatures .......................... 121(T)
temperature ......................... 67(F), 70
Ni18Cr15AlY, thermal expansion coefficients,
various temperatures .................... 121(T)
Ni18Cr16AlY
temperatures .......................... 122(T)
temperatures .......................... 121(T)
Ni19Cr10AlY, thermal expansion coefficients,
Ni19Cr12AlY

Index / 225

temperatures .......................... 122(T)
oxide life at 900C ......................... 102(T)
temperatures .......................... 121(T)
tested) .................................. 102(T)
tested) .................................. 102(T)
Ni20Cr5AlY
ductility vs. temperature .................. 199(F)
strain intensity vs. temperature ........... 199(F)
Ni20Cr6AlY
density ....................................... 181(T)
elongation ................................... 126(T)
heat capacity ................................ 181(T)
hot corrosion resistance, chromium-aluminum
content ratio ............................... 111
preparation of surface before coating
deposition ...................................46
temperature .................................. 181(T)
temperature ............................ 121(T)
Ni20Cr6AlY/Ni20Cr12AlY/ZrO2-8%Y2O3,
fatigue strength .............................. 141
Ni20Cr7AlY
elongation ................................... 129(F)
Ni20Cr9.5AlY, phase transformations, interaction
mechanism between overlay coating and
superalloy at high temperatures .... 88, 89(F)
Ni20Cr11Al, hafnium addition effect on service
life of protective oxides .................... 101
Ni20Cr11Al1Hf, oxide life at 900C ....... 102(T)
Ni20Cr11Al2Hf, oxide life at 900C ....... 102(T)
Ni20Cr11AlY, cobalt effect on oxidation ..... 103
Ni20Cr12AlY
adhesion strengths for various preparation and
deposition techniques ..................51(T)
blade elongation vs. service time .......... 22(F)
bond coat for thermal barrier coating, bend
testing .................................. 175(F)
bond coat on thermal barrier coating .... 171(F)
as bond coat, spalling of CMSX-4 superalloy
surface ..................................... 188
as bond coat tested for adhesive

strength ................ 187(F), 188189(F)
bond coat tested on gas-dynamic test
rigs ................................... 189, 190
appearance ............................ 203(T)
coating thickness distribution across turbine
blade middle section ................... 49(F)
cyclic life ........................................ 202
cyclic testing and fragmentation ...... 20, 21(F)
density ....................................... 181(T)
ductility vs. temperature .................. 199(F)
electric arc evaporation ..........................40
elongation ............................. 126, 129(F)
emissivity, coating ......................... 181(T)
emissivity factors ............................... 180
evaporation rate ..................................47
external deformed appearance after 500h
testing of turbine blades .......... 16, 17(F)
fatigue strength compared to that of thermal
barrier coating ............................ 141
fatigue strength of coated JS6U
superalloy ..........140(T), 141(F), 143(F)
fatigue strength of coated VJL12U superalloy
blades ............................ 140, 142(T)
fatigue strength of coated VLJ12U
superalloy ....................... 140, 143(F)
heat capacity ................................ 181(T)
heat treatment effect on beta-phase
quality ....................................77(T)
high-temperature corrosion ................. 11(F)
intensities of strains and stresses on coated
MAR-M-002 blade ................... 202(T)
on JS6U superalloy, heat resistance
testing with oxide structural
change ........................... 109, 110(F)
low-cycle fatigue ........149(F), 150(T), 151(F)
low-cycle fatigue testing ......... 149150(F,T)
microball peening .............................. 131
microstructure and distribution of elements
in ingot bath after 20 deposition
processes ........................... 47, 48(F)
multilayer coating system thermal
stresses ................................. 134(F)
no thermal fatigue cracking ........... 20, 21(F)
number of cycles to coating failure ..... 201(F)
oxidation depth of turbine blades .......... 16(F)
phase transformations, interaction mechanism
between overlay coating and superalloy at
high temperatures ................. 88, 92(T)
strain intensity .............................. 201(F)
strain-intensity range ...................... 201(F)
strain intensity vs. temperature ........... 199(F)
stress intensity .............................. 201(F)
surface preparation effect on adhesion strength
deposited by electric arc technique ..46(T)
temperature .................................. 181(T)

226 / Index
tensile stresses .................................. 130
systems ................................ 148(T)
thermal stress distribution on JS6U superalloy
blades ............................ 131, 132(F)
thermal stresses ....................130(F), 131(F)
thermal stresses on JS6F superalloy at different
orientations ...................... 133(F), 134
thermal stresses on JS6U
superalloy ....................... 132(F), 133
thermal stresses variations in coating and
JS6U superalloy ...................... 133(F)
systems ...................... 147(T), 148(F)
threshold recrystallization temperature ..... 131
weight change of overlay coating on JS6U
superalloy (isothermal test) ......... 110(F)
working temperature 1100 to 1150C,
protection capability reduction ...........16
x-ray pattern of overlay coating surface ... 76(F)
Ni20Cr12AlY/Al, thermofatigue testing of
coating/superalloy systems .. 147(T), 148(F)
Ni20Cr12AlY/chromium carbide
ash .................................... 115(F,T)
ash) ..................................... 116(T)
superalloy ............................. 109(T)
systems ................................ 148(T)
systems ...................... 147(T), 148(F)
Ni20Cr12AlY/ZrO2-8%Y2O3, as thermal
barrier coating for CMSX-4 single-crystal
alloy ........................................... 207
Ni20CrTi-Al system, diffusion layer .... 56, 57(F)
Ni21Cr9AlY
temperatures .......................... 122(T)
oxide life at 900C ......................... 102(T)
temperatures .......................... 121(T)
tested) .................................. 102(T)
tested) .................................. 102(T)
Ni22Cr6AlY
elongation ................................... 126(T)

ultimate strength ............................ 126(T)
Ni22Cr11Al1Hf
chemical composition ..................... 102(T)
corrosion depth of hot corrosion in GZT
ash ...................................... 111(F)
cyclic testing, hafnium and silicon
effects .................................. 102(F)
tested) .................................. 102(T)
tested) .................................. 102(T)
Ni22Cr11Al2Hf
ash ...................................... 111(F)
effects .................................. 102(F)
tested) .................................. 102(T)
tested) .................................. 102(T)
Ni22Cr11AlSi
ash ...................................... 111(F)
Ni22Cr11Al1Si
ash ...................................... 111(F)
effects .................................. 102(F)
Ni22Cr11Al2Si
effects .................................. 102(F)
Ni22Cr11AlY
effects .................................. 102(F)
ductility during bend testing .............. 129(F)
temperatures .......................... 122(T)
elongation ................................... 126(T)
temperatures .......................... 121(T)
temperature ......................... 67(F), 70
tested) .................................. 102(T)

Index / 227

tested) .................................. 102(T)
Ni22Cr12AlY, forces acting in coating
on JS6U superalloy versus coating
thickness ............................ 132(F), 133
Ni28Cr6AlY
coating life ...................................... 116
ash) ................................... 115(F,T)
ash) ..................................... 116(T)
superalloy ............................. 109(T)
Ni28Cr11AlY
oxide life at 900C ......................... 102(T)
tested) .................................. 102(T)
tested ................................... 102(T)
Ni36Cr5AlY, chemical composition .........10(T)
Ni40Cr10AlY, phase transformations, interaction
mechanism with superalloy at high
temperatures ..................... 9091, 93(F)
Coatings, specific types, Nickel-chromiummolybdenum-tungsten-cobalt alloys
Ni10CrMoWCo, oxidation constant ........ 108(T)
Coatings, specific types, Nickel-chromiumtungsten-molybdenum-cobalt alloys
Ni10CrWMoCo, fatigue strength of coated
JS6U superalloy ......................... 139(F)
Coatings, specific types, Nickel-cobalt-chromiumaluminum alloys
Ni6Co10Cr13AlTaReHf, thermal expansion
coefficients, various temperatures .... 123(T)
Ni6Co20Cr12AlY, thermal expansion
coefficients, various temperatures .... 123(T)
Ni6Co22Cr11AlY, weight loss (hot corrosion
test in GZT ash) ......................... 113(T)
Ni7Co12Cr17Al
alloying element effects at high
temperatures .............................. 116
coating service life ............................. 157
cyclic life, thermomechanical fatigue
testing ..........................155, 157(F,T)
superalloy ............................. 109(T)
Ni7Co12Cr17Al(Ni8Co12Cr7Al/Al5Si1.5Y)
ash) ..................................... 115(T)
electric-arc technique, microstructure ..... 75(F)
Ni8Co12Cr7Al
coatings life (hot corrosion test in GZT

ash) ..................................... 115(T)
coating service life ............................. 157
cyclic life, thermomechanical fatigue
testing ..........................155, 157(F,T)
Ni8Co20Cr12AlY
cyclic life ........................................ 202
density ....................................... 181(T)
heat capacity ................................ 181(T)
multilayer coating system thermal
stresses ................................. 134(F)
temperature .................................. 181(T)
Ni10Co15Cr8Al
chemical composition of structural
constituents ..............................69(T)
temperatures .......................... 124(T)
temperatures .......................... 123(T)
weight change (heat resistance isothermal
test at 1100C) ........................ 104(F)
test at 1200C) ........................ 104(F)
weight loss (hot corrosion in GZT
ash) ..................................... 112(T)
Ni10Co15Cr8AlY
ductility .......................................... 128
elongation, various temperature .......... 127(T)
phase transformations after high-temperature
testing ........................................71
temperature .............................. 72(F)
temperatures .......................... 127(T)
test at 1100C) ........................ 104(F)
test at 1200C) ........................ 104(F)
Ni10Co15Cr14Al
constituents ..............................69(T)

228 / Index
temperatures .......................... 124(T)
temperatures .......................... 123(T)
test at 1100C) ........................ 104(F)
test at 1200C) ........................ 104(F)
ash) ..................................... 112(T)
Ni10Co15Cr14AlY
temperature .............................. 72(F)
weight change (heat resistance isothermal test
at 1200C) ............................. 104(F)
ash) ..................................... 112(T)
Ni10Co20Cr12AlY
elongation ................................... 129(F)
on JS6U superalloy, heat resistance testing
with oxide structural change 109, 110(F)
temperatures .......................... 127(T)
Ni10Co22Cr11Al
temperatures .......................... 124(T)
temperatures .......................... 123(T)
Ni10Co22Cr11AlY
elongation and ductility ....................... 127
temperatures .......................... 127(T)
test at 1200C) ........................ 105(F)
ash) ..................................... 112(T)
Ni10Co30Cr8Al
constituents ..............................69(T)

temperatures .......................... 124(T)
temperatures .......................... 123(T)
test at 1100C) ........................ 104(F)
test at 1200C) ........................ 104(F)
ash) ..................................... 112(T)
Ni10Co30Cr8AlY
temperature .............................. 72(F)
temperatures .......................... 127(T)
ash) ..................................... 112(T)
Ni10Co30Cr14Al
constituents ..............................69(T)
temperatures .......................... 124(T)
temperatures .......................... 123(T)
test at 1200C) ........................ 104(F)
ash) ..................................... 112(T)
Ni10Co30Cr14AlY
temperature .............................. 72(F)
ash) ..................................... 112(T)
Ni18Co12Cr7Al, thickness of coatings on
JS6U superalloy ......................... 109(T)
Ni20Co15Cr11Al
temperatures .......................... 124(T)
temperatures .......................... 123(T)
Ni20Co15Cr11AlY

Index / 229

test at 1200C) ........................ 105(F)
ash) ..................................... 112(T)
Ni20Co20CrAlY, cobalt effect on oxidation .. 103
Ni20Co20Cr5AlY
in bond coat of thermal barrier coating ..... 187
cyclic testing of this inner layer
alloy ..................................185186
Ni20Co20Cr11AlY
ash .................................... 115(F,T)
ash) ..................................... 116(T)
superalloy ............................. 109(T)
ash) ................................... 115(F,T)
ash) ..................................... 116(T)
testing .................................. 108(T)
superalloy ............................. 109(T)
Ni20Co20Cr12Al, oxidation constant affected
by alloying elements ..................101102
Ni20Co20Cr12AlY
alloying element content variation according
to melting techniques ..................44(T)
bond coat deformation ..................183184
in bond coat of thermal barrier coating ..... 187
appearance ............................ 203(T)
creep-rupture strength ..... 134, 135(F), 136(F)
creep-rupture strength in air of coated JS26-DS
superalloy .................136137, 138(F)
cyclic life ........................................ 202
cyclic testing and fragmentation ...... 20, 21(F)
cyclic testing of this outer layer
alloy ..................................185186
density ....................................... 181(T)
elongation .......................127128, 129(F)
elongation of JS6U superalloy ..... 136, 137(F)
evaporation rate ..................................47
fatigue strength of coated JS6U
superalloy ............................. 140(T)
gas contents in ingots subjected to

extra deoxidation with carbon and
cerium ....................................44(T)
gas contents in various zones of ingot without
extra deoxidation .......................44(T)
heat capacity ................................ 181(T)
interaction zone of overlay coating
with superalloy with high-temperature
testing ............................... 88, 91(F)
microstructure after high-temperature testing
for stress-rupture strength ........... 138(F)
minimum calculated cyclic life .............. 203
temperature .................................. 181(T)
systems ................................ 148(T)
thermal stresses ....................130(F), 131(F)
thermal stresses on JS6F superalloy at different
orientation ....................... 133(F), 134
thermal stresses on JS6U
superalloy ....................... 132(F), 133
systems ...................... 147(T), 148(F)
thermomechanical fatigue testing, hysteresis
loop calculation ................. 152(F), 153
temperatures .......................... 127(T)
Ni20Co20Cr13AlY, phase transformations,
interaction mechanism with superallloy
at high temperatures ................. 91, 93(F)
Ni20Co22Cr8Al
temperatures .......................... 124(T)
temperatures .......................... 123(T)
Ni20Co22Cr8AlY
temperatures .......................... 127(T)
test at 1200C) ........................ 105(F)
ash) ..................................... 112(T)
Ni20Co22Cr11Al
temperatures .......................... 124(T)

230 / Index
temperatures .......................... 123(T)
ash) ..................................... 112(T)
Ni20Co22Cr11AlHfSi
oxide life .................................... 103(T)
weight change (heat resistance cyclic
test) ..................................... 105(F)
test) ..................................... 103(T)
weight gain (heat resistance isothermal
test) ..................................... 103(T)
ash) ..................................... 113(T)
Ni20Co22Cr11AlHf2Si
oxide life .................................... 103(T)
test) ..................................... 105(F)
test) ..................................... 103(T)
test) ..................................... 103(T)
ash) ..................................... 113(T)
test) ..................................... 105(F)
ash) ..................................... 113(T)
Ni20Co22Cr11Al2HfSi
oxide life .................................... 103(T)
test) ..................................... 105(F)
test) ..................................... 103(T)
test) ..................................... 103(T)
ash) ..................................... 113(T)
test) ..................................... 105(F)
ash) ..................................... 113(T)
Ni20Co22Cr11AlY
temperatures .......................... 124(T)
oxide life .................................... 103(T)

temperatures .......................... 123(T)
temperatures .......................... 127(T)
test) ..................................... 105(F)
test at 1200C) ........................ 105(F)
test) ..................................... 103(T)
test) ..................................... 103(T)
ash) ..................................... 112(T)
ash) ..................................... 113(T)
oxide life .................................... 103(T)
test) ..................................... 105(F)
test) ..................................... 103(T)
test) ..................................... 103(T)
ash) ..................................... 113(T)
Ni20Co22Cr14Al, thermal expansion coefficients,
Ni20Co22Cr14AlY
test at 1200C) ........................ 105(F)
ash) ..................................... 112(T)
Ni20Co30Cr11Al, thermal expansion coefficients,
Ni20Co30Cr11AlY
test at 1200C) ........................ 105(F)
ash) ..................................... 112(T)
Ni22Co22Cr11AlY, weight change
(heat resistance isothermal test at
1200C) ................................... 105(F)
Ni22Co22Cr12AlY, microstructure .......... 74(F)
Ni23Co28Cr9AlY
coating life ...................................... 116
ash .................................... 115(F,T)
ash) ..................................... 116(T)
superalloy ............................. 109(T)

Index / 231
Ni30Co15Cr8Al
constituents ..............................69(T)
temperatures .......................... 124(T)
temperatures .......................... 123(T)
test at 1100C) ........................ 104(F)
test at 1200C) ........................ 104(F)
ash) ..................................... 112(T)
Ni30C015Cr8AlY
ductility .......................................... 128
temperature .............................. 72(F)
temperatures .......................... 127(T)
ash) ..................................... 112(T)
Ni30Co15Cr14Al
constituents ..............................69(T)
temperatures .......................... 124(T)
temperatures .......................... 123(T)
test at 1100C) ........................ 104(F)
test at 1200C) ........................ 104(F)
ash) ..................................... 112(T)
Ni30Co15Cr14AlY
thermalized expansion coefficients vs.
temperature .............................. 72(F)
ash) ..................................... 112(T)
Ni30Co22Cr11Al
temperatures .......................... 124(T)
temperatures .......................... 123(T)
Ni30Co22Cr11AlY
weight change (heat resistance isothermal test
at 1200C) ............................. 105(F)
ash) ..................................... 112(T)
Ni30Co30Cr8Al
constituents ..............................69(T)
temperatures .......................... 124(T)
temperatures .......................... 123(T)
test at 1200C) ........................ 104(F)
ash) ..................................... 112(T)
Ni30Co30Cr8AlY
temperature .............................. 72(F)
ash) ..................................... 112(T)
Ni30Co30Cr14Al
constituents ..............................69(T)
temperatures .......................... 123(T)
test at 1200C) ........................ 104(F)
ash) ..................................... 112(T)
Ni30Co30Cr14AlY
sigma-phase occurrence typical ................71

232 / Index
temperature .............................. 72(F)
ash) ..................................... 112(T)
Ni37Co15Cr11AlY, phase transformations,
interaction mechanism with superalloy at
high temperatures ............... 9192, 94(F)
Coatings, specific types, Thermal barrier
coatings
RT31/RT33
characteristics of coating ......184(T), 187188
growth parameter ........................... 185(T)
service life ................................... 185(T)
thermally grown oxide thickness ........ 185(T)
TBC, thermofatigue testing of coating/superalloy
systems ......................... 147(T), 148(F)
TBC-4
characteristics of coating ............ 184(T), 187
growth parameter ........................... 185(T)
service life ................................... 185(T)
thermally grown oxide thickness ........ 185(T)
ZrO2-8%Y2O3 /Ni20Cr12AlY, ceramic-layer
cohesive ultimate strength ....... 204205(F)
ZrO2-8%Y2O3 /Ni20Co20Cr11/6Al0.25Y, service
lives of samples ................... 183(F), 187
ZrO2-8%Y2O3 /Ni23Co20Cr11/5Al0.26Y, service
lives ....................................... 182(F)
Coating strain ................................. 196, 197
Coating-thermal-stresses modeling ... 132134(F)
Coating thermomechanical fatigue
characteristic .................................. 199
Coating thickness/specimen cross section
ratio .............................................. 119
Coating thickness value ........................... 197
Cobalt
aluminizing .....................................27
content effect on beta-phase-depleted outer
zone and oxidation ....................104105
content effect on elongation and ductility ..... 127
Ni-Co-Cr-Al alloys ......................... 112
content effect on overlay coatings with hightemperature testing .. 9193(F), 94(F), 95(T)
content effect on thermal expansion coefficient
of Ni-Co-Cr-Al alloys ................121122
content effect on thickness of interaction zone
of overlay coating and superalloy 93, 95(F)
content growth during high-temperature testing of
diffused coatings ..............................89
content increase in coating effects on coating/
alloy system ............. 9293, 94(F), 95(T)
effect on heat resistance and oxide-layer
formation and growth ...................... 103
Cobalt-nickel-chromium-aluminum alloys. See
also Coatings, specific types, cobalt-nickelchromium-aluminum alloys ..................... 3
chemical compositions .......... 71, 72(T), 212(T)
temperatures ....................... 124, 125(T)
elongation, various temperatures ........... 128(T)
heat resistance testing ..................105106(T)

hot corrosion resistance ................112113(T)
magnetic transformation .................. 71, 74(F)
microstructures ............................. 71, 73(F)
normalized number of variable .............. 212(T)
as overlay coatings, and high-temperature
testing ..........................9193(F), 94(F)
phase compositions ........................ 71, 72(T)
phase compositions and structure changes .....87,
9193(F), 94(F), 95(T)
phase transformations ..............................71
regression equations for property
values ...................................214215
temperatures ....................... 124, 125(T)
variability of values ............................... 214
yield strength .................................. 127(T)
Cobalt-nickel-chromium-aluminum-yttrium
alloys, hot corrosion resistance in GZT ash.
See also Coatings, specific types, cobalt-nickelchromium-aluminum alloys .............. 113(T)
Coefficient of plasticity ............................ 196
Coefficients of transfer in flux for various
materials, electric arc evaporation ........40(T)
Colloxiline ......................................... 33, 34
Color, of nickel aluminide diffusion
coatings ........................................56(T)
Composition-property regression equations ..... 3
Composition-structure-property relationships .. 3
Compressive stresses ............................... 198
of ceramic coatings ............................... 184
in ceramic layer of thermal barrier
coatings .................................206207
Compressor pressure ratio .......................... 5
Concentration gradients .............................87
Condensation coatings. See Overlay coatings.
Confidence coefficient .............................. 121
Copper
cathode spot temperature and vapor
pressure .....................................40(T)
evaporation .................................40(T)
Corrosion penetration depth ..................... 100
Creep rates, of superalloys in air and GZT
ash ............................................ 136(F)
Creep-rupture strength
of superalloys at high temperatures in
air ..............................136137, 138(F)
of superalloys in media provoking hot
corrosion ................134136(F,T), 137(F)
Creep strains, of thermal barrier coatings ....... 205
Cruising conditions .................................... 5
Crystalline structure, of nickel aluminide diffusion
coatings ........................................56(T)
Crystallization interval, regression equation for
calculation of ................................... 213
Cyclic heat resistance testing ..................9798
Cyclic service life calculation, of coating .........20

Index / 233
Cyclic temperature variations, and flaws in

turbine engines .................................... 5
Cyclic testing, for heat resistance ...................97
D
Databases ................................................ 3
DBTT. See Ductile-brittle transition temperature.
Defects, causes in coatings ...........................10
Deformation patterns .............................. 151
Degreasing, to prepare blade surfaces for
overlay coatings ........................ 45, 4647
Degree of ionization, of electric arc evaporated
products ........................................40(T)
Delta-phase nickel aluminide (Ni2Al3) ............55
color ...............................................56(T)
density ............................................56(T)
media .......56(T), 57(F), 58(F,T), 59, 60(T),
61(F), 62(F), 63
in diffusion coating formed in moderate and
low-activity media ....................... 64, 67
hardness ..........................................56(T)
oxidation time ................................. 100(T)
specific volume ..............................128129
Density
of evaporated substance ............................39
of nickel aluminide diffusion coatings .......56(T)
of zone of coating (q) .........................1718
Diffusion aluminide coatings ........................ 1
annealing of powder mixture ......................30
application ............................................. 8
application by electric arc technique under
vacuum ....................................3537
application in circulating gaseous
phase ................................... 3132(F)
application in powder mixtures with halide
activators .............................. 2531(F)
application using slurry ................... 3235(F)
beta-phase nickel aluminide aluminum
content ........................................ 145
blade service conditions limitation ................ 2
changes in phase size ...................... 64(F), 66
characteristics dependent on temperature, time of
annealing, and specific mass of aluminum
layer .............................................61
chemical compositions .................... 10, 11(T)
composition ranges .................................30
distribution patterns of alloying elements ..6768
ductile-brittle transition temperature ....... 120(F)
electric arc deposition below 640C ..............55
fatigue strength of coated
superalloys ......................... 139140(F)
filling saturation mixture, preventing outpour ..31

formation at temperatures above
640C ............................... 5963(F,T)
formation in low-activity media ...... 6369(F,T)
formation in moderate-activity
media ................................ 6369(F,T)
formed in high-activity media ......... 5563(F,T)
fragmentation vs. time of testing ........ 20, 21(F)
hot corrosion resistance .......................... 136
inspection of ....................................3738
interaction mechanism with
superalloys ............................ 8586(F)
intermetallic compound zones grow as in
aluminizing .....................................80
kinetics of coating growth on nickel-base
superalloys ............................ 27, 28(F)
life expectancy increased with blade surface
temperature ............................ 1519(F)
and low-cycle fatigue of
superalloys ...................... 149150(F,T)
mass of deposited Al5Si1.5Y alloy vs. coating
thickness .................................... 61(F)
microstructure ........................... 5563(F,T)
microstructure changes during high-temperature
tests .................................. 7986(F,T)
modified with silicon or yttrium ................... 2
on nickel, phase transformations .............. 80(F)
phase composition ...................... 5563(F,T)
phase composition changes during hightemperature tests .................. 7986(F,T)
phase composition, factors determining .........55
preparation of blade inner cavity for coating ...31
specific mass of deposited layer ..................61
stages of technological process ...................30
on superalloys, heat resistance of coating/
superalloy system .............. 106108(F,T)
thickness dependence on temperature and time
of aluminizing process ..................6667
thickness related to applied slurry thickness ....34
thickness related to specific mass of aluminum
alloy deposited .................................62
zones of successively arranged aluminides .....61
Diffusion annealing
after aluminizing ....................................29
after diffusion coating application using
slurries ........................32, 3334, 35(F)
temperature effect on hardness in slurry
technique .............................. 34(F), 35
Diffusion layer ........................55, 56(T), 57(F)
formation below 640C ...........55, 56(T), 57(F)
Dilatometric analysis
of cobalt-nickel-chromium-aluminum
alloys ............................................71
to examine phase transformations of Ni-Cr-Al
alloys ............................................70
Dimethyloxolate .................................. 33, 34
Directionally solidified superalloys, creep-rupture
strength ..........................136137, 138(F)
Disproportionation reactions ..................2526
Disulfides ................................................ 5

234 / Index
Dopants ......................................... 179, 180

erbium, for zirconium oxide crystal lattice .... 179
gadolinium as ...................................... 179
neodymium as ...............................179180
nickel oxide as ............................... 179, 180
ytterbium, for zirconium oxide crystal
lattice ....................................179180
Ductile-brittle transition temperature (DBTT)
of diffusion coatings .......................... 120(F)
range ................................................ 198
Ductility, vs. temperature ....................... 199(F)
Duga-90 unit ........................... 42, 43(F,T), 44
E
EA. See Electric arc technique.
EB. See Electron beam evaporation technique.
EBR. See Electron beam remelting.
Efficiency ................................................ 5
Elasomat instrument ............................... 121
Elastic deformation ................................. 151
Elasticity modulus
of coating ........................................... 196
of Co-Ni-Cr-Al alloys .................. 124, 125(T)
of Ni-Cr-Al alloys ............................. 122(T)
of Ni-Co-Cr-Al alloys ........ 123, 124(T), 125(F)
and thermal expansion coefficients of coating
alloys ................................120124(T)
20800C, regression equation for calculation
of ..................................213, 215, 216
Elastic-plastic strain range in coating under
nonisothermal loading ....................... 200
Electric arc current (Iea) .............................40
Electric-arc (EA) evaporation technique ......... 2
to apply overlay coatings .................3941(T)
cycle stages for coating deposition ...............50
forming overlay coatings showing
phase transformations followed by volume
effects .......................................77(T)
parameters controlled by units of deposition ...52
requirements on preparation of blade
surfaces ................................4647(T)
units for coating deposition ............ 4244(F,T)
Electric-arc method ................................... 2
Electric arc (EA) technique
for applying diffusion coatings under
vacuum ....................................3537
for depositing diffusion coatings, phase
transformations after high-temperature
testing ....................................... 84(F)
deposition of bond coats ......................... 167
diffusion coatings above 640C ....... 5963(F,T)
diffusion coatings below 640C ....... 5559(F,T)
microstructure of diffusion coating ..........7576
Electric arc vacuum deposition, as coating method
for diffusion coatings ........................11(T)
Electron beam (EB) evaporation technique ...... 2
applied to ceramic layer of thermal barrier
coatings .........................................21
to apply overlay coatings .....................3839

to apply overlay coatings, phase
transformations ............................ 85(F)
boundary temperatures at which condensate
structure undergoes changes ........ 4749(F)
ceramic pellets for ................... 162165(F,T)
condensate structures deposited on substrates
having different temperatures ........... 74(F)
cycle stages for coating deposition ... 4750(F,T)
depositing condensates for TEC studies ....... 120
evaporation of special ceramics .................... 2
forming overlay coating showing
phase transformations followed by volume
effects .......................................77(T)
forming overlay coating subject to heat treatment
operation ...............................7677(T)
for manufacturing aircraft engine turbine
blades ......................................... 162
parameters controlled by units of
deposition .................................5152
strict requirements to surface
preparedness .......................... 4546(F)
and thermal shock testing of coating/superalloy
systems .......................... 148149(F,T)
units for coating deposition ............ 4142(F,T)
Electron beam remelting (EBR), to melt alloys for
coatings deposited by the EB technique ......44
Electron beam thermal barrier
coatings ............................ 161190(F,T)
Elongation
of Co-Ni-Cr-Al alloys ........................ 128(T)
of Ni-Cr-Al alloys ........................... 126(F,T)
of Ni-Co-Cr-Al alloys .............. 127(T), 129(F)
Emissivity factor .......................... 180, 181(T)
Engine-equivalent quick tests ......................10
Engine testing ..........................................10
Epsilon-phase nickel aluminide (NiAl3) ..........55
color ...............................................56(T)
density ............................................56(T)
media ............... 56(T), 57(F), 58(F,T), 59,
60(T), 61(F), 62(F,T), 63
in diffusion coating formed in moderate and
low-activity media ........................56(T)
hardness ..........................................56(T)
melting point of compound ........................59
specific volume ..............................128129
Equilibrium, between phases of a coating and a
superalloy .........................................86
Equilibrium pressure of vapor of substance
evaporated ........................................39
Equivalent quick tests ...............................10
Erbium, dopants for ZrO2 crystal lattice ......... 179
Erosion ................................................... 8
Erosion on gas flow attack .................. 183, 184
Error for model using elastic strains ........... 197

Index / 235
ES 1/3/60 (Leybold) unit ......................... 177

Evaporation
coefficient of electron beam evaporation ........39
temperature of .......................................39
time of ................................................39
Evaporation rate, electron beam
evaporation ...................................3839
Evaporator thermal efficiency .....................38
F
Fatigue life, overlay coating effect ............... 140
Fatigue strength
coatings effect on superalloys ......138142(F,T),
143(F), 144(F)
factors having effect on coated specimens .. 138
139
superalloys with diffusion coatings .. 139140(F)
superalloys with overlay
coatings ... 140141(F,T), 142(F,T), 143(F),
144(F)
FEM. See Finite-element method.
Fighters, temperature and stress fields
nonuniform ......................................... 5
Finite-element method (FEM) ...... 202, 203, 204,
205(F), 207
Firing temperature ................................. 165
Flash method ........................................ 177
Force in coating (P) ................................ 133
Fractography, to evaluate thermomechanical
fatigue test specimens ................... 155, 157
Fragmentation ....................1920(F), 198, 199
causes .................................................19
on ceramic layer of thermal barrier coating .....21
as initial stage of thermal barrier coating
failure ......................................... 203
thermomechanical fatigue criterion
consideration ................................. 202
Free aluminum, effect on thickness in high-activity
mixture for aluminizing ................ 2627(F)
G
Gadolinium, as dopant ............................. 179
Gamma double-prime phase nickel aluminide
(Ni3Al) .............................................. 2
aluminum content ............................. 119(T)
chromium solubility in Ni-Cr-Al system
alloys .......................................... 111
color ...............................................56(T)
controlling interaction mechanism between
overlay coatings and superalloys ......9396
density ............................................56(T)
media ........................ 61(F), 62(F,T), 63
in diffusion coating formed in moderate and lowactivity media .......63(F), 64(F), 6566, 67,
69(T)
elongation ...................................... 119(T)
formation beneath oxide layer in NiAl
oxidation ...................................... 101
hardness ..........................................56(T)
high-temperature strength ........................ 120
of Ni-Cr-Al system ....................... 87, 88)(F)
in overlay coatings at high-temperature tests ...87
oxidation time ................................. 100(T)
phase transformations in diffusion coatings after
high-temperature tests ............84(F), 85(F)
in phase transformations of overlay coatings
on superalloys ............89(F), 90(F), 91(F),
92(T), 95(T)
specific volume ..............................128129
two-way diffusion in diffusion coatings .... 80(F),
81(T), 82(F), 83(T), 84(F), 85(F)
yield strength .................................. 119(T)
Gamma-phase .........................................98
in aluminum-silicon-yttrium diffusion
coatings ..................................... 85(F)
in cobalt-nickel-chromium-aluminum
alloys ..........................71, 72(T), 87, 91
in nickel-chromium-aluminum alloys .. 65(T), 70,
76(F), 77, 87(F), 88(F), 89(F), 90, 91(F),
92(T), 93(F)
in nickel-cobalt-chromium-aluminum alloys
.... 68(T), 69(T), 70, 71, 87, 91, 94(F), 95(T)
Gamma-prime phase .................................97
in aluminum-silicon-yttrium diffusion
coatings ..................................... 85(F)
in nickel-chromium-aluminum alloys .. 65(T), 70,
76(F), 77, 88(F), 89(F), 91, 92(T), 93(F)
in nickel-chromium-molybdenum-tungstencobalt alloys ....................................82
in Ni-Co-Cr-Al alloys ...68(T), 69(T), 70, 71, 91,
94(F), 95(F)
Gamma-solid solution ................. 70, 76, 77, 87,
88, 89, 9396
controlling interaction mechanism between
overlay coatings and superalloys ......9396
formation, and phase transformations and internal
oxidation ............................... 17, 18(F)
in overlay coatings at high-temperature tests ...87
Gas aluminizing methods ............................ 2
Gas corrosion resistance ............................97
Gas dynamic forces ................................... 8
Gas-dynamic test rig, schematic diagram ... 145(F)
Gas-dynamic test rigs and
engines ............................. 189190(F,T)
Gaseous halides, in application technique of
diffusion coatings ....................... 3132(F)

236 / Index
Gas-flow test benches (gas rigs) ...................99

Gas temperature, at turbine entry ................... 5
Gas turbine fuel synthetic (GZT)
ash ................................. 99100, 111(F)
composition ......................................99(T)
creep-rupture strength of superalloys having hot
corrosion ................134136(F,T), 137(F)
for hot corrosion testing of Ni-Cr-Al
alloys ............................. 110111(F,T)
for hot corrosion testing of overlay coatings
...........109110, 112(F,T), 113(T), 114(F),
115(F), 116(F)
GZT ash. See Gas turbine fuel synthetic ash.
H
Hafnium
alloying addition effect on Ni-Co-Cr-Al
alloys ....................................101103
Ni-Co-Cr-Al alloys ......................... 112
overlay coatings ............................. 111
effect on heat resistance of Ni-Cr-Al
alloys ................................ 101, 102(F)
as replacement for yttrium, effect on oxides of
Ni-Cr-Al alloys .................... 101, 102(T)
substitution for yttrium effect on Ni-Co-Cr-Al
cyclic heat resistance ....................... 105
Halide activators
coating formation from powder mixtures .......37
compounds used in aluminizing ..................26
diffusion coatings in powder mixtures .. 2531(F)
Halides ................................................ 1, 2
amount at aluminizing ..............................27
Hall end-face plasma accelerator ............. 40, 41
Hard coatings, case for .......................... 198
Hardness, of nickel aluminide diffusion
coatings ........................................56(T)
Hardness testing, of diffusion aluminide
coatings ............................................38
Heat capacity, determination by pulse
method .....................................175176
Heat flow ................................................. 5
Heat of formation, of nickel aluminide diffusion
coatings ........................................56(T)
Heat resistance, metallographic determination ...98
Heat resistance, cyclic test 20900C
1000h, regression equation for calculation
of .............................................. 215
of .............................................. 215
of .............................................. 215
Heat resistance, isothermal test at 1200C
100h, regression equation for calculation of .. 214
300h, regression equation for calculation of .. 214
Heat resistance parameter ........................ 107
Heat resistance testing ...................... 9799(F)

of aluminides ......................... 100106(F,T)
overlay coatings ...................... 100106(F,T)
Heterophase diffusion theory statement .........64
High-activity media
description ............................................55
temperature range ...................................55
High-pressure (HP) turbines, nozzle vanes ....... 6
High-temperature corrosion ......................... 8
of overlay coatings .............................. 11(F)
High temperatures
blade working cross-sectional area reduced in
hot zone ...................................... 195
effect on diffusion aluminide coatings,
phase composition and
microstructure ..................... 7986(F,T)
protection against .................................... 2
High-temperature strength .......................... 7
coating effects on superalloys ...... 134138(F,T)
High-temperature testing ............... 7996(F,T)
of overlay coatings, microstructural
changes ............................. 8796(F,T)
of overlay coatings, phase composition
changes ............................. 8796(F,T)
Homogeneity region, of nickel aluminide diffusion
coatings ........................................56(T)
Hot corrosion ................................ 1, 3, 8, 10
causing defects in coating .........................10
of coating/superalloy systems in GZT
ash .......................134136(F,T), 137(F)
mechanism caused by aggressive compounds ..99
of Ni-Cr-Al alloys .................... 110111(F,T)
with salt addition to gas flow ................... 145
of superalloys, creep-rupture strength with GZT
ash exposure ...........134136(F,T), 137(F)
temperature range 600 to 1200C resistance ..... 9
temperature regimes, depth of
development ............................... 11(F)
thermal barrier coatings for protection ...........20
of turbine blades ........................... 1215(F)
of ZrO2-8%Y2O3 ceramics ....................... 184
700C, 200h, regression equation for
calculation of ................................ 215
calculation of ................................ 215
calculation of .......................... 214, 215
calculation of .......................... 214, 215
Hot corrosion resistance ........................ 10, 97
of coated superalloys ................ 113116(F,T)
of Co-Ni-Cr-Al alloys ..................112113(T)
effect of refractory elements, low solubility in
aluminides ......................................63
of Ni-Cr-Al alloys .................... 110111(F,T)
of Ni-Co-Cr-Al
alloys ................... 111112(F,T), 113(T)
of overlay coatings ................... 109113(F,T)
Hot corrosion resistance testing, 99100(T)
HP. See High-pressure turbines.

Index / 237
HT-1000 calorimeter (Setaram Company) .... 177

Hydroabrasive blasting, to prepare blade surfaces
for overlay coatings ..................... 45(F), 46
Hydroabrasive processing ........................ 185
Hydrocarbon combustion ............................ 6
Hydrogen, as impurity of jet fuels ................... 5
Hydrogen gas ........................................... 6
Hydroxide method ............................ 164, 177
Hysteresis loops ........................... 152(F), 153
I
Impurities
industrial, as hazard ............................... 6, 9
of jet fuels ..........................................56
Induction transducer ............................... 121
Infrared pyrometers, in Duga-90 unit .............44
Inspection
of diffusion aluminide coatings ...................37
of overlay coatings .............................5152
Interaction zone thickness (d) ......................90
Intercrystalline defects ..................45(F), 46(F)
Intercrystalline porosity ........................... 140
Ion transfer rate, of electric arc technique ........36
Iron
aluminizing ....................... 2728(F), 29
in deposits detected on gas-turbine engine blade
surfaces .........................................99
Islets of lighter phase, at diffusion layer/aluminum
interface .................................. 56, 57(F)
J
Jet fuel combustion .................................... 6
product composition ................................. 5
product effect on turbine blade surface layers ... 5
temperature of ........................................ 6
Jet fuels, impurities .................................... 5
Jet fuel sulfur ........................................... 5
L
Lamellar carbides precipitation ...................85
Laser flash method, to measure thermal
conductivity of thermal barrier coating
ceramic layer .......................... 175, 176(F)
LA-TMF test. See Load-adjusted thermomechanical
fatigue test.
Lead, as impurity of jet fuels ......................... 5
Leaders ......................................45(F), 46(F)
Life-expectancy calculations ............... 1415(F)
Liquid penetrant inspection, of

thermomechanical fatigue crack on coated
superalloy ............................. 155, 157(F)
Liquid-phase aluminizing method ................33
Liquidus temperature, regression equation for
calculation of ................................... 213
Load-adjusted thermomechanical fatigue test
(LA-TMF) ......... 150151, 152153(F), 156
for JS6U superalloy ........................... 200(F)
Low-cycle fatigue, of superalloys, coating
effect ......................... 149150(F), 151(F)
Low-cycle fatigue
testing ....................... 149150(F), 151(F)
Low-pressure plasma spray (LPPS) method
deposition of bond coats ......................... 167
for deposition of coatings on superalloys ..... 153
LPPS. See Low-pressure plasma spray method.
M
Macrosegregation, in ingot with overlay
coatings ........................................45(T)
Magnesium, as impurity of jet fuels ................. 5
Manganese, as impurity of jet fuels ................. 5
Mansons formula ..................... 200201, 202
MAP unit ................................... 4243(F,T)
Margin of safety, reduction of ......................10
Mass, changes not related to coating life
expectancies .......................................17
Material density (q) ..................................97
Material evaporated per second, electric arc
evaporation technique ...........................40
Maximum admissible gas temperature ........... 7
Maximum cycle temperature ..................... 200
Maximum relative tensile thermal stress of
thermal barrier coatings ................ 204(F)
Mean mass temperature, of gas at turbine
entry ................................................. 5
Mechanical mixing ....................... 163, 164(T)
Melting temperature, of nickel aluminide
diffusion coatings ............................56(T)
Mercaptanes ............................................ 5
Metal heat resistant coatings .......................21
Metallographic examination, to determine
coating thickness of diffusion aluminide
coatings ....................................... 37, 38
Microball peening ........................ 130(F), 131
of as-deposited coating ........................... 115
of bond coats for thermal barrier coatings .... 166
of CMSX-4 superalloy with coating ........... 153
fatigue strength of overlay coatings affected 140
of overlay coatings ........................... 75, 140
of thermal barrier coatings ....................... 185
Microcracking .........................................21
Microdots, and low-cycle fatigue ................. 149
Microdroplet phase, in electric arc
evaporation ........................................40

238 / Index
Microdrops ........................................... 140

on bond coat, caused by high yttrium contents of
liquid bath .................................... 161
on bond coat, causing defects in TBC ceramic
layer ........................................... 161
Microstructure, high-temperature test effects on
diffusion aluminide coatings ....... 7986(F,T)
Minimum cycle temperature ..................... 200
Molecular mass of substance evaporated ........39
Molybdenum
alloying NiAl with ................................ 101
solubility in gamma double-prime
phase compound after high-temperature
testing ...........................................82
Moment of the coating chemical and phase
composition change .............................98
Multilayer coatings .................................... 9
Multiple correlation coefficient (Rf) corrected on
the basis of degrees of freedom .............. 212
l-phase
molybdenum-base as precipitate ..................82
in nickel-chromium-molybdenum-tungsten-cobalt
superalloys ........................ 81(T), 82(F)
two-way diffusion in diffusion
coatings ............................ 81(T), 82(F)
N
Neodymium, as dopant .......................179180
Neutral coatings, case for ....................... 198
Nickel
aluminizing .....................................27
evaporation .................................40(T)
alloys .......................................... 106
of Co-Ni-Cr-Al alloys ...................... 124
degree of ionization .............................40(T)
mass-spectrometric analysis data for material
evaporation .................................40(T)
Nickel aluminide diffusion coatings
physical properties ......................... 55, 56(T)
saturating media classification ........... 55, 56(T)
NiAl2O4 spinel ..........................................83
formation with NiAl oxidation .................. 101
Nickel-chromium-aluminum alloys. See also
Coatings, nickel-chromium-aluminum
alloys ................................................ 3
bend testing ........................................ 128
chemical compositions .......... 65(T), 70, 212(T)
temperatures ............................. 122(T)
elongation, various temperatures ......... 126(F,T)
heat resistance ........................... 101, 102(T)
hot corrosion .......................... 110111(F,T)
microstructure ................................... 66(F)
phase composition and structure

changes ........................8791(F), 92(T)
phase compositions ........................ 65(T), 70
phase diagrams at high temperatures ......... 87(F)
regression equations for property values ...... 216
temperatures ............................. 121(T)
temperature ............................ 67(F), 70
ultimate strength ............................. 126(F,T)
weight loss due to hot corrosion ............ 110(T)
Nickel-chromium-molybdenum-tungsten-cobalt
alloys .................................. 81(T), 82(F)
Nickel-cobalt-chromium-aluminum alloys. See
also Coatings, nickel-cobalt-chromiumaluminum alloys ................................... 3
alloying effect on specific mass loss of hot
corrosion .................................. 112(F)
alloying effect on thickness of beta-phasedepleted outer zone ............... 103, 105(F)
bend testing ........................................ 128
beta-phase contents in alloy and coating
structures .............................. 70, 72(F)
chemical composition ........... 68(T), 70, 211(T)
constituents ............................ 69(T), 70
temperatures ............. 123, 124(T), 125(F)
elongation, various temperatures ........... 127(T)
hot corrosion resistance ... 111112(F,T), 113(T)
microstructures ............................. 69(F), 70
as overlay coatings, and high-temperature
testing ................. 9193(F), 94(F), 95(T)
phase composition and structure
changes ...........87, 9193(F), 94(F), 95(T)
phase compositions ........................ 68(T), 70
phase transformations ..............................71
regression equations for property
values ...................................213214
strength levels of cast alloys lower than modern
superalloys ................................... 128
thermal expansion coefficients ..... 122124(F,T)
temperatures ........................... 123(F,T)
temperature ............... 71, 72(F), 73, 74(F)
ultimate strength, various temperatures .... 127(T)
testing) .................................... 103(T)
weight gain (heat resistance, isothermal
testing) .................................... 103(T)
yield strength, various
temperatures ................... 127(T), 129(F)
Nickel-cobalt-chromium-aluminum-yttrium
alloys, service life ............................ 15(F)
Nickel oxide

Index / 239
as corrosion product penetrating the coating ....13

as dopants, effect on thermal
conductivity ............................ 179, 180
precipitate .................................... 178, 179
Nickel substrate, ductile-brittle
transition temperatures for diffusion coatings
50lm
thick .......................................... 120(F)
Nickel superalloys, as turbine blade material ...... 1
Nitrogen, as impurity of jet fuels .................... 5
Nondestructive testing, to detect cracks in
diffusion coatings ................................38
Normalized magnitudes of
variables ..............................211212(T)
Nozzle guide vanes .................................... 6
Nozzle vanes
of high-pressure (HP) turbine ...................... 6
temperature differences between vane back and
trailing edges .................................... 7
thermal inhomogeneity .............................. 6
Number of cycles to coating failure at
thermocyclic loading of blade .......... 201(F)
Number of cycles to crack detection under given
cyclic loading conditions .................... 199
Number of loading cycles ......................... 199
O
Operating diagram of an airliner engine .. 5, 6(F)
Oriented porosity fraction ........................ 178
Overlay coatings .................................2, 3, 9
application ............................................. 8
application techniques .................. 3852(F,T)
ceramic ............................................. 163
chemical composition of Ni-Cr-Al alloys with
hafnium and silicon additions ......... 102(T)
chemical compositions .................... 10(T), 74
cobalt-base, advantages .......................... 116
cobalt-nickel-chromium-aluminum ...... 71, 72(T)
crystallization intervals .........................49(T)
electric arc evaporation technique for
application .............................3941(T)
electron beam evaporation technique for
application .................................3839
fatigue strength of coated superalloys ....... 140
141(F,T), 142(F,T), 143(F), 144(F)
fragmentation vs. time of testing ........ 20, 21(F)
heat treatment effect on beta-phase
quantity ................................ 76, 77(T)
heat treatment effect on phase transformations
and volume effects ........................77(T)
high-chromium or cobalt-base, orientation ......74
high-temperature corrosion .................... 11(F)
high-temperature heat treatments ..... 7677(F,T)
hot corrosion resistance ....... 109113(F,T), 136
inspection of ....................................5152
interaction mechanism with superalloys ....9396

interaction zones ....................................79
iron-base, hot corrosion resistance ............. 116
life coefficient values ......................... 116(T)
life on turbine blades of JS6U
superalloys ............................... 116(T)
liquidus-solidus temperatures .................49(T)
microstructure changes at high-temperature
tests .................................. 8796(F,T)
microstructures ..................... 71, 74(F), 75(F)
nickel-cobalt-chromium-aluminum
alloys ........................ 69(F,T), 7071(F)
peening of finished coatings ......... 5051(T), 52
phase composition changes at high-temperature
tests .................................. 8796(F,T)
phase composition, factors determining .........55
phase compositions ..................... 6977(F,T)
preparing blade surfaces for coating 4547(F,T)
producing ingots and cathodes for
coatings ................................4445(T)
recrystallization .................................... 130
recrystallization annealing .........................51
strain-hardening processing ........................51
structural constitution, factors determining .....55
substrate temperature effect on stability and
hardness ........................................75
on superalloys, heat resistance of coating/
superalloy system .............. 108109(F,T)
thermomechanical fatigue testing .. 150158(F,T)
thickness of condensed substance ................39
x-ray structural pattern ................ 76(F), 77(T)
Oxidation ............................................. 1, 8
blade damage ........................................17
causing defects in coating .........................10
during cooling in aluminizing process ...........29
effect on creep-rupture strength of
superalloys ................................... 137
penetration depth ....................................98
resistance .....................................1, 10, 97
steady-state ...........................................98
of superalloys at testing and surface cracks ... 134
temperature range 600 to 1200C resistance ..... 9
temperature regimes, depth of
development ............................... 11(F)
thermal barrier coatings for protection ...........20
turbine blade operating at temperatures
above 1000C ......................... 1519(F)
Oxidation activation energy (Q) ................. 107
Oxidation constant (Kp) ........................ 97, 98
for diffusion aluminized coated
superalloys ............................... 108(T)
Oxide lifetime ...................................... 98(F)
Oxide protective life .............................. 98(F)
hafnium addition effect on nickel-chromiumaluminum alloys ............................. 101
of overlay coatings on JS6U superalloys .. 108(T)
Oxides, performing protective function on
blades ..................................... 17, 18(F)
Oxide spalling .........................................98

240 / Index
Oxygen ................................................... 6
P
Pack calorizing
as coating method for diffusion coatings ....11(T)
under vacuum, as coating method for diffusion
coatings .....................................11(T)
Palladium, as alloying element for diffusion
coating ............................. 115(T), 116(T)
Parabola exponent (n) .......................... 97, 98
Passenger aircraft engines, temperature and stress
fields ................................................ 5
Peening. See also Microball peening.
of Ni-Cr-Al and Ni-Co-Cr-Al overlay
coatings ..................................... 75(F)
overlay coatings ....................... 5051(T), 52
Phase compositions
determined from coating thickness ...............80
high-temperature test effects on diffusion
aluminide coatings ................ 7986(F,T)
Phase transformations, in diffusion coating on
nickel ........................................... 80(F)
Phonon scattering ................................... 179
Pillars ..................................................... 8
Pit corrosion .................................. 12, 13(F)
Plane-strain-state models,
generalized ..........................202, 203, 207
Plasma flux ionization ...............................40
Plasma method ......................................... 2
Plasma transfer ........................................ 2
Plasticity, coefficient of ............................ 196
Poissons ratio for the coating in case of its
elastic straining ................................ 196
Porosity, as result of diffusion processes ..........79
Potassium, in deposits detected on gas-turbine
engine blade surfaces ............................99
Powder metallurgy ..............................1, 2, 8
aluminide coating application drawbacks .......30
aluminide coating characteristics .................30
diffusion coating applications with halide
activators .............................. 2531(F)
Power of the electron beam (Pe) ...................38
Preoxidizing, of bond coat of thermal barrier
coatings .......................................... 167
Pulse method ......................................... 177
to determine thermal diffusivity and heat
capacity .................................175176
Q
Quantimet instrument, to determine NiAl phase
amount in coating structure .....................70
Quartz resonators, in Duga-90 unit ................44
R
Radiation heat exchange .......................... 176
Recrystallization .................................... 131
in overlay coatings ................................ 130
Recrystallization annealing
overlay coatings .....................................51
of thermal barrier coatings ....................... 185
Reduced modulus of elasticity in the strain
theory of plasticity ............................ 197
Refractory elements
content growth during high-temperature testing
of diffused coatings ...........................89
diffusive mobility ...................................68
low solubility in aluminides providing
hot corrosion resistance to diffusion
coatings .........................................63
not contained in NiAl compound of
diffusion coating in low-activity powder
mixtures ............................... 65(T), 67
phase precipitation of diffusion coatings on base
of ...........................................6667
Regression equations .........................212216
for Ni-Co-Cr-Al alloys, specific weight
measurements ................................ 103
Relative elongation, 7001100C, regression
equation for calculation of .................... 215
Reproducibility, of aluminizing coatings ..........29
Residual compressive stress, in aluminide
coatings .......................................... 129
Residual sum of squares .......................... 212
Resonance vibrations .............................. 7, 8
Rigidity of coatings ................................. 144
Rinsing, to prepare blade surfaces for overlay
coatings .................................. 45, 4647
Roentgen (x-ray) radiation .........................38
S
Sample specific weight gain ........................97
Sample specific weight loss .........................97
Scale removal method ............................. 100
SC-TMF test. See Strain-control thermomechanical
fatigue test.
Sea salts .......................................... 6, 9, 12
Secant modulus for the coating material ...... 196
Semidry molding technique ............ 164, 165(T)
Service life .........................................2, 3, 8
of coated turbine blades ............................. 5
coating chemical composition effects on
thermomechanical loading .......202203(T)
definition ........................................... 186
and deposition technique ......................... 190
determining coating composition and
thickness ........................................10
1710BL unit .......................................... 186
Sigma phase .......................................7071

Index / 241
in cobalt-nickel-chromium-aluminum
alloys ................................... 71, 72(T)
in nickel-cobalt-chromium-aluminum
alloys .......................................7071
Sign of first principal strain for maximum strain
intensity in coating ........................... 197
Silicification, with aluminizing ....................... 1
Silicon
addition effect on hot corrosion resistance of
overlay coatings ............................. 111
aluminizing ................................. 28(F)
alloying addition effect on Ni-Co-Cr-Al
alloys ....................................101103
as alloying element for diffusion coating ...... 116
as alloying element, heat resistance benefit ... 101
content effect on aluminide coatings ........... 108
content effect on diffused coating thickness with
slurry technique ........................... 35(F)
Ni-Co-Cr-Al alloys ......................... 112
content in diffusion aluminide
coatings ................................ 62(F), 63
effect on heat resistance of Ni-Cr-Al
alloys ................................ 101, 102(F)
substitution for yttrium effect on Ni-Co-Cr-Al
cyclic heat resistance ....................... 105
Silicon-and yttrium-alloyed coatings,
testing ..................................84(F), 85(F)
Siliconizing process ...................................28
Silver, coefficient of transfer in flux, electric arc
evaporation ....................................40(T)
Simulations .............................................. 3
Simultaneous deposition ................. 164, 165(T)
Slurry aluminizing, as coating method for
diffusion coatings ............................11(T)
Slurry technique ....................................... 2
for application of diffusion aluminide
coatings ................................ 3235(F)
for applying diffusion aluminide coatings, phases,
structures, and compositions .....62(F,T), 63
composition of powders ............................33
diffused coating thickness .........................34
explosion hazard ....................................35
and mechanical finishing afterwards .............34
surface roughness, hardness and diffused
coating thickness affected by silicon
content ................................. 3435(F)
Sodium
surfaces .........................................99
Sodium chloride (NaCl), causing hot
corrosion ...........................................99
Sodium salt .............................................99
causing hot corrosion ...............................99
Sodium sulfate (Na2SO4), causing hot
corrosion ...........................................99
Softening .............................................. 2, 8
Soft heat resistant metallic coatings ......... 198
Solidus temperature,, regression equation for
calculation of ................................... 213
Spalling
caused by unsatisfactory adhesive strength
level ........................................... 133
of ceramic layer ................................... 169
of ceramic layer, black color after check
annealing .........................170, 171(F,T)
of ceramic layer of thermal barrier
coatings ..........166, 182188(F), 190, 203,
206(F)
of ceramics on gas-dynamic test
rig ................................. 189190(F,T)
of ceramics on turbine blades with thermal barrier
coatings ................................ 2022(F)
of diffusion aluminide coatings ...................38
of diffusion coating/superalloy system during
isothermal test ............................... 107
as result of diffusion processes between coating
and alloy ........................................79
of thermal barrier coating ceramic layer,
criterion of ......................... 205209(F)
of ZrO2-15%Y2O3 ceramic ...................... 109
Specific mass gain ....................................38
Specific power .......................................... 5
Specific weight .......................................... 5
Specific weight change .. 97, 98(F), 104(F), 105(F),
110(F), 114(F)
Specific weight gain .................97, 100, 102(T),
103(T), 106(T)
measurement at aluminizing to determine coating
thickness ........................................38
Specific weight loss ..................97, 100, 110(T),
112(T), 113(T)
Specimen mass loss ................................. 100
Spinels ............................................ 83, 189
formation on bond coat surface ................. 182
formation on diffusion coating of coating/
superalloy system ........................... 107
in oxide layer of overlay coating/superalloy
system ......................................... 109
proportion in Ni-Co-Cr-Al alloys increased with
cobalt content ................... 103, 104105
Splashing ...............................................51
during electron beam evaporation technique ....47
Stabilization time of evaporation ..................39
Stabilized zirconia oxide-base ceramics ........... 2
Stark (statistical analysis regression
correlation) ..................................... 211
Steady-state oxidation ...............................98
Strain-control thermomechanical fatigue test
(SC-TMF test) .................... 150151, 152
Strain intensity .................................. 201(F)
in coating .............................. 196197, 198
maximum value of ................................ 197
minimum value of ................................ 197
vs. temperature ................................ 199(F)

242 / Index
Strain-intensity range
in the coating ................... 197, 199200, 202
at thermocyclic loading of blade ............ 201(F)
Strain range ...................................... 154(F)
in coating ................................. 153, 154(F)
of specimen at cycling loading .................. 199
Strains .................................................... 2
of coating and blade ........................195196
in nozzle vanes ....................................... 7
Stress concentrators ................ 140, 141, 143(F)
causing low-cycle fatigue ........................ 149
coating defects in overlay coatings ....... 141, 142
Stresses. See also Thermal stresses ............... 2, 7
Stress intensity in the coating .......... 196, 201(F)
Stress of separation ....................... 208(F), 209
Sulfides ................................................... 5
Sulfur ..................................................... 6
content effect on hot corrosion ............... 11, 12
content in corrosion products ................1314
content range of aviation fuel ...................... 5
surfaces .........................................99
Superalloys
for aircraft turbine blades, chemical
compositions .......................... 10, 11(T)
coated, hot corrosion resistance .... 113116(F,T)
coating effect mechanism on high-temperature
strength ............................. 137138(F)
with coatings, heat resistance ...... 106109(F,T)
creep-rupture strength in air at high
temperatures .................136137, 138(F)
creep-rupture strength in media provoking hot
corrosion ................134136(F,T), 137(F)
with diffusion coatings, fatigue
strength ............................. 139140(F)
with diffusion coatings, heat resistance
of ................................. 106108(F,T)
fatigue strength, coating
effects ......... 138142(F,T), 143(F), 144(F)
high-temperature strength coating
effects ............................ 134138(F,T)
interaction mechanism with overlay
coatings ....................................9396
low-cycle fatigue, coating effect on .......... 149
150(F), 151(F)
nickel-aluminum, formation of diffusion layer
below 640C ..................55, 56(T), 57(F)
with overlay coatings, fatigue
strength ... 140141(F,T), 142(F,T), 143(F),
144(F)
with overlay coatings, heat resistance
of ................................. 108109(F,T)
service life .......................... 155, 157, 183(T)
thermal fatigue and coating
effects ............................ 142150(F,T)
thermomechanical fatigue coating effects
on ................................. 150158(F,T)
Superalloys, specific types
CMSX-4
ceramic layer spalling .....................185(F),

187(T), 188, 207
microstructure of thermal barrier coating bond
coat on ........................... 186(F), 188
service life ....................................... 157
temperatures .......................... 121(T)
thermomechanical fatigue testing, cyclic life
with coatings .... 155(F), 157(F,T), 158(F)
thermomechanical fatigue testing of coated
superalloy systems .......... 153, 154(F,T),
155(T), 156(F,T), 157(F,T), 158(F)
Hastelloy X
cycles to spallation of thermal barrier coatings
on superalloy ..............184(F), 187188
physical properties of thermal barrier coated
samples ................................ 183(T)
service lifetime of thermal barrier coated
samples ................................ 183(T)
IN 625
physical properties of thermal barrier coated
samples ................................ 183(T)
service lifetime of thermal barrier coated
samples ................................ 183(T)
JS-type, ductile-brittle transition temperatures for
diffusion coatings 50lm thick ........ 120(F)
JS6F
cyclic testing with diffusion and overlay
coatings ............................. 20, 21(F)
fatigue strength ................................. 140
multilayer coating system residual
stresses ................................. 134(F)
thermal fatigue cracking
(fragmentation) .................... 19, 20(F)
thermal stresses variations with different
orientation ....................... 133(F), 134
JS6F-DS
systems ................................ 148(T)
thermofatigue testing ...... 146147(T), 148(F)
JS6F-SC, thermal shock testing of coating/
superalloy systems ...................... 148(T)
JS6K
aluminizing ........................................27
aluminizing from slurries, phase and chemical
compositions of outer zone ...... 62(T), 63
aluminizing in aluminum-silicon
mixture ................................... 28(F)
coating thickness related to aluminum reserve
in coating ....................................86
diffusion coating characteristics, 98% Al-Fe
alloy and 2% ammonium chloride .. 64(F),
65(T)
heat resistance and fatigue strength of coated
blades ...................................... 142
interaction kinetics with overlay coatings and
chromium content increase ...... 90, 92(T)

Index / 243
low-temperature aluminizing and composition

of coating surface zone ...............5758
oxidation constant with diffusion aluminized
coatings, isothermal tests ........... 108(T)
oxide life .................................... 106(T)
phase and chemical compositions of diffusion
coating formed in low-activity powder
mixtures ............................ 65(T), 67
phase and chemical compositions of diffusion
testing ...............................8283(T)
thermal fatigue resistance of diffusion coated
superalloy ............................. 145(F)
time-to-failure in air and in GZT
ash ................................ 135(T), 136
JS6U
aluminizing by EA method from alloy Al5SiY,
phase and chemical compositions ......59,
60(F,T), 61(F)
ceramic layer thickness effect on service
life ................................ 183(F), 186
coating life at 800C ....................... 116(T)
creep rates ................................... 136(F)
creep-rupture strength in GZT ash causing hot
corrosion ................... 134, 135(F), 136
cyclic life of overlay coated vs. diffusion
coated superalloy ......................... 146
cyclic testing with diffusion and overlay
coatings ............................. 20, 21(F)
elongation of coated and uncoated specimens
in GZT ash ...................... 136, 137(F)
fatigue strength of coated blades at 20C .. 140,
142(T)
fatigue strength of overlay coating/superalloy
systems ...140(T), 141(F), 142(T), 143(F)
fatigue strength of thermal barrier coating vs.
no coating ....................... 141, 144(F)
fatigue strength with diffusion coatings
........................................... 139(F)
forces acting in Ni22Cr12AlY coating vs.
coating thickness ............... 132(F), 133
blades ...................................... 142
heat resistance with diffusion coatings
(isothermal and cyclic tests) .. 107, 108(F)
high-temperature test results of interaction
kinetics with overlay coatings, chemical
composition effects ............... 93, 95(T)
interaction kinetics between overlay coatings
and superalloys, chemical composition
effects ............................... 93, 95(T)
life of diffusion coatings, hot corrosion testing
in GZT ash ............................ 115(F)
overlay coating changes of phase composition
and structure after high-temperature tests
............... 8788(F), 89(F), 90(F), 92(T)
overlay coating thicknesses and deposition
method ................................. 109(T)
oxidation constant with diffusion aluminized

coatings, isothermal tests ........... 108(T)
oxidation rates in diffusion coating .. 1516(F)
oxide life on uncoated superalloy, cyclic
testing .................................. 108(T)
oxide structural change with heat resistance
testing of overlay coatings/superalloy
system ............................ 109, 110(F)
testing ............................... 84, 85(F)
phase transformations, aluminized in 98%Al-Fe
alloy and 2% ammonium chloride
at 1000C ................................ 64(F)
with overlay coating at high
temperatures .......... 89, 9091(F), 93(F)
thermal stress distribution measured during
chemical removal of
coatings .......................... 131, 132(F)
thermal stresses from coatings
calculated ........................ 132(F), 133
thermal stresses in diffusion coating vs.
temperature ...................... 129(F), 130
thermal stresses in overlay coatings at
20C .................................... 130(F)
thermal stresses of coatings applied to .. 131(F)
thermal stresses variations in Ni20Cr12AlY
coating and superalloy .............. 133(F)
thermofatigue testing ................146147(T)
systems ...................... 147(T), 148(F)
thermomechanical fatigue tests of coated
superalloy ............................. 200(F)
thickness variations of diffusion coatings
during high-temperature testing ...... 86(F)
time to failure values in GZT synthetic
ash ....................................137138
uncoated, thermal fatigue present ............ 146
weight change of overlay coating/superalloy
system (isothermal test) ............. 110(F)
JS6U-DS, thermal shock testing of coating/
superalloy systems ...................... 148(T)
JS26
high-temperature test results of interaction
kinetics with overlay coatings, chemical
composition effects ............... 93, 95(T)
effects ............................... 93, 95(T)
temperatures .................. 9091, 93(F)
JS26-DS
creep-rupture strength in
air ...........................136137, 138(F)
low-cycle fatigue testing ............ 149150(F)

244 / Index
Superalloys (continued)
microstructure after high-temperature testing
for stress-rupture strength ........... 138(F)
systems ...................... 147(T), 148(F)
JS30, chemical composition ...................11(T)
JS30-SC, thermofatigue
testing ..................... 146147(T), 148(F)
JS32
density ....................................... 181(T)
JS32-DS
fatigue strength of coated superalloy with
overlay coatings .......................... 141
systems ...................... 147(T), 148(F)
MAR-M-002
intensities of strains and stresses within startmaximum-shut cycle ................. 202(T)
thermal barrier coating testing on gas-dynamic
test rigs and engines ........... 189(F), 190
thermal expansion coefficient and cyclic life of
coating ..................................... 203
thermomechanical fatigue testing, hysteresis
loop calculation ................. 152(F), 153
MAR-M-200
coated blade surface temperature at minimum
and maximum gas temperature .... 202(F)
cycles to coating failure of coated
blades ............................ 201202(F)
overlay coatings effect on strain intensity and
ductility vs. temperature ............ 199(F)
test rigs and engines ..................... 190
Nickel
chemical composition of diffusion
coatings ............................. 57, 58(T)
diffusion aluminide coating characteristics,
aluminized in 98% Al-Fe alloy and 2%
ammonium chloride ........... 62(T), 63(F)
phase transformation, aluminized in 98% Al-Fe
alloy and 2% ammonium chloride
at 1000C ................................ 64(F)
phase transformation diagrams after EA
deposition method ................ 57, 58(F)
Ni20CrTi
chemical composition of diffusion
coatings ............................. 57, 58(T)
diffusion aluminide coatings of ...57, 58(F), 59
alloy and 2% ammonium chloride ...65(T)
diffusion coating for ................... 5758, 59
formation of diffusion layer below
640C .......................55, 56(T), 57(F)
metallographic structure ................ 57, 59(F)
phase and chemical compositions on diffusion

testing ...............................8081(T)
phase composition changes at high-temperature
tests ..................................8081(T)
phase transformation diagrams after EA
deposition method ................ 57, 58(F)
Ni10CrMoWCo, phase and chemical
compositions on diffusion coatings after
high-temperature testing ........ 81(T), 82(F)
Ni10CrWMoCo
chemical composition ..........11(T), 57, 58(T)
alloy and 2% ammonium chloride ...65(T)
diffusion coating for ............... 5758, 59(F)
metallographic structure ................ 57, 59(F)
VJL12U
cobalt and chromium content effect on
thickness of interaction zone with overlay
coating .............................. 93, 95(F)
fatigue strength with overlay
coating ........................... 140, 142(F)
blades ...................................... 142
effects ............................... 93, 95(T)
interaction zone with overlay coatings after
high-temperature testing ......... 88, 91(F)
phase transformations, after high-temperature
testing .................................... 84(F)
temperatures .......... 9192, 93(F), 94(F)
systems ...................... 147(T), 148(F)
Surface area (S) .......................................97
Synthetic ash of gas turbine fuel (GZT) 99100,
109116(F,T), 134137(F,T)
composition ......................................99(T)
gas medium composition at testing ...........99(T)
Systematization ......................................... 3
T
Takeoff power, effect on temperature field ........ 5
Tantalum, cathode spot temperature and vapor
pressure ........................................40(T)
TBC-APS. See Air plasma spraying of powders.
TBC-EB. See Vapor condensation at EB
evaporation of ceramic pellets.
TBCs. See Thermal barrier coatings.
TC-3000H unit (Sinku-Riko Company) ....... 176
Tearing stresses

Index / 245
magnitude of ....................................... 184

test rigs ....................................... 189
Tear-off stresses .......................................21
TEC. See Thermal expansion coefficient.
Temperature
characterized by zero thermal stress level ..... 196
effect on coating and superalloy fatigue
strength .................................141142
gas flow at turbine inlet ............................. 6
related to rates of diffusion processes ............79
Temperature gradient, of nozzle vanes ............ 7
Temperature of evaporation .......................39
Temperature rise, rate of .......................... 164
Tensile stiffness of coating ........................ 197
Tensile stresses ................................7, 8, 198
relaxation in thermal barrier coatings .......... 206
TGO. See Thermally grown oxide.
Thermal barrier coatings
(TBC) ............................2, 3, 203209(F)
actual stresses in ceramic layer on blade
leading edge ........................ 207, 208(F)
adhesive strength .................................. 182
adhesive strength curve ................ 206(F), 207
blade elongation vs. service time ............. 22(F)
bond coat surface preparation ...166167, 168(T)
centrifugal load for blades ....................... 195
ceramic layer adjacent to bond coat, element
contents analysis .................. 170, 171(T)
ceramic layer condensation rate ....... 171, 172(F)
ceramic-layer cracking criterion ...... 203205(F)
ceramic layer damage causes ....180184, 206(F)
ceramic layer durability ............. 180190(F,T)
ceramic layer microstructure .......... 172, 173(F)
ceramic layer properties .......................... 162
ceramic-layer spallation criterion ..... 205209(F)
ceramic layer structure, factors governing .... 166
ceramic layer testing for adhesive
strength .......................187(F), 188189
ceramic layer thermal stresses ......... 172, 173(T)
ceramic layer, thickness variation ................21
correction factors for irradiation
losses ................................ 177, 178(T)
cracking probability ............................... 205
deposition technique features ....... 165175(F,T)
deposition technique stages ...................... 166
design, typical ................................. 161(F)
electron beam deposited ............. 161190(F,T)
emissivity ................................. 180, 181(T)
failure causes ...................................... 203
fatigue strength on superalloys .................. 141
influence on blade temperature field ........... 195
inner aluminide heat-resistant bond coat ...... 161
lab tests of service life .............. 184188(F,T)
life estimation ............................ 203209(F)
life expectancy .................................. 21, 22
maximum relative tensile thermal stress ... 204(F)
pellet evaporation and ceramic layer
formation ........................ 168175(F,T)
phase transformations ............................ 182
quality control ........................... 174175(F)

recrystallization processes in ceramic layers .. 183
requirements .......................................... 8
residual stresses in ceramic layer on blade leading
edge ................................. 207, 208(F)
ridges intruding into ceramic layer ......... 205(F)
service life ............................. 167, 172, 182,
183(F), 184188(F,T)
service life, characteristics governing .......... 166
sintering of ceramics .............................. 183
spalling of ceramic layer .......170, 171(F,T), 175
spalling of ceramics on turbine blades .. 2022(F)
steady-state temperature field at blade leading
edge ....................................... 207(F)
stresses facilitating ceramic-layer separation from
blade surface ............................. 206(F)
testing on gas-dynamic test rigs and
engines ........................... 189190(F,T)
thermal conductivity ........177, 178179, 180(F)
thermal diffusivity .....................177,178, 179
thermal expansion coefficients ....... 180, 181(T),
182(F)
thermophysical properties ...........175180(F,T),
181(F,T)
ultimate adhesion strength and stress of spallation
vs. time ............................. 207, 208(F)
Thermal barrier effect ...............................21
Thermal conductivity, of ceramic layer of thermal
barrier coatings ....................... 175, 176(F)
Thermal diffusivity, determination by pulse
method .....................................175176
Thermal energy .......................................38
Thermal expansion coefficients (TECs) ..... 19, 20
causing thermal stresses ............................20
of coatings, variations and their causes .... 133(F)
for Co-Ni-Cr-Al alloys ................. 124, 125(T)
difference between coating and blade
materials ...................................... 198
elasticity modulus of coating alloys ..120124(T)
mismatch between coatings and substrates for
compressive stresses ..................130131
of nickel-chromium-aluminum alloys ...... 121(T)
of nickel-cobalt-chromium-aluminum
alloys ............................. 122124(F,T)
range 20400C, regression equation for
calculation of ................................ 214
calculation of ................................ 214
calculation of ................................ 214
calculation of ................................ 214
calculation of ................................ 214
calculation of ................................ 215
calculation of ................................ 215
calculation of .......................... 213, 216

246 / Index

calculation of .......................... 213, 216
calculation of .......................... 213, 216
calculation of .......................... 213, 216
calculation of .......................... 213, 216
calculation of .......................... 213, 216
calculation of .......................... 213, 216
calculation of .......................... 213, 216
Thermal fatigue ........................................ 7
coating effects on superalloys ...... 142150(F,T)
of overlay coatings ................................... 9
of turbine blades ...................1920(F), 21(F)
Thermal fatigue cracking ...... 1, 9, 10, 1920(F),
21(F), 144
diffusion coatings .......................... 20, 21(F)
of hollow samples of coated superalloys .. 147(T)
overlay coatings ............................ 20, 21(F)
on vane surfaces ...................................... 7
Thermal fluxes .........................................21
Thermally grown oxide (TGO) .................. 180
causing spalling between ceramic layer and bond
coat ............................ 185(T), 188, 190
layer formation spending aluminum from bond
coat .................................. 187(T), 188
thickness of, versus test time .......... 183(F), 186
Thermal shock, as thermal stress
manifestation .................................... 148
Thermal shock testing ......................... 148(T)
Thermal strains, calculation of ................... 196
Thermal stress cracking, in
coatings ...........................1920(F), 21(F)
Thermal stresses ......................................21
calculation of ...................................... 128
in ceramic layer of thermal barrier coating,
cracked .................................... 204(F)
in diffusion aluminide coatings ....... 129130(F)
factors for generation of .......................... 128
initiating cracks or spalling ...................... 195
methods for studying those of coatings ........ 128
in nozzle vanes ....................................... 7
in overlay coatings ............................ 130(F)
test rigs ....................................... 189
and thermal fatigue cracking .................... 144
Thermodynamic potential ..........................27
Thermofatigue testing ................ 142150(F,T)
equipment schematic diagram ............... 145(F)
results ..................................... 146, 147(T)
specimen sketches ....................... 145146(F)
test conditions (temperature/time) .......... 146(T)
Thermomechanical cyclic loading ............... 151
Thermomechanical fatigue
coating effect mechanism .............. 155158(F)
criterion ................................ 199203(F,T)
of superalloys, coatings effects

on ................................. 150158(F,T)
Thermomechanical fatigue testing 150158(F,T)
experimental method ............................. 152
specimen loading conditions ................ 154(T)
test conditions ........................ 153155(F,T)
Thermomechanical life, prediction of ........... 200
Thickness of zone of coating (hi) ..................18
Thiophenes .............................................. 5
Thrust .................................................... 6
Time of evaporation ..................................39
Time of high-temperature test (s) .................90
Time of oxide lifetime (sp) ..........................98
Time to ceramic-layer spalling ............... 206(F)
Time-to-failure to stress ........................... 135
Tin
cathode spot temperature and vapor
pressure .....................................40(T)
Titanium, as impurity of jet fuels .................... 5
Transfer coefficient (l), of electric arc
technique ..........................................36
Tungsten
cathode spot temperature and vapor pressure
................................................40(T)
evaporation .................................40(T)
solubility in gamma double-prime phase
compound after high-temperature
testing ...........................................82
Turbine-blade coating protective
properties ............................ 97117(F,T)
Turbine blade coatings
aircraft-engine, causes ..................... 1022(F)
defects, causes .......................................10
principles for choosing for aircraft
engines ...................................... 810
requirements for ...................................... 8
selection parameters for aircraft engines ..... 9(F)
Turbine blade operation, parameters determining
operating conditions .............................. 5
Two-way diffusion ....................................80
U
UE-type coater
diagram of central chamber ............... 41, 42(F)
specifications .....................................41(T)
Ultimate strain criterion ................ 198199(F)
Ultimate strength
of coating material ................................ 201
alloys ...................................... 127(T)
of nickel-chromium-aluminum alloys .... 126(F,T)
alloys ............................ 127(T), 129(F)
7001100C, regression equation for
calculation of ................................ 215

Index / 247
Uniform corrosion .......................12, 13, 14(F)

Uniform hot corrosion, of engine blades running
at high temperature ..............................14
V
Vacancy sinkers .......................................79
Vacuum arc remelting, for melting alloys for
Vacuum-induction (VI) melting, to melt alloys for
Vanadium
content effect on hot corrosion ............... 11, 12
content in corrosion product .......................14
surfaces .........................................99
Vanadium oxide ........................................ 5
causing hot corrosion ...............................99
Vapor condensation at EB evaporation of
ceramic pellets (TBC-EB) ................... 162
VAR. See Vacuum arc remelting.
Variable elasticity parameters method ......... 197
Variable stresses, in blades ........................... 8
VI. See Vacuum-induction melting.
Vibration loads ......................................... 8
Vibrogrinding ....................................... 166
Vibropolishing ....................................... 166
to prepare blade surfaces for overlay
coatings ................................ 45(F), 46
Visual inspection, to detect diffusion coating
discontinuity ......................................38
W
Warping ................................................. 7
Warplane engines, temperature and stress fields
nonuniform ......................................... 5
Weight after corrosion product removal (mk) ..97
Weight with all its oxidation products (mox) ....97
X
X-ray diffraction analysis
of ceramic pellets and temperature
changes ............................. 169170(F)
of crystallite texture of bond coats ............. 166
of thermal stresses ....................130, 131, 132
Y
Yield strength
alloys ...................................... 127(T)
alloys ............................ 127(T), 129(F)
700900C, regression equation for calculation
of .............................................. 215
Ytterbium, dopants for zirconium oxide crystal
lattice .......................................179180
Yttrium
alloying addition effect on nickel-cobaltchromium-aluminum alloys .........101103
content effect on formation of beta-phase-free
zone ........................................... 105
alloys .......................................... 106
Co-Ni-Cr-Al alloys ......................... 113
Ni-Co-Cr-Al alloys ......................... 112
content in diffusion aluminide
coatings ................................ 62(F), 63
content requirements for bond coats of thermal
barrier coatings .............................. 161
oxidation in melt of overlay coatings ............44
replacement with hafnium effect on oxides of
Ni-Cr-Al alloys .................... 101, 102(T)
Yttrium oxides, formation with hot corrosion .. 113
Z
Zinc
evaporation .................................40(T)
Zirconia, heat capacity increase with temperature
rise ...................................... 177, 178(F)
Zirconia yttria oxide, temperature restrictions
above 1150 C due to high oxygen-diffusion
mobility ............................................22
Zirconium, alloying NiAl with .................... 101
Zirconium oxide dissociation, during ceramic
pellet evaporation, products and degree
of ............................................. 168(T)
Zirconium oxide-yttrium oxide ........ 162163(F)
application techniques ............................ 162
ceramic pellets, chemical and phase compositions
made by mechanical mixture of
powders ............................. 163, 164(T)
ceramic pellets, chemical and phase
compositions, made by simultaneous
deposition .......................... 164, 165(T)
ceramic pellets, phase composition, volume
% .......................................... 165(T)
ceramic pellets, quality-control system ........ 165
color change of ceramic layers and phase
composition ........................ 169, 170(T)
phase composition changes, ceramic
pellets ..................................... 164(T)
solid-solution synthesis techniques ............. 163

Publication title
Product code
06738G

Fax 1-440-338-4634
Mail


www.ameritech.co.uk
Neutrino Inc.
Telephone: 81 (0) 424 84 5550
ASM International.

Protective Coatings For Turbine Blades

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Protective Coatings For Turbine Blades

Uploaded by

Copyright:

Available Formats

2002 ASM International. All Rights Reserved.