Journal of Surface Engineered Materials and Advanced Technology, 2013, 3, 116-132
http://dx.doi.org/10.4236/jsemat.2013.31A016 Published Online February 2013 (http://www.scirp.org/journal/jsemat)
High Temperature Sintering and Oxidation Behavior in
Plasma Sprayed TBCs [Single Splat Studies] Paper
2—Relevance of Variation in Materials Systems of TBC
Components
Swarnima Deshpande
Center for Thermal Spray Research, Department of Materials Science and Engineering, State University of New York at Stony Brook,
New York, USA.
Email: swarnima_d@yahoo.com
Received December 1st, 2012; revised January 3rd, 2013; accepted January 11th, 2013
ABSTRACT
The TBC system’s response to thermal exposure at high temperature is discussed here. The relevance of the micro-
structural aspects of each component of the TBC system is emphasized. The top coat is a YSZ ceramic coating consist-
ing of a collection of splats on top of one another. The most important aspect of this layer is the inherent inter-splat and
intra-splat porosity which undergoes sintering during thermal exposure. This study investigates the effect of thermal
exposure on the microstructure and sintering behavior in single splats produced using different starting powders since
this has been shown to influence the basic microstructure of YSZ topcoat. The bond coat is an MCrAlY metallic coating
which serves as an Al reservoir and allows the formation of a protective alumina, Thermally Grown Oxide (TGO) layer
between the bond coat (BC) and the top coat (TC) layers. This oxide scale formed upon thermal exposure prevents fur-
ther oxidation of the underlying component (substrate) and thus provides protection. As such, the content of free Al in
the bond coat layer is of significance and makes it crucial to understand the influence of bond coat microstructure evo-
lution and oxidation involved during its formation. The interaction between the bond coat, the TGO and the top coat
layers is examined in this study to understand the high temperature behavior of the TBC system with regards to varia-
tions in the top coat and bond coat material systems used.
Keywords: TBC; HVOF Bond Coat; Top Coat; Thermal Exposure; Vacuum Environment; Oxidation;
TGO Imperfections; Sintering; Microcracks
1. Introduction
This study seeks to look into the influence of different
material systems used for TBC system components. Dif-
ferent YSZ powders can be used to spray the topcoat and
a variety of spray methods can be used to generate the
bond coats. Previous study involving thermal exposure
using single splats has indicated that the microcracks in
these splats provide a path for the oxygen to reach the
aluminum depleted areas in the underlying bond coat [1st
paper, JSEMAT 2013]. Hence it was thought necessary
to examine the TGO formation and interactions with YSZ
during thermal exposure as a function of the initial YSZ
splat layer and the as-sprayed bond coat microstructures.
High temperature and thermal cycling behavior of
TBC systems has been a subject of great interest. When a
TBC system is thermally exposed, it is transparent to
oxygen because of the abundance of oxygen ion vacan-
cies in ZrO2 and hence oxygen ingresses through the
YSZ top coat. The bond coat undergoes oxidation and
leads to the formation of Thermally Grown Oxide (TGO).
This bond coat is typically an MCrAlY alloy, designed as
a local Al reservoir, enabling α-alumina to form in pref-
erence to other oxides. The TBC Alumina is preferably
formed because of its low oxygen diffusivity and supe-
rior adherence [1]. The thickness of TGO increases with
thermal exposure and induces the strain energy for the
crack propagation during spallation [2].
After exposure the TGO has a convoluted morphology
with major imperfections, the most prominent being the
undulations of the original TBC/bond-coat interface. This
is outlined by the dark gray TGO layer. The thicker areas
of TGO are predominantly α-alumina containing veins of
yttrium aluminates. Other oxides occur in isolated do-
mains within the TBC next to the TGO and have a lighter
gray contrast. These are typically spinels comprising oxi-
des of Cr/Ni/Co often with associated internal porosity
Copyright © 2013 SciRes. JSEMAT
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
117
and these spinels when formed act as preferential sites
for failure [3].
Shillington and Clarke studied the Changes in TGO
during Oxidation [3]. Pieces of the same bond-coat al-
loy were oxidized for different times (64, 128, 256, 384,
512 and 640 h) at 1121˚C. It was seen that after 128 h,
the only phase detectable was α-alumina. After 256 h, the
first indications of the formation of spinel and α-chromia
were detected. Alumina was no longer continuous after
384 h and the oxide was principally a mixture of spinel
and α-chromia. Alumina was no longer detectable after
512 h.
Other studies have indicated that differences in the
YSZ powder used generate differences in splat mor-
phologies and porosity content of the TBC topcoat [4-6].
Also, bond coats (NiAl) sprayed by different processes
have been extensively examined to understand the oxida-
tion involved during spraying and its influence on micro-
structural evolution. It has been shown that Air plasma
spraying and Wire arc spraying involve different mecha-
nisms of oxidation during the entire spray process and
generate coatings with a much higher oxide content com-
pared to HVOF spray process [7]. This study aims to
examine the effect of these variations in the component
material systems on the behavior of the TBC system
upon subsequent thermal exposure.
2. Experimental Approach
Single splats of YSZ were collected on NiCrAlY/
CoNiCrAlY bondcoat surfaces in order to conduct ther-
mal exposure studies on the same. For the first set of coat-
ings, NiCrAlY coatings were air plasma sprayed onto In-
conel 718 superalloy substrates. Processing conditions for
the same are tabulated below in Table 1.
Zirconia powders with different powder morphologies
were used to spray splats onto the polished surfaces of
these MCrAlY bond coats using APS process with spray
parameters as indicated in Table 2. The PSZ powders
Table 1. Deposition parameters for APS sprayed NiCrAlY
Bond coat.
Gun Sulzer F4MB
Gun voltage 68 V
Gun current 500 A
Primary gas (Ar) 50 SLPM
Secondary gas (H2) 10 SLPM
Carrier gas (Ar) 3 SLPM
Spray distance 120 mm
Powder Feed rate 40 g/min
Table 2. Deposition parameters for APS sprayed YSZ splats.
Gun PT-F4MB
Gun voltage 65 V
Gun current 650 A
Primary gas (Ar) 40 SLPM
Secondary gas (H2) 8 SLPM
Carrier gas (Ar) 3 SLPM
Spray distance 100 mm
Powder Feed rate 10 g/min
Gun traverse speed 10 mm/sec
Substrate rotational speed 160 rpm
used were:
1. Fused and Crushed Zirconia & 2. HOSP Zirconia
(Plasma-Densified Hollow Spheres).
For the second set of coatings, splats were collected on
polished surfaces of HVOF and VPS sprayed CoNiCrAlY
bond coats obtained from Engelhard Surface Technolo-
gies to analyze the effect of the bond coat system.
Thermal exposure behavior of these splats in air was
studied using a Thermolyne 47,900 box furnace. The
splats were isothermally exposed in Air at 1100˚C for 8
hr and 24 hrs to see the effect of duration. Also, thermal
exposure was studied in vacuum to eliminate effect of
bond coat oxidation. For this purpose, the samples were
sealed in quartz tubes prior to treatment, up to a vacuum
of 105 torr. This set of samples was subjected to an in-
termediate vacuum heat treatment at 1100˚C for 2 hr fol-
lowed by a subsequent isothermal exposure in Air at
1100˚C for 24 hr.
Particular splats were identified and the microcrack
network, surface roughness, and splat lifting/spalling
before and after HT were compared. SEM Back-scattered
imaging (Leo 1550, FEG) was employed on splat cross-
sections to observe the splat dimensions, microcrack sin-
tering and effect of TGO growth after HT. Energy Dis-
persive Spectrometry (EDS) gave elemental composition.
Atomic Force Microscopy (AFM) was employed to
quantify surface roughness of splats.
3. Results and Discussion
The thermal exposure studies were conducted using sin-
gle splats on bond coat. Single splats as opposed to free-
standing coatings include the splat/substrate interfacial
interaction. Absence of thick top coat minimizes the fac-
tors introduced by TBC coating and behavior can be more
related to the bond-coat chemistry and microstructure.
Only isothermal heat treatments were carried out in order
Copyright © 2013 SciRes. JSEMAT
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
Copyright © 2013 SciRes. JSEMAT
118
3.1.1. Splat Interfaces in Coating to eliminate cyclic TGO elongation effects.
The number of interfaces in a coating is directly related
to the thicknesses of individual splats which in turn are
affected by flattening and spreading. Composite SEM
cross-sectional images of coatings for both powder mor-
phologies were used to observe the splat interfaces as
shown in Figures 2(a) and (b). Several visible splat
boundaries in these cross-sections are marked and splat
thicknesses are noted for each of these splats. Then all
these splat thickness values are used to compute the av-
erage splat thickness for each case.
3.1. Effect of Variation in YSZ Top Coat
Microstructure
Changing the feedstock translates into changes in the as-
sprayed splat dimensions which in turn have a distinct
effect on the top coat structure. Different morphology
YSZ powders, Fused and Crushed (FC) and Plasma Den-
sified Hollow Spheres (HOSP) when sprayed through the
plasma exhibit different flow behavior and different mel-
ting efficiencies and form splats with different dimen-
sions.
The average splat thickness for FC YSZ coating aver-
aged over 13 values was 4.14 μm and that calculated for
HOSP YSZ coating using 18 different values was 2.46
μm. Thus it is once again seen that fused and crushed
YSZ powders produced coatings with splats 1.7 times
thicker than in the case of HOSP YSZ. Hence it was in-
ferred that within a given thickness of say 250 μm, the
average number of interfaces created by HOSP coating
Cross-section images of splats in Figure 1, show the
diameters and thicknesses for FC and HOSP splats. Mi-
crocrack widths in each case are also measured. It can be
noted that the FC zirconia splats are approximately 1.75
times thicker compared to HOSP splats. This is further
addressed in the following section.
Fused and Crushe
d
HOS
P
2.61
m
177.9
m 110.0
m
0.49
m
(a)
Fused and Crushe
d
HOS
P
0.48 m
0.52
m 0.45 m
0.48
m
0.45
m
0.52 m
0.41 m
(b)
Figure 1. (a) Cross-sections showing splat dimensions; (b) Cross-sections showing microcrack widths.
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
119
(b1)
(b2)
(b3)
(a) (b)
Figure 2. (a) Composite image of FC YSZ coating at 5000×; (b) Composite image of HOSP YSZ coating (b1) and (b2) at
5000×, (b3) at 10,000×.
will be 103 - 104 whereas the number of interfaces in
case of a FC coating will be 61 - 62. Thus larger splat
thickness translates into lower number of interfaces per
given thickness of coating. This difference in the splat
structure of the two coatings is not insignificant when it
comes to coating properties such as thermal conductivity
and elastic modulus. The implication of these results has
also been discussed in more detail in another study [8].
Hence, in order to examine the influence of top coat
microstructure on the TBC system’s behavior under ther-
mal exposure, different feedstock of YSZ were used to
collect the splats and then these were subjected to similar
heat treatments.
The two TBC system samples consisted of superalloy
substrates, MCrAlY bond coats (BC) and top coats (TC)
made of single splats collected using FC or HOSP YSZ
powders. These were subjected to the various heat treat-
ments as mentioned in Section 2 and are categorized as
given below in Table 3. Henceforth the samples will be
referred to by the YSZ powder type and the Set number
indicating the type of thermal exposure.
Specific splats were identified for each sample (pow-
der type) and set (thermal exposure type) and the top
surface microstructures as well as cross-sections of these
were examined before and after thermal exposure. SEM,
AFM and EDS analysis were employed for observation
Copyright © 2013 SciRes. JSEMAT
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
120
Table 3. Different types of thermal exposure treatments on
FC and HOSP splats.
Samples
1. (Substrate/BC/FC TC)
2. (Substrate/BC/HOSP TC) Thermal exposure
Set 1 As-sprayed splats
Set 2 Isothermally exposed in Air for 8 hr
at 1100˚C
Set 3 Isothermally exposed in Air for 24 hr
at 1100˚C
Set 4
Isothermally exposed in Vacuum for
2 hr at 1100˚C and then Isothermally
exposed in Air for 24 hr at 1100˚C
and comparison of several effects of thermal exposure.
3.1.2. Splat Dimensions
The top surface microstructures of the single splats of FC
and HOSP YSZ taken at the same magnification before
and after different exposure times in air show that the
splat dimensions have not changed as an effect of ther-
mal exposure. The cross-section images of splats are
compared to show that the splat thicknesses also remain
the same after thermal exposure; 2.6 μm for FC and 1.49
μm for HOSP splats after 24hr HT and 8hr HT respect-
tively (Figure 3).
3.1.3. Microcrack Sintering
The microcrack network is observed in the top surface
SEM micrographs of individual splats. It is seen that this
network is not altered after thermal exposure at 1100˚C.
Most of the fine microcracks are retained when the splats
are thermally exposed in air. This is also confirmed by
measuring microcrack widths in the cross-section images
of splats before and after different thermal exposure times
(Figure 4).
However in case of samples of Set 4 that have under-
gone the intermediate vacuum thermal exposure, the mi-
crocrack network is modified. Most fine microcracks in
this case have started to sinter and are not as noticeable
in the top surface microstructures of splats. The cross-
section images also show that microcracks widths have
reduced significantly after the subsequent 24 hr Air HT
(Figure 5). This indicates that the intermediate vacuum
HT is responsible for instigating the sintering of these
microcracks. This may be related to the increase in lattice
spacing that was observed by Thornton et al. [9] when a
TBC coating was heat treated in vacuum. In vacuum,
there is no oxygen to replace that lost from zirconia in
the formation of bond coat oxide and less oxygen appears
to cause larger lattice spacing [9]. Increased lattice spac-
ing in turn could be initiating the sintering of micro-
cracks.
3.1.4. TGO Growth
Bond coat oxidation at high temperature, leads to the for-
mation of TGO and thereby creates another layer at the
interface between BC and TC. The TGO layer thickness
gradually increases with HT duration.
Figure 6 shows the development and growth of the
TGO layer between the bond coat and the top coat upon
thermal exposure in Air at 1100˚C. It is seen that for both
the powders, the samples show increased TGO thickness
with increasing duration of thermal exposure. For FC
YSZ samples, the TGO thickness increases from ap-
proximately 1.5 μm after 8 hr HT to 2.7 μm after 24 hr
HT. For the HOSP YSZ samples, the TGO thickness is
seen to increase from approximately 1.3 μm after 8 hr
HT to 2.0 μm after 24 hr HT. This TGO layer is mainly
constituted by Alumina since Al in the bond coat is pref-
erably oxidized but at some places Chromia is also for-
med as disclosed by EDS. Previous studies also show
that other oxides occur in isolated domains within the
TBC next to the TGO and have a lighter gray contrast.
These are typically spinels comprising oxides of Cr/Ni/Co
often with associated internal porosity [10].
3.1.5. Splat Surface Roughening
Figure 7 shows the splats and large magnification im-
ages (taken at 50,000×) of their top surfaces before and
after different thermal exposures. The grain structure on
the splat surfaces is thereby visible. In case of both FC
and HOSP splats it is observed that at the areas marked
by purple arrows, the columnar grains seem to have risen
upwards. A comparison of the splat surfaces between as-
sprayed and thermally exposed splats indicates that the
splat surface has roughened overall. This splat surface
roughness was quantified using AFM in case of HOSP
powder samples for as-sprayed splats and those exposed
for 8hr in Air at 1100˚C. The results of this measurement
are shown in Figure 8 below. The mean surface rough-
ness of the as-sprayed splat was approximately Ra = 11.9
nm and that of the thermally exposed splat was approxi-
mately Ra = 153 nm.
In order to examine the cause behind the increased
surface roughness, cross-sections of the splat were ob-
served and the TGO layer generated after 24 hr HT in Air
at 1100˚C was characterized. In most areas the TGO
consists primarily of Alumina, as in Figure 9(a), but in
some areas oxides like Chromia also form and exhibit a
lighter grey contrast as discussed in Section 3.1.4 above.
Studies have shown that such other oxides like spinels
when formed also act as preferential sites for failure [10].
Reason might be that the interfacial fracture resistances
of the TBC/α-chromia and the TBC/spinel interfaces are
lower than that of the TBC/α-alumina interface originally
present [11]. Figure 9(b) indicates the splat being lifted
Copyright © 2013 SciRes. JSEMAT
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
Copyright © 2013 SciRes. JSEMAT
121
FC YSZ splat HOSP YSZ splat
2.61 m2.61 m2.61 m
Set 3 (splat thickness)
Set 1 (splat thickness)
2.6 m2.6 m2.6 m
Set 3
Set 1
Set 2
Set 1 Set 2 Set 1
Set 3
Set 1
1.49 m1.49 m1.49 m1.49 m
Set 2 (splat thickness)
Set 1 (splat thickness)
Figure 3. Splat dimensions before and after HT in Air@1100˚C (for FC and HOSP splats).
FC YSZ spla
t
HOSP YSZ splat
Set 2 (microcrack network)
Set 1 (microcrack network)
0.52 m
0.48 m
0.45 m0.52 m
0.48 m
0.45 m
Set 2 (microcrack width)
Set 1 (microcrack width)
0. 48 m
0. 48 m0. 52 m
0. 48 m
0. 48 m0. 52 m
Set 2 (microcrack network)
Set 1 (microcrack network)
Set 3(microcrack network)
Set 1 (microcrack network) Set 3(microcrack network)
Set 1 (microcrack network)
Set 2 (microcrack width)
Set 1 (microcrack width)
0.48 m0.52 m
0.45 m0.41 m
0.48 m0.52 m
0.45 m0.41 m
0.30 m
0.39 m
0.30 m
0.39 m
Figure 4. Microcracks in splats before and after HT in Air@1100˚C (for FC and HOSP splats).
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
122
Set 4(microcrack network)
Set 1 (microcrack network) Set 4(microcrack network)
Set 1 (microcrack network)
FC YSZ splat HOSP YSZ splat
0. 09 m0. 09 m0. 09 m
Set 4 (microcrack width)Set 4 (microcrack width)
0.17 m0.17 m
Figure 5. Microcracks in splats before and after HT in vacuum + HT in Air@1100˚C (for FC and HOSP splats).
FC YSZ splat HOSP YSZ splat
Set 1 (interface between BC and TC)
1.3 m1.1 m
Chromium oxide
Aluminum oxide
1.3 m1.1 m
Chromium oxide
Aluminum oxide
2.7 m
Aluminum oxide
2.7 m
Aluminum oxide
Set 2 (TGO layer between BC and TC)
Set 3 (TGO layer between BC and TC)
Set 1 (interface between BC and TC)
Set 2 (TGO layer between BC and TC)
Set 3 (TGO layer between BC and TC)
Chromium oxide
1.27 m
0.74 m
Aluminum oxide
Chromium oxide
1.27 m
0.74 m
Aluminum oxide
Aluminum oxide
1.99 m
Aluminum oxide
1.99 m
Figure 6. TGO layer thickening with increasing duration of HT in Air@1100˚C (for FC and HOSP splats).
Copyright © 2013 SciRes. JSEMAT
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
Copyright © 2013 SciRes. JSEMAT
123
FC YSZ splat HOSP YSZ splat
Set 3
Set 1
Set 2
Set 1
Set 3
Set 1
Set 2
Set 1
Figure 7. Roughening of splat surface due to TGO layer thickening w i th ther mal exposure (for FC and HOSP splats).
HOSP YSZ splat
Section AnalysisRoughness Analys is
Mean Roughness (Ra) 11.852 nm
Section AnalysisRoughness Analys is
Mean Roughness (Ra) 11.852 nm
Secti on A nalysisRoughnes s Analys is
Mean Roughness (Ra) 153.08 nm
Secti on A nalysisRoughnes s Analys is
Mean Roughness (Ra) 153.08 nm
Set 2
Set 1
Figure 8. Roughening of splat surface due to TGO layer thickening with thermal exposure.
upwards at such a location.
Thickness imperfections in TGO enlarge in regions
where O2-diffusivity through TGO is exceptionally large
i.e. at locations where TGO contains oxides other than
alumina [1]. These TGO undulations must then push the
grains in the splats upward and cause splat lifting or may
be spalling. This is visible in Figures 9(c) and (d) and
explains the increase in surface roughness of splats as
measured in Figure 8.
3.1.6. NiO O utgrowth
A
nother effect of thermal exposure is the appearance of
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
124
(b) Non-Uniform TGO layer (Alumina + Chromia) splat being lifted upwards
(a) Uniform TGO layer (Alumina)
FC YSZ splat
(c) TGO layer (Alumina + Chromia) Uniform TGO layer (Alumina)
Cr
O
Cr
O
Cr
OAl
Cr
OAl
EDS analysis of Alumina and Chromia regions
of TGO
(for FC and HOSP splats)
(d) Thickness imperfection in TGO layer at the location of chromia formation
HOSP YSZ splat
Figure 9. TGO layer between bond coat and YSZ top coat generated after 24 hr HT in Air at 1100˚C Effect of Chromia
ormation at TGO/YSZ splat interface. f
Copyright © 2013 SciRes. JSEMAT
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
125
some outward oxide growth from the bond coat, visible
at larger microcracks on the top coat surface, as marked
by green arrows in Figure 10. This outgrowth was as-
sumed to originate from the bond coat oxide below the
splat. It was also observed when cross-section images
were examined. EDS analysis was conducted and these
outgrowing oxide grains gave an EDS pattern showing
Ni and O peaks, also shown in Figure 10.
APS NiCrAlY bond coat microstructure (Figure 11)
contains internal oxide chunks of alumina. These regions
may be locally depleted of free aluminum. Microcracks
in splats are seen to coincide with oxide points in the
bond-coat, Figure 11.
Several studies have shown that oxygen diffusion
through TGO along grain boundaries causes more TGO
growth at these boundaries [1,12,13]. Similarly at these
microcrack positions, there is a path for oxygen to reach
a BC area that is locally depleted of aluminum [7]. As
such, other oxides could form at these microcrack posi-
tions.
However a similar investigation of the Set 4 samples
that underwent the intermediate vacuum HT (Figure 12)
revealed no such oxide outgrowth. The surfaces of splats
indicated almost sintered microcracks and no outgrowth
in the top surface as well as the cross-section images.
The outward oxide growth is curtailed probably because
the intermediate vacuum treatment started the sintering
or sealing of most microcracks from their bottom end.
One possible explanation for this microcrack sintering
under vacuum has been discussed above in Section 3.1.3.
The oxide outgrowth in case of thermal exposure in
Air is also thought to be related to the bond coat mic-
rostructure and will be addressed in the following Sec-
tion 3.2.1.
3.2. Effect of Variation in MCrAlY Bond Coat
Microstructure
The bond coat microstructure is greatly influenced by the
spraying technique adopted for the coating formation.
Ni-5 wt% Al bond coatings obtained using four different
spraying techniques are compared in Figure 13. Wire arc
spraying, Air plasma spraying, HVOF spraying and Cold
spraying, each differ w.r.t their feedstock injection, mel-
ting methods, spraying parameters as well as oxidation
involved during spraying. As such they result in distinct
particle conditions during flight. Wire arc and APS gen-
erate lower particle velocities compared to HVOF and
Cold spray. HVOF sprayed droplets have lower tempera-
tures and Cold spray is a unique process in which entire
deposition takes place in the solid state [14].
As such these processes lead to distinct differences in
the microstructures of coatings produced. The porosity
and oxide content in the micrographs are estimated using
Image analysis, shown in Figure 14.
Both wire arc and APS coatings show significant oxi-
dation and porosity. HVOF coating depicts a much lower
“dark area” but a significant portion of this is oxide.
Cold-spray coatings indicate low porosity and almost no
oxide formation, probably due to high impact velocities
and low process temperatures.
Figure 15 summarizes the process of bond coat mi-
crostructure evolution by describing the key mechanisms
involved during spraying such as in-flight oxidation, sin-
gle splat formation, post-impact oxidation and splat-splat
linkage leading to coating buildup. Schematics for me-
chanisms and corresponding microstructural observations
are shown.
Oxide segregation during in-flight oxidation in case of
APS process causes splashing of oxide beneath the splat,
whereas in case of HVOF spraying, complete spreading
of splat before solidification generates very flat, disk sha-
ped splats having better contact with substrate. Less in-
flight oxidation and slow solidification in the HVOF
process generates an Al depletion region below the post
impact oxidation on splat surface. Post impact oxidation
in case of APS splats however, generates poorer wetting
and gives rise to lamellar porosity between the new splat
and oxide of the previous layer. Metal-metal contact in
both processes generates inter-splat coalescence.
Nano scale observations revealed oxide bands after
every splat in case of APS coating and more splat coales-
cence in HVOF coating. Phases in each coating were
analyzed showing fcc Ni as the primary phase in APS
and gamma-Ni (richer in Al) in case of HVOF. The
above observations are investigated and discussed in de-
tail elsewhere [7].
These observations make it very apparent that the
choice of bond coat will also play a role in the high tem-
perature behavior of the TBC system as the bond coat is
an integral component of such a system. Hence, in order
to examine the influence of bond coat microstructure on
the TBC system’s behavior under thermal exposure,
bond coats sprayed by different techniques were used.
Top coat splats were collected on these bond coat sur-
faces and the obtained TBC systems as a whole were
then subjected to similar heat treatments. Table 4 shows
a list of the samples used.
The top surfaces of splats and cross-sections of coat-
gs were compared before and after thermal exposure.
Once again different effects of thermal exposure were
observed upon examination.
3.2.1. Ni O Outgrowth—HT in Air
Figure 16 shows the as-sprayed splats and the splats af-
ter 8 hr HT in Air at 1100˚C on APS, HVOF and VPS
bond coat surfaces.
Copyright © 2013 SciRes. JSEMAT
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
126
FC YSZ spla
t
HOSP YSZ spla
Set 1 Set 2 (outward oxide growth)
Set 3(outward oxide growth) Set 1
Set 2 (outward oxide growth)
Set 1
Set 3(outward oxide growth)
Set 1
EDS analysis of Oxide outgrowth
Ni and O peaks
FC Set 2 (outward oxide growth) HOSP Set 2 (outward oxide growth)
Ni
Ni
ONi
Ni
O
Figure 10. NiO outgrowth through microcracks in splats after HT in Air @ 1100˚C (for FC and HOSPsplats).
O K
Al K
Ni L
O K
Al K
Ni L
Figure 11. Microcracks coinciding with oxide points.
Table 4. Different types of thermal exposure treatments on
FC and HOSP splats.
No. TBC system components Thermal exposure
1.
Single splats of YSZ on
APS, HVOF and VPS bond
coat
Isothermally exposed in Air
for 8 hr at 1100˚C
2. Thin YSZ coating on APS
bond coat
Isothermally exposed in Air
for 24 hr at 1100˚C
3. Thin YSZ coating on HVOF
bond coat
Isothermally exposed in Vacuum for
2 hr at 1100˚C & then Isothermally
exposed in Air for 24 hr at 1100˚C
In case of APS bond coat, outward oxide growth and
splat surface roughening is observed as discussed before
in Sections 3.1.6 and 3.1.5 respectively. But in case of
HVOF and VPS bond coats, no such oxide outgrowth is
seen and splat surface roughening is also minimal.
The HVOF and VPS bond coats have much lesser in-
herent oxidation as seen in Section 3.2. Lesser in-flight
oxidation was reported for HVOF compared to APS and
of course VPS process pretty much eliminates oxidation
during spraying. As such, the Al concentration in the BC
is uniform, there are no alumina chunks, and microcracks
cannot coincide with any Al depleted areas. This natu-
Copyright © 2013 SciRes. JSEMAT
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
127
Set 4 Sealed microcracks
No oxide outgrowth
FC YSZ splat HOSP YSZ splat
Set 1
Set 4 (no oxide outgrowth)
Set 1
Set 4 (no oxide outgrowth)
Figure 12. Microcracks in splats before and after HT in Vacuum + HT in Air@1100˚C (for FC and HOSP splats).
APS coating showing
thinner splats
Cold spray coating showing
very dense cross-section
Wire Arc coating showing thicker splats
HVOF coating showing very
few visible splat boundaries
Figure 13. Bond coat microstructures using different spray-
ing techniques.
rally precludes the abovementioned method of formation
of NiO.
3.2.2. Sintering Under High Temperature
Figure 17 shows TBC samples of thin YSZ coating on
APS BC and thin YSZ coating on HVOF bond coat sub-
jected to the 24 hr HT in Air at 1100 degrees.
Careful examination shows visible sintering of inter-
Wire-arc APS HVOF Cold
16.5
20.5
3.1 1.4
Wire-arc APS HVOF Cold
16.5
20.5
3.1 1.4
Wire-arc APS HVOF Cold
16.5
20.5
3.1 1.4
Figure 14. Porosity and oxide content in bond coat micro-
structures obtained using image analysis.
lamellar pores and intersplat boundaries in the APS BC
sample, as marked by yellow arrows. Most microcracks
are however still retained as indicated by red star sym-
bols.Whereas, upon inspection of the HVOF BC sample,
interlamellar pores and intersplat boundaries are seen to
be still present as shown by pink arrows and red stars
indicate that microcracks have started sintering.
Although this is not completely understood, one possi-
ble theory is that as observed before, during HT in Air,
NiO outgrowth occurs through microcracks of YSZ
splats on APS BC whereas; this is not the case with
HVOF bond coats. Since this basic hindrance to sintering
(blocking of microcracks) is now missing, the micro-
cracks can sinter.
If these same samples are subjected to intermediate
vacuum HT and then the 24 hr Air HT (Figure 18), then
however, most microcracks are seen to have sintered in
ase of both bond coat samples. This is in agreement c
Copyright © 2013 SciRes. JSEMAT
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
128
APS bond coat HVOF bond coat
Internal
oxide,
Al2O3
Internal
oxide,
Al2O3
Oxide
segregation
Al2O3
Oxide
segregation
Al2O3
SubstrateSubstrate SubstrateSubstrateSubstrateSubstrate
SubstrateSubstrateSubstrateSubstrate
SubstrateSubstrateSubstrateSubstrate SubstrateSubstrateSubstrateSubstrate
Oxide splashing
beneath splat
[1]
Oxide splashing
beneath splat
Oxide splashing
beneath splat
[1]
Complete spreading of splat before solidificationComplete spreading of splat before solidification
5 m5 m
[2]
Al depletion below oxide
[2]
Al depletion below oxide
Coalescence between two
metal splats with no oxide
in between
Coalescence between two
metal splats with no oxide
in between
Lamellar pore
between new
splat and oxide
of previous layer
[3]
Lamellar pore
between new
splat and oxide
of previous layer
[3]
2 m2 m2 m2 m
Substrate
Oxide growing on top of splat
SubstrateSubstrateSubstrateSubstrate
Oxide growing on top of splat
Oxide bands
after every splat
[4]
Oxide bands
after every splat
[4] More splat
coalescence
[5]
More splat
coalescence
[5]
[6]
Ni (200)
Ni (020)Ni (2 20)
fcc Ni
[6]
Ni (200)
Ni (020)Ni (2 20)
fcc Ni
Ni (200)
Ni (020)Ni (2 20)
Ni
(200)
Ni
(022)
Ni
(311)
Ni (200)
Ni (020)Ni (2 20)
Ni (200)
Ni (020)Ni (2 20)Ni
(111)
fcc NiNi
Ni
(200)
Ni
(022)
Ni
(311)
Ni
(111)
Ni
(200)
Ni
(022)
Ni
(311)
Ni
(111)
Ni
(200)
Ni
(022)
Ni
(311)
Ni
(111)
Ni
(111)
Ni
(311) Ni
Figure 15. Bond coat microstructure evolution—influence of spray process.
APS bond coat HVOF bond coat VPS bond coat
Set 1
Set 2 (splat surface roughening and
oxide outgrowth)
Set 1
Set 2 (no splat surface roughening and
no oxide outgrowth)
Set 1
Set 2 (less splat surface roughening,
no oxide outgrowth)
Figure 16. Effects on top coat single splats after 8 hr HT in Air@1100˚C (for APS, HVOF and VPOS bond coats).
Copyright © 2013 SciRes. JSEMAT
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
129
APS bond coat HVOF bond coat
*
*
Set 3 (interlamellar boundaries and microcracks)
Set 3 (interlamellar boundaries and microcracks)
*
*
*
Figure 17. Sintering in top coat after 24 hr HT in Air@1100˚C (for APS and HVOF bond coats).
Set 4 (interlamellar boundaries and microcracks)
APS bond coat HVOF bond coat
Set 4 (interlamellar boundaries and microcracks)
Figure 18. Sintering in top coat after 2 hr HT in Vacuum + 24 hr HT in Air@1100˚C (for APS and HVOF bond coats).
with the previous observations made in case of single
splats.
3.2.3. TGO Growth
The TGO formation in these samples with different bond
coats was also considered and is illustrated in Figure 19.
For the same duration of HT i.e. 24 hr in Air at 1100 de-
grees, APS bond coat shows faster TGO growth than
HVOF bond coat. The TGO thickness is seen to be 2.7
μm for APS bond coat sample and 1.7 μm for the HVOF
bond coat sample.
3.2.4. TGO Imperfections
When the elemental composition along the TGO in both
samples is compared, then for the APS bond coat sample,
many locations along the TGO length show formation of
Copyright © 2013 SciRes. JSEMAT
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
130
other oxides like Chromia.However, for the HVOF bond
coat sample, although some locations show TGO thick-
ening, chromia formation is not visible. This is clearly
shown in Figure 20.
Once again this is related to inherent oxidation levels
in both bond coats. APS bond coats have undergone
more oxidation during their formation and hence have
lesser free Al available for TGO formation. Al depletion
2.7 m2.7 m
1.7 m1.7 m
APS bond coat HVOF bond coa
t
Figure 19. TGO growth after 24 hr HT in Air@1100˚C (for APS and HVOF bond coats).
Set 4 (TGO layer at BC/TC interface)
APS bond coat HVOF bond coat
Set 4 (TGO layer at BC/TC interface)
Locations of Chromia formation
T
GO thickening but no Chromia formation
Figure 20. Thickness imperfec tions in TGO—thicke ning and for m ation of othe r oxide s (for APS and HVOF bond coats).
Copyright © 2013 SciRes. JSEMAT
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
131
in BC causes other oxides to form, as also shown by
Tolpygo and Clarke [15,16]. HVOF bond coats have
more free Al retained in the coating and hence they have
a longer way to go before Al depletion occurs.
Fewer instances of chromia formation means fewer
occurrences of thickness imperfections in the TGO layer
and hence reduced splat surface roughening in case of
HVOF and VPS bond coats. This further explains the
observations in Figure 16, Section 3.2.1.
4. Conclusions
In this study, the following effects of thermal exposure
were considered and compared as an outcome of varia-
tions in top coat splat dimensions and bond coat micro-
structures.
Microcracks sintering in splats.
TGO growth at interface between top coat and bond
coat.
Splat surface roughening.
NiO outgrowth occurring through microcracks in
splats.
Effect of intermediate Vacuum HT during thermal
exposure.
It was seen that in case of top coat, different YSZ
feedstock give different initial splat dimensions in the
as-sprayed splats but the splats show the same high tem-
perature behavior when subjected to similar heat treat-
ments. Intermediate vacuum heat treatment alters the
microcrack sintering behavior observed for Air heat
treatment and also prevents the NiO outgrowth upon
thermal exposure. But these effects are also similar for
both YSZ powders. So, the top coat microstructure with
respect to single splats does not have an influence on
high temperature behavior of the system.
In case of bond coats, however, when different spray-
ing techniques are used, the inherent oxidation levels in
the bond coat microstructures are different and this does
influence the behavior of the system upon high tempera-
ture exposure.
The free Al available in the BC for TGO formation
dictates the occurrence of chromia formation and hence
determines the extent of TGO thickness imperfections
leading to splat surface roughening or spallation.
Absence of Alumina chunks in the HVOF and VPS
bond coat microstructures prevents the described method
of formation of NiO outgrowth through microcracks and
stimulates microcracks sintering which is not observed in
APS bond coats under similar thermal exposure.
5. Acknowledgements
We would like to thank Glenn Bancke, AnirudhaVaidya,
John Gutleber and Li Li (CTSR) for preparation of the
specimens and spraying diagnostics.
REFERENCES
[1] A. G. Evans, D. R. Mumm, J. W. Hutchinson, G. H.
Meier and F. S. Pettit, “Mechanisms Controlling the Du-
rability of Thermal Barrier Coatings,” Progress in Mate-
rials Science, Vol. 46, No. 5, 2001, pp. 505-553.
doi:10.1016/S0079-6425(00)00020-7
[2] A. Nusair Khan and J. Lu, “Behavior of Air Plasma
Sprayed Thermal Barrier Coatings, Subject to Intense
Thermal Cycling,” Surface and Coatings Technology,
Vol. 166, No. 1, 2003, pp. 37-43.
doi:10.1016/S0257-8972(02)00740-5
[3] A. Rabiei and A. G. Evans, “Failure Mechanisms Associ-
ated with the Thermally Grown Oxide in Plasma Sprayed
Thermal Barrier Coatings,” Acta Materialia, Vol. 48, No.
15, 2000, pp. 3963-3976.
doi:10.1016/S1359-6454(00)00171-3
[4] A. Kulkarni, S. Sampath, A. Goland and H. Herman,
“Porosity-Thermal Conductivity Relationships in Plasma
Sprayed Zirconia Coatings,” In: C. C. Berndt, Ed., Ther-
mal Spray: Surface Engineering via Applied Research
(ITSC 2000), ASM International, Materials Park, Mont-
real, 8-11 May 2000, pp. 1061-1066.
[5] A. Kulkarni, Z. Wang, T. Nakamura, S. Sampath, A. Go-
land, H. Herman, A. J. Allen, J. Ilavsky, G. Long, J.
Frahm and R. W. Steinbrech,“Comprehensive Micro-
structural Characterization and Predictive Property Mod-
eling of Plasma-Sprayed Zirconia Coatings,” Acta Mate-
rialia, Vol. 51, No. 9, 2003, pp. 2457-2475.
doi:10.1016/S1359-6454(03)00030-2
[6] S. Deshpande, A. Kulkarni, S. Sampath and H. Herman,
“Application of Image Analysis for characterization of
Porosity in Thermal Spray Coatings and Correlation with
Small Angle Neutron Scattering,” Surface and Coatings
Technology, Vol. 187, No. 1, 2004, pp. 6-16.
doi:10.1016/j.surfcoat.2004.01.032
[7] S. Deshpande, S. Sampath and H. Zhang, “Mechanisms
of Oxidation and Its Role in Microstructural Evolution of
Metallic Thermal Spray Coatings—Case study for NiAl,”
Surface and Coatings Technology, Vol. 200, No. 18-19,
2006, pp. 5395-5406. doi:10.1016/j.surfcoat.2005.07.072
[8] Z. Wang, A. Kulkarni, S. Deshpande, T. Nakamura and H.
Herman, “Computational Approaches to Estimate Proper-
ties of Thermally Sprayed Porous Coatings,” Acta Mate-
rialia, Vol. 51, No. 18, 2003, pp. 5319-5334.
doi:10.1016/S1359-6454(03)00390-2
[9] J. Thornton, D. Cookson and E. Pescott, “The Measure-
ment of Strains within the bulk of Aged and As-Sprayed
Thermal Barrier Coatings Using Synchrotron Radiation,”
Surface and Coatings Technology, Vol. 120-121, 1999,
pp. 96-102. doi:10.1016/S0257-8972(99)00340-0
[10] A. Rabiei and A. G. Evans, “Failure Mechanisms Associ-
ated with the Thermally Grown Oxide in Plasma Sprayed
Thermal Barrier Coatings,” Acta Materialia, Vol. 48, No.
15, 2000, pp. 3963-3976.
doi:10.1016/S1359-6454(00)00171-3
Copyright © 2013 SciRes. JSEMAT
High Temperature Sintering and Oxidation Behavior in Plasma Sprayed TBCs [Single Splat Studies] Paper 2—
Relevance of Variation in Materials Systems of TBC Components
132
[11] E. A. G. Shillington and D. R. Clarke, “Spalling Failure
of a Thermal Barrier Coating Associated with Aluminum
Depletion in the Bond-Coat,” Acta Materialia, Vol. 47,
No. 4, 1999, pp. 1297-1305.
doi:10.1016/S1359-6454(98)00407-8
[12] T. Xu, M. Y. He and A. G. Evans, “A Numerical As-
sessment of the Durability of Thermal Barrier Systems
That Fail by Ratcheting of the Thermally Grown Oxide,”
Acta Materialia, Vol. 51, No. 13, 2003, pp. 3807-3820.
doi:10.1016/S1359-6454(03)00194-0
[13] A. M. Karlsson, J. W. Hutchinson and A. G. Evans, “The
Displacement of the Thermally Grown Oxide in Thermal
Barrier Systems upon Thermal Cycling,” Materials Sci-
ence and Engineering, Vol. 35, No. 1-2, 2003, pp. 244-
257. doi:10.1016/S0921-5093(02)00843-2
[14] S. Deshpande, S. Sampath, P. I. Gouma and H. Herman,
“Microstructural Characterization across Length Scales of
Thermal Sprayed Ni5wt%Al Coatings,” In: B. R. Marple,
Ed., Thermal Spray: Advancing the Science and Applying
the Technology (ITSC 2003), ASM International, Material
Park, Orlando, 5-8 May 2003, pp. 1419-1428.
[15] R. Panat, S. Zhang and K. J. Hsia, “Bond Coat Surface
Rumpling in Thermal Barrier Coatings,” Acta Materialia,
Vol. 51, No. 1, 2003, pp. 239-249.
[16] V. K. Tolpygo and D. R. Clarke, “Surface Rumpling of a
(Ni,Pt)Al Bond Coatinduced by Cyclic Oxidation,” Acta
Materialia, Vol. 48, No. 13, 2000, pp. 3283-3293.
doi:10.1016/S1359-6454(00)00156-7
Copyright © 2013 SciRes. JSEMAT