8/8/2019 Salt Fluxing Degradation Final Report- Gary Bywater
http://slidepdf.com/reader/full/salt-fluxing-degradation-final-report-gary-bywater 1/14
University of Manchester
Student ID: 76041250
Materials Performance Module
Degradation Mechanism Final Report
Salt Fluxing Hot Corrosion Mechanism of Nickel-Base
Superalloy Gas Turbine Blades
Dr T J Marrow
March 2010
8/8/2019 Salt Fluxing Degradation Final Report- Gary Bywater
http://slidepdf.com/reader/full/salt-fluxing-degradation-final-report-gary-bywater 2/14
8/8/2019 Salt Fluxing Degradation Final Report- Gary Bywater
http://slidepdf.com/reader/full/salt-fluxing-degradation-final-report-gary-bywater 3/14
University of Manchester
Student ID: 76041250
Hot Corrosion
Oxidation is the most important hot corrosion mechanism; metals/alloys will be oxidized when
heated to elevated temperatures or when operating in highly oxidizing environments such as a
combustion atmosphere. It is important to know that upon oxidation of a gas turbine blade an oxide
layer forms which can be protective or non protective depending on its chemical morphology and
adherence to the substrate. Initially an oxide layer which forms is sacrificial, regenerative, and
protective against other hot corrosion attack such as sulphidation, carburization, and as the concern
of this report molten salt fluxing attack. The alloy additions of chromium and aluminium allow the
superalloy to from a protective chromium/aluminium oxide layer.
Over time however due to the various hot corrosion mechanisms the oxide layer can either be
degraded exposing the substrate or, can be chemically changed to no longer be protective and will
result in a degrading porous oxide formation under the surface layer of the superalloy. When hot
corrosion of the turbine blade is discussed within this report it is assuming there is no specialised
surface coating present other than the protective oxide layer which is being degraded to expose the
superalloy for direct hot corrosion attack. The role of coatings and how they protect against hot
corrosion will be discussed later in this report.
The salt fluxing mechanism of hot corrosion involves the dissolution of the protective oxide layer at
the oxide/salt interface resulting in non-protective precipitates (such as chromium sulphides). Oxide
layer dissolution occurs by either the combination of oxides with O2-
to form anions (via basic
fluxing) or by decomposition of oxides into cations and O2-
(via acidic fluxing) depending on the salt
composition. Salts high in SO3 are acidic and basic when low in SO3. Sodium sulphate is the
dominant salt of hot corrosion due to its high thermodynamic stability. Molten salt hot corrosion
occurs predominantly in the range of 700-900˚C involving sulphur from jet fuel which reacts with
sodium chloride (NaCl) originating from ingested air during combustion in the combustor to form
molten sodium sulphate (Na2SO4) which deposits on the turbine blade. Firstly the protective oxide
layer is attacked and once degraded the superalloy substrate is attacked. Other salt contaminants
can be formed from the environment; oxygen and sulphur will combine with sodium, potassium,
vanadium, and chlorine within the gas turbine engine to create sodium chloride (NaCl), vanadium
oxide (V2O5), and potassium sulphate (K2SO4) (2) (3). Several chemical reactions can occur to form
sodium sulphate.
2NaCl + SO3 + 1/2O2 Na2SO4 + Cl2
2NaCl + SO3 + H2O Na2SO4 + 2HCl
2NaCl + SO2 + O2 Na2SO4 + Cl2
Hot corrosion can be separated into two forms, type I hot corrosion known has high temperature
hot corrosion (HTHC), and type II hot corrosion known as low temperature hot corrosion (LTHC).
HTHC occurs in the temperature range of 850-950˚C where pure sodium sulphate is above its
melting temperature. Sodium sulphate will form a mixture with the other salt contaminants such as
sodium chloride; this lowers the melting temperature as a eutectic is formed and broadens the
molten salt attack range. In this attack sulphur is released from the reduction of sodium sulphate
which diffuses into the superalloy to form chromium sulphides as in the reaction below (3).
8/8/2019 Salt Fluxing Degradation Final Report- Gary Bywater
http://slidepdf.com/reader/full/salt-fluxing-degradation-final-report-gary-bywater 4/14
University of Manchester
Student ID: 76041250
Cr2O3 + O2-
2CrO2-
This reaction depletes the superalloy matrix of chromium allowing the oxidation of the base metal to
be accelerated and a porous non-protective oxide layer will form. It is found that
chromium/titanium sulphide particles deposit in the depleted matrix zones. As the reaction
proceeds released sulphur diffuses deeper into the substrate to form more sulphides inducing
further corrosion. The further the oxidation attack penetrates the larger the loss of structural
material will occur and failure will be inevitable. HTHC is a form of intergranular attack and is basic
fluxing. Figure 1 below shows the near surface morphology of HTHC attack.
Figure 1: High temperature hot corrosion (HTHC) (4)
Type II LTHC involves acidic fluxing and occurs approximately in the range of 600-800˚C (there is a
cross over range of HTHC and LTHC forming a transition attack). This form of corrosion again
involves sodium sulphate however eutectic mixtures are formed with nickel sulphate (NiSO4)
generating a eutectic mixture with a much lower melting temperature than that of HTHC. LTHC
sulphide eutectic mixtures are dependent upon the partial pressure of SO3 in a gaseous phase and
typically show no chromium depleted matrix and little intergranular attack. With a high partial
pressure of SO3 a high rate of attack is the result causing pitting of the surface. It is found that
chromium and titanium sulphides form a continuous layer. Figure 2 below shows surface
morphology of LTHC attack.
Figure 2: Low temperature hot corrosion (4)
It is common that a combination of HTHC and LTHC can occur resulting in a surface morphology
depicted in Figure 3.
8/8/2019 Salt Fluxing Degradation Final Report- Gary Bywater
http://slidepdf.com/reader/full/salt-fluxing-degradation-final-report-gary-bywater 5/14
University of Manchester
Student ID: 76041250
Figure 3: Transition hot corrosion (4)
Influence of superalloy composition and microstructure
Considering the gas turbine blade independent of surface coating, the composition and
microstructure are directly responsible for the severity of hot corrosion attack (the former being
more influential due to significance of electrochemistry). Generally, the addition of alloying
elements such as aluminium, chromium and silicon are used to form protective oxide layers to resistattack from hot corrosion. Increased chromium content will improve HTHC resistance, with a deeper
chromium matrix reservoir to deplete (see Figure 4). However, excess amounts of chromium will
result in the precipitation of a topologically close-packed (TCP) phase which embrittles the alloy,
reducing ductility and the high temperature strength (1) (5).
Figure 4: Corrosion rate of various superalloys with chromium content (1)
An example of alloy development is shown in Figure 5. Pratt & Whitney improved a single crystal Ni-
base superalloy for gas turbine blade application by increasing the aluminium content from 5-
5.6wt% and reducing the titanium content to close to zero. The reason for this is it was found that
titanium ions introduced vacancies into alumina lattices increasing ionic mobility’s making the
previous superalloy more susceptible to fluxing.
8/8/2019 Salt Fluxing Degradation Final Report- Gary Bywater
http://slidepdf.com/reader/full/salt-fluxing-degradation-final-report-gary-bywater 6/14
University of Manchester
Student ID: 76041250
Figure 5: Pratt & Whitney data of improved alloy, PWA1484
Critical alloy additions to be aware of are molybdenum (Mo) and tungsten (W). They are capable of causing catastrophic self-sustaining degradation as they form oxides which react with sodium
sulphate to form acidic salts high in SO2. This is a classic example of an acidic fluxing mechanism
forming compounds such as Na2MoO4 which posses a high solubility of the protective Al2O3 and
Cr2O3 layers. Molybdenum and tungsten are useful for enhancing mechanical properties but there is
a clear trade off in properties for hot corrosion resistance and vice versa (1) (2) (5).
Table 3: Role of alloying elements in superalloys (6)
The addition of rare earth elements have been found useful. Hafnium (Hf), lanthanum (La), and
yttrium (Y) bind strongly with sulphur which is responsible for sulphidation attack leading to
oxidation attack, as mentioned these mechanism are closely related to salt fluxing. Figure 6 shows
example test data of rare-earth metal doping.
8/8/2019 Salt Fluxing Degradation Final Report- Gary Bywater
http://slidepdf.com/reader/full/salt-fluxing-degradation-final-report-gary-bywater 7/14
University of Manchester
Student ID: 76041250
Figure 6: Weight change data due to cyclic oxidation of a CMSX-4 single-crystal superalloy with a peak temperature of
1093C, and with variants of the superalloy doped with rare-earth elements La, Ce, and Y.
Ni-base superalloys are highly alloyed components. The hot corrosion behaviour is very complex
and there is the constant trade off of mechanical properties for hot corrosion/oxidation resistance
when varying composition and the behaviour is made even more complex when considering the
numerous mechanisms of hot corrosion attack and yet again, the behaviour of different superalloy
composition to each mechanism. From Figure 7 it can be noticed that later day generation single-
crystal superalloys have a lower hot corrosion resistance than their elders. This is due to the
development of surface coating technology to protect the superalloy from oxidation/hot
corrosion/thermal degradation and allows the design of the superalloy to focus on resisting the
mechanical stresses of operation. This leads to the focus of this report, the role of surface coatings.
Figure 7: Cyclic oxidation tests at 1100˚C of 1hour intervals of various uncoated single-crystal superalloys (1)
8/8/2019 Salt Fluxing Degradation Final Report- Gary Bywater
http://slidepdf.com/reader/full/salt-fluxing-degradation-final-report-gary-bywater 8/14
University of Manchester
Student ID: 76041250
Surface Coatings
In order to improve the hot corrosion resistance of gas turbine blades specialised protective surface
layers are applied. There are many types of surface coatings available however the main three types
are: thermal barrier coatings (TBC), diffusion coatings, and overlay coatings. TBC’s are used to
insulate the turbine blade from the high operating temperatures by a few hundred degrees
centigrade. As the melting point of Ni-base superalloys are in the vicinity of 1450˚C which is also
their operating temperature, without TBC’s their safe application would not be possible. Thermal
protection is not the concern of this report so will not be touched on any further.
Diffusion and overlay coatings are used to protect the superalloy from environmental oxidation and
other various corrosive attacks. Diffusion coatings involve the powder deposition (and subsequent
diffusion) of aluminium onto (and into) the surface of a superalloy. The aluminium reacts with the
nickel to form intermetallic compounds of Ni3Al, NiAl and Ni2Al3. Heat treating increases adhesion
and results in inter-diffusion with the substrate generating an aluminium rich layer on the surface of
the component. With the high surface concentration of aluminium the superalloy substrate is now
protected from oxidation as the aluminium rich layer forms an aluminium-oxide (Al2O3) scale, a thick
continuous protective layer to the harsh oxidising conditions (1) (6).
Overlay coatings are an improvement in oxidation protection of superalloys than diffusion coatings
due to a more chemically stable oxide scale generated. The oxide scale is more chemically stable as
the deposited powder is prealloyed to be independent of the substrate alloy (some interdiffusion
will occur). High temperature overlay coatings are of the form MCrAlX, where M denotes nickel (Ni)
or cobalt (Co), Cr is chromium, Al is aluminium and X denotes a minor proportion element added to
enhance adherence of the oxide layer to the substrate: yttrium, hafnium and silicon are used.
Similar to the diffusion coating the aluminium forms an aluminium-oxide scale to resist oxidation
whereas the chromium protects against other various hot corrosion mechanisms (1) (6).
Table 4: Composition of commonly used MCrAlY overlay coatings in wt% (6)
Figure 8 below shows the significant synergistic improvement in surface coating protection of
overlay coatings with the addition of silicon and hafnium.
8/8/2019 Salt Fluxing Degradation Final Report- Gary Bywater
http://slidepdf.com/reader/full/salt-fluxing-degradation-final-report-gary-bywater 9/14
University of Manchester
Student ID: 76041250
Figure 8: Progressive degradation of EB-PVD and plasma sprayed overlay coatings, showing synergistic benefit of silicon
and hafnium additions (1)
The method of deposition greatly influences the effectiveness of the surface coating. Diffusion
coatings are deposited using pack-cementation or chemical vapour deposition (CVD). Pack-
cementation involves the component to be immersed in the powder mixture to be deposited and is
heated within a sealed vessel in a protective atmosphere. The pack powder consists of the coating
element source, an activator (such as NaF, NaCl, or NH4Cl), and an inert filler material often alumina
to prevent the source from sintering. The deposition requires the activator to react with the
aluminium source to for aluminium halides which deposit onto the surface of the substrate alloy to
react and release aluminium. A flaw of pack cementation is that it is used to coat surfaces which are
brought into direct contact with the pack, complex cooling channels in gas turbine blades would not
be able to be coated and thus left vulnerable to oxidation/hot corrosion attack, hence the use of
CVD. CVD is similar to pack-cementation however the aluminisation takes place in the gaseous
phase (hence vapour deposition), requiring the reaction of AlCl3 powder in argon with hydrogen gas
to take place to form aluminium monochloride which reacts with the superalloy substrate to release
aluminium at the surface (1).
Overlay coatings (and TBC’s) are deposited by means of electron beam physical vapour deposition
(EB-PVD) and plasma spraying. The process uses an electron beam to vaporise an ingot of coating
material, once the coating material has evaporated it is deposited onto the surface of the superalloy
substrate. The key difference of EB-PVD to CVD is that no chemical reactions take place and it is a
much more expensive process to operate (1). Plasma spraying uses the energy from a thermally
ionised gas to melt and propel fine metal or oxide particles from powder form onto a surface so they
adhere and agglomerate to a produce a coating. Plasma sprayed layers have a characteristic ‘splat’
surface morphology. The advantage of plasma spraying is that the powder composition closely
matches that of the deposited coating, this is not always the case with EB-PVD due to evaporation
8/8/2019 Salt Fluxing Degradation Final Report- Gary Bywater
http://slidepdf.com/reader/full/salt-fluxing-degradation-final-report-gary-bywater 10/14
University of Manchester
Student ID: 76041250
rates. A disadvantage of plasma spraying is the difficulty to produce dense layers without some
layer porosity.
Component testing
Gas turbine blades of different aircraft will operate under varying conditions, when concerning hotcorrosion the important lifetime factors are the amount of fuel air contamination and the blade
operating temperature. A key example of how salt contaminants influence turbine blade life is
shown below in Figure 9. An exponential decrease in life is caused by a small increase in parts per
million (ppm) of sodium salt contaminant. In this example the coating used is a platinum-aluminium
coat, where platinum is uniformly electroplated followed by a CVD layer of aluminium.
Figure 9: Effect of sodium sulphate hot corrosion on turbine blade life at 870˚C (7)
The surface coating of platinum-aluminium intermetallic is highly corrosion resistant. The example
above is the conventional method of taking lifetime data directly. The blades were run within the
same test machine side by side in severe corrosive conditions. An interim evaluation after 11,300
hours of service (consisting of 289 starts) showed the uncoated blade had 0.005inch corrosion attack
over 50% of the airfoil with approximately 0.01inch penetration at the base of the airfoil. The
coated blade on the other hand had no visual evidence of attack other than two small roughened
regions. Another example is through the work of Sidhu et al (8) where superalloys of composition
shown in Table 5 were tested uncoated and coated with an high velocity oxy-fuel sprayed (HVOF aform of plasma spraying) Ni-15.3Cr-3.1B-4.8Si-4.2Fe-0.6 wt% coating in a molten salt environment
consisting of Na2SO4 and V2O5 in the ratio of 40:60 wt% at 900˚C. HVOF is a thermal spraying
technique producing coats with lower porosity, higher hardness, superior bond strength, and lower
decarburisation than other spray techniques. The given composition of salt contaminants provides a
eutectic with a low melting point of 550˚C and is a rather aggressive environment.
Table 5: Sidhu et al superalloy compositions
8/8/2019 Salt Fluxing Degradation Final Report- Gary Bywater
http://slidepdf.com/reader/full/salt-fluxing-degradation-final-report-gary-bywater 11/14
University of Manchester
Student ID: 76041250
The coated and uncoated alloys were subject to a cyclic regime of 50 cycles, each cycle being 1 hour
of heating at 900˚C. A thickness of 3-5mg/cm2
of salt contaminant mixture were deposited with a
camel hair brush on the preheated samples. The micrographs (Figure 10) show clearly that the
uncoated samples suffered a significant amount of spallation and sputtering of the surface due to
molten salt corrosion and clearly indicate the presence of cracks. The use of x-ray diffraction (Figure
12 & Figure 13) outlines the main phases present on the surface before and after oxidation/hot
corrosion. By clearly identifying the corrosion products the attack mechanism can be identified for
the coated and uncoated superalloys.
Figure 10: SEM analysis for the uncoated and coated samples after 50 cycles; a)uncoated Superni 600, b) coated Superni
600, c) uncoated Superni 601, d) coated Superni 601, e) uncoated Superfer800H, f)coated Superfer800H
8/8/2019 Salt Fluxing Degradation Final Report- Gary Bywater
http://slidepdf.com/reader/full/salt-fluxing-degradation-final-report-gary-bywater 12/14
University of Manchester
Student ID: 76041250
Figure 11: SEM analysis of the as sprayed NiCrBSi coating
Figure 12: X-ray diffraction patterns for uncoated superalloys after 50 cycles; a) uncoated Superni 601, b) uncoated
Superfer 800H, and c) uncoated Superni 600
The x-ray diffraction peaks show that the uncoated corroded Superni 600 has the main phases of
NiO, Fe2O3, NiCr2O4, Ni(VO3)2, FeV2O4, and FeV occurring. Corroded Superni 601 shows similar
behaviour but peak intensities of NiO and Ni(VO3)2 are low.
8/8/2019 Salt Fluxing Degradation Final Report- Gary Bywater
http://slidepdf.com/reader/full/salt-fluxing-degradation-final-report-gary-bywater 13/14
University of Manchester
Student ID: 76041250
Figure 13: X-ray diffraction patterns for coated superalloys after 50 cycles; a) coated Superni 600, b) coated Superni 601,
c) coated Superfer 800H
Figure 14 below shows clearly the uncoated Ni-base superalloys performed better than the Fe-base
superalloy yet the same surface coating resulted in practically equal oxidation/hot corrosion
resistance. This is an excellent outline of the significant benefits of surface coatings and how the
superalloy composition influences oxidation/hot corrosion of the uncoated superalloy.
Figure 14: Weight gain of specimens for up to 50 cycles
8/8/2019 Salt Fluxing Degradation Final Report- Gary Bywater
http://slidepdf.com/reader/full/salt-fluxing-degradation-final-report-gary-bywater 14/14
University of Manchester
Student ID: 76041250
Bibliography
1. Reed, Roger C. The Superalloys Fundamentals and Applications. s.l. : Cambridge University Press,
2006. 978-0-521-85904-2.
2. Hot corrosion of some superalloys and role of high-velocity oxy-fuel spray coatings- a review.
Sidhu, T S, Agrawal, R D and Prakash, S. s.l. : Elsevier, 2005, Surface & Coatings Technology, Vol.
198, pp. 441-446.
3. Lai, George Y. High-Temperature Corrosion of Engineering Alloys. s.l. : ASM International, 1997. 0-
87170-411-0.
4. An investigation of blades failures in combustion turbines. Viswanathan, R. s.l. : Pergamon, 2001,
Engineering Failure Analysis, Vol. 8, pp. 493-511.
5. Hot corrosion in gas turbine components. Eliaz, N, Shemesh, G and Latanision, R M. s.l. : Elsevier,
2002.
6. Campbell, F C. Manufacturing Technology for Aerospace Structural Materials. s.l. : Elsevier, 2006.
978-1-85-617495-4.
7. Boyce, Meherwan P. Gas Turbine Engineering Handbook. s.l. : Elsevier, 2006. 978-0-7506-7846-9.
8. Hot corrosion behaviour of HVOF-sprayed NiCrBSi coatings on Ni- and Fe-based superalloys in
Na2SO4- 60% V2O5 environment at 900C. Sidhu, T S, Prakash, S and Agrawal, R D. s.l. : Elsevier,
2006, Acta Materialia, Vol. 54, pp. 773-784.
9. Bernstein, Henry L. Materials Issues For Users Of Gas Turbines. San Antonio, Texas : Gas TurbineMaterials Associates.
10. On the surface preparation of nickel superalloys before CoNiCrAlY deposition by thermal spray.
Bardi, U, et al. s.l. : Elsevier, 2004, Vol. 184, pp. 156-162.
11. The effect of EB PVD coatings on structure and properties of nickel-base superalloy for gas
turbine blades. Tchizhik, A A, et al. St. Petersburg, Russia : Elsevier, 1996, Surface & Coatings
Technology, Vol. 78, pp. 113-123.
12. Vapour aluminide coating of internal cooling channels, in turbine blades and vanes. Smith, A B,
Kempster, A and Smith, J. s.l. : Elsevier, 1999, Vols. 120-121, pp. 112-117.
13. Hot corrosion of materials: a fluxing mechanism? Rapp, Robert A. s.l. : Pergamon, 2002,
Corrosion Science, Vol. 44, pp. 209-221.
14. Chemistry and Electrochemistry of Hot Corrosion of Metals. Rapp, Robert A. s.l. : Elsevier, 1987,
Materials Science and Engineering, Vol. 87, pp. 319-327.
15. Hot corrosion behaviour of AIP NiCoCrAlY(SiB) coatings on nickel base superalloys. Wang, Q M, et
al. s.l. : Elsevier, 2004, Surface & Coatings Technology, Vol. 186, pp. 389-397.