Document Sample

      Hariyati Purwaningsih1, Sulistijono2, Lukman Noerochim2, Cartha Kharisma3
                    Materials and Metallurgical Engineering Department
                              Industrial Faculty of Technology
                   Institute of Technology of Sepuluh Nopember (ITS)
                     Campuss ITS Keputih Sukolilo Surabaya 60111
             Phone/Fax : +62 31 5997026; email:
                    Materials and Metallurgical Engineering Department
                              Industrial Faculty of Technology
                   Institute of Technology of Sepuluh Nopember (ITS)
                     Campuss ITS Keputih Sukolilo Surabaya 60111
                                Phone/Fax : +62 31 5997026
         Graduate Student of Materials and Metallurgical Engineering Department

This research summarized the microstructure analysis of alumina coating which
deposited to substrate Al-Si alloy by use of PFS (Powder Flame Spraying) technique, and
thermal cyclic where the out put specimen from flame spraying is brought to the boil
inside the furnace at 600oC with several cycles at 20, 50, and 100, and each cycle is held
a weighing. XRD, SEM (Scanning Electron Microscopes) and EDS (Energy Dispersive
Spectrometry) were applied to characterized phases and microstructures. In order to find
out the hardness which it formed at each layer and interface is using the micro hardness
tester. This research has ascertainable, that the bond which is formed at interface has a
high hardness where as thermal cyclic had an effect on TGO figuration and varying
phase. The TGO figuration, as effect of thermal cyclic, caused a reduction of the bonding.
It formed micro porosity which to make a failure at TBC system

1. Introduction                                 Coating (TBC system). In many research
        Diesel engine was the machine           had mentioned that thermal barrier
that working by fuel hypodermic of air          coating application at diesel engine aim
which have compressed so that have              to increase performance of machine,
own the high temperature and pressure           make-up of machine age, and efficiency
application from 300°C until 500°C,             of fuel usage (Beardsley, 1997), as
with the pressure equal to 2492 KPA. On         research done by Ramaswamy (2000)
its operation oftentimes meet the               that is use 8YPSZ and Mullite
constraints caused by the existence of          (3Al2O3.2SiO2) as thermal barrier
some component from diesel engine               coating material of diesel engine piston,
experience of the damage of effect on           what have been known able to give the
high temperature operation, so that its         protection to oxidation and degradation.
performance become less be optimal                       Some problems arise with the
        To increase resilience to high          thermal barrier coating method, that is
temperature required a veneering process        abrading of ceramic coat (Top Coat) and
use the system of Thermal Barrier               its failure because of thermal stress in its
bearing by forming TGO (Thermal                       In this study using specimens 4,
Growth Oxide) at interface between top        where 1 specimen without treatment and
coat and bond coat. Thermal Growth            the other 3 specimens applied to thermal
Oxide have different characteristic with      cycles, respectively - each is 25, 50, and
top coat and bond coat. This coat is          100 cycles. Cyclic thermal treatment
heterogeneous and imperfect, so that          performed at a temperature of 600°C,
caused crack on the surface of top coat,      issued each made 1 hour and weighed to
interface top coat-bond coat and also at      determine weight changes that occur.
TGO itself (Kristanto, 2008). So, it is       Observations began on specimens
required a furthermore research to            without treatment. As in Figure 3.1.a is
analysis of microstructure and phase          the initial condition of the coating
formed at interface between top coat and      specimen with flame spraying process
bond coat.                                    that has not been given heat treatment.
                                                      Macrostructure analysis was
2. Method                                     observed a corrugated surface and not
        The design and experimental           smooth and the color of specimen is grey
procedures can be explained as follows:       with little black spots on its surface. This
Substrate is Al-Si super alloy and Al2O3      condition is caused prior to coating with
powder for coating ceramic material.          the flame spraying process, sand blasting
The properties of Al2O3 ceramic               conducted in order to obtain the
powders, among others; M = 101.94 g /         substrate surface area that allows
mol; Specific Surface Area = 120-190          occurred bond between the substrate-
m2 / g; Stamping density = 950-1100 g /       bond coat and bond coat-top coat. Where
l. And also use Ni-Al powder for bond         a coating material powder particles are
coat material. The process of thermal         sprayed at high temperatures with a high
cycle include heating the specimen at         speed also, as a result uneven surface of
600oC for 1 hour and cooling down at          the specimen. In the specimen is a small
room temperature for 15 minutes, this         area of the specimen that is not coated,
cycle uses the following standard cycle       due to block by the holder during the
testing in high temperature corrosion.        coating process, as in Figure 3.1.
One cycle in this study equal to 1 hour               After      25     cyclic    thermal
holding time for heating up and hold 15       treatments, the substrate had melt,
minutes for cooling down. The duration        starting from not coated area. The
of the heating is done until peeling of top   growing cycle, the conditions are also
coat happened at specimen surface, this       increasingly rough surface, and the color
is indicates that the failure crack in TBC    changed to blackish specimens, as
systems. Checks carried out each cycle,       shown in Figure 3.1.a. Melting causes
by considering specimens with analytical      cracks at substrate side, it is also due to
balance. Then SEM and XRD                     differences in coefficient thermal
characterization was applied to obtain        expansion of each layer, the crack length
the phase and microstructure changed.         with increasing the heating cycle, until
                                              100 cycles, as shown in Figure 3.1.d,
                                              due to fire and fall off the substrate, with
3. Result and Discussion                      consideration that the thermal cyclic was
3.1 Macrostructure Analysis                   stopped.
                                          (a)                (b)

                                          (c)             (d)

                     Figure 3.1 Macrostructure analysis for (a). as sprayed, (b) 25 cycles, c. 50
                     cycles, (d). 100 cycles
Weigth (gr)                                                              Weigth (gr)

                           Cycle (hour)
                                                                                          Cycle (hour)
                              (a)                                                                        (b)
                                     Weigth (gr)

                                                          Cycle (hour)

              Figure 3.2 Result of microbalance analysis to obtained mass changed after (a) 25 cycles,
                                          (b) 50 cycles and (c) 100 cycles
        In this study also conducted          Ellingham diagram element Al is the
weighing at each cycle, to be known           easiest to oxidize, the oxidation reactions
weight changes of specimens due to            that form a protective layer of Al2O3.
cyclic thermal, to facilitate the analysis            After thermal cyclic, so the cycle
of weight change data plotted in a graph,     into 25 phases are formed almost the
as shown in Figure 3.2, and given the         same as before the thermal cycle, the
trend line to see the pattern of the graph.   phase γ-Al2O3 and NiO except that
Based on the graph in Figure 3.2 above,       there is little change, ie there was a shift
can be identified due to cyclic thermal       position slightly left 2θ, where the
causes severe changes in the fluctuations     highest peak of cycle 25 is occupied γ-
in the specimen, all the graphs have          Al2O3 phase with a cubic crystal system
common in the first cycle until some          in the 2θ = 44.470°, while referring to
point cycle increase in weight, due to the    the XRD results on specimens prior to
occurrence of oxide, since oxygen react       treatment with the highest peak position
with elements in the layer. But in the        of 2θ = 44.48167. But overall there was
next cycle, decrease in weight of the         no change crystal phase and the system,
specimen, it is because at around cycle at    due to the small shift differences. While
20 cycles occurs cracks in the specimen,      NiO phase after 25 cycles on the graph
and also specimens fall burning, until the    (in Figure 3.3.b) occupies the other
specimen weight decrease significantly,       highest peaks at 2θ = 43.273°, which
until the cycle to 100.                       when compared with the XRD results on
                                              specimens before treatment, NiO phase
3.2 X-ray Diffraction Analysis                top the view position 2θ = 38.62, it can
         Figure 3.3 is the result of x-ray    be known that there was a shift 2θ value,
diffraction analysis prior to cyclic          which is significantly more right, these
thermal. Where the results of tests on        shifts can result in a change of crystal
specimens prior to cyclic thermal formed      system. It was appropriate when search
γ-Al2O3 phase with cubic crystal system,      using software analysis, it is known that
and another phase is formed NiO with          the crystal system changes from cubic to
cubic crystal system. Results obtained in     rhombohedral.
the XRD test, which reflects that the                 In the cycle to 50 based on the
phase formed at the interface top coat        results of XRD analysis (Figure 3.3.c),
and bond coat is the result of oxidation      the phase also formed γ-Al2O3 and NiO,
of elements contained in the top coat and     but there is a change compared to the
bond coat. This is consistent with            highest peak position in 25 cycles, where
research conducted Ogawa (2003), that         the XRD results at the 50th cycle is the
at the interface between top coat and         highest peak position is the position of
bond coat TBC systems will be formed          2θ = 44.486°, which shifted more right
TGO (Thermal Oxide Growth), where             than cycle to 25, and prior treatment.
the TGO formed because of oxidation of        Due to the change is not too significant,
porosity with elements that are contained     the crystals formed system is not much
in the top coat and bond coat, which          different end result was γ-Al2O3 phase
starts from the porosity at the interface.    with a cubic crystal system. NiO phase
Furthermore the oxygen reacts with the        while the intensity increased as shown in
element to forming the top coat               Figure 3.3.c there are several other peaks
oxidation, because based on the               that are occupied by such NiO phase in
the position of 2θ = 2θ = 43.29 and              different from the 25 cycle , so also no
37.24 are NiO phase with rhombohedral            change of crystal. system.
crystal system, this position is not much

                                                            A : Alumina-cubic
                                                            N : NiO-cubic
                                                            R : NiO-
                               A             N
                           N           N




Figure 3.3 X-ray Diffraction Analysis for (a) as sprayed, (b) after 25 cycles (c) after 50
cycles (d) after 100 cycles
       After 100 Cyclic thermal               of the substrate reach all parts of the
treatment, there is not a significant         specimen. Thus expected that element
change, where the dominant phase is           of the substrate diffusion by porosity at
Al2O3 with cubic crystal system, which        the interface, so that elements of the
can be seen on the XRD results at Figure      substrate react with oxygen to form SiO
3.3.c, γ-Al2O3 phase with a crystal           at interface. To clarify the phases and
system cubic occupies the highest peak        microstructure formed on the specimen,
in 2θ = 44.885°. When compared with           it can be observed through SEM/EDS
the XRD results before treatment, and at      (Scanning Electron microscopic/Energy
25 and 50 cycles there was a slight           Dispersion Spectrometry) analysis
change in the position of the highest
peak, which shifts more right. While          3.3 Microstructure Analysis using
NiO phase obtained at the position of 2θ      SEM/EDX
= 43.29°, with a rhombohedral crystal                In    this     study     conducted
system, there was no change 2θ position.      observations of the interface between top
However, the XRD results for 100              coat and bond coat by using SEM / EDS
cycles in a new phase of SiO with the         (Scanning Electron microscopic/Energy
cubic crystal system in the position of 2θ    Dispersion Spectrometry), to determine
= 57.43. Based on analysis results, it is     phase and microstructure formed on the
because the thermal treatment occurred        specimen before and after thermal cyclic
melted substrate, which started in not        done.
coating area, until the 100 cycle, melted

                        TC                   BC

Figure 3.4 Microscope Electron analysis to as sprayed specimen. It shown porosity and
also TGO area

       Figure 3.4 is the result of            between the top coat and bond coat
microscope electron before thermal            porosity will occur as a result of the
treatment cycle, it shown porosity at top     spraying process. The existence of
coat. This is because there are grains of     porosity caused the separation or
alumina is not completely melt during         irregularity of the surface layer of
the coating using a flame spraying,           ceramic with a bond coat, as well as
because it is done in areas with no           between the bond coats to the substrate.
vacuum so some gases were trapped             Surface irregularity of flame spray
during the spraying process. According        interface, resulting in the emergence of
to research conducted by Chen (2004),         impurity and porosity which resulted in
whose explained that the interface area       the formation of surface defects, such as
particle adhesion inclusion and Sand          the interface regions top coat and bond
Blasting results, especially due to the       coat, so it looks like fissures that cut
melt ceramic particles are fired with a       through the interface areas, where there
certain speed and grinding on the bond        is TGO evenly spread, where the growth
coat surface, as shown above, a layer         of TGO give effect to the failure and
covering the top coat such as a bond coat     peeling of Thermal Barrier Coating.
layer, so that the boundary between the       Where the growth of TGO is depends on
top coat and bond coat is not too clearly     time,     temperature     and     chemical
visible, it shown in a different color, and   composition that is in the top coat and
porosity line blackish color. Shown in        bond coat. Many studies have stated that
Figure 3.4, the specimens are not             the oxidation of the bond coat is the
subjected to thermal treatment already        most important thing in a decrease in
started happening at the interface top        performance of the system at high
coat and bond coat, but oxidation occurs      temperatures. As in this study, the crack
is still thin and not very visible, because   occurred due melt substrate, where the
the intake of oxidation during spraying.      melting of the substrate provides a stress
Oxidation which occurs not destructive,       in the layers of the others, it can be seen
because it formed a thin layer of             in Figure 3.1, cracks occur on the side of
protective nature.                            the specimen, which continued to spread
          After 25 thermal treatment it was   in all parts of the specimen.
found that in areas along the interface
between top coat and bond coat will be               After the 100th cycle, cyclic
formed TGO (Thermal Oxide Growth)             thermal treatment is stopped. That's
and alumina thin layer of intermixed          because state specimens are not possible.
zone area (area that contains a blend of      Melted Substrate began in the not
elements contained in the top coat and        coating area and continued to other parts
bond coat ) are shown with a slightly         of the specimen, as in the previous
blackish color on the interface. As           explanation on macrostructure analysis.
research conducted by Quadakkers, et al       At the 100th cycle and TGO formed
(2005) explained that during the heat         more and more, in general, this layer
treatment (thermal cycle) on the              exists between top coat and bond coat
specimens using the TB, resulting in the      because the area contains a lot of
emergence of TGO (Thermal Oxide               porosity and provides a way for the
Growth) in the interface between top          oxide to oxidize the elements contained
coat and bond coat. This region is the        in the bond coat, but is not possibility
weakest area and the beginning of the         TGO formed inside the bond coat. As
formation of cracks. The existence of         research conducted by Chen (2004),
porosity in the ceramic layer as a result     found TGO trapped inside the bond coat
of oxidation, as shown in Figure 3.5.a        due partly to avoid the bay area bond
porosity is formed along the cracks.          coat surface, so the oxygen is oxidized
          Porosity in the specimens           and reacts with the elements contained in
increased with increasing time of             the bond coat. According to research
thermal cyclic treatment, as well as in       conducted Sulistijono (1998), Al content
the cycle to 50 levels of porosity in the     in the specimen which reached 8.27%
top coat and bond coat on the area, as        should be able to protect the metal
shown in Figure 3.6. Porosity occurs at       underneath the aluminum oxide formed
during the thermal cyclic. Oxide can be            that form the interface between top coat
formed Al2O3 as TGO layer (Thermal                 and bond coat, are O, Al, Ni. The
Oxide Grown) provides protection by                composition of the elements are not
reacting with oxygen through the TGO               equal to the before and after thermal
layer (Thermal Oxide Grown). Because               cycle.
react with Al, so ideally there is no
oxygen enter into the bond coat layer
(according to the Elingham diagram).
But in fact, there are other oxides, to find
out what oxides are formed, its existence
can be detected using EDS (Energy
Dispersion Spectometry).
        Table 3.1 is summary result of
SEM / EDS, it can be seen the elements


                                                                           Resin at porosity


                     (a)                               (b)

              Figure 3.5 SEM analysis on specimen after (a) 25 thermal cycles and (b) 50
                         cycles. It shown porosity, crack and TGO area

       Table 3.1 Comparison of element quantity of specimen

     ELEMENTS        AS SPRAYED      25 Cycle     50 Cycle     100 Cycle
                       (% Mass)     (% Mass)     (% Mass)      (% Mass)
         C                 -             -            -          3,97
         O               34,15        38,43        21,47         32,93
         Al              26,27        35,10         0,11         21,92
         Ni              39,57        26,47        78,42         38,64
         Si                -             -            -          2,55
      Phases          As sprayed     25 Cycle     50 Cycle     100 Cycle
                      (% Mass)      (% Mass)     (% Mass)      (% Mass)
        Al2O3              49,64      66,31        0,21          41,41
        NiO                50,36      33,69        99,79         49,16
        SiO2                 -          -            -           5,46
          Table 3.1 shows that almost all     reaching 78.42% at 99.79 on Ni and
the elements that formed at the interface     NiO. This can be caused because at the
changes after cyclic thermal, these           time of other elements to form bonds,
changes vary from each cycle, elements        not react to form Ni oxide or oxide
and compounds. Before cyclic thermal,         formation can be said NiO late. As a
obtained for the elements O was 34.15%        result of the bond coat Nickel Ni from
because of the oxidation process during       the substrate accommodating before
the spraying process. In cycle 25 there is    forming the nickel oxide resulted in the
an increase due to the occurrence of          mass fraction of Ni grows. According to
oxidation processes during the cyclic         research conducted Chen, et al (2004)
thermal oxidation which occurs both in        explains that after the heat treatment of
the furnace and the moment when issued        specimens will be formed oxide - a new
free air. Increased % of mass is in           oxides such as (Cr, Al) 2O3, Ni (Cr, Al)
accordance with the weighing done on          2O4, which is heterogeneous NiO / non
visual observation, where the heavy           uniform and cause layer has a low
curve in each of his cycle until the Cycle    adhesivity.
25 has increased. However, severe                      . In the cycle to 100, based on
decline in the next cycle, until the 100th    test results of SEM / EDS acquired
cycle, this would fit in Table 3.1 it can     elements Si and SiO compounds, which
observe the cycle to 50 and 100% mass         are not found at previous cycle, this is
of O has decreased, it is because of          because the condition of the specimen
cracks in the specimen due to substrate       that has undergone melting in the
melting.                                      substrate, where it reaches the overall
        Almost the same Al element to         melting of the specimen in this cycle,
the condition O elements, which               and based on SEM observation test
increased in % mass after cycle 25,           (Figure 3.5) where the substrate melts
before the cyclic thermal 26.27% to           filled at porosity and top coat bond coat.
35.10%, due to the formation of bonds                  From SEM analysis at Figure 3.6
with oxygen to form Al2O3, formed a           note that after the thermal spraying
thin layer of protection. Of course, these    occurred many bonding at the interface
conditions are also accompanied by an         between top coat and bond coat such as:
increase in mass% Al2O3 formed                mechanical interlock, chemical bonds
compounds in the cycle up to 25, more         (oxidation reactions), and diffusion.
time on the cyclic thermal oxidation rate     Mechanical interlock between the
is also greater. This causes the content of   surface of the top coat and bond coat
Al and eventually thinned out (smialek,       caused some of them. First, at the time
2006). The statement according to Table       of the shooting of ceramic particles, not
3.1, where the number % of the mass of        all particles in the melt state and the
Al and Al2O3 on the cycle decreased to        shooting occurred at high speed so that a
50 up to 100 cycles.                          portion of the bond coat surface that
        Whereas for Ni element in the         does not deform uniformly. Second, the
cycle 25 has decreased, it could be due       condition of the bond coat surface is not
to the time of the shooting on the area       flat so it had a high roughness, so that
covered with a thin layer of test is more     the particles are fired top coat and follow
dominant Al2O3. With increasing time,         the contour of the bond coat. Between
the next cycle occurs increased % mass,       ceramic particles and the other one has a
different cooling rate so that the bond         interlock can improve adhesivity
will be formed interlock mechanical             between top coat and bond coat, because
layer. According to research conducted          the area of overlap between top coat and
by Chen, et al (2004), mechanical               bond coat expanded.

             Chemical                                  Mechanical
             bonding                                    Interlock

                               Top Coat       BC

         Figure 3.6 Microscope electron analyses at interface layer between top coat and
                                       bond coat.

        Chemical bonds or result of                    and bond coat to form oxide at
oxidation reaction occurs at the interface             interface layer
area between the bonds coats based                  3. At interface layer found chemical
super alloy with oxygen is trapped in                  bonding resulted from oxide
porosity. The oxidation reaction formed                reaction      and      mechanical
a variety of metal oxides at different                 interlocked bonding
forms. At the top coat and bond coat
Al2O3 and partially formed within the           5. Reference
TGO formed many kinds of oxides as              Chen, W. R., Wue X, Marple B. R.,
described previously. Diffusion process            Patnaik, P. C., 2004. “ Oxidation and
occurs between the surfaces of the top             crack nucleation/growth in an air-
coat with a bond coat surface with atoms           plasma-sprayed thermal barrier
reactivity.                                        coating with NiCrAlY bond coat”.
                                                   Institute for Aerospace Research,
4. Conclusion                                      National Research Council of Canada,
From this research it can conclude that:           1200 Montreal Road, Bidg M-13,
    1. Thermal growth oxide layer had              Otawa, Ontanio, Canada, K1A0RG
       form at the interface between top        Kristanto. 2008. “Studi Fasa dan
       coat ceramic and bond coat Ni-               Mikrostruktur pada Interface Top
       Al that formed alumina (Al2O3)               coat 8YSZ dan Bond coat
       phase                                        NiCoCrAlY Akibat Thermal
    2. Thermal       cycle     at    high           Fatique”. Surabaya: Jurusan Teknik
       temperature operation caused                 Metalurgi dan Material- Institut
       diffusion of element on substrate            Teknologi Sepuluh Nopember
Ogawa . K, Gotoh .N, 2003. “ The            Proquest Science Journals pg 729
  influence of thermal barrier top       Smialek, J. L., jan 2006. “Mousture-
  coating on the initiation and growth       induced delayed spallation and
  of thermally cycled thermal barrier        interfacial hydrogen embrittlement
  coatings”. Sweden : Lund University.       of alumina scales, JOM. Pp 29-35.
Ramaswamy,P.,2000,“Thermo                Sulistijono, 1998, “ Pelapisan keramik
  mechanical fatique characterization        pada sudu turbin gas untuk
  of zirconia (8%Y2O3-ZrO2) and              meningkatkan ketahanan korosi pada
  mullite thermal barier coating on          temperature tinggi “. Surabaya :
  diesel engine components: effect of        Jurusan Teknik Mesin-Institut
  coatings on engine performance”,           Teknologi Sepuluh Nopember.

Shared By: