There are three principal laboratory testing techniques to determine cavitation rates:
· ultrasonic cavitation testing
· cavitating jet testing
· venturi cavitation testing.
The cavitation rate is usually given in terms of weight loss per time period. However, the rate can also be reported in terms of a change in thickness per time period or a volume loss per time period.
The ultrasonic (vibratory) method of cavitation testing uses a magnetostrictive or piezoelectric device to produce a high-frequency (generally 20 kHz) vibration in a test specimen immersed in a liquid (Figure 4). During one half of each vibration cycle, a low pressure is created at the test specimen surface, producing cavitation bubbles. During the other half of the cycle, bubbles collapse at the specimen surface. It is a simple, relatively fast, and inexpensive technique and has been the most widely used technique for cavitation testing (March and Hubble 1996). A standard test procedure for ultrasonic cavitation testing has been approved by the American Society for Testing and Materials (ASTM) as Standard G 32 (ASTM 1992). The technique has been modified by placing the test specimen a small distance below the tip of the ultrasonic probe (Schwetzke and Kreye 1996).
Figure 4. Ultrasonic cavitation testing: (A) ASTM G 32 (B) Modified method (Schwetzke and Kreye 1996).
Results of ultrasonic vibratory cavitation testing for polymer coatings on concrete were reported to not correlate well to the field cavitating conditions. The ultrasonic test apparatus was not able to reproduce in the laboratory the same type adhesion failures that frequently occurred for polymer coatings under field conditions (Cheng, Webster, and Young 1987).
The cavitating jet method for cavitation testing uses a submerged cavitating jet to erode a test specimen placed in the jet's path (Figure 5). This technique is relatively compact and provides a higher range of cavitation intensities than do the ultrasonic probe method or the venturi method.
The cavitating jet test methodology was found to provide consistent, reproducible results for a given operating condition. The relative cavitation rate, referenced to a standard material, provides a good method for comparing materials that have a wide range of properties (March and Hubble 1996).
The TVA has used the results obtained from laboratory cavitating jet testing to select weld materials for field demonstrations. Weld materials that had higher cavitation resistance compared to welded stainless steel in the laboratory also performed better than stainless steel in the field (Karr et al. 1990). The cavitating jet laboratory test results for weld alloys were found to correlate well with field experience.
Therefore, based on results reported in the literature, the cavitating jet test is better than the ultrasonic cavitation test at predicting the field performance of materials.
Figure 5. Schematic of Cavitating Jet testing apparatus (March and Hubble 1996).
A venturi-type cavitation testing machine is shown in Figure 6. An uncoated steel test panel served as the con-trol specimen. The inlet pressure was maintained at approximately 60 psi, produc-ing a water velocity of app-roximately 70 ft/sec through the venturi throat. This generated a sustained, moderately cavitating envi-ronment. This test required that the panels be removed on a regular basis from the test apparatus, inspected, weighed, and returned to the test apparatus until failure was observed (Baker 1994). The venturi cavitation method was found to require long times to complete the test-as many as 2078 hours-so it was deemed inappropriate for this research.
Cavitation barrier coatings were applied in June 1989 to the backside of one blade of a Kaplan Turbine Unit at Rocky Reach Dam, Unit #13, Chelan County Public Utility District (PUD), Washington. Approximately 45 sq ft along the outer edge of the blade was coated with Tribaloy® T-400 and an urethane top coat.
Chelan County Public Utility District (PUD) personnel repaired all the previous cavitation damage, restoring the blade's shape and contour. Chelan County PUD personnel, assisted by a contractor, grit blasted the surface to be coated. The contractor set up the cavitation barrier equipment and applied Tribaloy® T-
400 coating using HVOF equipment. A urethane coating was brush-applied over the Tribaloy coating.
Figure 6. Venturi cavitation testing apparatus (Baker 1994).
Two problems were encountered with the application of the cavitation barrier coatings:
1. High-velocity equipment was not designed to be taken into turbines, so there were problems with fuel gases, degassing and powder feed.
2. When the first coating of Tribaloy was applied several problem areas were noted and the entire coating was found unsatisfactory. The coating was removed and a new coating applied.
Based on further experience in the field, these problems can be overcome by implementing a number of changes to the procedure:
1. The proper attention to preparation of the surface is required. In the case of turbine blades, grit blast the area to be coated one day before the application of the coating, followed by the use of heat blankets on the top of the blades for approximately 12 hours to remove moisture from the surface and prevent condensation. Grit blast the surface to be coated to white metal finish just before spraying. After application of the Tribaloy coating, the urethane coating would be sprayed (rather than brush-applied) to improve thickness and finish.
2. Fuel bottle heaters are now available to help maintain fuel temperature in a cool environment. Insulating the fuel lines will also help maintain the fuel gasses.
3. The Metco Diamond Jet HVOF equipment has since been modified to address powder feed problems.
Inspection in June 1990 found approximately 15 sq ft of cavitation barrier coating in basic contact on the bottom portion of the blade although the urethane coating had come off in large pieces. The HVOF Tribaloy® T-400 coating appeared intact in an area of mid-blade. However, the areas of the blade most vulnerable to high cavitation had no remaining cavitation barrier coating. (This area is the outer tip of the blade, approximately 4 x 8 ft). Some minor cavitation damage to the underlying metal was noted-approximately 8 x 3 in.-with a waviness of the blade surface in the area of the cavitation.
It was concluded that the Tribaloy® T-400 applied by HVOF and coated with urethane had an impeding effect on the cavitation. Improvements in equipment, technique, and experience levels would be expected to provide better results (Lontz 1992).
The Bureau of Reclamation conducted a study for USACERL to determine the cavitation resistance of inorganic and ceramic coatings applied over steel substrates. Testing was conducted in a Venturi-type cavitation testing machine (see Figure 6). An uncoated steel test panel served as the control specimen. The inlet pressure was maintained at approximately 60 psi producing a water velocity through the throat of approximately 70 ft/sec. This generated a sustained, moderately cavitating environment. A criterion for coating failure was established for coated panels as the time when 1 to 2 percent or more of the coating had been removed down to the substrate. The test panels were inspected at regular intervals to determine time of failure (Baker 1994).
Two sets of cavitation results are presented in Table 4. The first set contained a mild steel control sample and two coated samples: Panel 11, metallized coating (Stellite Tribaloy® T-400) and an organic topcoat (total 50 mils); Panel 12, 24 mils Stellite Tribaloy® T-400 and 10 mils organic topcoat of a reinforced epoxy (Belzona Superglide®6). Belzona Superglide® is a two-component nonmachinable-grade material consisting of a silicon steel alloy blended within high molecular weight reactive polymers and oligomers.
The second set of results consisted of three sets of samples: Panel 21, stainless steel; Panel 22, stainless steel plus 10 mil Stellite Tribaloy® T-400 applied by wire feed thermal spray; and Panel 23, stainless steel plus 10 mil Stellite Tribaloy® T-400 + 20 mil organic topcoat of a reinforced epoxy (Belzona Superglide®).
The metallized coatings were ranked according to time to first damage. The best performer, with a time to first damage of 565 h, was Panel 12: 24 mils Tribaloy® T-400 + 10 mils of a reinforced epoxy (Belzona Superglide®). Second best, with a time to first damage of 386 h, was Panel 23: 10 mils Tribaloy® T-400 and 20 mils of a reinforced epoxy (Belzona Superglide®). Third best, with a time to first damage of 218 h, was Panel 11: metallized coating (Tribaloy® T-400) and organic topcoat (total 50 mils). The fourth best, with a time to first damage of 186 h, was Panel 22: 10 mils Tribaloy® T-400. The organic topcoat, a reinforced epoxy (Belzona Superglide®), was found to extend the life of the metallized coating (Tribaloy® T-400). Although the topcoat was found to fail early, it did provide added protection when present. The reinforced epoxy (Belzona Superglide®) topcoats were found to be superior to polyurethane topcoats (Baker 1994).
The results of Baker showed that the time to failure of stainless steel was 2075 hours, the time to failure of mild steel was 1038 hours, and the time to failure of the metallized Tribaloy® T-400 was 545 hours. The time to failure during cavitation testing of the metallized Tribaloy® T-400 coating was found to be less than either the carbon steel or the stainless steel.
Problems encountered during the testing included:
1. Water flow across the panels was not uniform.
2. The depth of the testing surface in the cavitating water stream was inconsistent. Samples of mild steel showed that panels placed deeper in the water stream sustained more severe cavitation damage than the control panel.
3. Long exposure times were required to complete the test-as long as 2078 hours, limiting the number of samples that may be tested in a reasonable period.
The results obtained using the Venturi cavitation testing apparatus provided valid insights into the material systems tested, but the long testing periods required made the technique inappropriate for this CPAR research.
Results of Soares, Souza, Dalledon, Baurque, and Amado 1994
Tests were performed on thermal spray coatings with both liquid impingement and vibratory cavitation devices. Some of the best coatings were tested further in a 6 meter Francis hydroelectric turbine with a previous history of severe cavitation. The materials investigated and the erosion and cavitation resistance results are shown in Table 5. The cavitation rate was given as a change in thickness of the coating (_m/h).
Coatings number 1 - 5 and 7 - 10 were tested in a liquid impingement erosion test apparatus in accordance with ASTM Standard G 73. The erosion resistance of samples 1, 2, 3, and 9 were of a similar order of magnitude as the SAE 1020 steel reference material. Samples 4, 5, 7, and 8 failed the test as the coating came off the substrate. The cavitation resistance of coated samples, measured using a vibratory testing apparatus in accordance with modified ASTM Standard G 32, was generally lower than the carbon steel reference material. The cavitation resistance of the ceramic-loaded polymer, sample 13, was significantly lower than for the thermal sprayed metal or ceramic coatings.
Thermal spray and polymeric coatings were applied in a turbine at the Gov. Bento Munhoz hydroelectric project of COPAL (Companhia Paranaense de Energia, or Energy Company of Parana [Brazil]). Coatings number 1, 2, and 6 were applied over stainless steel weld layers in areas of medium cavitation. Polymer coatings number 14 and 15 were applied in areas of low to medium cavitation in the same turbine. After 1500 hours of operations it was observed that coatings 1, 2, and 6 were gone to various degrees, with there being more area of coating 6 and less area of coating 1 gone. The polymeric coatings 14 and 15 were completely gone in areas where the substrate was stainless steel, but in the area of carbon steel the coatings were relatively well retained. In these protected areas the intensities of cavitation were lower. During the same time of operation, the carbon steel regions without coatings, subjected to low or medium cavitation, did not show any indication of cavitation.
Soares et al. (1994) concluded that despite their elevated hardness and/or abrasion resistance, the best thermal sprayed coatings were at best only similar to carbon steel (SAE 1020 or AWS 309 stainless steel) based on the cavitation resistance as evaluated in the laboratory tests. Additionally, since these coatings can be applied only to a very small thickness (i.e., 0.5 mm), they found little or no advantage compared to conventional welded layers for turbine blades. An additional problem of poor adhesion was observed during the field tests in the hydroelectric turbine: the sprayed layers simply peeled off after a few months of operation (Soares et al. 1994). Based on laboratory and field data the researchers concluded that thermal spray coatings were not suitable in severe cavitation applications.
Results of March and Hubble, 1996
Cavitation testing of mostly weld materials and some other coating materials was conducted at the Tennessee Valley Authority (March and Hubble 1996). The cavitating jet test apparatus was used at 4000 psi (Table 6). Weld overlay material including Ireca, Nitronic 60, Stellite® 6, Stellite® 21, Stoody 6, and Stoody 2110 with one coating Imperial Clevite WC-204 were found to have substantially lower cavitation rates than the 308 stainless steel reference panel. The cobalt-containing austenitic stainless steel, Ireca, had a relative cavitation rate of 0.02 times that of 308 stainless steel-the lowest rate among all the materials tested.
In addition to weld alloys, this work included testing of thermal spray coatings such as Hardco spray 110, Hardco spray Stellite 21, Plasmadynne plasma spray Stellite 21, and several elastomeric materials including Devcon pump repair epoxy, Belzona ceramic reinforced epoxy, and a nylon coating. In general, the coatings displayed higher relative cavitation rates compared to 308 stainless steel, with rate values ranging from 11 to 67 times that of the reference panel. However, the relative cavitation rate of one coating-Imperial Clevite WC-204-was 0.3 times that of the reference. Coatings were also susceptible to mechanical damage and bond failure under the test conditions (March and Hubble 1996).
Source: March and Hubble 1996.
Based on the results of March and Hubble (1996), advanced weld alloys such as Ireca alloys (marketed as Hydroloy® 913) provided superior cavitation resistance and were recommended for use in areas of severe cavitation. TVA in 1988 successfully tested Hydroloy® 913 (the commercial form of the Ireca alloy) on the runner and crown of a hydroelectric pump/turbine at Raccoon Mountain, Chattanooga, TN. Following inspection in 1990, after 6782 hours of operation, the turbines blades repaired with Hydroloy® 913 had significantly less cavitation damage than blades repaired with 309L stainless steel (Karr et al. 1990).
Results of Schwetzke and Kreye, 1996
Cavitation experiments were performed using a vibratory apparatus according to ASTM G 32, modified to place the test specimen 0.5 mm below the vibrating steel disc of the ultrasonic horn. Tests were conducted for up to 5 hours. The steady-state cavitation rates of the coatings tested are given in Table 7. For the cermet (metal ceramic alloy) and oxide coatings tested, the mass loss versus exposure time revealed an almost constant erosion rate between 1 and 5 hours of testing.
Coatings investigated included stainless steel (316L), self-fluxing nickel-based alloys (NiCrFeBSi, type 60), tungsten carbide-cobalt (WC-17 Co), chromium carbide-nichrome (Cr3C2-25 NiCr), and chromium oxide (Cr2O3). The results demonstrated that HVOF-sprayed coatings of NiCrFeBSi, WC-17 Co, Cr3C2-25 NiCr, and Cr2O3 exhibited erosion rates as low as that obtained from bulk specimens of stainless steel (AISI 321 or 316 L). However, the cavitation rates of plasma sprayed cermet coatings were about an order of magnitude higher than the erosion rate of the best HVOF coatings (Schwetzke and Kreye 1996). A similar high difference of the erosion rates of plasma sprayed as compared to HVOF-sprayed cermet coatings has recently been reported for the removal of those coatings by high-pressure water jets (Kreye et al. 1995).
HVOF coatings of NiCrFeBSi, WC-17Co, Cr3C2-25 NiCr and Cr2O3 exhibited rather high resistance to cavitation and were recommended for consideration as a protective surface layer against cavitation (Schwetzke and Kreye 1996). This study provides support for the use of these materials in the repair of hydroelectric turbine components such as draft tube liners.
Spray Process |
System |
Fuel |
Material |
Hardness (VHN 300 g) |
Cavitation Rate (mg/h) |
HVOF |
JP-5000 |
Kerosene |
Stainless Steel 316 L |
263 |
6.8 |
HVOF |
Jet Kote |
Propane |
NiCrFeBSi type 60 |
674 |
4.3 |
HVOF |
JP-5000 |
Kerosene |
NiCrFeBSi type 60 |
767 |
4.7 |
HVOF |
Top Gun |
Hydrogen |
Tribaloy® T-400 |
579 |
20.4 |
HVOF |
Top Gun |
Hydrogen |
Tribaloy® T-700 |
589 |
12.4 |
Plasma |
A-3000 S |
Ar / H2 |
WC-Co 88-12 |
764 |
74.8 |
HVOF |
Top Gun |
Propane |
WC-Co 88-12 |
1178 |
11.9 |
HVOF |
Top Gun |
Propane |
WC-Co 83-17 |
1376 |
5.8 |
HVOF |
Jet Kote |
Propane |
WC-Co 83-17 |
1052 |
30.0 |
HVOF |
Jet Kote |
Propane |
WC-Co 83-17 |
1127 |
23.4 |
HVOF |
Jet Kote |
Ethylene |
WC-Co 83-17 |
1243 |
22.8 |
HVOF |
DJ 2700 |
Ethylene |
WC-Co 83-17 |
1399 |
7.2 |
HVOF |
JP-5000 |
Kerosene |
WC-Co 83-17 |
1420 |
6.3 |
Plasma |
A-3000 S |
Ar / H2 |
Cr3C2-NiCr 75-25 |
722 |
59.5 |
HVOF |
Top Gun |
Propane |
Cr3C2-NiCr 75-25 |
1021 |
17.6 |
HVOF |
Jet Kote |
Propane |
Cr3C2-NiCr 75-25 |
978 |
13.9 |
HVOF |
DJ 2700 |
Ethylene |
Cr3C2-NiCr 75-25 |
1134 |
5.5 |
HVOF |
JP-5000 |
Kerosene |
Cr3C2-NiCr 75-25 |
1220 |
3.8 |
Plasma |
A-3000 S |
Ar / H2 |
Al2O3-TiO2 97-3 |
772 |
52.8 |
HVOF |
Top Gun |
Acetylene |
Al2O3-TiO2 87-13 |
972 |
24.7 |
Plasma |
A-3000 S |
Ar / H2 |
Cr2O3 |
1322 |
6.6 |
HVOF |
Top Gun |
Acetylene |
Cr2O3 |
1210 |
2.9 |
Bulk material: Stainless Steel X6 CrNiTi 18 10 (type 321) |
226 |
5.5 | |||
Bulk material: Stainless Steel X2 CrNiMo 17 13 2 (type 316 L) |
165 |
6.0 | |||
Source: Schwetzke and Kreye 1996.
Results of Musil, Dolhof, and Dvoracek 1996
The wire arc spray (WAS) process of functional and multilayered coatings was successfully used for the repair of vanes on reversible Francis turbines (Musil, Dolhof, and Dvoracek 1996). The two-wire arc spray process employs the spraying of two different wire materials to create a mixed or graded coating structure. NiAl and Cr stainless steel were used for the two-wire arc spraying. NiAl (95% Ni - 5% Al) is widely used in the power industry. Wire sprayed NiAl coatings have shown higher bond strengths than plasma sprayed coatings and also maintain their high bond strength at greater thicknesses (Unger and Grossklaus 1992). High-chromium stainless steel was selected as the spray material for the functional top-coat. Due to the severe cavitation damage, with some pit depths greater than 25 mm, the deposition of very thick coatings was required. Damaged materials were removed and the surface cleaned and grit blasted before application of the repair coating.
Thick multilayered coatings deposited by WAS were evaluated for the repair of vanes on a Francis turbine. Three types of functional graded coating were evaluated: (A) a duplex of high Cr stainless steel with NiAl bond coat, (B) bond coat, graded NiAl -Cr stainless steel coatings with a Cr stainless steel top coat, and (C) multilayered graded NiAl-Cr stainless steel coatings with a Cr stainless steel topcoat (Figure 7). The alternating layers in the NiAl-Cr stainless steel multicomponent graded coating were approximately 1.5 mm thick. Laboratory analysis showed that the multilayered graded NiAl-Cr stainless steel coatings (Figure 7C) yielded the best results with the lowest residual stress.
Figure 7. WAS coatings (A) Duplex of high Cr stainless steel with NiAl bond coat, (B) Bond coat, graded NiAl -Cr stainless steel coatings, and Cr Stainless steel, (C) Multilayered bond and graded NiAl-Cr stainless steel coatings and Cr stainless steel topcoat. Source: Musil, Dolhof, and Dvoracek 1996.
Repair was performed on large eroded areas (1-3 m2) of the vanes on a Francis turbine. Localized cavitation damage with pit depth of 30-35 mm maximum was repaired by sprayed materials. Multilayered graded NiAl-Cr stainless steel coatings (Figure 7C) were applied by the WAS process to stationary wicket gate supports in four hydroelectric power stations located in the Czech Republic. The main steps in the repair process were:
· examination
· alumina blasting
· hand working with power tools and chemical cleaning
· alumina blasting
· local WAS application of extremely damaged parts
· hand working with power tools and blasting
· WAS application of functional multilayered graded coatings
· application of special seals
· hand working with power tools and special seal application.
The seal material was not specified. After 30-36 months of continuous operation, the coatings applied by WAS showed better performance in comparison to the original carbon steel (Musil, Dolhof, and Dvoracek 1996). This demonstrated the successful use of thermal spray coatings for the repair of hydroelectric components and provides additional support for their use. However, for severe cavitation damage, the authors of the current study recommend weld repair. As will be shown, advanced iron-based weld alloys such as D-CAV®, NOREM®, CaviTec®, or Hydroloy® 914, may be considered for the repair of severe cavitation damage.