White Papers & Presentations Erosion Resistance of Tungsten Carbide Braze Cladding in Coal-Fired Power Plants Download complete paper (849 KB)  Abstract As an alternative to weld overlay wear resistant coatings, brazed tungsten carbide cladding provides increased erosion protection with less weight added. When used correctly, these claddings have been shown to be superior in fly ash conveyance fans and other industrial applications where erosion by solid particles and corrosion are the principal mechanisms of failure. Erosion resistance of depends not only on the particular material selected but also on the method of application, impact angle, velocity and eroding media. This presentation will discuss several standard materials and test conditions performed per the ASTM G76, “Standard Test Method for Conducting Erosion Tests by Solid Particle Impingement Using Gas Jets”. Plain carbon steel, tungsten, tungsten carbide composites, aluminum oxide and chrome carbide weld overlay were examined for erosion resistance at various angles and particle velocities. The principal conclusions reaffirm three basic concepts: At 90° impact angles there are no “outstanding” performers. As the angle of impact decreases the commonly employed materials begin to segregate into two groups (positive and negative correlation) based on wear mechanism and performance. The rule of mixtures applies to performance in that composite or metal matrix material erosion directly correlates the proportions of reinforcing particle and matrix material. Introduction When evaluating the relative erosion resistance of materials, a number of factors must be considered. The obvious factors are temperature, velocity of the impacting particles, their size and shape, and the impacting angle. These factors can be controlled in standardized testing but combining their range of variability to comprehensively evaluate performance is limited. Standardized testing procedures, such as ASTM G76 (1), reduce a number of the variables with the intent of providing a common baseline for comparison. The issue with such standardized tests is that the differences in erosion rates can be diminished to the point where differentiation is difficult or misleading. Short of full scale tests, this leaves variation of the standardized tests as the only means to more accurately mimic actual conditions and rank erosion resistance in a meaningful way. Constitutive models, based on intrinsic properties (2) are used to predict erosion resistance and have been making progress over the last ten years but are still immature. Fluid dynamic models used to predict flow patterns of particle laden streams have become quite good and are commonly used for critical applications (3,4). However, the constitutive and fluid model methodologies still require measured values for erosion rates for the properties table and for validation of the results. This work describes some of the measurements made by using the ASTM G76 standard to measure and rank the erosion resistance of infiltration brazed tungsten carbide composite claddings. The ranking was made with respect to several common materials, while varying impact angle to better approximate actual conditions within a power plant environment. Infiltration Brazed Tungsten Carbide Cladding Infiltration brazing, as defined here, means to fill by capillary action with molten filler metal, a porous coating or structure that has a melting point higher that the filler metal. While there are many means for applying the carbide and braze in preparation for infiltration braze coating, the principal method discussed here involves a non-woven preformed cloth. The particles used in this process are sized and mixed to provide a stable, dense coating. Figure 1 illustrates the process steps from powder mixing to brazing. Figure 2 shows the resulting microstructure of the cladding and the interface. The constraints on this process are straight-forward. Infiltration brazing is performed in a vacuum furnace. Large parts as measured by volume, weight or area, are often segmented prior to cladding to avoid limitations of furnace size. If parts tendency to distort or cannot be straightened or machined then this presents a barrier. Currently there are no means for field application of these types of coatings. Substrates such as aluminum or low melting alloys cannot currently be protected by this process.     Figure 1: Process schematic for infiltration brazing tungsten carbide cladding.    Figure 2: Optical micrograph showing substrate/cladding interface.   Erosion Testing The test method employed for this work was ASTM G76. This test method covers determination of erosion rate by solid particle impingement in a gas stream. Actual erosion conditions involve particle sizes, velocities, angle of attack as well as various operating environments. Consequently, a single test is usually insufficient for evaluating expected performance over all conditions. Test apparatus, Figure 3, was rather straight-forward. The nozzle’s inside diameter was 6.4 mm (0.25 in.) by 152 mm (6 in.) long. The length to diameter ratio was 24:1, minimum recommended in the ASTM G76 procedure was 25:1.     Figure 3: Schematic of erosion test apparatus. As a first step, calibration of particle velocity was performed. The ASTM procedure provides test data at 30 and 70 m•s-1 (98-230 ft•s-1) for 1020 steel at a 90° impact angle for interlaboratory tests. The standard deviation and deviation between laboratories (provisional) are also given in the standard. The standard recommends the use of a rotating double-disk, laser velocimeter or high speed photography as methods to measure particle velocity at the sample. This work relied exclusively on velocity calibration by duplication of the standard data given in the ASTM G76 procedure. The material used for calibration was 1018 steel, annealed, 71 HRB. Results and Discussion Figure 4 shows the results from weight loss measurements at 30 and 70 m•s-1 (98-230 ft•s-1) for 1018 steel. The dashed lines represent the limits of variation between laboratories, the solid lines represent the standard deviation within laboratories performing the test. The open circles show results for 70 m/s, the open diamonds represent the measurements at 30 m/s. Weight loss measurements were within the inter-laboratory deviations of ASTM G76, but slightly high and outside with respect standard deviation, Table 1.     Figure 4: Examples of erosion versus time for 1018 steel at 30 m/s and 70 m/s (98-230ft•s-1), 50 µm alumina, 2.1 g/min. , 90° impact angle.   The steady state erosion rate is equal to the slope of the mass loss versus time plot shown in Figure 4. The average erosion rate is calculated by dividing the erosion rate (g/min) by the abrasive flow rate (g/min) then dividing by the material’s density (g/cm3) :    (1) Average erosion rate measurements were within the inter-laboratory deviations of ASTM G76, but slightly high and outside with respect to standard deviation, Table 1. Table 1: Average Erosion Rate (mm3/g) Measured and ASTM G76 Standard Rate for 1018 Steel. Measured Standard 30 m•s-1 (98 ft•s-1) 70 ms-1 (230 ft•s-1) 30 m•s-1 (98 ft•s-1) 70 ms-1 (230 ft•s-1) 2.89 32.02 2.73 28.16 0.55 STDEV 1.47STDEV 0.47STDEV 0.97STDEV Calibration of erosion rate for the steel standard showed reasonable consistency with the ASTM procedure. While the continuation of measurements using alumina for ranking the erosion resistance of other materials appeared useful, changing the eroding media to fly ash or magnesia pointed out some differences in behavior as well as limitations of the experimental apparatus. Differences in erosion behavior could be generally categorized by broad descriptions of the type of wear. In general, soft materials (a decidedly relative term) erode by plowing or cutting when impacted at low angles. Relatively soft materials can be distinguished from relatively hard materials by the fact that their erosion rate tends to increase at lower angles of impact. Increasing erosion rate with decreasing impact angle is often referred to as negative correlation, referring to the slope of the erosion versus impact angle. In Figure 5, the erosion rates for a variety of materials are shown. The negative correlation aspect of erosion rate is illustrated by the behavior of 1018 steel over the 30° to 90° impact angle range. There is also a negative correlation for the weld overlay and the AR400 material at impact angles greater than 60°. Below 60° impact angle, the weld overlay, AR400 and the infiltration brazed tungsten carbide claddings show a positive correlation although the slope of the latter decreases with increasing impact angle. Another observation worth mentioning was that the materials tested, in some instances, seemed to converge near 90° impact angle. In the range of 0.04 to 0.050 mm3/g what were otherwise distinctive levels in performance tended to diminish. Exceptions to this observation were the tungsten sample and the weld overlay, which showed the lowest and highest erosion rates at 90°, respectively.     Figure 5: Erosion rate of various materials measured at 230 ft•s-1 (70 m•s-1) using 50 µm alumina, 2.1 g/min., tube diameter 0.25 in. (6.4 mm).   A comprehensive evaluation of over 200 materials , subjected to 90° impact with alumina, 558 ft•s-1 (170 m•s-1), 5g/min, was performed by Hansen (5) in 1979. What he clearly showed was that nearly all metals and metallic alloys, except tungsten and molybdenum, had similar erosion resistance at room temperature for 90° impingement. Hansen observed only a 30 percent improvement within the range of materials tested at 90° impingement. He concluded that if a metallic component fails prematurely by erosion, at 90°, better performance could not be obtained from the substitution of another metallic alloy regardless of hardness. Alternatively, tungsten carbide cermets show improved erosion resistance in a manner related to the binder content. Uuemyis and Kleis (6) verified a mechanism which eroded the metallic binder from around carbide grains. By maximizing the area of carbide presented to the eroding media (minimizing binder content), erosion resistance could be increased by a factor of three or four. As a minimum, and as much as ten to fifteen times as a maximum, compared to cermets with high binder content (>40 volume percent) or metallic alloys, respectively. As a measure of erosion performance related to cladding microstructure, Lindsley and Marder (7) showed the dependence of erosion resistance on a modified Hall-Petch relationship: property = initial property value + k•(microstructural parameter)-1/2   (2) The property of interest in this case is erosion resistance. The initial value is functionally related to an intrinsic material property, the proportionality constant k is often related to a micro-structural parameter, and the micro-structural parameter itself which is measured as the mean free path length between carbide particles. The effectiveness of this relationship in representing the property of interest with regard to infiltration brazed tungsten carbide composites is shown in Figure 6. While the correlation was not high, R2≈ 0.6, there is, clearly, an increase in erosion resistance as particle spacing decreases.     Figure 6: Erosion rate of infiltration brazed tungsten carbide as a function of mean free path (particle spacing decreases left-to-right). Test conditions: 70 m/s using 50 µm alumina, 2.1 g/min., tube diameter 6.4 mm (0.25 in.).   Example Applications Fan blades are the principal victims in fly ash erosion although deflector rings, piping and other ancillary equipment are often high-wear items once the principal problem has been solved. Figure 7 (a) shows a heavily eroded section of weld overlay in a centrifugal fan. The leading edge of centrifugal fans is commonly protected with some sort of erosion resistant material. The protection often is in the form of a replaceable liner. Figure 7 (b) shows an erosion resistant liner after service equivalent to that seen in Figure 7 (a). The difference in erosion between the chrome carbide weld overlay in Figure 7 (a) and the infiltration braze coated liner in Figure 7 (b) is significant because of the reduced wear rate. Because of the observed erosion rate for the infiltration brazed cladding, compared to the erosion observed on the weld overlay, service life for the brazed cladding would be expected to be five times longer than the weld overlay coating. Two blade liners that ran for 14 months in the same fan are shown in Figure 8. The liner on the left, Figure 8a, was coated with >6 mm (0.25 in.) of chrome carbide weld overlay. The figure on the right, Figure 8b, shows an infiltration brazed tungsten carbide liner with 0.25 to 0.30 mm (0.010-0.012 in.) after the same exposure as the chrome carbide liner. The relative wear rate was approximately 20:1. In a side-by-side comparison of several different wear resistant technologies (infiltration brazed tungsten carbide cladding, tungsten carbide oxy-fuel HVOF & plasma spray and chrome carbide weld overlay) TVA, Kingston checked the performance of sample blades after 60 days in Unit 9. All blades on the fan were exposed to high erosion over this short run. Tungsten carbide coating options (thermal spray, plasma spray, HVOF) appear to have been penetrated in the 60 day, high erosion test, Figure 9a, 9b. The infiltration brazed tungsten carbide clad blades lost a maximum of 0.25 mm (0.010 in.), Figure 9d, compared to the chrome carbide weld overlay’s maximum loss of 3.8 mm (0.150 in.), Figure 9c. At the coating thickness applied, the infiltration brazed cladding would be expected to last from 4 to 15 times longer that the sprayed tungsten carbide coatings or the chrome carbide weld overlay, respectively.      Figure 7: Centrifugal fan, (a), showing wear of chrome carbide weld overlay after approximately four months of service. Centrifugal fan liner coated with infiltration brazed tungsten carbide, (b), after approximately four months service. Coating was 0.762 mm (0.030 in.) thick, maximum wear of coating was 0.152 mm (0.006 in.)   Figure 8: Fan liners after 14 months exposure to fly ash erosion. The chrome carbide weld overlay liner, (a) was coated >6 mm (0.25 in.), infiltration brazed tungsten carbide liner, (b) showed 0.25-0.3 mm (0.010-0.120 in.) of wear.           Figure 9: Four of the blades from the TVA Kingston, Unit 9, side-by-side, 60 day comparison run. Areas of penetration are circled. Blade (a) was a 0.5 mm (0.020 in.) thick tungsten carbide plasma spray coating, (b) was 0.25mm (0.010 in.) HVOF tungsten carbide coating, (c) was 4 mm (0.150 in.) chrome carbide weld overlay. The infiltration brazed tungsten carbide cladding is bottom right (d), 1 mm (0.040 in.) on the blade, 2 mm (0.080 in.) on the pad. The infiltration brazed tungsten carbide blade, (d) showed 0.25 mm (0.010 in.) of wear. Conclusions Particle erosion of alloys and composite coatings is strongly influenced by impingement angle of the eroding particles. Variable fluid flow, particle loading, temperature and geometry further complicate the direct transfer of standard testing to engineering applications. The ASTM G76 standard does provide reasonable correlation of laboratory data to field performance in that the magnitude of volume loss per unit mass of eroding media delivered permits ranking of relative erosion rates. Based on ASTM G76 testing and results from field testing, infiltration brazed tungsten carbide claddings eroded at a rate 1/15th to 1/4 that of chrome carbide weld overlay in the 30° to 90° range of impingement angles, respectively. In general, all the materials tested showed less relative difference in erosion at 90° impingement angle. The erosion rate of infiltration brazed tungsten carbide composites is a function of the carbide spacing (mean free path). This dependence has also been characterized on the basis of matrix material indicating that as matrix fraction increases, erosion resistance decreases. An area not studied here concerns the frangibility of the eroding particles. If the particles disintegrate on impact then the mechanisms for surface erosion will change. This change in eroding mechanism due to particle fracture will lead to variations in erosion rate depending on the ductility of the substrate and the interaction with fragmented particles. Additional work should be performed in the following areas: Characterization of erosion rate using alumina in the ASTM G76 test and comparing that with data obtained under the same conditions using a variety of fly ash. This would help verify the magnitude of the acceleration believed to exist when using alumina as opposed to fly ash. The average erosion rate using alumina appears to be two to three orders of magnitude higher than that obtained using fly ash. The possibility of producing a composite sub-structure in the matrix of infiltration brazed claddings may provide additional erosion resistance by mitigating what appears to be the current primary wear mechanism for these coatings. Written by: Don Buchlolz, Vice President of Technology, Conforma Clad Inc. Chris Harley, Senior Applications Engineer, Conforma Clad Inc. © 2003 Conforma Clad Inc. All rights reserved. “Standard Test Method for Conducting Erosion Tests by Solid Particle Impingement Using Gas Jets”, Annual book of ASTM Standards, Wear and Erosion; Metal Corrosion“, v. 03.02,West Conshohocken, PA, , pp. 311-317, 1999. R. Pieters, “Surface Nitriding of Ti-6Al-4V Alloy in Nitrogen Atmosphere Using PW and CW Nd:YAG Lasers”, Colorado School of Mines,Ph.D. Thesis, Golden CO., 1999. Y. D. Jun, W.Tabakoff, “Numerical Simulation of a Dilute Particulate Flow (laminar) Over Tube Banks”,J. Fluid Eng. (Trans of the ASME), 116:770-777, December 1994. Y. Hamabe, K. Toda, M Yamamoto, “Numerical Simulation of Sand Erosion Phenomena in Particle Separator”, European Congress on Computational Methods in Applied Sciences and Engineering, Barcelona, September 2000. J.S. Hansen, “Relative Erosion Resistance of Several Materials”, Erosion: Prevention and Useful Appliactons, ASTM Symposium, Vail CO, pp.148-162, Oct. 1977. K. Uuemyis, I. Kleis, V. Tumanov and T. Tiideman, translated from Poroshkovaya Metallurtiya, no. 3, (135) Appeared in Siviet Powder Metallurgy and Metal Ceramics, pp. 248-250, July 1972. B.A. Lindsley, A.R. Marder, “Solid Particle Erosion of an Fe-Fe3C Metal Matrix Composite”, Metallurgical and Materials Transactions, v. 29A, pp.1071-1079, March 1998.