Edward Valves Quick-Closing Equiwedge Isolation Valves... Ongoing Qualification. V-Rep 82-2

Transcription

1 Edward Valves Quick-Closing Equiwedge Isolation Valves... Ongoing Qualification V-Rep 82-2

2 By E.A. Bake, Research Manager and J.B. Gallagher, Senior Development Engineer, Flow Control Division, Rockwell International First Published 1982 Following nearly a decade of development, production, and qualification of balanceddisk globe valves for critical applications in nuclear power plants, Rockwell undertook a program in the mid-seventies to develop quick-closing versions of the new Equiwedge gate valve. The program involved extensive analyses and testing to prove the valve design and materials combinations; even more, the program involved development and qualification of a broad line of stored-energy actuators to provide the force, speed, and reliability demanded for this service. Introduction The actuator qualification work was guided by requirements published in such standards as IEEE-323, IEEE-344, and IEEE-382 (considering the changing requirements of each new draft). Although these standards did not apply directly to the qualification of the Equiwedge gate valve itself, criteria similar to those normally involved for actuators were applied in the program to qualify the valve. While the program can be considered successful in that over 100 large Equiwedge valves with Rockwell actuators have been sold for main steam or feedwater isolation service, the program is never complete. Requirements of customers and regulating agencies, domestic and international, escalate regularly. This article can be considered as a progress report on the ongoing work to prove reliability of the quick-closing Rockwell Equiwedge gate valve. Background Valves used to isolate potential pipeline ruptures in nuclear power plants have some of the most demanding reliability requirements yet faced by the valve industry. For 14 years, Rockwell has been a major supplier of such valves for water and steam systems in nuclear plants all over the world. Initially, all such valves were built in the Flow Control Division plant in Raleigh, North Carolina; however, our French subsidiary, Rockwell Valves S.A. (RVSA), has also been building these valves in more recent years. Valves built in either the U.S.A. or France have been furnished for applications in pressurized water reactor, boiling water reactor, heavy water reactor, and breeder reactor power plants. While some of the larger Rockwell -supplied special isolation valves have been for other applications (some involving quickopening rather than quick-closing performance), the vast majority of these valves are intended to isolate the nuclear containment in the event of a line rupture, loss of coolant accident, or seismic event to protect the safety of critical nuclear plant components and systems. The valve assemblies (valve, actuator, and functional accessories) are referred to as safety related and active because operability is necessary to ensure safety during or following certain specified accidents. Historically, most codes and standards related to valves were focused on proof of pressure boundary integrity. Proof of functional integrity involved little more than operational tests and seat leakage tests often under rather ideal laboratory or manufacturing plant conditions. With the growing recognition of the importance of operability of active safety-related valves in nuclear plants, there has been a corresponding growth in the demand for qualification of such valves. The word qualification has taken on a meaning beyond that given in most dictionaries; generally, it refers to proof or assurance that a valve assembly will operate as specified, even when exposed to conditions which would exist during severe accidents in nuclear power plants. Some things can be qualified by analyses alone, but other things require testing (or a combination of analyses and tests). However, more emphasis has been placed on testing rather than analysis for this proof. Even when equipment is qualified by analysis, because it is too large for a meaningful test, it is generally required that some partial testing be performed to verify the analysis. These new requirements are reflected in part by IEEE- 382 (1980). Although the basic intent has not been changed, the 1980 edition, when compared to the 1972 edition, is much more specific in the types and levels of the environmental parameters, The current drafts of IEEE-323 and IEEE-344 also reflect a similar trend. Today, equipment qualification has become a continual program. The constantly changing requirements have resulted in original reports being supplemented with partial type tests or analyses. To stay abreast of these everchanging requirements and in anticipation of future requirements, Rockwell is constantly performing qualification tests on its valves and actuators. This 2

3 A N O L K M J VALVE ACTUAR E D VALVE ACTUAR B 1) HIGH PRESSURE RESERVOIR, 2) BALL VALVE, 3) EXHAUST MUFFLER, 4) DELAY TIMER, 5) PRESSURE SWITCH, 6) RUPTURE DISK ACTUATION SOLENOID VALVES, H 7) RUPTURE DISK PRESSURE REGULAR, 8) UPSTREAM PRESSURE TRANSDUCER, 9) VALVE CENTER CAVITY PRESSURE TRANSDUCER, 10) DOWNSTREAM PRESSURE TRANSDUCER 11) LINEAR POTENTIOMETER 12) VALVE STEM STRAIN GAGE, 13) VALVE ACTUAR HYDRAULIC PRESSURE TRANSDUCER, 14) VALVE ACTUAR GAS PRESSURE TRANSDUCER, 15) AIR RESERVOIR PRESSURE TRANSDUCER, 16) RUPTURE DISK ASSEMBLY I TEST VALVE RUPTURE DISK ACTUATION MANIFOLD P F G C PRESSURE INLET PRESSURE INLET FIGURE 1: Flow Interruption Test Schematic Size 16 Rockwell Equiwedge Gate Valve will ensure that the equipment offered by Rockwell for commercial nuclear service is adequately qualified not only to meet today s requirements but also those which may arise in the future. Rockwell has published a long series of technical articles on analyses and test programs related to its quick-closing isolation valves, but this article will focus on recent extensions of qualification work on Rockwell Equiwedqe gate valves combined with Rockwell Type A actuators. Earlier Rockwell quick-closing valves, based on balanced-disk versions of Flite-Flow globe valves, are still preferred by some users; 3

4 A N R M I O L HYDRAULIC CONTROL PANEL K J E VALVE ACTUAR D B 1) HIGH PRESSURE RESERVOIR, 2) BALL VALVE, 3) EXHAUST MUFFLER, 4) DELAY TIMER, 5) PRESSURE SWITCH, 6) RUPTURE DISK ACTUATION SOLENOID VALVES, 7) RUPTURE DISK PRESSURE REGULAR, H 8) UPSTREAM PRESSURE TRANSDUCER, 9) VALVE CENTER CAVITY PRESSURE TRANSDUCER, 10) DOWNSTREAM PRESSURE TRANSDUCER, 11) BENDING PLATE LOAD CYLINDER (4 REQ), 12) FRAME LOAD CYLINDER, 13) VALVE ACTUAR GAS PRESSURE TRANSDUCER 14) VALVE ACTUAR HYDRAULIC PRESSURE TRANSDUCER, 15) VALVE STEM STRAIN GAGE, 16) RUPTURE DISK ASSEMBLY, 17) LINEAR POTENTIOMETER, 18) AIR RESERVOIR PRESSURE TRANSDUCER BY PASS LINE TEST VALVE RUPTURE DISK ACTUATION MANIFOLD P F G C PRESSURE INLET PRESSURE INLET FIGURE 2: Flow Interruption Test Schematic Size 28 Rockwell Equiwedge Gate Valve however, Equiwedge valves have been selected by an increasing number of users since they have been offered for these services. Test Summary Three major test programs were conducted at the Flow Control Division Test Station, located near Pittsburgh, PA, in the high pressure air flow system. The first test program [1] 1 was conducted with a size 16 Equiwedge. and a prototype A-230 actuator. This series of tests was performed to 4

5 show that a large production Equiwedge could close repeatedly against high differential pressures and high flow rates and still maintain good seat tightness. In the second test program, two valve and actuator combinations were used. The first combination was the same size 16 Equiwedge and A-230 actuator used in the first test program. The second valve/actuator combination was a size 2:8 Equiwedge with an A- 290 actuator. This second program was conducted to demonstrate that the Equiwedge gate valve can close reliably against line rupture flow while simultaneously subjected to pipe and upperstructure loads. These loads represented the forces which would act on the valve during either a line rupture or a seismic event. Although the second program successfully demonstrated the capabilities of the Equiwedge gate valve and Type A actuator, no quantitative measure of the individual effects (e.g., stress and deflection) due to the simulated faulted conditions was made. A third test program was, therefore, conducted to supplement the second program. This testing was conducted on the same size 28 Equiwedge and A-290 actuator. Measurements covered both the individual and combined effects of external loads and line rupture flow on the performance, stress state, and deflection of the valve. As in the case with the second program, this third program also proved the reliability of the valve/actuator combination in closing under severe conditions of loads and flow. Seat Tightness As stated above, the first test program was conducted to show that a large production Equiwedge could repeatedly close against high differential pressures and high flow rates, and still maintain seat tightness. To accomplish the testing, a size 16 Equiwedge was mounted in the high pressure air flow system as shown schematically in Figure 1. In this system, air is stored in the high pressure reservoir (item 1) up to a maximum pressure of 1500 psi (103 bar). At the time this testing was performed, the reservoir had a storage volume of over 300 ft 3 (8.5 M 3 ). With the test valve open, the high pressure air was prevented from flowing to the atmosphere by the ball valve (item 2) located downstream. To establish flow through the test valve, the ball valve was opened in less than 0.5 seconds, releasing the stored high pressure air to the atmosphere. When the ball valve opened, a limit switch located on its actuator sent an electrical signal to a timer. The timer would delay approximately 0.75 seconds, allowing maximum flow to develop through the test valve, and then deliver a signal to the test valve actuator to initiate valve closure. The size 16 Equiwedge was closed against a series of the simulated line ruptures with initial reservoir pressures ranging from 330 psi (23 bar) to 1440 psi (99 bar). The terminal pressures after the valve seated ranged from 70 psi (5 bar) to 1200 psi (83 bar). The terminal pressure is approximately equal to the differential pressure across the valve just prior to seating. Because of the limited high pressure storage volume, partial stroke tests were performed to enable the valve to close against high differential pressures. The tests were conducted with the valve initially 100% open, 50% open and 25% open. After closing against 10 simulated line ruptures, 17 one-minute seat leakage tests were performed with air at pressures ranging from 5 psi (0.3 bar) to 1500 psi (103 bar). The maximum recorded leakage was 6.5 scfh (0.18 std.m 3 /hr) at a test pressure of 1500 psi (103 bar). Line Rupture and Static Load Testing The second test program was specifically designed to address valve closure during a guillotine line rupture or a seismic event. During such faulted conditions, the valve must not only be capable of closing against large flow rates, but it must also be able to withstand any forces such as pipe end loads which result from the accidents. The principal valve used for this testing was a size 28 Equiwedge. This was a full-scale Main Steam Isolation Valve (MSIV) built exclusively for tests. However, to justify some basic assumptions made during the size 28 valve testing, additional tests were performed on the size 16 Equiwedge. The two major objectives of this test program were to demonstrate that the designs of the Equiwedge gate valve and Type A actuator meet the following criteria: 1. The valve body and upperstructure have the structural strength and stiffness required to withstand the forces from a line rupture or seismic event so that the valve can close and seat without binding. 5

6 Fa = AXIAL PIPE LOAD Mh = HORIZONTAL BENDING PIPE LOAD Fu = VALVE UPPERSTRUCTURE LOAD F u F a M h FIGURE 3: Directions of the Applied Loads 6 2. The actuator can operate during the faulted conditions with sufficient capability to close the valve within the specified time limit (typically 5 seconds) and to produce the required stem thrust to effectively shut off the flow through the valve. When a gate valve is mounted in a piping system and subjected to a line rupture or seismic event, it will experience the following two types of loading: 1. Forces applied to its body due to stresses in the attached piping. 2. Inertia forces due to the acceleration of its upperstructure. Since both these types of forces are due to the dynamic nature of the postulated accidents, the absolute force-magnitudes will vary from a maximum value to zero in an oscillatory, though not necessarily harmonic, manner. The valve normally will not experience the full value of a given force for the entire duration of the faulted condi- 6

7 BENDING PLATE P CYLINDERS A B FRAME CYLINDER E BENDING PLATE BOTM CYLINDERS C D T F G U 21 U27 U22 H M U23 I J U24 U25 K L U26 U28 N O U29 Q P R ARE SOLENOLD-OPERATED VALVES ARE PRESSURE-REDUCING VALVES. SOLENOID VALVES ARE SHOWN ENERGIZED. S FIGURE 4: Loading Cylinder Hydraulic Schematic tion. Assuming the maximum force magnitudes are known, a test can be conducted by applying these forces statically to the valve for the entire test duration. Since the valve is held in the maximum deformed configuration, which it would experience for only a portion of the actual accident, this type of testing conservatively demonstrates the structural integrity of the valve and ability to close and seat without binding. Therefore, to achieve the first objective, the approach of static load testing was selected. Because of the interactions between the electric, hydraulic and pneumatic systems, the operation of the Type A actuator, like any other valve actuator, is far too complex to fully test using the static load method. It is generally accepted that only actual dynamic testing is sufficient to prove operational reliability of a valve actuator. In our program, this did not present a problem because the Type A actuator had been previously qualified to IEEE-382 [2]; therefore, its operation 7

8 = STRAIN GAGE = DIAL INDICAR (DIAL INDICARS WERE MOUNTED GROUND REFERENCE STRUCTURES.) FIGURE 5: Strain Gage and Dial Indicator Locations for the Size 28 Static Load and Flow Interruption Tests under dynamic conditions representative of the two faulted conditions was verified. Because actuator operability was proven under dynamic conditions, the second test objective could be met by showing that the actuator can reliably close the valve against line rupture conditions while the valve was subjected to static loading. To realistically simulate line rupture conditions, a valve should close and seat against a differential pressure of more than 1000 psi (69 bar). However, because of the large stroke and flow area of the size 28 Equiwedge, an extremely large storage volume of high pressure air would be required to maintain the desired differential pressure if full-stroke closing tests were performed. Therefore, it was decided that the tests performed with the size 28 Equiwedge would be partial stroke tests which would enable the valve to seat against the desired differential. This approach did not compromise the test results because it can be shown by analysis that the closing forces on a gate 8

9 TABLE NO. 1 SIZE 28 FLOW INTERRUPTION TEST EXTERNAL LOAD COMBINATIONS (SECOND TEST PROGRAM) Horizontal Upperstructure Axial Load Bending Load Load Case kips (kn) ft kips (kn m) kips (kn) (569) (411) (569) (2160) (569) TABLE NO. 2 SIZE 28 FLOW INTERRUPTION TEST INITIAL AND FINAL SYSTEM PRESSURES (SECOND TEST PROGRAM) Initial Final Pressure Pressure Test Condition psi (bar) psi (bar) Baseline (100) (61) Upperstructure Load (100) (74) Upperstructure and Bending Loads (100) (65) Upperstructure and Axial Loads (103) (63) Upperstructure and Bending Loads (102) (51) (100% Valve Travel) TABLE NO. 3 SIZE 28 FLOW INTERRUPTION TEST EXTERNAL LOAD COMBINATIONS (THIRD TEST PROGRAM) Horizontal Upperstructure Axial Load Sending Load Load Case kips (kn) ft-kips (kn) kips (kn) (1000) (2100) 3 46 (205) (1000) (2100) (205) TABLE NO. 4 SIZE 28 FLOW INTERRUPTION TEST INITIAL AND FINAL SYSTEM PRESSURES FOR THE TESTS WITH FLOW (THIRD TEST PROGRAM) Initial Final Pressure Pressure Test Condition psi (bar) psi (bar) Baseline (99) (81) Axial Load (103) (79) Horizontal Bending (99) (79) Upperstructure Load (99) (79) Axial, Horizontal Bending, and Upperstructure Loads (101) (81) 9

10 valve are relatively small until approximately the last 25% of stem travel. But, to substantiate this assumption, several line rupture tests were conducted with the size 16 Equiwedge to experimentally measure closing force as a function of stem travel. These tests were conducted at full stroke, and the portion of stem travel where the closing forces are significant was determined. Although the size 16 Equiwedge is smaller than the size 28 Equiwedge, a large reservoir of high pressure air was still required to have the desired differential at seating. To overcome this problem, Rockwell Flow Control added another 300 ft 3 (8.5 m 3 ) of storage volume to the high pressure air flow system, thus making a total available air storage of over 600 ft 3 (17 m 3 ) at 1500 psi (103 bar). Another reason for performing the preliminary testing on the size 16 Equiwedge was to determine the best method of simulating the line rupture. The test set-ups shown in Figures 1 and 2 had two different methods of simulating line rupture. The first method was to open the ball valve (item 2), located downstream of the test valve, in approximately 0.5 seconds or less. This method was used in our first test program described above. The second method was to use two rupture disks in series (item 16). The two disks in series each had a specified breaking pressure of 1080 psi (74 bar). With the full system pressure of 1500 psi (103 bar) on the upstream disk, rupture was prevented by maintaining 750 psi (52 bar) between the two disks. This pressure was maintained by de-energizing the solenoid valves (item 6) during system pressurization and setting the pressure regulator (item 7) at 750 psi (52 bar). To simulate the line rupture, the solenoid valves were energized, admitting the full system pressure 1500 psi (103 bar) to the space between the disks; this caused the downstream disk to rupture. The loss of pressure between the disks resulted in a differential pressure approaching 1500 psi (103 bar) across the upstream disk, so it also ruptured. Flow was then initiated through the valve. The process which led to valve closing was initiated by a signal from the pressure switch (item 5). The switch actuated because of the decrease in pressure at the discharge pipe downstream from the valve, During the size 16 Equiwedge testing, this sent an electric signal to the actuator, thus initiating valve closure. During the size 28 Equiwedge testing, the switch sent a signal to the delay timer (item 4). After a delay of approximately 1 second, valve closure was initiated. Two flow interruption tests were performed on the size 16 Equiwedge. The first test simulated line rupture using the ball valve, and the second test simulated line rupture using the rupture disk assembly. initial pressures were 1480 psi (102 bar) (test with the ball valve), and 1440 psi (99 bar) (test with the rupture disks). The terminal pressure for both tests was 920 psi (63 bar). Except for minor differences, the two sets of test data indicated that there were no significant differences in either the flow through the valve or in the closing forces. The differences in the two sets of data were the result of a shock wave which, the system experienced due to the bursting of the rupture disks. However, the shock wave effects dissipated in approximately 0.4 seconds and the test data showed the same subsequent characteristics as noted in the test with the ball valve. The rupture disks did establish a more instantaneous flow and, therefore, were more representative of an actual line rupture. With the exception of a baseline test, all the testing performed on the size 28 Equiwedge simulated line rupture by the rupture disks. The major result of the testing of the size 16 valve was that it proved the closing forces, which the valve actuator must overcome, do not become significant until the valve is nearly closed. Analysis of the data, obtained from the stem strain gage (item 12, Figure 1), showed that the required closing force did not exceed the stem blowout force (i.e., stem area x valve internal pressure) until the valve was over 70% closed. The stem blowout force represents about 15% of the total seating force. In order to have a significant margin in the second portion of this program, it was decided to conduct the tests on the size 28 Equiwedge by closing the valve through the last 50% of its closing travel. Prior to performing the testing of the size 28 Equiwedge, substantial renovation of the existing facility had to be undertaken. Even though partial stroke tests were going to be performed, the existing reservoir capacity was again increased by another 300 ft 3 (8.5 m 3 ). This brought the total reservoir volume to over 900 ft 3 (25.5 m 3 ), which provided sufficient air volume to maintain approximately a 1000 psi (69 bar) differential pressure just prior to the 10

11 valve seating. A hydraulic load system was also installed so that the forces shown in Figure 3 could be applied to the valve during the test. Figure 4 shows the hydraulic schematic for the load system. The first four items are hydraulic cylinders which were used to apply loads at the valve ends. Item 5 is a hydraulic cylinder which applied a load to the valve upperstructure. Fluid pressure was directed to either the rod end or piston head end of the cylinders by using the directional control valves (items 16, 17 and 18). The hydraulic pressure was regulated by using the pressure-reducing valves (items 11, 12, 13, 14 and 15). By controlling the direction and magnitude of the pressure in the cylinders, the forces shown in Figure 3 could be applied to the valve in any combination. Because the force output of a hydraulic cylinder is directly proportional to the fluid pressure, pressure gages (items 20, 21, 22, 23 and 24) were used to verity the loading on the valve. The size 28 Equiwedge was mounted and instrumented in the high pressure air flow system as shown schematically in Figure 2. To verify that the system and associated instrumentation were operational, a baseline test was conducted. For this test, line rupture was simulated with the ball valve and the test valve was closed through its full stroke. A flow interruption test was conducted with each of the three load cases shown in Table 1 applied to the valve. In these tests, line rupture was simulated with rupture disks and the valve was closed through its last 50% of full stroke. A fourth test was conducted with rupture disks in which the valve was closed through its full stroke while subjected to load case 3. Table 2 lists the initial and final pressures in the system. In all the tests, the size 28 Equiwedge closed and effectively shut off flow within the 5-second time limit. Additional Line Rupture and Static Load Testing Although the objectives of the second test program were met, a third program was conducted to further evaluate the performance of the Equiwedge gate valve and Type A actuator. The second program demonstrated that the valve/actuator combination can successfully perform during a line rupture of seismic event. However, the testing was qualitative rather than quantitative in that no data were taken to show effects of applied loads on valve stress and deflection. The third program had the same basic objectives as the second program. But in addition, it was also designed to measure the individual effects of the loads and flow, and to evaluate more precisely their combined effect on valve actuator performance. The program consisted of four distinct parts: 1) Baseline Test-The valve was closed against internal pressure with no external loads or flow. 2) Static Load Tests-The valve was closed against internal pressure with various combinations of pipe end loads and upperstructure loads applied to it. These tests were conducted with no flow through the valve. 3) Line Break Flow Test-The valve was closed against simulated line rupture flow with no external loads applied to it. 4) Static Load and Line Break Flow Tests- The valve was closed against simulated line rupture flow with various combinations of pipe end loads and upperstructure loads applied to it. The same three types of loads addressed in the second program (refer to Figure 3) were also addressed in this program. The valve was subjected to the four load cases shown in Table 3 (developed from a customer specification) in two test series. For the first three cases, the valve was subjected to each load individually. For the fourth case, all three loads were applied to it simultaneously. This sequence of loading was performed twice, During the first series, the valve was closed against internal pressure but no flow. During the second series, the valve was closed against internal pressure and flow. In the tests with no flow, the valve was closed through its full stroke; in the tests with flow, the valve was closed through the last 50% of its full stroke. All the testing was performed on the size 28 Equiwedge, and line rupture was simulated with rupture disks. The test set-up was basically the same used in the second program (refer to Figure 2), except strain gages and dial indicators were added to measure stresses and deflections at critical sections of the valve and piping system; refer to Figure 5 for locations. The results from these tests agreed very well with the analytical predictions. Under all 11

12 FIGURE 6: Flow Interruption Test of Size 28 Rockwell Equiwedge Gate Valve. conditions of load and flow, the valve and actuator closed properly. The various test conditions did not adversely affect the closing time, required closing force, or seat leakage. The strain gage data provided good information on the stress state of the valve. The deflection data in conjunction with the strain gage data fully determined the condition of the valve while under load. Conclusion A line rupture event is an extremely complex phenomenon with many variables, such as the compressibility and thermodynamics of the system fluid. Basically, there are too many variables to allow analytical prediction of all effects that this type of accident will have on valve performance. However, with the experience acquired FIGURE 7: Test Set-up for Size 28 Rockwell Equiwedge Gate Valve at Flow Control Division Test Station. through three major programs, Rockwell has obtained a high degree of confidence in the design of valve/actuator combinations which may have to perform during a line rupture accident. These test programs have confirmed, for example, that the seating and guiding surfaces in the Equiwedge gate valve body are designed to withstand 12

13 Flow Control Division Edward Valves Flowserve Corporation has established industry leadership in the design and manufacture of its products. When properly selected, this Flowserve product is designed to perform its intended function safely during its useful life. However, the purchaser or user of Flowserve products should be aware that Flowserve products might be used in numerous applications under a wide variety of industrial service conditions. Although Flowserve can (and often does) provide general guidelines, it cannot provide specific data and warnings for all possible applications. The purchaser/user must therefore assume the ultimate responsibility for the proper sizing and selection, installation, operation, and maintenance of Flowserve products. The purchaser/user should read and understand the Installation Operation Maintenance (IOM) instructions included with the product, and train its employees and contractors in the safe use of Flowserve products in connection with the specific application. While the information and specifications contained in this literature are believed to be accurate, they are supplied for informative purposes only and should not be considered certified or as a guarantee of satisfactory results by reliance thereon. Nothing contained herein is to be construed as a warranty or guarantee, express or implied, regarding any matter with respect to this product. Because Flowserve is continually improving and upgrading its product design, the specifications, dimensions and information contained herein are subject to change without notice. Should any question arise concerning these provisions, the purchaser/user should contact Flowserve Corporation at any one of its worldwide operations or offices. For more information about Flowserve Corporation, contact or call USA FLOWSERVE CORPORATION FLOW CONTROL DIVISION Edward Valves 1900 South Saunders Street Raleigh, NC USA Toll- Free Telephone Service (U. S. and Canada) Day: After Hours Customer Service US Sales Offices Phone: Facsimile: Facsimile: Visit Our Website Flowserve Corporation, Irving, Texas, USA. Flowserve and Edward Valves are registered trademarks of Flowserve Corporation. V-Rep /03 Printed in USA