Furnace Cycle Research Articles

Deposition and Oxidation of Oxide-Dispersed CoNiCrAlY Bondcoats

CoNiCrAlY powder and nano-size alumina powder were milled by a high-energy-attrition ball-mill, and an oxide-dispersed powder was produced with a mixed structure of metal and alumina in each particle. The oxide-dispersed bond coat powder was deposited by HVOF. Pores, however, were observed in the coating since the alumina was deposited without sufficient melting. Isothermal oxidation tests were carried out for the bond coat specimens at a temperature of 1373 K up to 1000 h in air. As a result, oxidation proceeded inside the coating, since oxygen penetrated through pores formed in the dispersed alumina. However, the authors find that another deposition using higher power levels led to a bond coat without pores. A commercially available oxide-dispersed CoNiCrAlY powder was also deposited by HVOF and VPS, and isothermal oxidation tests were performed. The analysis clarifies that the HVOF bond coat exhibited the thinnest thermally grown oxide than those of the VPS bond coat and conventional metallic bond coat. Furnace cycling tests were conducted using the specimens with an additional ceramic thermal-barrier coating. The specimen with the bond coat sprayed by VPS using commercial oxide-dispersed powder showed almost same number of cycles to delamination compared with the specimen with the conventional metal bond coat.

Impact of superalloy composition, bond coat roughness and water vapor on TBC lifetime with HVOF NiCoCrAlYHfSi bond coatings

The performance of thermal barrier coating (TBC) specimens with superalloy 1483 and X4 substrates and high velocity oxy fuel (HVOF)-NiCoCrAlYHfSi bond coatings was evaluated in dry and wet air in 1h furnace cycles at 1100°C. Compared to a prior study of NiCoCrAlYHfSi bond coatings, the HVOF bond coatings were rougher Ra=8μm, closer to commercial HVOF coatings. For TBC coatings on the X4 substrates cycled in air with 10% water vapor, increasing the roughness from Ra=5 to 8μm had little effect on the average TBC lifetime. The presence of 10% water vapor reduced the average TBC lifetime for 1483 by 13%, however, in both wet and dry conditions, the lifetimes for this substrate were 30–50% below those observed for X4 substrates. This may be attributed to the lower Al content and higher Ti content in 1483. Roughness and residual stress in the alumina scale were measured on specimens without a ceramic top coating, but no effects were observed due to the change in substrate or the addition of water vapor.

Experimental and Finite Element Study of an Air Plasma Sprayed Thermal Barrier Coating under Fixed Cycle Duration at Various Temperatures

Cyclic furnace tests were conducted on Air Plasma Sprayed (APS) Thermal Barrier Coating (TBC) samples which were coated with NiCoCrAlY on a nickel‐based superalloy at three temperatures. From these tests the following were determined: oxidation kinetics and activation energy, bond coat rumpling rate, and cyclic failure lives. All failures occurred in the top coat consistent with engine experience. The measured rumpling rate and oxidation kinetics were used as input for a finite element model that utilized a highly realistic top coat constitutive model that included creep and relaxation behavior as well as tension/compression flow stress asymmetry consistent with available experimental data. The modeling results based on these measured inputs show that the primary sources of stress and strain in descending order are rumpling, oxide growth and finally thermal expansion mismatch between the top coat and underlying metal. Three important implications of these results are; predicted stresses are too low to predict failure based on single crack fracture mechanics, cycling increases accumulated inelastic strain in a fixed time interval providing an additional reason why cycling is damaging and finally to get realistic stresses and strains it is necessary to accurately characterize the geometry changes associated with rumpling.

Solidification of Lead Tin and Lead Telluride Eutectic Alloys in Microgravity

The Brazilian Microgravity Program is mainly based on experiments carried out on sounding rockets. A solidification furnace, capable of producing temperatures up to 900 °C, was developed to process metal and semiconductor alloys in microgravity environment. This paper describes a solidification experiment made in this furnace during a parabolic flight, with two eutectic alloys. The behavior of the eutectic alloys PbSn with 26.1 Pb at. % composition, and PbTe with 10.9 Pb at. % composition were presented and compared with laboratory solidifications carried out in the same furnace and thermal cycle. It was concluded that the formation of dendritic structures in PbSn alloy is related to the presence of sedimentation and convective flow during solidification, and the size of these structures is connected to the solidification time. Thus, in the microgravity alloy, there was no formation of dendritic structures and the profile of solute distribution remained constant throughout the sample. For the PbTe eutectic alloy the microgravity conditions have not caused significant changes compared to the earth solidified sample.

Thermal Fatigue Behavior of Air-Plasma Sprayed Thermal Barrier Coating with Bond Coat Species in Cyclic Thermal Exposure.

The effects of the bond coat species on the delamination or fracture behavior in thermal barrier coatings (TBCs) was investigated using the yclic thermal fatigue and thermal-shock tests. The interface microstructures of each TBC showed a good condition without cracking or delamination after flame thermal fatigue (FTF) for 1429 cycles. The TBC with the bond coat prepared by the air-plasma spray (APS) method showed a good condition at the interface between the top and bond coats after cyclic furnace thermal fatigue (CFTF) for 1429 cycles, whereas the TBCs with the bond coats prepared by the high-velocity oxygen fuel (HVOF) and low-pressure plasma spray (LPPS) methods showed a partial cracking (and/or delamination) and a delamination after 780 cycles, respectively. The TBCs with the bond coats prepared by the APS, HVOF and LPPS methods were fully delaminated (>50%) after 159, 36, and 46 cycles, respectively, during the thermal-shock tests. The TGO thickness in the TBCs was strongly dependent on the both exposure time and temperature difference tested. The hardness values were found to be increased only after the CFTF, and the TBC with the bond coat prepared by the APS showed the highest adhesive strength before and after the FTF.

Effect of water vapour content on thermal barrier coating lifetime

Furnace cycle testing was conducted in air with 10, 50 and 90 vol.-% water vapour and compared to prior results in dry O2. The first series of experiments examined Pt diffusion and Pt modified aluminised bond coatings on second generation superalloy N5 at 1150°C with commercially vapour deposited yttria stabilised zirconia (YSZ) top coats. Compared to dry O2, the average lifetimes with Pt diffusion coatings were unaffected by the addition of water vapour, while the Pt modified aluminide coating average lifetime was reduced by >50% with 10% water vapour, but less reduction was observed with higher water contents. The second series of experiments examined MCrAlY and MCrAlYHfSi bond coatings and air plasma sprayed YSZ top coatings on superalloy X4 cycled at 1100°C. Compared to dry O2, the addition of 10% water vapour decreased the lifetime of MCrAlY by 28%. Higher average lifetimes were observed with Hf in the bond coating.

Assessment of Cyclic Lifetime of NiCoCrAlY/ZrO2-Based EB-PVD TBC Systems via Reactive Element Enrichment in the Mixed Zone of the TGO Scale

The chemical composition of the alumina-zirconia mixed zone (MZ) of an electron beam physical vapor deposited thermal barrier coating (EB-PVD TBC) system is affected by service conditions and by the interdiffusion of elements from the substrate alloy below and the zirconia top coat. Three NiCoCrAlY bond-coated Ni-base substrates with YPSZ or CeSZ EB-PVD TBCs were subjected to a cyclic furnace oxidation test (FCT) at 1373 K (1100 °C) in order to provide experimental evidence of a link between chemistry of the MZ, the substrate alloy, the ceramic top coat, and the time in the FCT. Energy dispersive spectroscopy of the MZ revealed preferred accumulation of Cr, Zr, Y, and Ce. The concentration of the reactive elements (RE = Ce + Y + Zr) was related to the respective average lifetimes of the TBC systems at 1373 K (1100 °C). The RE content in the MZ turned out to be a life-limiting parameter for YPSZ and CeSZ TBC systems which can be utilized to predict their relative lifetimes on the individual substrates. Conversely, the TBC failure mechanisms of YPSZ and CeSZ TBC systems are dissimilar.

Effect of water vapor on the 1100 °C oxidation behavior of plasma-sprayed TBCs with HVOF NiCoCrAlX bond coatings

With the goal of investigating the reported detrimental effect of water vapor on thermal barrier coating (TBC) performance, furnace cycle experiments were conducted in dry O2 and air with 10 and 50% water vapor at 1100°C. The TBC systems evaluated were air plasma-sprayed (APS), yttria-stabilized zirconia (YSZ) top coatings with high velocity oxy fuel (HVOF)-deposited NiCoCrAlY or NiCoCrAlYHfSi bond coating. Average TBC lifetime was reduced by ~30% in air with 10% water vapor compared to cycling in dry O2, using 1h cycle durations. Superalloy substrates with Y and La additions also were investigated but showed no statistical change in the average TBC lifetime compared to the base CMSX4 superalloy. In all cases, the bond coating with Hf and Si additions increased YSZ lifetime by 20% or more. Experiments that increased water vapor to 50% showed no further decrease in TBC lifetime. Increasing the cycle frequency to 100h resulted in a large increase in TBC lifetime, especially for the NiCoCrAlYHfSi bond coatings. Co-doping the NiCoCrAl bond coat with Y and Hf was beneficial to TBC lifetime, but did not mitigate the detrimental impact of water vapor.

Fractographic and microstructural study of isothermally and cyclically heat treated thermal barrier coatings

The fracture surfaces from adhesion tested thermal barrier coatings (TBC) have been studied by scanning electron microscopy. The adhesion test have been made using the standard method described in ASTM 633, which makes use of a tensile test machine to measure the adhesion. The studied specimens consist of air plasma sprayed (APS) TBC deposited on disc-shaped substrates of Hastelloy X. The bond coat (BC) is of NiCoCrAlY type and the top coat (TC) consists of yttria partially-stabilised zirconia. Before the adhesion test, the specimens were subjected to three different heat treatments: 1) isothermal oxidation 2) thermal cycling fatigue (TCF) and 3) burner rig test (BRT). The fracture surfaces of the adhesion tested specimens were characterised. A difference in fracture mechanism was found for the different heat treatments. Isothermal oxidation gave fracture mainly in the top coat while the two cyclic heat treatments gave an increasing amount of BC/TC interface fracture with increasing number of cycles. Some differences could also be seen between the specimens subjected to burner rig test and furnace cycling.

Effect of increased water vapor levels on TBC lifetime with Pt-containing bond coatings

To investigate the effect of increased water vapor levels on thermal barrier coating (TBC) lifetime, furnace cycle tests were performed at 1150°C in air with 10vol.% water vapor (similar to natural gas combustion) and 90vol.%. Either Pt diffusion or Pt-modified aluminide bond coatings were applied to specimens from the same batch of a commercial second-generation single-crystal superalloy and commercial vapor-deposited yttria-stabilized zirconia (YSZ) top coats were applied. Three coatings of each type were furnace cycled to failure to compare the average lifetimes obtained in dry O2, using the same superalloy batch and coating types. Average lifetimes with Pt diffusion coatings were unaffected by the addition of water vapor. In contrast, the average lifetime of Pt-modified aluminide coatings was reduced by more than 50% with 10% water vapor but only slightly reduced by 90% water vapor. Based on roughness measurements from similar specimens without a YSZ coating, the addition of 10% water vapor increased the rate of coating roughening more than 90% water vapor. Qualitatively, the amount of β-phase depletion in the coatings exposed in 10% water vapor did not appear to be accelerated.

Corrosion Behavior of Fe-Cr Alloys in Ar-10%O<sub>2</sub> Atmosphere

The corrosion of four Fe-Cr alloys (Cr: 9, 12, 20, and 25%) with different Cr contents were subjected to cyclic oxidation furnace in Ar-10%O2 atmosphere at 950oC for 120 h. The samples after oxidation were investigated by using X-ray diffraction (XRD), optical metallography and SEM/EDS. The results indicated that increasing Cr content reduced the oxidation rate and the scale cracking. Alloys with less Cr content occurred breakaway oxidation due to formation of Fe-rich oxide scale. However, a thin protective chromia scale was observed on alloys with Cr > 20%. The corrosion behaviour on Fe-Cr alloy was then discussed based on experimental observation.

Failure Mechanism for Thermal Fatigue of Thermal Barrier Coating Systems

Thick thermal barrier coatings (TBCs), consisting of a CoNiCrAlY bond coat and yttria-partially stabilized zirconia top coat with different porosity values, were produced by air plasma spray (APS). The thermal fatigue resistance limit of the TBCs was tested by furnace cycling tests (FCT) according to the specifications of an original equipment manufacturer (OEM). The morphology, residual stresses, and micromechanical properties (microhardness, indentation fracture toughness) of the TBC systems before and after FCT were analyzed. The thermal fatigue resistance increases with the amount of porosity in the top coat. The compressive in-plane stresses increase in the TBC systems after thermal cycling; nevertheless the increasing rate has a trend contrary to the porosity level of top coat. The data suggest that the spallation happens at the TGO/top coat interface. The failure mechanism of thick TBCs was found to be similar to that of conventional thin TBC systems made by APS.

Process repeatability and sources of error in indirect SLS of aluminium

PurposeThe purpose of this paper is to investigate the accuracy and repeatability of the indirect selective laser sintering of aluminium process.Design/methodology/approachThis work characterised the shrinkage of indirect SLS aluminium parts during the various stages of production. Standard scale parts were measured using a Giddings and Lewis co‐ordinate measuring machine in both the green and infiltrated condition.FindingsThe experiments conducted show that most accuracy is lost during the furnace cycle and that the greatest loss of accuracy occurred in the Z dimension. Additionally the position of parts within the part bed in both X, Y and Z is shown to influence accuracy, with smaller parts being built closer to the edge of the bed later in the build. These results have been interpreted as being a result of the phenomenon of “Z‐growth”. Finally, the research shows that the overall accuracy of the indirect selective laser sintering of aluminium process is comparable with many existing processes such as investment casting.Originality/valueBefore any new material can be accepted, there is a need to not only fully characterise the dimensional accuracy attainable, but also to gain a thorough understanding of the processes that contribute to the inaccuracies. This paper addresses this need.

Study of failure of EB-PVD thermal barrier coating upon near-α titanium alloy

The cracking failure of a conventional thermal barrier coating (TBC), consisting of a near-α titanium substrate, a NiCoCrAlY bond coat (BC), and a 8 wt.% yttria-stabilized zirconia ceramic layer deposited by electron beam-physical vapor deposition (EB-PVD) method, was studied by cyclic furnace testing and isothermal exposure. The scanning electron microscope, electron probe microanalysis, and microhardness indentation were used to probe the failure mechanism. It is found that due to the mismatch of the coefficient of thermal expansion, the as-deposited BC is suffered the long-term tensile creeping at room temperature. During the high-temperature exposure, the TBC locally rumples, bringing in-plane tensile stress at the shoulders, and out-of-plane tensile stress at the peak of the rumpled BC, where primal cracks are originated. During the cooling period, the ridges of substrate pulled by the local rumpling of the BC blocks the contracting of the BC, originating new cracks in planar BC, and aggravating the original cracks. Furthermore, the oxidation products pushed into the BC and the 8YSZ enlarges the TBC and cracks the substrate along the weakest diffused grain boundaries. The cracking failure related to the diffusion of the BC to the substrate is also discussed.

Processing and mechanical properties of porous 316L stainless steel for biomedical applications

Highly porous 316L stainless steel parts were produced by using a powder metallurgy process, which includes the selective laser sintering(SLS) and traditional sintering. Porous 316L stainless steel suitable for medical applications was successfully fabricated in the porosity range of 40%-50% (volume fraction) by controlling the SLS parameters and sintering behaviour. The porosity of the sintered compacts was investigated as a function of the SLS parameters and the furnace cycle. Compressive stress and elastic modulus of the 316L stainless steel material were determined. The compressive strength was found to be ranging from 21 to 32 MPa and corresponding elastic modulus ranging from 26 to 43 GPa. The present parts are promising for biomedical applications since the optimal porosity of implant materials for ingrowths of new-bone tissues is in the range of 20%-59% (volume fraction) and mechanical properties are matching with human bone.

Thermal Cycling Behaviour of Lanthanum Zirconate as EB‐PVD Thermal Barrier Coating

Thermal cycling tests with two different EB‐PVD thermal barrier coatings (TBC) were performed in a furnace cycle test. The results of these tests showed an increase of endurable cycle number when pyrochloric La2Zr2O7 was used as TBC. 1865 cycles were reached with La2Zr2O7 and 1380 cylces with 7 weigth‐% yttria stabilised zirconia (YSZ) EB‐PVD TBC. Additional investigation was made with scanning electron microscope (SEM) to investigate morphology and to determine chemical composition by electron dispersive x‐ray spectroscopy (EDS) analysis. X‐Ray diffraction was performed to analyze structural constitution of deposited coatings.

Monitoring Delamination Progression in Thermal Barrier Coatings by Mid‐Infrared Reflectance Imaging

Mid‐infrared (MIR) reflectance imaging is shown to be a reliable diagnostic tool for monitoring delamination progression in thermal barrier coatings (TBCs). MIR reflectance imaging utilizes the maximum transparency of TBCs in the 3–6 μm wavelength region to probe below‐surface delamination crack propagation that is typically hidden from visible wavelength inspection. The image contrast that identifies delamination progression arises from the increased reflectance produced by a large component of total internal reflection at the TBC/buried‐crack interface. Imaging was performed at a wavelength of 4 μm to take advantage of the relatively high transmittance of plasma‐sprayed 8 wt% yttria‐stabilized zirconia (8YSZ) TBCs along with a desirable relative insensitivity to potentially interfering absorptions by atmospheric constituents at that wavelength. A key advantage of MIR reflectance imaging over competing techniques is that it is sensitive to delamination progression even at very early stages before delamination cracks start linking together; therefore, TBC health assessment can be achieved throughout the life of the TBC well before TBC failure is imminent. Examples are presented to demonstrate monitoring delamination progression by MIR reflectance imaging in 8YSZ TBC‐coated specimens subjected to furnace cycling to 1163°C. The experimental results were in good agreement with reflectance values predicted by a four‐flux Kulbelka–Munk approximation applied to the extreme cases of a completely adherent and a completely detached TBC. Practical considerations, including potential interfering effects from surface contamination, sintering, and erosion are discussed.

Effect of thermally grown oxide (TGO) microstructure on the durability of TBCs with PtNiAl diffusion bond coats

The role of pre-oxidation surface treatments on the oxide microstructure and the failure mechanism of multi-layer thermal barrier systems based on Pt-modified NiAl bond coats and electron beam deposited thermal barrier coatings (TBCs) have been studied. The primary pre-oxidation experimental variable was the partial pressure of oxygen in the pre-oxidizing atmosphere at constant temperature and bond coat composition. The durability of TBCs deposited on surfaces following different pre-oxidation treatments were measured and compared using furnace cycling tests. The oxide layers corresponding to different levels of TBC performance were characterized microstructurally, chemically, and compositionally using scanning electron microscopy (SEM), transmission electron microscopy (TEM), and X-ray photoelectron spectroscopy (XPS) techniques. TBC performance was enhanced by the formation of a surface oxide having a coarse-grained columnar structure during the pre-oxidation process. Increased TBC durability was consistent with a slower oxide growth rate during exposure of the TBC to high-temperature, cyclic conditions, as was observed for this particular pre-oxidation condition. An oxide microstructure having fewer through-thickness transport pathways (grain boundaries) should also result in slower lateral oxide growth rates, consistent with a slowed rate of ratcheting as was observed in the pre-oxidized samples that had the best TBC performance. The desired surface oxide grain structure was achieved by pre-oxidizing the bond coat prior to TBC deposition at an intermediate partial pressure of oxygen.

On the failure mechanisms of thermal barrier coatings with diffusion aluminide bond coatings

The mechanisms governing the failure of multi-layer thermal barrier systems based on Pt-modified nickel aluminide bond coats and electron beam deposited thermal barrier coatings (TBCs) have been studied. The primary experimental variable is the morphology of the bond coat surface prior to application of the TBC, at constant multi-layer chemistry. The durability of these systems in a furnace cycle test has been measured and compared. The failure mechanisms, as well as the thickening of the thermally grown oxide (TGO), have been characterized for each of the surface morphologies. The major findings are that the durability is enhanced by removing imperfections on the surface of the bond coat, as well as by surface pre-treatments that diminish subsequent TGO thickening and by incorporating a reactive-element in the substrate that strengthens the bond coat upon inter-diffusion during manufacture. These effects are consistent with the expectations of a TGO instability mechanism, driven by a combination of growth and thermal expansion misfit strains in the TGO. The grain structure of the bond coat also affects failure through its influence on the TGO instability sites.

Damage evolution in an electron beam physical vapor deposited thermal barrier coating as a function of cycle temperature and time

Failure of thermal barrier coatings (TBCs) deposited on a single-crystal superalloy with a grit-blasted platinum modified nickel aluminide [β-(Ni, Pt) Al] bond coat has been studied as a function of thermal cycling temperature and time. One-hour cyclic furnace tests were conducted at 1100 °C, 1121 °C and 1151 °C, and 24-h tests were run at 1121 °C. It was found that all the samples tested in the 1-h cycle failed in the TBC, near the TBC/TGO interface, due to progressive cracking beginning at ∼20% life fraction. In contrast, the 24-h cyclic test samples failed at the TGO/bond coat interface. Thus, a life prediction for this TBC will ultimately require the use of two independent damage mechanisms and failure will be predicted on the basis of whichever occurs first during the TBC cyclic life. A single-valued relation was found between the rumpling amplitudes and the oxide thickness, independent of temperature and cycle time, consistent with oxidation being rate controlling.