THE UNIVERSITY OF MICHIGAN College of Engineering Department of Mechanical Engineering Cavitation and Multiphase Flow Laboratory Technical Report No. 01357-12-T EFFECTS OF PRESSURE AND TEMPERATURE VARIATION IN VIBRATORY CAVITATION DAMAGE TEST BY: Frederick G. Hammitt Financial Support Provided by: National Science Foundation Grant No. GK-1889 June 1969

ABSTRACT Vibratory cavitation damage tests (20 kHz, 2 mil double amplitude) have been conducted in water on SAE-660 bronze, over the temperature range 55 F to 2500 F and with 1 to 4 static NPSH supplied by variation in cover gas pressure. The effects upon damage of variation of temperature at constant static NPSH and variation of static NPSH at constant temperature are presented. Observations on the possible effects of gas content are included, and a full comparison with the existing literature is made in all respects. *ii

TABLE OF CONTENTS ABSTRACT.................. 11 LIST OF TABLES............... iv LIST OF FIGURES............... v I. INTRODUCTION......... 1 II. EXPERIMENTAL PROGRAM........ 2 A. Experimental Facility.2.... 2 B. Tests Performed.... 4 C. Experimental Results.5.... III. DISCUSSION OF RESULTS.......... 8 A. Comparison with Previous Pertinent Results. 8 B. Additional Points Concerning Present Test Re sults................10 IV. CONCLUSIONS................ 12 ACKNOWLEDGMENTS........... 14 BIB LIOGRAPHY...15 iii

LIST OF TABLES Page Table I. Mechanical Properties SAE 660 4 Table II. Test Conditions 17 Table III. Maximum Damage Rates 18 iv

LIST OF FIGURES Page Fig. 1. Schematic of Ultrasonic vibratory horn 19 Fig. 2. Cumulative Damage vs. Test Duration of SAE-'660 20 at 250 F and 2 atm. NPSH Fig. 3. Effect of Temperature at 1 atm. NPSH 21 Fig. 4. Effect of NPSH at 1500F 22 Fig. 5. Effect of NPSH at 250 F 23 Fig. 6. Damage at 60 minutes vs. Temperature 24 Fig. 7. Damage at 60 min. vs. Vapor Pressure 25 Fig. 8. Damage at 60 minutes vs. NPSH 26 Fig. 9. Damage at 60 minutes versus NPSH based on Actually 27 Damaged Area Fig. 10. Damage Rate vs. Temperature for 304 SS in Open 28 Beaker Test Fig. 1 1. Effects of NPSH on specimen damage pattern for 29 a water temperature of 150 F Fig. 12. Effect of NPSH on specimen damage pattern for a 30 water temperature of 250 F Fig. 13. Weight Loss versus Temperature for Cooling 31 Fig. 14. Effects of various fluids on specimen damage pattern 32 for 316SS v

I. INTRODUCTION Cavitation damage is encountered with almost any liquid pressure or temperature, and in many cases of present-day importance, with fluids other than water. Laboratory tests, however, must use experimental facilities which are usually standardized to their own most appropriate test conditions. In the case of the conventional vibratory facility, e. g., the test fluid is usually water at approximately room temperature. The pressure, when the horn is not operating, is usually one atmosphere, but the liquid pressure seen by cavitation bubbles during their life cycle is in the form of an approximately half sine wave giving a maximum collapse pressure of many atmospheres. The precise value differs according to the facility, since it depends upon frequency and amplitude. The gas content, another important parameter, is usually an equilibrium value characteristic of the vibrating horn in an open beaker, and somewhat less than the saturated content for one atmosphere at the water temperature. The gas content in field conditions can have any value depending upon many parameters. In any laboratory test there are unavaidable minor variations of pressure, temperature, and gas content. Hence, to achieve precise and repeatable results from laboratory tests it is necessary to know the magnitude of the effect of such variations. A recent study conducted in various laboratories under the guidance of ASTM (1) has done much to help in this regard. However, even more important, in order to use laboratory data to predict field results, it is necessary to investigate effects upon cavitation damage i variations 1

in pressure, temperature, and gas content from the test conditions. Such a study will also assist in increasing the basic understanding of cavitation damage mechanisms, and is important for this reason. The present paper concerns the effects upon damage of pressure and temperature variation in a relatively conventional vibratory facility, and will hopefully provide some of the still required inf o rmation. Very briefly, it is to be expected that the variation of pressure and temperature will affect cavitation damage through several mechanisms, which are more or less important in different cases, e. g.: 1. Change in fluid-dynamic behavior due to change in effective NPSH seen by bubbles. 2. "Thermodynamic " effects upon bubble growth and collapse due to the fact that the growth and collapse begin to vary significantly from isothermal behavior as the temperature is raised. 3. Change in dissolved gas content due to temperature variation. 4. Change in material properties due to temperature variation. II. EXPERIMENTAL PROGRAM A. Experimental Facility The vibratory horn used for these tests is mounted by an "0" ring arrangement into a sealing flange so that the gas space between liquid surface and flange can be pressurized as desired. The tank containing the cavitating fluid is mounted into a heating tank in which cooking oil was used as a heat transfer fluid to maintain the cavitating fluid at the required temperature. The arrangement is shown 2

schematically in Fig. 1. The horn, transducer, and driver are commercial units providing a nominal 2 mil double amplitude at 20 kHz. The correlation between horn amplitude and power setting to the unit was frequently checked using a Fotonic Sensor. Distilled water was used for all the tests with pHk me 7. 4. Typical measurements for cold water tests indicate a gas content somewhat less than saturation. No measurements were made in the hot water tests, but it is presumed from previous measurements on cold water tests that after a few minutes of cavitation, the gas content would be reduced to somewhat less than saturation for the test temperature. Thus the total gas in the high temperature tests would be substantially less than for the room temperature tests, though the entrained portion might be relatively greater due to dissolution effects during the heating process. This variation in gas content may have had some influence on the test results, though recent tests by Hobbs (2) indicates that at least for cold water the effect of total gas content variation over most of the subsaturated range is small. For future tests of this type a facility modification should be made so that tests at constant gas content over a range of temperature variation could be made. All test speci-nens were of SAE 660 bronze. Its room temperature mechanical properties, as measured in our laboratory, are listed in Table I. The temperature range employed (55-250 F) is not enough to cause significant variation in these properties, and corrosion should be negligible. Hence no variation due to these factors need be considered. Sonifier Converter Model J, Branson Instruments, Inc. Fotonic Sensor Model KD-38, Mechanical Technology, Inc.

Table I - Properties of Test Material (SAE 660 Bronze) Yield Strength = 24. 3x10 psi (Y. S.) Tensile Strength =45. 2x10 psi (T. S.) Hardness = 1890 BHN Elastic Modulus = 12. 8x10 psi (E) Elongation = 23 % Area Reduction = 25% B. Tests Performed The tests were divided into constant temperature sets for varying NPSH and constant NPSH sets for varying temperature. Common parameter conditions were used in evaluating the effect of these two primary variables, but the matrix of conditions was not completely filled. Table II shows the conditions used. Four temperatures, ranging from 55 to 250 F and selected for approximately equal vapor differentials, and four NPSH values ( v35 -144 ft.) were used. A total of 9 separate parameter combinations was tested, using 17 specimens. Thus in most cases the results are the average of two specimens. All tests were continued to at least 60 minutes. MDP after 60 minutes was used in the curves to show effects of NPSH and temperature variation. In addition, maximum damage rates (MDPR were computed (Table 3) for each of the 9 parameter conditions. The individual curves are typical for our facility, showing an increasing weight loss rate up to a maximum which usually persists over a reasonable duration, followed by a falling rate. Fig. 2 is a typical weight loss and MDP vs. time curve.,NPSH = (pressure above vapor pressure)/density; horn static. Since suppression pressure was actually maintained constant for runs at different temperature, NPSH for these runs varies slightly due to slight.. Langes in density. MDP = Mean depth of Penetration = (volume loss)/(specimen face area). MDPR - MDP rate. 4

No runs were made for pressures greater than that required for 4 atm NPSH at 250 F (73. 9 psig) because of erratic horn operation at this condition, and limitations in vessel strength. It is likely that the horn will become power limited at sufficiently high NPSH. As NPSH is increased, cavitation occurs during a decreasing portion of the negative pressure part of the sine wave, so that a larger portion of the full ideal sine wave is realized. The power requirement is thus larger. It is not certain that this difficulty was responsible for the erratic behavior observed at 4 atm. NPSH. For all tests the horn was operated at 2 mil double amplitude and 20 kHz. A power meter was used for measuring amplitude, but it was frequently checked with the Fotonic Sensor. C. Experimental Results Fig. 3 shows weight loss (and MDP) vs. duration with the 4 temperatures as curve parameters at 1 atm (nominal) NPSH. The 55 and 250 F curves are nearly identical, as are the 150 and 230 F curves, although the terminal MDP for the intermediate temperatures are about twice those for the extreme temperatures. Fig. 4 is similar with nominal NPSH as curve parameter; all tests at 1500F. Fig. 5 is analogous to Fig. 4, but for 250 F. For both of these figures, there is a strong increase in terminal MDP, and rate, for increase in NPSH. However, the rate of increase is reduced between 3 and 4 atm. NPSH at 250 F (4 atm. data was not obtained at 150 F). The same results can be consolidated in a manner more suitable for showing the overall pattern if 60 min. MDP values (i. e. average MDPR for a one hour test) are used rather than complete MDP vs. time curves. The maximum rates could be used as well, al'though for practical purposes the average rate may be more meaningful. Fig. 6-9 are plots of this type. Fig. 6 is a composite curve showing average one hour MDPR as a function of temperature 5

(vapor pressure also shown as abscissa) with nominal NPSH as curve parameter. An open-beaker curve is included for comparison. The shape of this curve was taken from data previously generated in our laboratory using a 304 stainless steel specimen (Fig. 10). The stainless steel data was normalized to coincide with that from the 1 atm. NPSH SAE-660 test of the present series at the maximum damage temperature of the open beaker test (about 1200F). This is reasonable, assuming the specimen material, does not affect the curve shape, since at this temperature the vapor pressure is almost negligibly small. Thus NPSH values for the two tests are about the same (34 vs. 30.4 ft.). A comparison of the open-beaker test with that for 1 atm. nominal NPSH (for which the curve is reasonably well defined with 4 data points) shows that the open-beaker and constant NPSH curves are of similar shape, with the maximum damage point occurring at a higher temperature for the constant 1 atm. NPSH curve (about 1900F vs. 120 F). Although only 2 data points each exist for the 2 and 3 atm. nominal NPSH curves, and one point for the 4 atm. curve, it is assumed that their shapes are similar. The curves are drawn to indicate an increasing maximum damage temperature for increasing NPSH. This assumption is made on the theoretical grounds that the thermodynamic restraints, which increase with temperature, would become relatively less important as the collapsing head seen by the bubbles (increases with NPSH) is increased. This curve shape is also consistent with our available data points, as well as most of the curves of Peters and Rightmire (9) at constant pressure. Fig. 7 is entirely analogous to Fig. 6 except that the abscissa is proportional to vapor pressure instead of temperature (which is also shown along the abscissa). Since the bubble collapse pressures and velocities are a direct function of pressure rather than temperature, this seems a logical method of presentation. The 6

curves do indeed appear better behaved when plotted in this fasb ion, giving a better basis for extrapolation to higher vapor pressures and temperatures. Fig. 8 shows average MDPR as a function of nominal NPSH (suppression pressure also shown on abscissa) for 1500 and 250~F, the two temperatures for which sufficient data points were available. The open-beaker test point (Fig. 6 and 7) is included at its own NPSH at 150 F, as part of that curve. The increase of average MDPR with increase in NPSH is approximately linear at 150~F, but the rate of increase decreases somewhat as NPSH is increased. o0 1.18 For 250 F, over the whole range from 1 to 4 atm. AMDPR I/ — ANPSH However, the rate of increase decreases at higher NPSH. From 1 to 2 atm. the exponent is 1. 62, from 1 to 3 atm., n= 1. 37, and from 2 to 4 atm., it is 0. 745. In all cases there appears to be a threshold NPSH of 1/4 to 1/2 atm. below which damage would be close to zero. Fig. 9 is entirely analogous to Fig. 8 except that the average MDPR is based on the actual damaged area which decreases for increasing NPSH as discussed later. In this case the increase of damage with NPSH is more than linear in all cases. For 150 F, _A MDPROC ANPSH 1.15 over the whole range, and the exponent at 250 F is approximately 1. 27 throughout. The changing damage patterns with NPSH and temperature are shown in Fig. 11 and 12 which are low-magnification photographs of the damaged surfaces for 150 and 250 0F respectively. 7

III. DISCUSSION OF RESULTS A. Comparison with Previous Pertinent Results The present paper, with the exception of Peters and Rightmi re (9), is to the writer's knowledge the first publication of the results of a combined investigation of temperature and pressure effects upon cavitation damage. However, there have been several previous studies reporting upon effects of the variation of either separately. The effect of temperature variation in an open-beaker test, i. e. constant pressure rather than constant NPSH, has been included in several papers (e. g., 3,4, 5, 6, 7,8, 9). Generally a curve very similar to Fig. 10 results showing a maximum damage rate at about 120 0F and monotonically decreasing damage for either higher or lower temperatures (Fig. 10 from ref. 6 is included for convenience). Two recent Russian publications (4, 5) have shown a minimum rate at a relatively low temperature (105 F), and then an increasing rate for lower temperatures. There is a continuous decrease in damage as temperature is increased beyond the maximum damage temperature,(140 F), as is usual. Fig. 11 shows the shape of this curve (4). A third Russian paper (8) reports a monotonic decrease in damage with temperature from about 70-200 F. No results on temperature effect fox constant NPSH (rather than constant pressure) have been previously reported to the writer's knowledge. The results of Peters and Rightmire (9) for temperature variation at constant pressure appear to show rather conflicting results. The effect of pressure variation at a single temperature has been reported by several (e, g. 2, 9, 10, 11). With the exception of the first reported investigation of this sort, Peters and Rightmire (9), all studies have shown a strong increase of damage rate with increased pressure. In their test at 100 C they found a 8

complex curve which increased to a maximum, as NPSH was increased, at about 1 atm. NPSH, a minimum (factor ofv 2) at 1-1/3 atm. NPSH, a second maximum (about same magnitude as first) at 2 atm., and thereafter a decrease by a factor of -o 10 at 2-1/2 atm. Their facility was of the low frequency type (6. 5 kHz, 3. 65 mils), and hence differed substantially from the others for which results are reported (ranging from 15 to 25 kHz). No explanation for their unusual result is apparent other than through a changing balance between the countering mechanisms of increased bubble collapse intensity at higher pressures along with reduced number and size of bubbles. With a given unit it is obvious that cavitation damage will disappear for high enough pressure since no cavitation bubbles will be formed. The present tests show that the change in MDPR is approximately proportional to the corresponding change in NPSH over the range of 1 to 4 atm. NPSH. On the other hand the tests of Hobbs (2), Plesset (10)in cold water, and Young and Johnson (11) in 800 F sodium show that damage variation is proportional to about the square of NPSH variation. This is a relatively accurate description of the Hobbs and Plesset data. However, the exponent in the Young-Johnston tests varies between about 1. 0 and 2. 7 depending upon the material and interval selected. In the Peters-Rightmire test (9) with 1000C water there was very little damage increase with pressure increase, but rather a decrease in a portion of the pressure range. Another parameter which may importantly influence damage rate is gas content. It is the entrained gas portion which is presumably important, but this may in general be expected to vary with total content. Extreme range effects seem obvious. Near-zero gas would appreciably increase liquid tensile strength and reduce the number and size of cavitation bubbles, though increasing the collapse violence for those bubbles which were formed. At the other e xtreme, large amounts 9

of gas would increase the number of bubbles but cushion their collapse so that little damage would be done. The above hypotheses are consistent with Hobbs tests (2) where a substantial dimunition of damage for a gas content below about 20% saturation, and a gradual dimunition as gas content is increased above about 70% saturation is indicated. Quite similar results were also found by Hansen and Rasmusssen (12) using a rotating disk apparatus. Somewhat different results are shown in papers by Bebchuk and Rozenberg (8) and Sirotyuk(13). Tests are reported by Bebchuk using a 8 kHz magnetostrictive horn in cold water. He reports a more than ten-fold monotonic decrease in damage as gas content in increased from 10% to 100% saturation. Sirotyuk compares the damage rates for "degassed" and "ordinary" water as a function of temperature. The degassed water is always most damaging, but the ratio decreases from 4 to about 3 as temperature is raised from 20 to 800C. In his experiment, the specimens are placed in a cavitation field generated by a 28. 5 kHz transducer. As already mentioned it is believed that the gas content in the present tests remains a fairly constant portion of saturation ( A70%) at the test water temperature, which differs in the various tests. Since the solubility of gas in water decreases with increasing water temperature, the total gas content in the higher temperature tests is probably less than in those at lower temperature, though the entrained gas portion may conceivably vary inthe opposite direction. However, it is impossible to isolate the effect of gas content on damage in the present tests. B. Additional Points Concerning Present Test Results In the vibratory test, an increase in NPSH affects the fluid-dynamics of the cavitating regime in two ways. It decreases the extent of the bubble cloud, concentrating it toward the center of the specimen, since the pressure oscillation induced by the horn is reduced 10

near the specimen outer radius. It also increases the driving head for bubble collapse, rendering individual collapses more damaging. The present results show a somewhat decreasingly strong increase in MDP as NPSH is increased. However, it is obvious for an individual horn that this trend could not continue to very high NPSH, since eventually the horn oscillation would not be sufficient to cavitate the fluid at all. Fig. 11 and 12, consistent with the sodium data of Young and Johnston (11), show the increasing extent of the outer undamaged annular region as NPSH is increased in the present water tests. Additional radial damage striations, with annular regions of relatively light damage succeeding heavier damaged regions at larger radius are also shown. This feature, and the relatively periodic nature of the circumferential damage pattern, are suggestive of a very short sonic wave-length in the two phase mixture in this region. In the present tests an increase in NPSH was brought about by an increase in pressure. In previous tests in this laboratory, the effects of NPSH variation were investigated through a change of fluid density with pressure maintained constant. For this purpose tests were conducted on fluids ranging from mercury (13. 6 g/cc) to molten lithium (0. 5 g/cc). As NPSH was increased in this fashion (Fig. 14) by a factor of about 27, the same effect of a concentration of damage toward the center was observed. Thus the range of NPSH variation in the density-varying tests was much greater than in the present tests, where the effective range of NPSH is really quite moderate, although its precise definition is complex. The variation in static pressure which was used (3 atm. ), is only,- 7%o of the half sine wave pressure amplitude for the 2 mil,20 kHz operating condition (,' 43 atm.). Note that if NPSH is changed by density change, both static and oscillating portions are affected in equal proportion. 11

IV. CONCLUSIONS For a typical 20 kHz, 2 mil conventional vibratory cavitation damage test in water at 150 and 250 F, it has been shown that damage rate increases about in proportion to the corresponding increase of static NPSH (measured with horn stationary) from 1 to 4 atm. NPSH. A more rapid increase of damage with increase in NPSH has been shown by Young and Johnston (11) for molten sodium, and Hobbs (2) and Plesset (10) for cold water. They show a damage increase roughly with the square of NPSH increase. This exponent varies considerably with material and NPSH interval for the sodium tests (11). In the present tests there appears to be a threshold NPSH of 1/4 to 1/2 atm., below which damage would be nearly zero. This observation is consistent with the other tests reported (2, 10, 11). It has also been demonstrated that an increase in NPSH, brought about either by pressure increase or density decrease, causes a decrease in the radial extent of the undamaged outer rim and an accumulation of damage toward the center. The effect of temperature increase at pressures above one atmosphere has been shown to be similar to that at one atmosphere, i. e. maximum damage at an intermediate temperature, and damage decreasing monotonically as temperature is either increased or decreased from this value. The decrease at high temperature is ascribed to the increasing importance of thermodynamic restraints on bubble collapse, but the mechanism for the decrease at low temperature is not clear. It may involve an increasing gas content at low temperature. The general existence of a monotonically decreasing damage rate for temperatures below the maximum damage temperature is disputed by two recent papers (4, 5). 12

An examination of previous literature indicates in general that damage decreases significantly in an atmospheric test for an increase in total gas content in the range of 70 - 100%o saturation at standard temperature and pressure. There is some evidence that it also decreases for very low gas contents (2,12), which the present writer ascribes to the increased water tensile strength under these conditions and the resulting dimunition in size and number of bubbles. The decrease at low gas content is disputed by two previous investigators (8, 13). The present tests cannot isolate gas content variations from temperature and NPSH changes. 13

ACKNOWLEDGMENTS Financial support for the accumulation of most of the experimental data used in this paper was provided by the Worthington Corporation, Harrison, N. J. Most of the theoretical work required for the formulation of the paper was supported by the National Science Foundation under Grant No. GK-1889. The writer is also indebted to Messrs. D. J. Kemppainen, C. L. Kling, D. O. Rogers, and E. E. Timm for the formulation and supervision of the test program and to Messrs. G. Cgranowski, B. Anderson, and D. Getz for the performance of the tests, all research assistants in the Cavitation and Multiphase Flow Laboratory of the Mechanical Engineering Department, University of Michigan. 14

BIB LIOGRAPHY 1. C. Chao, F. G. Hammitt, C. L. Kling, D.O. Rogers, "ASTM Round-Robin Test with Vibratory Cavitation and Liquid Impact Facilities of 6061-T 6511 Aluminum Alloy, 316 Stainless Steel, Commercially Pure Nickel", ASTM Symposium on Characterization and Determination of Erosion Resistance, Atlantic City, N. J., June 1969. 2. J. M. Hobbs and A. Laird, "Pressure, Temperature and Gas Content Effects in the Vibratory Cavitation Erosion Test", ASME 1969 Cavitation Forum, Evanston, Ill., June 1969. 3. R. E. Devine and M.S. Plesset, "Temperature Effects in Cavitation Damage", Rept. 85-27, Div. Engr. and Appl. Sci., Cal. Inst. of Tech., April 1964. 4. B. K. Makarov and A. V. Kortnev, "Thermodynamic and Statistical Methods for Investigation of Ultrasonic Cavitation", FTD-HT-23-1242-67 (trans. from Acoustics and Ultrasound, Tekhnika Publisher, Kiev, 1966). 5. L. M. Dimenshtien and D. V. Litnak, Metalloved. Term. Obrab. Metal. 4, 59-62, 1968. 6. R. Garcia and F. G. Hammitt, "Cavitation Damage and Correlations with Materials and Fluid Properties", Trans. ASME, J. Basic Engr. D, 89,4, 753-763, Dec. 1967. 7. A. Thiruvengadam and H.S. Prieser, "Cavitation Damage in Liquid Metals", Rept. No. TR 467 (NASA CR-72035), Hydronautics, Inc., Nov. 1965. 8. A.S. Bebchuk and L. D. Rozenberg, "Dependence of Cavitation Erosion on the Solubility of a Gas Under a Liquid", Soviet Physics Acoustics, 6, 4, 496-498, April-June, 1961. 9. H. Peters and B. G. Rightmire, "Cavitation Study by the Vibratory Method", Proc. 5th International Congress Appl. Mech., 614, 1938. 10. M. S. Plesset, "Cavitation Damage Measurements with Nickel, Aluminum, and Stainless Steel for the ASTM Round-Robin", Rept. No. 85-42, Div. Engr. and Appl. Sci., Cal Inst. of Tech., Oct. 1967. 15

11. S. G. Young and J. R. Johnston, "Effect of Cover Gas Pressures on Accelerated Cavitation Damage, " NASA TMX-52414, 1968, to be published Proc. ASTM. 12. B. W. Hansen and R. E. H. Rasmussen, "Cavitation Damage Experiments in a Rotating Disc Apparatus Especially with Regard to the Gas Content of Water", J. Ship Research, 1968, 12(2), 83-88. 13. M.G. Sirotyuk, "The Influence of Temperature and Gas Content in Liquids on the Cavitation Process", Acoustics Journal, 12, 1, 87-92, 1966. 16

TABLE II - TEST CONDITIONS Temperature Vapor Vessel Suppression NPSH*'< T ( F) Pressure Pressure Pressure P - P p (psi) P (psig) p - (atm) r( 55 0.21 0.2 1 34.0 150 3.72 3.7 1 34.6 18.4 2 69.3 33.1 3 104.0 230 20.78 20.8 1 35.6 250 29.82 29.8 1 36.0 44.5 2 72.0 59.2 3 108. 0 73.9 4 144.0 "Suppression pressure (pressure above vapor pressure) and NPSH are different only by the factor of specific weight, y(T). Because the specific weight varies somewhat with temperature a constant suppression pressure does not necessarily indicate constant NPSH. It can be seen from Table I, however, that the variation is very small; and for this analysis it is ignored. The terms suppression pressure and NPSH are used interchangeably, i. e., NPSH may be given in atmospheres or feet. 17

TABLE III - MAXIMUM DAMAGE RATES Condition Peak Wt. Loss Peak MDPR Rate 55~F 1 atm..73 mg/min 1. 28 mil/min 150 1 (Fig. 11) 1.60 2.80 150 2 (Fig. 11) 3.95 6.94 150 3 (Fig. 11) 7. 20 12. 62 230 1 1.14 1.98 250 1 (Fig. 12).85 1.49 250 2 1.76 3.09 250 3 (Fig. 12) 5.43 9.51 250 4 (Fig. 12) 8.59 15.05 18

BRANSON 250 WATT POWER SUPPLY r-, ULTRASONIC DRIVER THERMO COUPLE l TO OVERPRESSURE LINE FLANGE SUPPORT — EXPONENTIAL HORN CONTAINMENT VESSEL TEST SPECIMEN OIL BATH 2691a Fig. 1. Schematic of Ultrasonic vibratory horn 19

6 200 180 5 160 140 4 120 0 3 1S/:!I, OlI U 80 2 60 40 0 2511 & 2520 F]n~~~~~~ ~~~O 2194 20 F / (72 ft. NIRSH) I I I I I 2671 15 30 45 60 75 90 Time (min. ) Fig. 2. Cumulative Damage vs. Test Duration of SAE-660 at 250 F and 2 atm. NPSH 20

6 200 180 5 160 140 4 o 120o bo 100 -, 3:~~~~~~~~~~~ ~~~~230oF ~ 80 1500F 2 60 40 250 F 20 — ~~~~~~~~55OF 20 _ _ (34-36 ft. NPSH) [[[[I 1 2683 15 30 45 60 75 90 Time (min.) Fig. 3. Effect of Temperature at 1 atm. NPSH 21

200 3 atm (104.0 ft. NPSH) 180 5 160 2 atm'69.3 ft. N P H) 140 4 120 068 0 100 3 80 1 am (34. 6Effect ft. NPSH) -2 60 40 20 2684 15 30 45 60 75 90 Time (min.) Fig. 4. Effect of NPSH at 150~F 22

200 180 4 atm (144 ft.) 160 140 3 atm (108. 0 ft.) 120 1o | / / / 2 atm (72 ft. ) 100 U 80 60 40 1 atm(36 ft. ) 20 2685 15 30 45 60 75 90 Time (min.) Fig. 5. Effect of NPSH at 250 0F 23

4 atm 6 (144 ft. ) 200 180 atm 5 (104-108 ft. ) 160 140 4 120 o 0 c4 /l11("6 9 - 72 f t. 3%o 100 O 80 uF 1 iatm 634a36 ft. 2 60 40 \ Open Beaker - I I I 1' 1 2688 0 50 100 150 200 250 Temperature (OF) I I I I I I I I I I I I I.1.25.50 1 2 3 4 6 8 10 15 20 25 30 35 Vapor Pressure (psia) Fig. 6. Damage at 60 minutes vs. Temperature 24

4 atm. (144 ft.) 6 200 m (104-108 ft. ) 180 5 160 140 1 "o, -m ft 100 12 _ L I I I I' I I I I I I 60 1200 180 190 200 210 220 230 240 250 260 1 att o (34-36 ft.F) S0 40 \ Open Beaker Fig. 7. Damage at 60 m. vs. Vapor Pressuresia) 25

180 160 500F 140 b O -4 120 0 4~~~~~~~~ ~20 F 20 80 60 40 X Open beaker test 20 o I I I I2687 0 30 60 90 120 150 Approx. NPSH (ft.) I I I I 0 1 2 3 4 Suppresion Pressure, p- Pv (atm) Fig. 8. Damage at 60 minutes vs. NPSH 26

8 7 I /, 4 0 30 60 90 150 Approximate NPSH (ft.) 2816 0 1 2 3 4 Pressure, p - Pv (atm.) Fig. 9. Damage at 60 minutes versus NPSH based on Actually Damaged Area 27

.16 Pv=1. 69 psi NPSH = 30.4 ft..14 - i.12 Pv=0. 82 ps' NPSH=31.3 t. kh ~~~ |I~~~ /~~ ANPSH = 26.8 ft..06 _ Pv=0. 21 psi NPSH = 33.4 ft..042 from Reference 1 p =7. 51 psi NPSH = 17. 9 ft. 0 50 100 150 200 2689 Temperature (OF) Fig. 10. Damage Rate vs. Temperature for 304 SS in Open Beaker Test 28

Specimen No. II-1-M Pressure: 1 atm. NPSH Duration: 60 min. Weight Loss: 59.5 mg. 2675 Specimen No.: N-5 Pressure: 2 atm. NPSH Duration: 90 min. Weight Loss: 144 mg. 2676a Specimen No. N-7 Pressure: 3 atm. NPSH Duration: 90 min. Weight Loss: 219 mg. 2( 7a Fig. 11. Effects of NPSH on specimen damage pattern for a water temperature of 150~F. 29

Specimen No.: N-3 Pressure: 1 atm. NPSH Duration: 90 min. Weight Loss: 42.5 mg. Specimen No.: 2522 Pressure: 3 atrm. NPSH Duration: 90 min. Weight Loss: 130 mg. Specimen No.: N-10 Pressure: 4 atm. NPSH Duration: 90 min. Weight Loss: 123 mg. Fig. 12. Effect of NPSH on specimen damage pattern for a water temperature of 250~F. 30

100 90 Cavitation Erosion of 80 Aluminum Foil Makarov and Kortnev 70 60 50 40 30 20 10 0 I I I 2817 0 20 40 60 80 Temperature (~C) Fig. 13. Weight Loss versus Temperature for Cooling 31

(1) 12 tfour Exposure (2) 6 Ilour Exposure Pb-.Bi at 500QF Pb.Bi at l 500OF (3) 12 tour Exposure (4) 12 Hlour Exposure Mercury at 500F M cury at 7 f'ercury at: 7~.F (5) 36 IHour Exposure (6) 10 1Hour EIxposure Water at 70OF i.thium at 50 Fig, 14. Effects of various fluids on specimen 2818 damage pattern for 316SS 32