Arc fusion of self-fluxed nickel alloys

Arc Fusion of Self-Fluxed Nickel Alloys

Paulo Villani Marques

Dept. Mechanical Engineering. Federal University of Minas Gerais. Brazil

[e-mail]

Roseana da Exaltação Trevisan

Dept. of Manufacturing Engineering. State University of Campinas. Brazil

[e-mail]

Self-fluxed nickel alloys are usually flame fused after thermal spraying. However, due to the practical aspects of high temperatures reached during flame fusing, large structures such as the hydraulic turbines for power generation, can not be efficiently coated. An alternative is to fuse the sprayed coating with a gas tungsten electric arc. In this case, heating is much more intensive and substrate temperature during and after the fusing operation is much lower, thus reducing the possibility that any problem will occur. In this work, coatings of self-fluxed nickel alloy fused by flame and gas tungsten arc were evaluated as protection of hydraulic turbines against cavitational damage. Several tests were performed, including the ASTM ultrasonically vibration-induced cavitation, optical and scanning electronic microscopic metallography, and hardness tests. The results showed that the arc-fused coating presented better cavitation damage resistance, probably due to its finer microstructure. A field application of this new technique is also described. A self-fluxed Ni alloy was flame sprayed in critical regions of Francis-type hydraulic turbine blades and fused by a gas tungsten arc after spraying. The blades will be inspected during the next two years.

Keywords: Wear, cavitation damage, thermal spraying, coating, nickel alloy, hydraulic turbine

Introduction

Thermal sprayed coatings have been applied in several ways, specially for electrical conduction or resistance, protection against wear, restoration of damaged surfaces, heat and high temperature resistance and corrosion resistance (Matejka & Benko, 1989).

Self-fluxed nickel alloys, in form of powder or rod, have been used mainly for protection against wear and corrosion and must be fused after spraying. Fusion is intended to increase density and to allow oxides to separate from the coating. An adequate fusion technique may be used to obtain optimal characteristics for the coating. If the material is insufficiently heated, fusion and oxide elimination will be incomplete, resulting in poor bonding and excessive porosity. Overheating may produce shrink holes in melted phases, contraction and distortion, and eventual dilution of the substrate. It is important to heat the coating and finish the fusion fast. This prevents excessive fluxing action, which may cause losses of cooper and silicon and segregation of slag inside the coating.

The melting temperatures of self-fluxed nickel alloys are between 1000 and 1100° C (AWS, 1985), and the temperature of substrate near the coated surface during fusion is about 900° C (Marques, 1996). Fusion is usually done with one or more oxy-fuel torches. However, when applied to very thick parts, large regions will be heated due to the low efficiency of the power source. In addition, the elevated consumption of energy and the slow and extensive heating make it difficult for personnel to be present near the working area. It can also introduce distortions as a consequence of the lowering of yield strength of the substrate as the temperature increases, which can occur in post-welding heat treatments, for example.

Therefore, an adequate technique must be used to avoid problems due to overheating. An alternative is to use a more intense power source, as an electric arc, which allows more efficient heating with lower energy consumption. A properly controlled gas tungsten arc (GTA) may be a satisfactory choice.

In this work, self-fluxed nickel alloy coatings were deposited on a low carbon steel substrate and fused with an oxy-acetylene flame and with a GT arc. The coatings fused by these two different techniques were evaluated with respect to their microstructure and cavitation erosion resistance. The application of this new technique of fusion to Francis-type hydraulic turbines is also described.

Experimental Procedure

The cross section of carbon steel pieces with diameter 16 mm and a length of 15 mm were coated with a self-fluxed Ni alloy with the chemical composition shown in Table 1. Surface preparation and flame spraying were done as instructed by the suppliers of the equipment and materials, all of which are available commercially.

The coatings were fused with oxy-acetylene flame and gas tungsten arc. Recommendations by the alloy supplier, such as a reducing flame, were followed for conventional fusion using the flame. In the arc fusion, an electronically controlled DC power source was used. An AWS WTh2 electrode with a diameter of 1.2 mm was used to supply a 40-60 A straight polarity- current to an arc which was kept at a distance of 2 mm approximately. All the operations were performed manually. After fusion, the coatings were tested to evaluate their characteristics. The final thickness of a transverse section of the fused coatings was measured and results of 0.37 and 0.46 mm were obtained for flame- and arc-melted coatings, respectively. This difference may be attributed to a dilution effect (see Discussion).

Hardness was evaluated using Rockwell C 150 kg and Vickers 500 g scales, in agreement with the ASTM E-18 and E-384 standards. Adhesion was evaluated by a tensile test according to the ASTM C-633 standard. Cavitation resistance was estimated by an ultrasonically induced cavitation test, in conformity with the ASTM G-32 standard. Transverse sections of the coatings were examined by optical and scanning electron microscopy. Chemical composition was determined by energy dispersive spectroscopy (EDS) in an electronic microprobe and the coatings were submitted to x-ray diffraction analysis.

Results and Discussion

Tables 2 and 3 show the results of the Vickers and Rockwell hardness test measurements, respectively. As can be observed, the Vickers hardness values were approximately the same for both fused coatings. On the other hand, the Rockwell C hardness values for the flame-fused coating was lower than that for the arc-fused coating. This can be explained by the difference in thickness of the two types of coating. The hardness of the thinner coating is influenced by the hardness of the substrate (lower than the hardness of the coating). The magnitude of Rockwell C hardness of the arc-fused coating agrees with that specified by the powder manufacturer.

Table 4 shows results of the tensile adhesion test. These showed that adhesion of the coatings to the substrate is very good in both cases, with all the specimens tested failing through the adhesive layer. The bond material shows a performance beyond expectation, with a resistance of about 30 MPa.

Tables 5 and 6 show results of the chemical analysis. Chemical compositions determined for the coatings under the as-sprayed and flame-fused conditions were similar to those obtained for the powder. However, the chemical composition determined for the arc-fused coatings "suggests" some deterioration of the coating.

For the flame-fused coating, the substrate temperature measured by a K-type thermocouple welded to the sample was up to 900° C in a region located 5 mm below the coated surface. This temperature does not seem to be high enough to cause melting. According to the suppliers, the nominal melting range of the nickel alloy is about 1100 ° C and the melting temperature for the substrate is around 1500 ° C. So, even if some overheating occurs, no melting of the substrate should be expected during the coating fusion.

The coatings fused by the arc had a 20% reduction in Ni content, while Fe content increased 20% with respect to the powder's chemical composition. These results can not be attributed to experimental error in the chemical analysis. In fact, they suggest the melting of the substrate and occurrence of some sort of dilution. This hypothesis is supported by the fact that the temperature of the arc (~6000° C) is much higher than that of the flame (~3000 ° C). Also the results of the SEM and EDS analysis, presented and discussed later, reinforce the hypothesis.

The results of cavitation tests of flame- and arc-fused coatings represented as a cumulative mass loss versus time curve were very similar, as shown in Fig. 1. However, one can say that the performance of arc-fused coatings is slightly better than that for flame-fused coatings, because the curves of accumulated mass loss for this type of coating are below the curve for the flame-fused coating. This is probably due to the refined microstructure obtained with the arc-fused layer, as shown in Figs. 2 and 3. In the Figs can be seem basically a two-phase microstructure with a dendritic aspect. The difference in dendrites sizes will be discussed later.

Fig. 4 presents the diffractograms obtained. Comparing the difractograms, it is possible to observe that the patterns obtained are very similar; this indicates that there were no significant changes in the microstructure of the alloy used under the several conditions studied. However, it may be remembered here that all the recognized phases are FCC structures, with very close lattice parameters, which agrees with spectra shown in Fig. 4. These results agree with the results presented in Table 7, where all the phases recognized were of the FCC type, by X-ray diffraction techniques.

Without etching, the appearance of the fused coatings was very uniform, with no indication of precipitates (Figs. 2 and 3). The flame-fused coatings presented a relatively high level of porosity, as mentioned previously, with small and medium pores uniformly distributed throughout the layer (Fig. 3-a). The arc-fused coatings presented also small pores, and others with a large diameter (about 10-15mm), as shown in Fig. 5. This later can be attributed to localized overheating, as indicated in the literature (AWS, 1985).

While the chemical analysis suggests the occurrence of dilution in the case of arc fusion (Tables 1, 6 and 7), the direct observation of the transverse section of the samples does not indicate this (Fig. 5). The results of the cavitation test do not show any significant difference in erosion rate either, as can be seen in Fig. 6. In fact, more detailed observation showed that the increase in Fe content occurred only in some well-defined regions of the coating and at different depths below the coated surface. These variations are probably due to changes in operational parameters, such as travel speed and arc length.

Despite the fact that the cavitation test of the arc-fused coatings showed a mass loss similar to that obtained from the flame-fused coatings (Fig. 1), preferential erosion in some areas (regions rich in Fe, for example) may have occurred, which was compensated by lower erosion in others with better resistance. SEM observations of the eroded surface showed the same pattern of wear in both types of samples, and EDS analysis presented equally abraded areas in Fe and Ni rich regions, depending only on their locations on the test probe (Table 6).

This latter result suggests that alloys with a lower Ni content, which are therefore less expensive, may present a good performance in preventing cavitational damage. However, although these alloys can be arc fused after spraying, special care must be taken to avoid dilution, which can make their performance worse. Fusion after spraying destroys the lamellar structure of the sprayed coating, yielding a more refined structure with dendritic characteristics. The more intense the fusion heat, the finer is the final structure, as seen in Figs. 2 and 3.

With this performance, fused coatings seem to be competitive with welded stainless coatings used nowadays in the maintenance of hydroelectric plants (Araújo, 1996). However, due to the size of the equipment, flame fusion is not feasible. The use of a GT arc seems to make the couple process/material viable for this application.

However, arc fusion is a delicate operation. It can cause excessive dilution and poor resistance of the coating to cavitation attack. On the other hand, in this work, local changes in chemical composition do not affect the microstructure or the general performance of the coating, making it easy to apply this technique to protecting hydraulic turbines against cavitation erosion.

Based on results obtained in this work, the procedure developed was applied to the maintenance of a cavitation-damaged Francis hydraulic turbine used for power generation in a plant belonging to the Energetic Company of Minas Gerais (Companhia Energética de Minas Gerais - CEMIG) in Brazil. Regions of a distributor blade and of a rotor vane were sprayed with the alloy used in this work and arc fused under field conditions. This machine has 13 vanes, a 297 MW power generation capacity at 138.5 rpm and a 239.5 m3/s water flow. The equipment will be inspected during the next two or three years, depending on the machine’s availability, to verify any improvement.

Conclusion

The arc fusion of self-fluxed alloy coatings may be an alternative in the field application of such coatings to large equipment where conventional flame fusion is difficult or even impossible. In this work, although some dilution occurred during the operation, this did not seem negatively affect the structure and/or properties of the coating. Depending on the application, in some cases an improvement in performance could be observed.

Acknowledgments

The authors would like acknowledge the cooperation of Fundação para o Amparo da Pesquisa do Estado de São Paulo - FAPESP, METCO do Brasil and OGRAMAC Metalização in this work.

Manuscript received: August 1999. Technical Editor: Alisson Rocha Machado.

• American Society for Testing And Materials, Philadelphia. ASTM C-633, Standard test method for adhesion or cohesive strength of flame-sprayed coatings. Philadelphia, 1974. 7p.
• American Society for Testing And Materials, Philadelphia. ASTM E 18, Standard test method for Rockwell hardness and Rockwell superficial hardness. Philadelphia. 1998. 14p.
• American Society for Testing And Materials, Philadelphia. ASTM E-384, Test method for microhardness of materials. Philadelphia. 1997. 19p.
• American Society for Testing And Materials, Philadelphia. ASTM G-32, Standard method of vibratory cavitation erosion test. Philadelphia, 1985. 6p.
• American Welding Society. Thermal Spraying - Practice, theory, and applications. Miami: AWS, 1985. 184p.
• Araújo, N. B. M. et alli. Cavitational damage of welded coatings in hydraulic turbines. In: Encontro Nacional De Tecnologia Da Soldagem, 22, 1996, Blumenau. Anais..., 1996. São Paulo: 1996, ABS. pp. 313-323 (in Portuguese).
• Matejka, D. and Benko, B. Plasma Spraying of Metallic and Ceramic Materials. Chichester: John Wiley & Sons, 1989. Cap. 5: Selection of powders for plasma spraying in industrial applications, pp.126-148.
• Marques, P. V. Cavitational damage in flame thermal sprayed coatings. Campinas: Mechanical Engineering School - UNICAMP, 1996. 160 p. Ph.D. thesis. (in Portuguese).

Publication Dates

• Publication in this collection
  13 Dec 2000
• Date of issue
  2000

History

• Received
  Aug 1999