SCT-19679; No of Pages 9 Surface & Coatings Technology xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat
Microstructural characteristics of the built up layer of a precipitation hardened nickel based superalloy by electrospark deposition M. Ebrahimnia a, F. Malek Ghaini a,⁎, Y.J. Xie b, H. Shahverdi a a b
Department of Materials Science and Engineering, Tarbiat Modares University, P.O. Box 14115-111, Tehran, Iran State Key Laboratory for Corrosion and Protection, Institute of Metal Research, Chinese Academy of Sciences, Shenyang 110016, China
a r t i c l e
i n f o
Article history: Received 24 May 2014 Accepted in revised form 19 August 2014 Available online xxxx Keywords: Nickel Crack Electrospark deposition Inconel EBSD
a b s t r a c t Buildup of precipitation hardened nickel base superalloys by electro spark deposition due to the low heat input of the process has many attractions. Characterization of the microstructure of the ESD built up layer of IN738LC over an as-cast base metal is accomplished in this work. The grain structure and solidification texture of the coating are investigated by orientation imaging microscopy (OIM), optical and scanning electron microscopy. It is shown that the deposited layer is formed mainly through epitaxial nucleation and growth on the base metal structure while discontinuities acting as nucleation sites produce fine grains with independent orientations. It is shown that such independent grains can have a significant role in improving the integrity of the ESD built up layer, since they can act as crack arrest sites and make the coating more resistant to the propagation of liquation and solidification fissures. Moreover, it is found that nanosized γ′ precipitates exist in the coating indicating the high tendency of γ′ for precipitation even in the extremely high cooling rates involved in the ESD process. Hardness measurements indicated a higher hardness for the built up layer which is attributable to the finer microstructure of the coating. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Electrospark deposition has many applications in rebuilding and coating conductive materials. Because of its very low heat input, it is ideal for rebuilding metals which are susceptible to heat affected zone (HAZ) cracking like nickel based superalloys [1–3]. Nickel based superalloys e.g. IN738LC have many applications in aerospace industries due to their excellent high temperature mechanical properties [4]. An exceptional combination of high-temperature strength and oxidation resistance is exhibited by the IN738LC, which has a γ′ precipitationstrengthened nickel-based superalloy. It has been widely used in today's heavy-duty gas turbines for hot gas path components such as blades, vanes and heat shields, as previously reported by Hays [5]. IN738LC can encounter several obstacles when being welded with conventional fusion welding processes e.g. laser and GTAW [6,7]. Liquation cracks and solidification cracks are two main defects when welding or cladding this alloy with a filler metal with the same composition as the base metal. Liquation of γ′ particles, γ-γ′ eutectic, MC carbides and Ni–Zr–B constituents in grain boundaries of HAZ, are responsible for liquation cracks in IN738LC [8,9]. Although there were many attempts to
⁎ Corresponding author. Tel.: +98 21828844388; fax: +98 2182884390. E-mail addresses:
[email protected] (M. Ebrahimnia),
[email protected] (F. Malek Ghaini),
[email protected] (Y.J. Xie),
[email protected] (H. Shahverdi).
eliminate the liquation cracks in this alloy, few have been found to be effective regarding this issue [10–12]. Electrospark deposition of NiCoCrAlYTa alloy on directionally solidified nickel-based superalloy with 8 mm thickness shows high potential of using Ni base filler alloy for rebuilding industrial components [13]. However, because of hot cracking susceptibility of precipitation hardened alloys, most filler metals for electrospark deposition were designed with lower mechanical properties than those of the base metal. In a recent study by the author it was shown that IN738LC can be built up by electro spark deposition (ESD) process using a similar filler metal and then subjected to pulsed laser fusion processing to improve the integrity of the ESD deposited layer [1]. It was found that the ESD material has more resistance to liquation and solidification cracking than the cast base metal. ESD deposits consist of individual splats which are connected together by metallurgical bonding due to the partial melting of electrode and substrate. In this process, a round conductive electrode is electrically charged (positive pole) and is rotated closely relative to a stationary conductive substrate (negative pole), resulting in intimate contacts and discharges. Because of short duration high-current electrical pulses, sparking occurs between electrode and substrate. As a result of sparking, multiple small molten splats form and get detached from the electrode, then contact the surface and deposit on top of each other [14]. The mechanism of coating in this process has been the subject of a number of previous researches [15–21]. The objective of the current investigation is characterizing the microstructure of the ESD built up layer IN738LC over an as-cast
http://dx.doi.org/10.1016/j.surfcoat.2014.08.045 0257-8972/© 2014 Elsevier B.V. All rights reserved.
Please cite this article as: M. Ebrahimnia, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.08.045
2
M. Ebrahimnia et al. / Surface & Coatings Technology xxx (2014) xxx–xxx
component in order to establish the metallurgical features that can contribute to its improved resistance to liquation and solidification cracking.
300 g and 15 s dwell time. TESCAN electron microscope and Carl Zeiss field emission electron microscope were used for microstructural study. 3. Results and discussion
2. Materials and methods As-cast Inconel 738LC was used with the following chemical composition (wt.%): C 0.10, Cr 15.73, Co 8.38 W 3.02, Mo 2.16, Nb 0.70, Fe 0.12, Al 3.4, Ti 3.45, Ta 1.80, Zr 0.04, B 0.01, Ni balanced. Electrodes were cut from the same casting in the form of a round pillar with 4 mm diameter and 5 cm in length by electrical discharge machining (EDM)-wire cut machine. Electro spark deposition of IN738LC was accomplished using an ESD machine developed at Tarbiat Modares University. The base metal was 100 × 50 × 5 mm3 rectangular plates machined from the as-received cast billets. In order to help the operator locating the rotating electrode on the target, a rectangle 10 mm × 10 mm with 0.5 mm depth was machined out from the base plate. ESD deposition was performed using a hand held gun with a co-axis argon shield gas with a flow rate of 15 L/min. Initially a number of tests were performed to establish a suitable ESD process window. The selected process parameters were: electrode rotation speed 2500 rpm, voltage 100 V, pulse frequency 250 Hz, and duty cycle 2.4%. A number of samples were taken for metallographic characterization of the as deposited electrospark buildup layer of Inconel 738LC. For electron microscopy characterization, the specimens were etched electrolytically in 5% oxalic acid solution at 6 V for 5 s. For electroetching of the IN738LC metallographic samples, they were connected to the positive pole and a piece of stainless steel sheet was connected to the negative pole of a direct current power source. An EPMA-1610 (Shimadzu, Japan) electron probe microanalyzer (EPMA) was used to analyze the chemical composition of ESD coating. The EPMA was operated at an accelerating voltage of 15 kV and a beam size of 1 μm and beam current of 10 nA for optimal spatial resolution for chemical composition analysis of deposited layer. Elemental map was acquired according to the elemental composition at an area of 15 μm × 15 μm with time of 2 h. The orientation imaging microscopy (OIM) technique based on electron backscattered diffraction (EBSD) was used as an efficient method to study complicated microstructures formed during solidification after ESD processing. For observing the grain structure of ESD layer and its association with base metal grain structure, EBSD analysis was used. For this purpose, the samples were subjected to the standard metallographic preparation procedure starting with grinding on SiC grit papers (up to 2000), followed by polishing in diamond particle suspension (3 and 2.5 and 1 lm size) and then electropolishing in (HNO3 30% + 70% ethanol vol.%) solution At 25 V for 3 s. This procedure resulted in good EBSD signal from the electrospark deposited layer and the substrate. TSL backscatter diffraction system installed inside a Philips XL 30 scanning electron microscope was used to collect OIM data. EBSD data and grain orientation measurements were obtained by TSL OIM Collection 5 and analyzed by TSL 5.31 OIM analysis software. The samples were mounted in the scanning electron microscope such that their surface normal was at 70° to the electron beam direction. The orientation of the crystal lattice at each predetermined sample surface point was then automatically determined in a sample coordinate system. Typical OIM scan in this experiment consists of 118745 points. Grain boundary lines presented in this work were constructed electronically using the criterion that a grain boundary between two points exists when the crystal misorientation angle between these two points exceeds 5°. Scanning was done with 1 μm step size. Vickers microhardness measurement was carried out using a Leco LM 247AT microhardness machine. A hardness profile was carried out on the ESD layer from top of the layer toward the base metal/ESD layer interface. The load of microhardness measurement was set at
Fig. 1-a shows an optical image of the ESD coating deposited on the IN738LC base metal (BM). ESD zone (ESDZ) has a thickness about 500 μm and consists of many tiny splats which were deposited subsequently to form a thick coating. Fig. 1-b shows the interface area with higher magnification. The different solidification products in the as cast base metal are labeled to help show different elemental segregations between as cast and ESD coating. Arrows show different constituents in the base metal microstructure mainly in the grain boundaries of the base metal like γ-γ′eutectic, MC carbides and terminal solidification products adjacent to the ESD zone which are the results of heavy elemental segregation in the as-cast base metal. It is worth mentioning that the cooling rate in the ESD process is about 105 K s− 1 which is much higher than that in the conventional fusion welding processes [1]. Because of the extreme cooling rate involved, there is no heavy elemental segregation in this process and this results in the prevention of formation of grain boundary terminal solidification constituents, which are the sources of liquation cracking in the cast alloy [1]. Electron probe microanalysis (EPMA) results also confirmed that there is very little elemental segregation in the ESD layer. Fig. 2 shows
Fig. 1. Microstructure of ESD zone (ESDZ) and base metal (BM). a) 500 μm thickness ESD coating on the IN738LC base metal, b) arrows show different constituents in base metal microstructure mainly in the grain boundaries of base metal like γ-γ′eutectic, MC carbides and terminal solidification products adjacent to the ESDZ.
Please cite this article as: M. Ebrahimnia, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.08.045
M. Ebrahimnia et al. / Surface & Coatings Technology xxx (2014) xxx–xxx
3
Fig. 2. EPMA analysis of ESD coating involving some splats and grain boundaries, different maps show approximately uniform elemental distribution in ESD coating (e.g.: elements Al, Ti, C, Cr, Nb.
different elemental maps of a small area of the ESD layer. The scanned area involves some splats which are deposited on top of each other and different shaded vertical lines are identified as grain boundaries. These maps show approximately uniform elemental distribution in the ESD layer except at some imperfections like inclusions, porosities and splat interfaces. The element map of Al and Ti in the ESD zone show uniform elemental distribution in the scanned area. However, EPMA analyses in the laser fusion zone of IN738LC do indicate a selective partitioning of various elements between dendrite cores and interdendritic regions [22]. It is seen that in the laser fusion zone elements Co, Cr, W, and Ni segregated into the dendrite core, while the interdendritic regions are enriched by Ti, Nb, Ta, Zr, and Mo [22]. Fig. 3 shows a single splat of molten metal which detached from the electrode and transferred to the substrate. When a single droplet impacts a surface, the phenomena of fingering (a quasi-periodic
unevenness of the boundary of the splat) and splashing (a breakingoff of such fingers from the splat into the air or along the surface) occur [23] (see Fig. 3). In Fig. 3-a, “A” shows some splashes and “B” shows the fingering marks. The fingers which break off during splashing are referred to as secondary droplets. Such phenomena occur when multiple droplets impact successively on a surface as well, though this case is largely unexplored. The dimensions of the splats are not the same, in Fig. 3 they have an estimated diameter of 400 μm and about 20 μm thickness on average. The outward splash marks around the splat periphery can be regarded as evidence of the speed by which the molten droplets detached from the electrode have impacted the substrate. The molten droplet gains speed as it travels through the magnetic field produced by the spark discharge. The impact force forms a hole at the center of
Please cite this article as: M. Ebrahimnia, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.08.045
4
M. Ebrahimnia et al. / Surface & Coatings Technology xxx (2014) xxx–xxx
Fig. 3. Structure of one splat and ESD coating result of many deposited splats. a) A single impact of ESD process droplet showing splash “A” and fingering “B” marks on the substrate as seen from the top, b) Microstructure of ESD layer from top view showing sequence of deposition of ESD splats in the same plane of view.
splat and pushes the molten metal to the periphery of the hole, so a valley occurs as a result of this phenomenon. As can be seen in Fig. 3-a, the splats do not have a uniform thickness. Because of the variation in thickness of the splats, the cooling rate would vary to some extent at different points. The volume of molten metal in each splat is very small compared to the mass of the cool substrate and that leads to extreme cooling rates. The molten splat freezes instantly without having time for the molten metal waveform produced as the result of impact to damp out and dissipate to form a smooth surface, resulting in an irregular shape as shown in Fig. 3-a. One can imagine the final ESD layer to form by deposition of many splats like the single splat shown in Fig. 3-a (see Fig. 1-a). Buildup of the ESD coating by multiple impacts of splats as seen from the top is shown in Fig. 3-b. Because of the irregular shape of every splat which deposits on top of each other, one can see different circular shapes which are at the same plane of view. In this figure, although there are different splats from number 1 to 5 with possibly different times of deposition, all of them can be seen in one plane of top view section. It means splat 1 which is deposited after splat 2 to 5 has boundaries with these splats. In this regard, the microstructure of splat 1 can be affected by different previous splat microstructures. It is worth mentioning that some of these circular shapes in Fig. 3-b can be some splashes of large splats which were deposited in irregular shapes.
Fig. 4-a shows the ESD microstructure in a cross section of the coating perpendicular to the ESD growth direction. This figure is actually a scanning electron microscope image of Fig. 3-b with high magnification. The dashed lines indicate the boundaries of 3 different droplets. In this figure at the surface of one splat, irregular boundaries which are deeply etched can be observed. Arrows indicate some of these boundaries. Fig. 4-b shows these boundaries with higher magnification. Moreover, cross sections of fine γ phase cellules with a diameter range of 300–500 nm can be identified in this micrograph. Because of the very fine cellular microstructure which is less than 1 μm, EPMA elemental maps do not show any patterned segregation in the ESD zone (Fig. 2). Therefore, it is very hard to investigate elemental segregation in cellular core and intercellular region. However, as mentioned before, elemental maps confirmed approximately uniform elemental distribution within ESD coating. The microstructure of the ESD layer consists of many splats. If droplets (about 300 μm in diameter) are significantly smaller than the substrate cast grain size (about 4 mm width in cross section), many neighboring splats have deposited on the same grain of the substrate. Therefore, the ESD zone above one grain of base metal has mostly one crystallographic direction and unlike laser fusion zone, there would be less chances for the formation of high angular grain boundaries in the ESD zone. However Fig. 4a–b shows boundaries which seem to be grain boundaries. Fig. 4-c shows the transverse cross section of the ESD coating and the molten splat interfaces. As a result of irregular thickness and position of splats above each other, one can see irregular boundaries and different splat thicknesses in this figure. Moreover, there is complete metallurgical bonding between different splats as a result of very shallow dilution between them. The very fine cellular microstructure of every splat as a result of the very high cooling rate involved in this process can be seen in this image. The authors have calculated the cooling rate of this process elsewhere [1]. Solidification of the γ phase will start at the interface with the substrate or previous splat or from the peripheral areas and then proceed outward in the shape of very fine cellules. Because splats are deposited on top of each other, the growth direction of γ phase is mostly upward. This can be seen in Fig. 4-c and with higher magnification in Fig. 4-d. Considering Fig. 4-c, one can observe some boundaries between cellular bundles which are etched deeply (arrows show these boundaries). Theses boundaries are the same boundaries which were observed in Fig. 4a–b. The comparison of cellular direction at opposite sides of these boundaries reveals the slightly different orientation of cellular growth direction. These small variations in direction of cellular microstructure in one splat result in the subsequent directionality of the next splats which get in contact with it. One can see these boundaries which propagate in other splats in Fig. 4-c. As a result of this solidification mechanism, although highly epitaxial cellular microstructure is observed in ESD coatings, the boundaries between bundles of γ cellules are also observed in these coatings. However, for accurate identification of these boundaries and their relation with the base metal microstructure, it is needed to investigate the coating with electron backscattered diffraction (EBSD) analysis. Fig. 4-d shows higher magnification of ESD cellular structure with no other arm branches. The microstructure consists of γ rods. According to solidification microstructure definition these rods have a cellular structure. Then the microstructure is cellular. This figure clearly shows the microstructure of the ESD layer which is completely cellular without primary or secondary arms. This suggests higher GL/VL ratio for ESD deposits in comparison with laser weld which typically has dendritic microstructure. GL/VL is the critical controlling parameter which determines the solidification microstructure. GL is the thermal gradient in the liquid at the liquid–solid interface and VL is velocity of the liquid– solid interface or solidification rate. When GL/VL ≥ ΔT/DL the plane front is stable. If GL/VL falls below this critical value due to a drop in gradient and/or an increase in solidification velocity, instabilities
Please cite this article as: M. Ebrahimnia, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.08.045
M. Ebrahimnia et al. / Surface & Coatings Technology xxx (2014) xxx–xxx
5
Fig. 4. ESD coating microstructure in different cross sections. a) SEM image of splats forming coating from top view, the arrows show some irregular boundaries in ESD zone, b) higher magnification of irregular grain boundaries of different grains in ESD zone, c) transverse cross section of an ESD layer produced by subsequent impact and accumulation of splats deposited on top of each other, arrows show some boundaries between cellular bundles, d) higher magnification of ESDZ cellular structure.
develop and a progression of solidification morphologies will occur and constitutional undercooling occurs. ΔT is the solidification range of the alloy and DL is the diffusion coefficient of the solute in the liquid. With decreasing GL/VL, the microstructure tends to be cellular/dendritic and a further decrease in this ratio leads to the formation of dendritic solid/liquid interface. A number of investigations characterize the cellular/dendritic transition [24] involving non-steady-state heat flow conditions. Ii is well known that the improvement on the material's performance intimately depends on the thermal parameters imposed during solidification, which affects the morphology and the scale of the microstructural array and the segregation [24–26]. Unlike the laser welding process, the ESD process is a super low heat input welding process. So the microstructure of the ESD coating has unique features. Solidification microstructure as a result of extreme rapid cooling consists of epitaxial cellular γ phase [1]. In contrast to laser weld deposits, where the microstructure of IN738LC typically has a dendritic microstructure, the ESD material has a fine cellular microstructure [6,22,27]. Although the microstructure of the ESD layer is observed by electron microscopy and optical microscopy, the EBSD is the most important and the most accurate tool for observing and investigating the microstructure and grain structure of the ESD layer. As with conventional SEM and optical imaging techniques, EBSD maps can be used to convey visually the basic characteristics of the material's microstructure with 2D information about grain size and shape. Moreover, because the phase and orientation of each pixel in the map is also known, EBSD data processing software can generate an enormous variety of additional visual and analytical information, including overall preferred orientation (texture), prevalence and distribution of grains in specific orientations, phase distribution, state of strain and local variations in
residual strain, and character and distribution of grain boundaries [28,29]. Fig. 5-a, shows an inverse pole figure map of the ESD layer on the IN738LC base metal. The scanning area includes small parts of two base metal grains and the ESD layer deposited above them. The dashed lines show the interface between coating and substrate. G1 and G2 are two different grains from the base metal. The IPF image shows crystallographic growth direction. Base metal grains have green and yellow color which shows crystallographic directions between b101 N and b111 N directions and between b001N and b 101N directions aligned with the sample normal respectively. At the interface of the coating and base metal and some distance into the coating, the ESD coating and base metal have the same crystalline direction and there are no grain boundaries between them. However not far from the interface, slightly different growth directions can be identified and there are many grain boundaries appearing. Except the blue dots which are attributed to bad or no signal from porosities or interfaces, the grains in the ESD coating are mostly elongated in the b 001 N and b 101 N directions which are visible as pink-red and green-yellow in color respectively. The texture of directional solidification of the ESD coating can be observed in this figure. So the solidified ESD grains are highly elongated in the direction from the bottom to the top of the coating. EBSD analysis reveals directional solidification of ESD coating which can be of interest for nickel based superalloys. It is commonly known that directional solidification microstructures (DS) have higher creep rupture ductility and creep life than conventional equiaxed microstructures. The main characteristic of DS structures as opposed to the equiaxed ones found in conventional castings is that the primary dendrites and grain boundaries are aligned with the main heat transfer, or growth direction [30].
Please cite this article as: M. Ebrahimnia, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.08.045
6
M. Ebrahimnia et al. / Surface & Coatings Technology xxx (2014) xxx–xxx
Fig. 5. Visual EBSD results of ESD coating. a) Inverse pole figure map of ESD coating (G1 and G2 are two different grains of substrate), b) grain size distribution chart of ESD coating, c) misorientation angle distribution chart of ESD coating.
Fig. 5b shows grain size distribution of ESD coating. As it can be seen that 10–20 μm grains have the highest area fraction in the scanned area. For preventing the error in grain size measurement the grains at the edges of the scan were excluded. The misorientation angle distribution for the ESD grains is demonstrated in Fig. 1(b). It can be seen that the misorientation angle between the grains is distributed over a wide range of 6.5–63.5° with a mean value of 33°. A grain boundary map can be generated by comparing the orientation between each pair of neighboring points in an OIM scan. A line is drawn separating a pair of points if the difference in orientation between the points exceeds 5° known as the tolerance angle. The grain boundary misorientation distribution can be seen in the image quality map (IQ map) combined with the grain boundary map (Fig. 6-a). From this figure most grain boundary rotation angles lie between 30° and 60°. Fig. 6-b shows misorientation profile of line I which is shown in Fig. 6-a. This profile shows misorientation degree of 2 different grains G1 and G2. According to this profile two grains have about 45° misorientation. This misorientation also exists in adjacent ESD coating which grow at the top of these grains. Fig. 6-c is the misorientation along Line II in Fig. 6-a. This line indicates misorientation between base metal and coating at the interface. This profile shows very little misorientation between grain G1 and ESD coating above it which is less than 5° and are belong to one grain. One can say that a very fine grain structure of ESD coating forms from some areas in the coating. It is interesting to know how these misorientations form and lead to grain boundary formation. A close investigation of the ESD coating can reveal some useful information about this phenomenon. Although some areas with perfect matching growth direction can be seen in the ESD coating, as Fig. 5 indicates the entire coating does not have this single grain structure. Fig. 7 shows two different areas of ESD coatings. In this figure arrows show some discontinuity in ESD coating. It is obvious there is a discontinuity between subsequent
droplets microstructures at the imperfection place. The imperfection in coating structure which is a result of lack of fusion of previous surface by subsequent droplet, is a place for nucleation and growth of cellular γ phase with random orientation independent from previous droplet microstructure (arrows show these nucleation sites in Fig. 7). When a small molten droplet contacts the surface of the substrate, the cellular growth direction follows the substrate preferential crystallographic direction mainly [001]. If the next deposited droplet makes good bonding with the previous droplet, this growth direction will be maintained to form a single grain. However, if there are some discontinuities between droplets or between coating and substrate, the cellular structure will nucleate and grow randomly mainly following heat sink direction. So, many boundary shape lines can be observed in these areas and continue in 3D format to all the subsequent droplets and entire coating microstructure. Because ESD coating usually include some imperfections like voids and lack of fusion, one can say the imperfections could be the main sources of grain formation in the ESD coating. The substrate grain structure is another main origin for different orientations of ESD coating growth direction. One can say, there would be a good chance for formation of one grain in the ESD coating wherever continuous deposition of subsequent layers without discontinuity happens. This can be of interest for deposition on the single crystal substrate like single crystal nickel base superalloys. However, unlike most welding processes e.g. laser welding and gas tungsten arc welding, ESD coating consists of many tiny splats which may have different condition of fusion and solidification. Grain boundaries in the ESD coating form due to very little difference in the growth direction of cells in the ESD layers. This difference forms very fine grain boundaries in initial splats and with thickening ESD layer and formation of subsequent splats, these boundaries continue to the subsequent splats. Therefore, even with a single crystal substrate it seems to be very hard to have a single crystal ESD coating, though this area is not investigated yet.
Please cite this article as: M. Ebrahimnia, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.08.045
M. Ebrahimnia et al. / Surface & Coatings Technology xxx (2014) xxx–xxx
7
Fig. 6. IQ map combined with grain boundaries map of ESD coating. a) Image Quality map with grain boundary lines, b) misorientation profile I shows misorientation between G1 and G2, c) misorientation profile II shows misorientation of base metal and coating along path II.
Microhardness measurement was used for evaluation of the ESD layer. Although there are some relations between hardness and material mechanical properties e.g. yield strength, ultimate strength and stressstrain flow curve in the literature [31], we consider that hardness value stands alone as a mechanical property of the ESD layer and compare it with base metal hardness. The hardness values of the ESD layer and the base metal are shown in Fig. 8. Hardness measurement was performed along two different profiles and trend lines in average hardness values in the ESD coating layer as a function of the distance with the interface with the base metal are shown as solid lines in Fig. 8. When using hardness profile
Fig. 7. Different discontinuities in ESD coating lead to new grain formation. Arrows show new grain formation from lack of fusion (LOF) between two droplets and grain boundary formation in ESD as result of base metal grain boundary (GB).
measurement it shows increases in hardness values moving from the base metal to the base metal/ESD coating interface toward the ESD coating. The average hardness of ESD coating was 390 HVN ± 8 using 300 g force. Measurement of base metal micro hardness shows lower hardness than ESD coating. The average hardness of the base metal was 335 HVN ± 7. However, the dendrite core and interdendrite region and also the grain boundaries in which carbides are present, have different microhardness values in base metal. One can say that the higher hardness of the ESD coating can be related to the finer microstructure of the coating. As verified with EBSD results, ESD deposition has very fine cellular microstructure and different cellular growth directions lead to formation of very fine grains or sub grain boundaries. These fine cellular and grain microstructure tend to increase hardness. It is appreciated that more slip systems are
Fig. 8. Microhardness measurements of the ESD layer and base metal (300 g force), the solid line shows the average values alongside upper and lower hardness values.
Please cite this article as: M. Ebrahimnia, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.08.045
8
M. Ebrahimnia et al. / Surface & Coatings Technology xxx (2014) xxx–xxx
usually operative near the grain boundaries. Therefore, the hardness usually will be higher near the boundary than in the center of a grain. As the grain diameter is reduced more of the effects of grain boundaries will be felt. Thus, the strain hardening of a fine grain size metal would be greater than that in a coarse-grain polycrystalline aggregate [31]. Montazery et al. investigated the relation between hardness and γ′ particle size [32]. According to their investigation as the γ′ size decreased the hardness increased subsequently. However, their measurement of hardness was accomplished at the micro sized γ′ precipitate substrate. Very fine γ′ particles can be sheared by dislocation and the very large particles can by passed by dislocations. So, there might be an optimum size of γ′ particles which leads to highest hardness. Some references imply higher volume fractions of γ′ with an optimum particle size of approximately 0.1 μm in diameter, lead to higher strength levels [33,34]. In this experiment, as-cast IN738LC has γ′ size of 230 nm and 500 nm in diameters at dendrite core and interdendritic regions respectively. It is interesting to investigate ESD coating for observation of possible γ′ precipitates. Although γ′ precipitation for restoring full mechanical properties needs standard aging heat treatment, some researchers have found very fine γ′ precipitates in weld zone of laser welding process which is considered to be a very low heat input welding process [10,35]. Moreover, Ojo et al., have characterized γ′ particles in TIG welding process of IN738LC [36]. The γ′ size in the range of 150– 200 nm have been found in TEM images of the interdentritic region and some smaller size in core dendrites in TIG weld region in this alloy. Because of lower heat input of ESD process than laser welding and TIG processes, weld metal cooling rate is higher than that in laser welding and TIG. Observation of the ESD layer using a scanning electron microscope shows some indications of very fine precipitates (Fig. 9). By using high resolution field emission electron microscope, these fine particles were assumed to be γ′ precipitations. The size of the γ′ particles was determined to be about 10–25 ± 2 nm in diameter. Aside from the fine directional grain structure of ESD coating, comparing the microstructure of ESD coating and as-cast IN738LC one can find high elemental segregation in the as-cast microstructure. These segregations can be seen as forms of different constituents like γ-γ′ eutectic, carbides and Zr–B compounds in as-cast base metal. Studying the ESD microstructure reveals less elemental segregation and no indication of large size carbides or eutectic compounds, which is desirable for eliminating liquation crack when welding or remelting these coatings with fusion welding processes like laser welding.
Fig. 9. The microstructure of cellular γ phase. Indication of γ′ precipitation in cellular microstructure of ESD layer.
4. Conclusion Electrospark deposition of IN738LC using matching electrode was performed. The deposited coating was investigated for understanding different features of IN738LC electrospark deposited coating. 1– Having epitaxial growth from substrate, ESD coating has mainly the same crystalline orientation with base material. This can lead to formation of deposited layer without producing new grain at the base metal/ESD coating interface. 2– Intrinsic imperfection and discontinuities of ESD coating e.g. lack of fusion and porosities can be nucleation sites of new grains of the ESD coating. This implies that ESD coating on single crystal material possibly would not be single crystal. 3– Although ESD coating consists of many individual splats which are deposited on top of each other, EBSD analysis shows that the entire ESD coating consists of very fine grains which grow directionally toward the surface of coating. These grains consist of very fine cellular γ phase with nanosized γ′ precipitates within them. 4– Rapid cooling rate involved in ESD process hinders heavy elemental segregation and results in a fine grain microstructure. Therefore, grain boundary terminal solidification constituents e.g. γ-γ′ eutectic and large carbides cannot be found in the microstructure of ESD coating. Moreover, very fine grains with a high amount of grain boundary area in ESD coating can be considered more resistant to liquation and solidification cracks. Acknowledgment This research was a collaborative effort of the Tarbiat Modares University in Tehran and the State Key Laboratory of Corrosion and Protection, Institute of Metal Research in China. The work is supported by the National Natural Science Foundation of China (Grant No. 50901081). References [1] M. Ebrahimnia, F. Malek Ghaini, H.R. Shahverdi, Sci. Technol. Weld. Join. 19 (2014) 25–29. [2] J. Durocher, N.L. Richards, J. Mater. Eng. Perform. 20 (7) (2011) 1294. [3] J. Durocher, N.L. Richards, J. Mater. Eng. Perform. 16 (6) (2007) 710. [4] J.N. DuPont, J.C. Lippold, S.D. Kiser, Welding Metallurgy and Weldability of Nickelbase Alloys, JohnWiley & Sons Inc., New Jersey, 2009. [5] C. Hays, J. Mater. Eng. Perform. 17 (2008) 254–259. [6] A.T. Egbewande, R.A. Buckson, O.A. Ojo, Mater. Charact. 61 (2010) 569. [7] S.M. Mousavizade, F. Malek Ghaini, M.J. Torkamany, et al., Scr. Mater. 60 (2009) 244. [8] O.A. Ojo, N.L. Richards, M.C. Chaturvedi, J. Mater. Sci. 39 (2004) 7401–7404. [9] A.T. Egbewande, H.R. Zhang, R.K. Sidhu, A. Ojo, Metall. Mater. Trans. A 40A (2009) 2694. [10] M. Zhong, H. Sun, W. Liu, et al., Scr. Mater. 53 (2005) 159. [11] M.F. Chiang, C. Chen, Mater. Chem. Phys. 11 (2009) 415–419. [12] T. Böllinghaus, H. Herold, Hot Cracking Phenomena in Welds, Springer-Verlag, Berlin Heidelberg, 2005. 4–6. [13] M.C. Wang, W.F. Wang, Y.J. Xie, J. Zhang, Trans. Nonferrous Met. Soc. China 20 (2010) 795–802. [14] C. Changjun, W. Maocai, L. Yiming, W. Dongsheng, J. Ren, J. Mater. Process. Technol. 198 (2008) 275–280. [15] S.K. Tang, T.C. Nguyen, Zhou, Weld. J. 89 (2010) 172-s. [16] S. Frangini, A. Masci, Surf. Coat. Technol. 204 (2010) 2613–2623. [17] C. Luo, S. Dong, X. Xiong, N. Zhou, Surf. Coat. Technol. 203 (2009) 3333–3337. [18] A.V. Ribalko, O. Sahin, K. Korkmaz, Surf. Coat. Technol. 203 (2009) 3509–3515. [19] I.V. Galinov, R.B. Luban, Surf. Coat. Technol. 79 (1996) 9–18. [20] J. Liu, R. Wang, Y. Qian, Surf. Coat. Technol. 200 (2005) 2433–2437. [21] Y.J. Xie, M.C. Wang, Surf. Coat. Technol. 201 (2006) 691–698. [22] R.K. Sidhu, O.A. Ojo, M.C. Chaturvedi, Metall. Mater. Trans. A 38A (2007) 858. [23] M. Xue, Y. Heichal, S. Chandra, J. Mostaghimi, J. Mater. Sci. 42 (2007) 9–18. [24] W.R. Osório, D.M. Rosa, L.C. Peixoto, A. Garcia, J. Power Sources 196 (2011) 6567–6572. [25] D.M. Rosa, J.E. Spinelli, W.R. Osorio, A. Garcia, J. Power Sources 162 (2006) 696–705. [26] L.C. Peixoto, W.R. Osório, A. Garcia, J. Power Sources 192 (2009) 724–729. [27] L. Li, J. Mater. Sci. 41 (2006) 7886–7893. [28] T. Maitland, S. Sitzman, in: W. Zhou, Z.L. Wang (Eds.), Scanning Microscopy for Nanotechnology, Techniques and Applications, Springer, 2007, p. 522. [29] A. Farnia, F. Malek Ghaini, K.V. Ocelı´, J.Th.M. De Hosson, J. Mater. Sci. 48 (2013) 2714–2723.
Please cite this article as: M. Ebrahimnia, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.08.045
M. Ebrahimnia et al. / Surface & Coatings Technology xxx (2014) xxx–xxx [30] M.J. Saari Henry, The processing of gas turbine engine hot section materials through directional solidification, (PhD Dissertation) Carleton University, 1999. [31] G.E. Dieter, Mechanical Metallurgy, McGraw-Hill Book Company, London, 1988. [32] M. Montazeri, F. Malek Ghaini, O.A. Ojo, Weld. J. 92 (2013) 258-s. [33] M.J. Donachie, S.J. Donachie, Superalloys: A Technical Guide, 2nd ed. ASM International, Ohio, 2002.
9
[34] J.J. Jackson, M.J. Donachie, M. Gell, R.J. Henricks, Metall. Trans. A 8A (1977) 1615–1620. [35] M. Gumann, Epitaxial laser metal forming of a single crystal superalloy, (Phd Dissertation) École Polytechnique Fédérale de Lausanne EPFL, 1999. [36] O.A. Ojo, N.L. Richards, M.C. Chaturvedi, Scripta Mater. 51 (2004) 683–688.
Please cite this article as: M. Ebrahimnia, et al., Surf. Coat. Technol. (2014), http://dx.doi.org/10.1016/j.surfcoat.2014.08.045