High Temperature Corrosion studies on Pulsed Current Gas Tungsten Arc Welded Alloy C-276 in Molten Salt Environment

Alloy C-276 is widely used in the power plant environment due to high strength and corrosion in highly aggressive environment. The investigation on high- temperature corrosion resistance of the alloy C-276 PCGTA weldment is necessary for prolonged service lifetime of the components used in corrosive environments. Investigation has been carried out on Pulsed Current Gas Tungsten Arc Welding by autogenous and different filler wires (ERNiCrMo-3 and ERNiCrMo-4) under molten state of K2SO4-60% NaCl environment at 675oC under cyclic condition. Thermogravimetric technique was used to establish the kinetics of corrosion. Weight gained in the molten salt reveals a steady-state parabolic rate law while the kinetics with salt deposits displays multi-stage growth rates. PCGTA ERNiCrMo-3 shows the higher parabolic constant compared to others. The scale formed on the weldment samples upon hot corrosion was characterized by using X-ray diffraction, SEM and EDAX analysis to understand the degradation mechanisms. From the results of the experiment the major phases are identified as Cr2O3, Fe2O3, and NiCr2O4. The result showed that weld fabricated by ERNiCrMo-3 found to be more prone to degradation than base metal and ERNiCrMo-4 filler wire due to higher segregation of alloying element of Mo and W in the weldment


Welding of Alloy C-276
The material selected for the study was alloy C-276 which was procured in the rolled sheet form of 4mm thickness. Welding was carried out in Pulsed Current Gas Tungsten Arc Welding (PCGTA) mode with and without filler wire. The chemical composition for base metal and filler wire (ERNiCrMo-3 and ERNiCrMo-4) are reported in Table 1. The various welding parameter employed for this studies are in Table 2. Table 1 Chemical composition of base metal and filler wires.

Sample preparation for hot corrosion test
Composite (30 mm × 10 mm × 4 mm) employed for the hot corrosion test were sliced from the welded plates by keeping weld as center using wire cut electrical discharged machining. The composite region consists of Weld Zone (WZ), Heat Affected Zone (HAZ) and Base Metal (BM). The coupon used for corrosion test were mirror polished down to 1 μm by SiC paper and alumina powder before the corrosion test. The hot corrosion studies were conducted for autogenous and filler wire case and the results are compared.

Molten salt corrosion test
The experiment was conducted for 50 cycles of which each cycle consists of 1 h heating at 675 o C in the silicon carbide tube furnace followed by 20 min of cooling in air. simulating to create a real case for hot corrosion testing. During the corrosion tests, the weight change measurements were taken at the end of each cycle using an electronic weighing balance machine, which has a sensitivity of 1 mg. The spalled scale was also retained during the measurement of the weight change to determine the total rate of corrosion. Efforts were made to formulate the kinetics of corrosion and the same is presented in this paper. The samples after corrosion tests were subjected to characterization studies by SEM/EDAX, XRD for a surface of the scale. Figure 1(a-d) show the macrograph of hot corrosion samples (i) as salt coted condition (ii) 5 th cycle (iii) 25 th cycle (iv) 50 th cycle of autogenous, ERNiCrMo-3 and ERNiCrMo-4 PCGTA weldments. During cyclic hot corrosion study after a few cycles the color of the weldment in all the cases (autogenous, ERNiCrMo-3 and ERNiCrMo-4) was turned to be black color from the white color salt coated. Yellow color small spots are found in the autogenous weldment ( Fig. 1b (ii)) after a few cycle and remained same for the end of 50 th cycles.

Visual Examination and Thermo-gravimetric analysis
After 18 th cycle the color of the scale become thicker and start to spalled out with increase in further cycles and color also changed to blue color and remain same for the end of cycles. Similarly, in ERNiCrMo-4 the color become blue after 10 th cycles and also the thicker scale starts to form after the 10 th cycle and the scale starts to spalled out further increasing the cycles. The scale formed on ERNiCrMo-3 weldment is very high compared to the autogenous and ERNiCrMo-4 and scale spalled also extensively high from the 20 th cycle. The comparative thermo-gravimetric analysis results are shown in Fig. 2. The weight gained in ERNiCrMo-3 is found to be very high as compared to the ERNiCrMo-4 and autogenous weldments. Very negligible weight gain found to be in base metal as compare to the other weldments. Figure 3 shows the parabolic rate constants, Kp for base metal, autogenous weld, ERNiCrMo-3 and ERNiCrMo-4 weld are 0.02, 0.05, 0.17 and 0.16 mg 2 /cm 4 /h respectively. Figure 4 shows the XRD analysis of weldments. XRD analysis was carried out for all the specimens at the end of the 50 th corrosion cycles. The major phases formed in the weldments are Cr2O3, NiCr2O4, Nio, Fe2O3, FeS and MnO2.

SEM/EDAX
The SEM/EDS analysis of autogenous welded zone as well as weld interface indicate the presence of continuous and smooth scale crystals dispersed throughout their scales (Fig 5). In the both cases, the scale consists mainly of NiO and contains small quantities of oxides of Fe and Cr. SEM micrograph along with EDAX measurements at some selected points for the oxide scale of ERNiCrMo-3 weldment exposed to hot corrosion in the molten salt at 675 0 C after 50 cycles is depicted in Fig. 6. However, the SEM micrograph of scale on weld reveals a uniform fine grained crystals consisting mainly of NiO, the scale on weld interface consists of distorted porous and non-protective granules with comparatively higher amount oxides of Cr, Mo and Fe. The presence of small quantities of these oxides in the top scale indicates the diffusion of these elements from the substrate.
SEM micrographs for ERNiCrMo-4 weld and weld interface after exposure to K2SO4-60%NaCl environment at 675 o C are shown in Fig. 7 along with EDS results. SEM micrographs indicate nonprotective nature of scale in the weld as well as weld interface. The main constituent of the scales for these two cases is NiO as can be seen from the EDS results at certain points in the respective cases. The white crystalline phase in the weld is rich in NiO with small oxide quantities of Fe, Cr and Mo. A massive and dense scale has been observed for weld interface which is rich in NiO (black matrix).

Discussion
Thermogravimetric data has been compiled in Fig. 2 in the form of a plot between weight gain per unit area expressed in mg/cm 2 and number of cycles for hot corroded base metal and all the weldments. Further, the weight gains square (mg 2 /cm 4 ) data were plotted as a function of time (number of cycles) as shown in Fig. 3 to establish the rate law for the hot corrosion. It is clear that in spite of some fluctuations in the data, the data conforms to the parabolic rate law to an acceptable limit for all the weldments and base metal. The curve showing that the initial rate of weight gained is higher in all the cases which slowly decrease.
It has been observed that the rate of weight gain was relatively high in all the cases during the initial period of exposure, which may partially be attributed to the formation of some catalyst which serves as oxygen carrier to the specimen, therefore it will lead to rapid oxidation of the basic elements of the alloy C-276 to form protective oxide scales. Simultaneously, the protective scale is destroyed or eliminated by molten salts and consequently the metal surface is exposed to direct action of aggressive environment as has been indicated in the current investigation also. Arivazhagan et al. [22] as well as Devendranath et al. [27] have suggested simultaneous growth of oxides and their dissolution in molten salt in their study. After a period of high corrosion rate it has been observed that the rate of corrosion tends to be almost uniform with the further progress of study. This can be ascribed to the slower growth rate of Ni and Cr scale. Furthermore, the formation of NiCr2O4 in the scales with the progress of study might have also contributed to slower oxidation rate as has been suggested by Arivazhagan et al. [22].
The parabolic rate constant of ERNiCrMo-3 shows higher compared to the other weldments. It is also well evident from the thermogravimetric analysis ERNiCrMo-3 obtained the higher weight gained compared to the other weldments. As reported in the Introduction section the proper selection of welding process and parameter may reduce the microsegregation. The same authors [28][29] also reported that PCGTA ERNiCrMo-3 shows the higher segregation compared to autogenous and ERNiCrMo-4 PCGTA weldment. The higher segregation is caused by large difference between chemical compositions of base metal and filler metal. Alloy C-276 has the nominal chemical composition Ni-16Cr-16Mo-5Fe-4W, whereas nominal chemical composition of ERNiCrMo-3 is Ni-22Cr-10Mo-1Fe. The variation in the chemical composition between the base metal and filler wire (ERNiCrMo-3) leads to higher segregation with higher corrosion compared to other weldments. It is observed from the present study weight gain autogenous weldment has proved to be the most corrosion resistant in the aggressive environment under study. It is also noted that the autogenous weldment has slightly less corrosion resistance as compared to base metal in the same environment. X-ray diffractograms of the scales for base metal and all the weldments are given in Fig. 5. XRD analysis shows the major predominant phases inferred from all the three weldments are Cr2O3, NiCr2O4, Nio, Fe2O3, FeS and MnO2. The presence of spinel NiCr2O4 has also been confirmed in all the specimens and these X-ray diffraction results are well supported by SEM/EDAX results. The studies conducted by Arivazhagan et al. [22] and Devendranath Ramkumar et al. [27] on similar studies also endorse formation of identical phases.
It is evident from the diffraction patterns that autogenous, ERNiCrMo-3 and ERNiCrMo-4 have shown formation of similar phases after hot corrosion in the given environment. NiO and NiCr2O4 are indicated as the main phases in all the weldments. The protection of weldments and base metal from hot corrosion  The weldment employed by under alloyed filler has indicated accelerated corrosion in K2SO4-60%NaCl environment at 675 ºC in comparison to that of autogenous weldment and base metal. On the basis of cumulative weight gain data for 50 cycles, corrosion rate of the specimen under study can be arranged in the following order: The superior corrosion resistance shown by autogenous weldment as well as base C-276 alloy might be ascribed to the formation of NiO and Cr2O3. Whereas ERNiCrMo-3 weldment underwent extensive spalling and sputtering of its scales. In addition to the above mentioned mechanism, the accelerated hot corrosion of this weldment might have also been occurred due to the presence of higher amount of molybdenum as suggested by Peters et al. [30]. This tendency may be attributed to the different values of thermal expansion coefficients of the oxide scale on different zones of the weldment as reported by Arivazhagan et al. [22]. However, as the environment of study in this case contains molten salt, therefore the elements such as Na, Cl and S and oxygen may diffuse along these splats or pores, instead of continuous scale on autogenous and base metal.
EDS analysis for ERNiCrMo-3 and ERNiCrMo-4 weldment indicates that sulphur has penetrated into the substrate alloys and there it is co-existing with Fe and might form FeS. The oxidation of this FeS might yield a protective scale of NiO and Cr2O3 as per the views of Arivazhagan et al. [22]. However, the absence of metallic chlorides in the scale, there is traces of Cl in the EDS result might be evaporate die to low melting eutectic. Once sulphide is formed, the diffusion of Ni and Fe through sulphide is faster than in oxide resulting in accelerated attack. The formation of non-protective oxides of Ni and Fe is the cause for introduction of large amounts of sulphur from potassium sulphate into the alloy, which essentially leads to the degradation. The autogenous weldment have been found uniform scaling behavior across the microstructural gradients in weldments when exposed to molten salt environments, whereas the scale forming on the under alloyed weldment was not uniform.
In all the three cases the formation of FeO is found to be absent in the oxide scale irrespective of the test temperature as suggested by Moretimer and Sharp [31]. The absent of FeO is due to the higher concentration of Cr content. From SEM the major concentrated element present is Ni in all the cases. Ni undergoes oxidation preference to alloy C-276. The next major concentrated element to Ni is Cr. The Cr act to protect the surface layer as CrO, CrO2, Cr3O. Furthermore, the FeS revealed in the Autogenous PCGTA weldment, which indicates the existence of both oxidation and sulphidation in surface and subsurface reaction. In other cases, the FeS is found to be very small peak observed in XRD, the scale may spall out due to low melting eutectic.
The hot corrosion test was performed for 50 cycles, it is found that the ERNiCrMo-3 weldment was observed as more corrosion than compare to other weldments, which also can be seen in the large degree of scaling and spattering on the weldment (ERNiCrMo-3). It is evident that the thickness of the scale is more in ERNiCrMo-3 than the other two weldments after being exposed to 50 cycles in K2SO4-60% NaCl.

Conclusion
From the study the major conclusions obtained are listed as follows.
1. The weight gain of salt coated welded specimen follows a parabolic rate law during hot corrosion 2. Major phases are identified as Cr2O3 and NiCr2O4 on scale over weldment may due to presence of higher amount of Ni and Cr in the weldment and base metal 3. The corrosion rates for the investigated friction welded dissimilar metals based on the overall weight gains after 50 cycles in all the environments could be arranged in the following order: ERNiCrMo-3 > ERNiCrMo-4 > Base metal. 4. PCGTA fabricated by ERNiCrMo-3 shows higher corrosion compared to the other weldment. The higher segregation leads to more corrosion in the weldment compared to base metal and ERNiCrMo-4