ML20249C013

Rev 0 to TR-ECCS-GEN-011, ECCS Suction Strainer Hydrodynamic Test Summary Rept

ML20249C013
Person / Time
Site:	Brunswick
Issue date:	06/11/1998
From:	Andersson L, Ashley G, Minichiello J DUKE ENGINEERING & SERVICES
To:
Shared Package
ML20249B959	List: BSEP-98-0125, Submits Supplemental Response to NRC Bulletin 96-003, Potential Plugging of ECC Suction Strainers by Debris in Bwrs. Amended Sections Delineated by Rev Bar in Right Hand Margin.No New Regulatory Commitments Contained in Ltr ML20249B969 ML20249B975 ML20249B988 ML20249C004 ML20249C009 ML20249C013
References
TR-ECCS-GEN-011, TR-ECCS-GEN-011-R00, TR-ECCS-GEN-11, TR-ECCS-GEN-11-R, NUDOCS 9806250177
Download: ML20249C013 (47)

Text

_ _. _ _ _ _ _ _ _ _ - _ _ _ _

1 Duke Engineering & Services Technical Document Cover Sheet l i

l l l l l DOCUMENT NUMBER: TR-ECCS-GEN-011 REV: 0 l

l DOCUMENT TITLE: ECCS Suction Strainer Hydrodynamic Test Summary Report PROJECT NAME: ECCS Suction Strainer Projects WID NUMBER: A16800 CLIENT: Internal l

SOFTWARE USAGE (Retam usagefrom pnor revasions. Ifsail appbcable) i

• Pre-Use Software Name Version Ilardware Platfonn/ Description of Functions, Features, Modules, Verification Operating System Libmries, Modeling Techniques

• Review Sonware Capabilities. Review Open Emr Notices. Ensure InstaHayon Test Completed and Access Control Satisfied; per DPR-3.5  ;

I I

DESIGN VERIFICATION METIIOD DESIGN 1.(EVIEW CRITERIA SOFDVARE REVIEW CRITERIA El U6 M H2 U6 M l

[ DesignReview [O Desir,ninput correctly Selected O E Software CapabilitiesReviewed O AltemateCalculation T O Armmptions Adequate / Reasonable O 7 OpenErrorNoticesReviewed O Qualification Testing O Assumptions Noted for Verification O 7 Software used correctly 7 O AppropriateDesignMethods O 7 SoftwareResultsDocumented I O DesigninputsIncorporatedinDesign O 7 Key Program Features Re ..rded I C Reasonable Output for the Inputs O E Interfacing Organizations Specified jIIh Bjly Preparer Verifier Approver biWW *,%l

• gnature g g W ((gj Printed Name ' John C. hUnichiello 'L. Olof Andersson Gregory R. AslNy

//l$[ Glllff$ OhY98 9806250177 900619324 PDR ADOCK 05 G

REVISION DESCRIPTION SHEET REVISION NUMBER PAGES REVISED AND DESCRIPTION O Initial Issue i

1 i

i I.

(

TR-ECCS-GEN-011, REV. O ii L--____--__--_____ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - - - - - . _ _ _ - . __

DUKE ENGINEERING & SERVICES,INC.

CERTIFICATION OF ENGINEERING REPORT This signature certifies that the above engineering report has been reviewed by me, the undersigned, and is correct, complete, and in total compliance with the applicable requirements.

l 1

i %

Signature: ~  !

G Name: dcwoty L' A5ata y Registered Professional Engineer

• '4s,',h,I,IN0,,,,,

Q9*f./ l /

Date: feb'/9 t, No. & State: 00 2 -of 9(,/g-TR-ECCS-GEN-011, REV. O iii

DUKE ENGINEERING & SERVICES, INC.

. COMPANY STATEMENT l

This report contains information belonging to Duke Engineering & Services, Inc. (DE&S) and is also an unpublished work protected by the copyright laws of the United States of America. Any l reproduction in full or in part of use of the information contained herein is strictly prohibited I. without the express prior written permission of DE&S,215 Shuman Blvd., Suite 172, Naperville, j Illinois 60563, USA.

DE&S assumes no responsibility for liability or damage of any kind which may result from the use, by others, of the information contained in this report.

l Controlled Copy No.1 Date ofIssue: 6/11/98 Issued To: Richard Tripp CP&L Brunswick Plant i TR-ECCS-GEN-011, REV. O iv l

wa-___________-_______________-___

1 EXECUTIVE

SUMMARY

This report presents the results of a hydrodynamic test program conducted by Duke Engineering u & Services (DE&S) to investigate the behavior oflarge capacity stacked disk Emergency Core Cooling System (ECCS) strainers subjected to accelerated fluid flow fields. The purpose of the test program was to generate the data required to support empirically-based values for the effective hydrodynamic (inertial) mass coefficient, C.. Test results were obtained by allowing the test strainer to vibrate in still water and comparing the resulting frequency to the vibration of the strainer in air.

The experimental investigation was designed and managed by DE&S and Dr. David Williams and performed at the EPRI NDE Center Test Tank in Charlotte, NC. The tests were performed using Performance Contracting, Inc. (PCI) Sure-Flow stacked disk ECCS strainers either the same as or similar to those which will be installed in BWR nuclear power plants.

This report describes the test program, test instrumentation, data reduction, and presents the results of the PCI stacked disk strainer hydrodynamic tests. The significant conclusions drawn  ;

from the reduction of the data recorded during the tests are as follows:

• For lateral vibration, the resultant effective hydrodynamic mass coefficient, C., is substantially lower than that for an impervious smooth cylindrical body of same major dimensions. A typical value for cylinders in water is 2.0, including both the displaced fluid (a mass coefficient of 1.0) and the entrained fluid (an added mass coefficient of L 1.0). For the strainers tested, the C. varied from 0.38 to 0.5. Including experimental statistical variation and a 10% margin, the recommended C. for design is less than 0.65 based on the volume displaced by the strainer envelope.

• For axial (longitudinal) vibration, the resultant effective hydrodynamic mass coefficient, 4 C., is substantially lower than that for an impervious smooth cylindrical body of same major dimensions, using an effective volume for the end plates. A typical value for short cylinders (UD ~ 2) in water is about 1.6, including both the displaced fluid (a mass coefficient of 1.0) and the added fluid associated with the end plates (an additional L

f mass coefficient of about 0.6). For the strainers tested, the C. varied from 0.36 to 1.12, based on a spherical reference volume equalin diameter to the disk OD. The recommended value of C. in the axial direction varies from 0.5 to 1.4, for use with L solid end flange covers (or blocked flow such as strainers attached to a ramshead L instead of straight-through flow) and the reference spherical volume based .on the l strainer disk diameter. The high end value is applicable only to strainers where the free end has a cover plate, effectively blocking flow from the core tube outward toward the pool.

, These conclusions are applicable in the lateral or axial direction to PCI Sure-Flow stacked disk Strainers which are installed at BWR nuclear plants. This report discusses in detail those i

parameters which significantly influence the conclusions and which must be evaluated to determine their applicability.

TR-ECCS-GEN-011, REV. O v L _ _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ .

l l

l TABLE OF CONTENTS i

SECTION DESCRIPTION PAGE LIST OF TABLES vii LIST OF FIGURES viii INDEX OF NOTATIONS AND VARIABLES ix l l

1.0 INTRODUCTION

1.1 Background 1 1.2 Approach 2 1.3 Objective 2 1.4 Report Layout 3 2.0 SCOPE OF TEST PROGRAM 2.1 General Description of Experiments 4 2.2 Test Specimens 4

3.0 DESCRIPTION

OF TEST EQUIPMENT 3.1 Test Facility 7 3.2 Test F4uipment 7 3.3 Test Instrumentation 7 3.4 Data Acquisition 8 4.0 TEST PROCEDURE 4.1 Measurement of Strainer Mass and Dimensional 14 Properties 4.2 Transducer Calibration and Data Acquisition Check 15 4.3 Test Matrix and Setup 15 4.4 Sequence of Test Operations 16 4.5 Test Data Collection 18 4.6 Test Data Analysis 18 4.7 Test Monitoring 18 5.0 RESULTS 5.1 Approach to Results Interpretation 22 5.2 Free Vibrations in Air (Lateral and Axial Tests) 22 5.3 Free Vibrations in Water (Lateral and Axial)' 23 5.4 Effective Hydrodynamic Mass Coefficients (Lateral and 25 Axial) 6.0 APPLICATION OF TEST RESULTS 30

7.0 CONCLUSION

S 36

8.0 REFERENCES

37 TR-ECCS-GEN-011, REV. O vi

L 1

l LIST OF TABLES l

TABLE DESCRIPTION PAGE 3.1 Support Hardware Used for a Typical Test 9 4.1 Weights of Covers, Bolts, Nuts, and Additional Mass on Rods 19 During a Typical Test Series 4.2 Accelerometer to FFT Analyzer Data Acquisition Calibration 19 4.3 Planned Test Matrix for a Typical Strainer In Air And Water 20 Lateral or Axial Tests 4.4 Typical Test Series 21 5.1 Quick-Release Lateral Test Results - Vertical Oscillation 27 Frequency 5.2 Quick-Release Axial Test Results - Vertical Oscillation 27 Frequency 6.1 Strainer Hydrodynamic Added Mass Coefficients 35 I

4 l

l 4

I I

TR-ECCS-GEN-011, REV. O vii

l l

LIST OF FIGURES FIGURE DESCRIPTION PAGE 2.1 Typical Strainer Test Specimen with Dimensions 6 3.1 Test Tank Plan View 10 3.2 Measurements of Tank and Fluid During Typical Lateral 11 Submerged Tests 3.3 Typical Strainer Configuration During Lateral Natural 12 Frequency Testing in Water 3.4 Typical Strainer Configuration During Axial Natural Frequency 13 Testing in Water 5.1 Typical Lateral Test in Air (Acceleration vs Time) 28 5.2 Typical Lateral Test in Air (Power Spectrum) 29 i

i l

TR-ECCS-GEN-011, REV. O viii l

h INDEX OF NOTATION AND VARIABLES l

c speed of sound in the fluid C. effective hydrodynamic mass coefficient CT core tube outside diameter D characteristic dimension of the body f . average frequencyin water t f.w average frequency in air g acceleration due to gravity M.w mass during the test in air

.M mass during test in water Naa number ofdisks OAL overall length, not including the cover on the flange end OD_ outside diameter of the added cover plate ODaa outside diameter of the disks

. ODn flange outside diameter OD,, gap cylinder outer diameter t, thickness of the added cover plate tn flange thickness Va(lateral) reference volume for lateral loads equal to the volume displaced ley the strainer envelope Va(axial) reference volume for axial loads equal to a sphere with the disk i diameter Wa d tyPi cal width of each disk w, p typical distance between disks W.;, weight during the test in air Waya,. weight of water." participating" in the free vibration Xo harmonic displacement amplitude y- weight density ofwater at the test temperature v kinematic viscosity of water at the test temperature l

l l

I i.

I i

i.

TR-ECCS-GEN-011, REV. O ix

1.0 INTRODUCTION

1.1. Background The installation and use oflarge-capacity passive suction strainers is a modification to allow BWR operators to comply with the recent USNRC requirements for Emergency Core Cooling Systems (ECCS)(Reference 1). The subsequent analysis for qualification of the new installation requires calculation of hydrodynamic loading typical for submerged structures in a BWR suppression pool.

The lateral hydrodynamic mass for the structural analyses of smooth cylindrical bodies is traditionally based on an effective hydrodynamic mass coefficient, C,,, of 2.0 (added mass coefficient of 1.0) to be applied to the reference volume of the cylinder displacement.

Values for axial vibration have ranged from 1.6 to 2.0 and higher, considering the cylindrical reference volume in addition to the spherical reference volume at the strainer end caps. Until recently theses values were used for hydrodynamic mass of strainers, by default, in the absence of any relevant empirical data. In view of the perforated nature of the strainer, it is believed that 2.0 (lateral) or 1.6 (axial) is a conservative value. This has been demonstrated for lateral vibration in recent tests of prototypical Performance Contracting, Inc., (PCI) Sure-Flow stacked disk bolt-on (cantilevered) strainers (Reference 2).

The objective of the subject tests (Test Plan, Reference 3) was to determine the effective added mass for a series of PCI Sure-Flow w tacked s disk ECCS strainers. Most of the strainers tested are larger than the previously tested (Reference 2) PCI strainers and have a different disk and gap width. The measured submerged " effective mass" for the perforated strainer consists of the steel mass and the total effective hydrodynamic mass. The latter includes the " contained" and the " entrained" hydrodynamic mass. For a nonperforated body the " contained" mass is the enclosed water mass if the body is flooded and the

" entrained" mass is the "added" mass. Compared to a nonperforated body the equivalent added mass for the perforated body is the total measured hydrodynamic mass less the

" contained" mass.

The subject test program was undertaken by Duke Engineering & Services, Inc. (DE&S).

The experimental investigation was designed and managed by DE&S with the assistance of Dr. David Williams and performed at the EPRI NDE Center in Charlotte, NC (EPRI).

Test instrumentation, data acquisition equipment, and equipment operation was provided by DE&S.

This report is a summary version ofReference 18. In this report, references to proprietary data will be provided by referring to the appropriate Sections of Reference 18. Portions of this data may be available only with the permission of the entity which contracted the specific test.

TR-ECCS-GEN-011, REV. 0 1

1 l' l.2. Approach

! The subject test program was designed to obtain an added mass (and C,. value) by application of basic dynamics related to the mass of a vibrating body. Authors such as Blevins (Reference 10) show that a body vibrating in a fluid has an added mass associated with its vibrating, due to the fact that the fluid near the body must be accelerated by the body. By measuring the frequency of a vibrating object in both air and water, with no change in the support stiffness, the panicipating mass of fluid can be determined.

For a smooth (impervious) cylinder, the accelerating flow pressure effects (classic Froude-Krylov forces) need to be applied to the added hydrodynamic mass forces (and the .

contained fluid mass, if any) to obtain the total inenial force. This is the basis of hydrodynamic inertial mass coefficient, C., equaling added mass coefficient, C., plus one, or C. = C, + 1 -

Unlike the situation with the smooth cylinder, for the subject perforated strainers the pressure gradient component is negligible since the integral of the pressure differential over the surface is insignificant. Further, the hydrodynamic mass obtained from free vibration in otherwise still fluid is a measure of both the contained effect and entrained mass. Thus the composite hydrodynamic inertial mass coefficient measured in the free vibration tests is the effective hydrodynamic mass coefficient, C., for use in qualification calculations.

Note, however, that the drag component of the total in-line force will be higher for the perforated member than for an otherwise smooth cylinder. Measured drag coefficients, C4, for the prototypical strainers were approximately 1.1 compared to 0.6 for the comparable smooth cylinder (Reference 6).

'l.3. Objective The objective of the test program was to generate the data required to develop an empirically-based value for the hydrodynamic added mass coefficient (and indirectly an effective hydrodynamic mass coefficient, C,,, and associated reference volume) for use in qualification calculations related to installation of replacement ECCS suction strainers.

l Both lateral hydrodynamic loads (flow normal to the longitudinal axis of the strainer) and axial hydrodynamic loads (flow along the longitudinal axis of the strainer) were the focus of this test program.

l TR-ECCS-GEN-011, REV, 0 2 L

t--------_------ _ _ - - - _ - - - - - - - - - - - - - - - - - - - _ - - - - - - - - - - - - .- - - - - - - - - - - - - - - - - - - - - - - - -

1 A. Report Layout

. The scope ofinvestigation including a description of test specimens is summarized in l

Section 2. The facility and test equipment are described in Section 3 and Section 4 provides an outline of the test procedure. The test results are presented in Section 5 and' discussed in Section 6. Findings and conclusions from the tests are summarized in Section 7.

TR-ECCS-GEN-011, REV. 0 3

2.0 ' SCOPE OF THE TEST PROGRAM 2.1. General Description of Experiments The test program consisted of free vibration tests (pluck tests) of five PCI designed Sure-Floww tacked s disk ECCS Strainer test specimens. The test strainers were subjected to free vibration tests in air and water (submerged in nominally still water), with the strainer axis oriented normal to the oscillating direction for lateral vibration and parallel to the oscillating direction for axial vibration. From each set of tests, the effective hydrodynamic added mass of each test strainer can be derived from the ratio of the free vibration oscillating frequencies in air and water for each direction ofvibration. Three strainers were subjected to both directions of vibration. One strainer was subjected to lateral vibration and one to axial vibration only. For the strainer subjected only to axial vibration, the strainer was tested in two configurations; in one, the flanged inlet opening was covered with a %" thick plate; in the other, the inlet was covered in standard 1/8" (40% open area) perforated plate. For one specimen, both ends of the core tube had solid end covers.

2.2. Test Specimens The test specimens for this program consisted of the PCI Sure-Flow stacked disk ECCS strainers designed by PCI. Figure 2.1 shows an outline of a typical strainer. Appendix F of Ref.18 provides photographs of typical specimens. The strainers tested varied in length, disk diameter, internal core tube diameter, internal stiffeners, and end plate configuration.

The specimens are detailed in the PCI drawings, References 4a - 4d, and the Reference 6 report, page 8. Each test strainer was weighed (traceable to NIST) prior to testing and significant dimensions verified.

The volume displaced by each strainer envelope, including the cover and flange, is used as the reference volume for lateral vibration and is found as follows:

Va(lateral) = x* (OAL*CT' + Naa*waa*(ODaa 2 2

-CT ) + (Naa - 1)*w,,'(OD,,2 - CT2 )

2 2

+tn*( ODn - CT ) + t *OD 2}/[4*1728]

l TR-ECCS-GEN-011, REV. 0 4 u - _ _ -----___ _ _ __ _ _ _ _ _ - - i

i~

l where, OAL = Overall length, not including the cover on the flange end l CT = Core tube outside diameter

= Number of disks .

N4a waa = Typical width ofeach disk ODaa - = Outside diameter of the disks t

w, y = Typical distance between disks OD,, = gap cylinder outer diameter

= Flange thickness -

ta ODa = Flange outside diameter .

t- = Thickness of the added cover plate

- OD- = Outside diameter of the added cover plate The volume of a sphere with a diameter equal to the OD of the disks is used as the reference volume for axial vibration and is found as follows:

Va(axial) = 4 *x*(ODaa')/(3 *8* 1728)

' The important parameters varied during this series of tests were:

e Overall length to disk OD (OAL/ODaa): 0.8 - 2.0 Core tube OD to disk OD (CT/ODaa): 0.44 - 0.67 e Disk OD (ODaa): 36 - 45 in .I e Lateral reference volume (Va(lateral)):  : 21'- 42 ft' Axial reference volume (Va(axial)): 28 ft' .

• Perforated plate hole sizei 3/32 - 1/8 m  !

. . Core tube end plate conditions: both perf. plate; both solid covers; one end perf and one end solid 1

1 TR-ECCS-GEN-011, REV, 0 5 e _ __ - _-_- - - _ _ --__- -

i 1

l l

l

! i FIGURE 2.1 Typical Strainer Test Specimen With Dimensions hE pt 4 i

~ \

1 lnt A e i e parv . .ru su.-a j t t

W FDE3 AM .58 e

/ II- e

-t. b st

. ILTI i l g k 5l li t -tci3 , . -r- -

i i 1 '

l iI .

L@ g{

!@- @U dk/ ,

Me \

k I b.K@il I  !!  ! i  ! ,

I I i il l l I li i! i I .I $g ll 1 ii i e i !!  !! l' j lg .

l i i il  !! i l i !!  !! ! I 3[. 3 5

! ll  !! ! l 1

l" ji i:

w I  :

. i !!. 4 : i  ! ii ii ! i e, l l! !! !I

• i !!  !! il t

!  !!  :- i i  !; ii : i l i l [ [ '. .. .

. ,, w E i !! ii ! ' i !! i.l 1: $*

h b i i il  :' i l 1 i: li ! i l l

i ii  !! ! i i !! it i!

!:  :; I

! !i ! i l

! !!  :' ! i :-  ! il N s I i. i !!  :! ' i'  :

1: i. I i li _ si i . li Li il q- -

i l(@ !' i i 11:  !! !  ! l 7 4,

> >.. M

[  ! si ' ;; ! i ii ii i s  !  ! ;i  : l  : ii  : : i 33-

[

i :; .-* ' .:  : 4 Y I

)) -

-7

~3 , _1  ; .  :. .. : , .

: 4 0 \

il 0

{0

.dh

~ ~

bF 6 ,e i

i i

= = o..w .m

, Ie E E

..w.....

i l ilr g gg gg

.-.....-..u.... <

4 I i . - . . . ..a 1

. m a.. . - .a L TR-ECCS-GEN-011, REV. 0 6 L _ _ _ _ _ _________

3.0 DESCRIPTION

OF TEST EQUIPMENT 3.1. Test Facility -

The tests were performed in accordance with the Test Plan (Reference 3) at the EPRI NDE Center located at Charlotte, North Carolina. All tests took place in the EPRI test tank so that the same arrangement was used for testing in air and in water. The tank is 20 ft diameter and approximately 21 ft deep with a beam spanning across the top of the tank as shown on the Reference 5 drawing. Figures 3.1,3.2,3.3, and 3.4 show schematic drawings of the test tank and a typical test specimen mounted in the tank, with important dimensions shown.

3.2. Test Equipment The test suppon fixture consisted of a standard pipe suppon spring hangar attached to each end of the strainer (for lateral testing) or the top of the end disk or cover plate (for axial testing) as shown in the schematic arrangement, Figures 3.3 and 3.4. Each of the two springs was a standard pipe support variable spring, a Grinnell Fig. B268, Type A (Reference 7). Attachments to the tank cross beam and the lugs on the test specimen were made with standard pipe support hardware: beam clamps and clevises, respectively. Table 3.1 lists the hardware used for a typical set of tests.

3.3. Test Instrumentation Instmmentation consisted of an acceleration transducer (accelerometer) mounted rigidly to either each end of the strainer (lateral vibration) or to the cover near the two lift points (axial vibration), as shown in Figures 3.3 and 3.4. The accelerometers were PCB Piezotronics, Inc. Model 326A13/040AC piezoelectric accelerometers, Serial Numbers 7370 and 7371. Calibration for these accelerometers is provided in Appendix B of Ref. ,

18. Because the natural frequency of the test specimen in water was predicted to be on the order of 1 Hz, the accelerometers were specially calibrated to 0.5 Hz, as shown in Appendix B of Ref.18.

In addition to monitoring the acceleration, the temperature of the tank water was also measured after the test, in order to establish a basis for viscosity and density calculations.

The temperature was measured with a calibrated thermocouple (Fluke Manufacturing Co.

Inc, Model 80PK-2A) and digital thermometer (Fluke Manufacturing Co. Inc, Model 52, Serial No. 4555214). Calibration data for this equipment is provided in Appendix B of Ref.18.

l i

iTR-ECCS-GEN-011, REV. 0 7 f

L

f 3.4. Data Acquisition

. Transducer output consisted of analog signals. The data was input directly to an Ono Sokki Model 920 Dual Channel FFT Analyzer for frequency analysis. In accordance with the Test Plan and Test Procedure, the accelerometer through FFT analyzer signal was checked both pre and post test (see Section 4.2 below) against a known result.- This showed the accelerometers and FFT Analyzer operated properly throughout the test.

The tests used to determine frequency consisted of 40-second sample time history. This gave a frequency resolution of 0.025 Hz (1/40 cycles per second). In addition to the tests for frequency, other tests were run with an 8-second sample. These tests allowed the time history to be enhanced in the time scale. The frequency resolution for these tests, however, was only 0.125 Hz (1/8 cycles per second), so these tests were not used in the subsequent frequency evaluations.

i i

l TR-ECCS-GEN-011, REV. 0 8

TABLE 3.1 Support Hardware Used for a Typical Test Item No. Grinnell Figure Number Description (Strainer #1) 1 292 UFS Beam Clamp, Size 1 2 B-268 Spring, Size 8, Type A

3 146 5/8" Rod, 2'-0" Long 4- 146 5/8" Rod, 6'-0" (Note 1) Long 5 299 Size 2 clesis with (Note 2) pin, S/8" tap size 6 299 Size 2 clesis with pin, 5/8" tap size Note 1: The weight per foot for 5/8" rod is 0.84 lb/ft [Ref. 7]

Note 2: The weight of a Size 2 clevis is ~1 lb [Ref. 7]

Note 3: The weights for the rods and clevises are used in the final calculation for frequency to

. determine the total mass affecting the springs. The weights for the items above the springs are not needed for calculations.

4 I

l l

l l

TR-ECCS-GEN-011, REV, 0 9

I FIGURE 3.1 Test Tank Plan View S12x31 existing beam Top of beam approx. 8 inches below top of tank EPRI Test Tank 20 foot diameter, 3/8" thick,21 feet high l

TR-ECCS-GEN-011, REV. 0 10 l

._ _ _ _____.__u

L FIGURE 3.2 Measurements Of Tank And Fluid During Quad Cities Lateral Submerged Tests (Typical For All Axial And Lateral Tests) n A A 20" Beam

___Y_

- 78" Spnns Hanger Water Y Line A 58"

~ 254 "

_{_l 22.5" Strainer Test Specimen y-g 94.5" Y V i

TR-ECCS-GEN-011, REV. 0 11 i

. . . ]

l l

l FIGURE 3.3 Typical Strainer Configuration Durine Lateral Natural Freauency Testine In Water (see Table 3.1 for the Bill of Materials)

(Not to scale) a==. .:a-I

.___._..________ -.________ ._-______.~ ___ __________--

N N L'?- M; -

v 7M t n

'wW '

S% fmi% I I 1 1 N wrices  !

~ _ _ ~

A i

s b;7 *:ct.r e et ta o @

d EETi-?  % g

/

E me t4

- te) s . __ ...... -

N r d e

. _D__ _ . .__ .__ _ ___ .__ ____. .___D. c smaata I

n

,_ , B

_ ____i......

big ic[I i

i O G-t ELEVATION - DURING " PLUCK" TEST TR-ECCS-GEN-011, REV. 0 12

FIGURE 3.4 Typical Strainer Configuration Durine Axial Natural Freauency Testine In Water (see Table 3.1 for the Bill of Materials)

(Not to scale)

.=:.- .:c j

- - --- -----n---e .. - . - - - . - . - - -

u .N m C O T *) (TW NO m c

'[% .

(*Y:$

(~o - i i i i l \ wEic-Ts ('vo'

-p __. , -

A i l

(TYp) a ,. i i .

' 3.

i 8

i i E E

' I I \

l

. - . - . - . 4- - - - - +- . - C STR AIN ER 1 I n I I I l I I l 1 E I i- ,

I f

, , 4 c,.2 i i 1

, A**I.E C YE'E3

" [IU EWit

=EM-$2E:"

O C+

y ELEVATION - DURING " PLUCK" TEST l

l j TR-ECCS-GEN-011, REV. 0 13 l

l l

l E-______________._____________ _ _ _ _ _ . _

l l.

4.0 - TEST PROCEDURE 4.1. Measurement of Strainer Mass and Dimensional Properties As explained in Section 2, each test strainer had been weighed prior to shipment to EPRI for testing, with weights ranging from 950 - 2600 lbs. Significant geometric dimensions, as indicated in the outlines ofReference 4a - 4d and Reference 6, page 8, were checked at EPRI with no discrepancies noted. Appendices C (lateral tests) and D (axial tests) ofRef. I8, the Test Log sheet labeled " Pretest Data Observations", provide the record of the dimensional data noted at the test site.

Three different sets of Grinnell springs were used for the testing._ Each set had a different spring constant and was chosen so that the mass would be approximately centered during the testing in water. Since the springs are stock springs, they have a minimum and a maximum load reading. Therefore, in order to ensure that sufficient travel was available for the testing in water, when buoyancy would reduce the sprung weight, DE&S added weights to the rods below the springs, but high enough so that they would not be under water during the in-water tests. These weights thus served two purposes:

. For strainers with the center of gravity (CG) shifted to one side, due to the presence of the flange and cover, the weights could be added to one rod so that the springs would have approximately the same reading.-

+ For strainers with the CG centered, as it typically was in axial testing, the weights could be uniformly distributed to the two rods and provide each spring with sufficient load so that the displacement during in-water testing did not have to be reduced from the desired target.

Table 4.1 lists the weights used during a typical test. Ten 10 lb weights were purchased from a' national su;, ply house, with certification paper on each weight, as shown in Appendix B of Ref.18. These weights were supported on the rods with %" thick steel plates above and l below the weights, with clamping nuts so the weights would not rattle. The support plates weighed 5 lb each, as verified by comparison to the 10 !b weights. Table 4.1 also includes the weight of the added cover and nuts and bolts used to attach the cover to the flange.

p TR-ECCS-GEN-011, REV. 0 14

l

-4.2. Transducer Calibration and Data Acquisition Check To establish a benchmark for and verify proper functioning of the recording equipment, the accelerometers were attached to a ponable 1-g shaker and the results recorded. The shaker was a PCB Piezotronics, Inc., Model 394C06 Hand Held Calibrator, Serial No.1359, calibrated as shown in Appendix B ofRef.18. The operating frequency of the ponable shaker is 159.2 Hz, as shown in Appendix B ofRef.18. Each accelerometer was mounted to the

' portable shaker, the output measured by the FFT analyzer, and recorded in the Test Log

- (Appendices C and D of Ref.18). Table 4.2 shows the results of the input to output calibration, both pre-test and post-test-in-water for a typical set of strainer tests. The results indicate that the accelerometers and FFT analyzer functioned properly throughout the testing, with no significant change in results and negligible error (0.5%) between measured and input frequency. The results were consistent across each set of tests. For a specific test, see the Test Log in Appendices C and D of Ref.18.

4.3. Test Matrix and Setup A typical test matrix is shown in Table 4.3. The test specimen was tested by pulling down on the center of the specimen and recording the oscillating accelerometer signals after quick-l release of the tensioned system.

l Since the purpose of the test was to measure frequencies in air and water, any reasonable -

- deflection could be used. However, it is important in water that the initial deflection be large 4

enough so that the velocity produces a Reynolds number above 10 , which is considered the lower bound for relatively constant C,.. Thus, the deflection in air is not critical, but that in i . water is. For that reason, it is important that the deflection in water be about 1", if possible.

E ~ Therefore, while the spring readings in air may be toward the low (bottom) end of the spring l.

scale, and the in-air deflection limited to prevent the springs from bottoming out, the readings in water were more toward the middle of the spring scale. This is the primary reason for the difference between the planned in-air and in-water applied deflections. Where possible the springs were deflected as much as 1" in air or 1-1/4" in water.

The test specimen was installed with the tank empty. It was suspended from the tank top l beam with the spring supports as shown in Figures 3.3 and 3.4. The approximate location of l each spring travelindicator was typically in the center or slightly below. Both springs had  !

about the same reading. This indicates the strainer weight was approximately uniform along l its length, so that the dominant mode would be a transnational mode (heave) as opposed to a pitching mode (one side up and the other side down). The accelerometers were rigidly j mounted to the test specimen oriented so as to measure vertical response. For lateral  !

vibration, the accelerometers were mounted to each end of the strainer. For axial vibration, l they were both mounted on one of the end plates, at the opposite ends of a diameter. l l

l TR-ECCS-GEN-011, REV. 0 15

I ..

1 i

l 4.4. Sequence oftest Operations The test series consisted of the following separate stages:

(a) Preliminary Free Vibration Testing / Data Acquisition Preliminary free vibration testing was done in air to ensure all equipment was working properly and the responses were as expected. Based on the fact that each Strainer tested was part of a series of tests in which the same equipment was used, this phase of testing was usually kept to one test. The response data (acceleration of the oscillating spring-supported strainer) was recorded by the data acquisition system during the test. The following operations were performed for the preliminary free ,

vibration tests:

. Applying a vertical downward load to the center of the test specimen using a tensioned i line system as shown in Figure 3.3 or 3.4. The load corresponded to a displacement of )

approximately % to 1 inch in each support spring.

. Checking the approximate location of the travel indicator in each spring'.

• Quick-releasing the load so that the test specimen will undergo free oscillation.  ;

I e Processing and plotting the data as histories of accelerations for the test duration and power spectra of the accelerometer signals.

. Comparing the records of the acceleration response at each end of the strainer to determine whether the response is in phase i.e. the oscillation response is primarily vertical translation.

(b) Free Vibration Testing / Data Acquisition in Air 1

Response data (acceleration of the oscillating supported strainer) was recorded by the data j acquisition system during the test. The free vibration test was to be repeated a minimum j of three times, the actual number depending on the repeatability of results. The test i included the following steps:

l, TR-ECCS-GEN-011, REV. O t6 c ._ . _ _ _ ._________________________-________-___a

l e Assigning each test a unique number and recording the pertinent data for that test (approximate travel indicator location, comments) on the Test Log. Appendices C and D of Ref.18 show copies of the completed Test Logs. The scheme used for the test l- number follows that recommended in the Test Procedure (Ref.14) is NN-A (or W)X, where, e NN = the Test Series (1 = lateral,4 = axial)

. A = Air and W = Water e X = the test number in that series e Checking the travel indicator of each spring to verify that it will permit sufficient displacement (approximately centered or slightly below).

• Applying a vertical downward load to the center of the test specimen using a tensioned line system as shown in Figure 3.3 or 3.4. The applied displacement should be equal to or greater than the Test Matrix Target Deflection.

• Checking the approximate location of the travel indicator in each spring; the spring ,

travel indicator should have been between the midpoint and the bottom ofits travel l range.

• Quick-releasing the load so that the test specimen would undergo free oscillation.

• Processing and plotting the data as histories of accelerations for the test duration and i power spectra of the accelerometer signals.

. Reviewing the test data for consistency and repeatability.

1 l l . (c) Free Vibration Testing / Data Acquisition in Water Following completion of the free vibration testing in air, the tank was filled with water to submerge the strainer by approximately 6' (to the centerline of the strainer) as shown in

. Appendices C and D of Ref.18. The water temperature was recorded in the Test Log, after the water testing. The temperature was ~55'F in all tests.

The same sequence of operations as described in (b), Free Vibration Testing / Data Acquisition in Air, was performed for each free vibration test in water.

TR-ECCS-GEN-011, REV. 0 17 L_ = __ _ _:_-__ _________ ___

i.

4.5. Test Data Collection Each test run was labeled with a unique number for test results data identification purposes. l The measured data for each test run was recorded by the data acquisition system and reviewed immediately after the test. Before beginning the in-water tests, the strainer was " bounced" several times on the springs in an attempt to evacuate all the air bubbles from the test )

specimen. Table 4.4 shows the actual test data for a typical test series.

4.6. Test Data Analysis Preliminary analysis of the acquired data consisted of plotting all measured responses as a function of time using the FFT analyzer. The preliminary results were analyzed at the test site I

using all the data for that strainer and that series (excluding the tests which had an 8-second sample, which were performed for information only).

4.7. Test Monitoring I All tests were performed and witnessed by qualified test engineers: Mr. John Minichiello of DE&S Resources and Dr. David Williams, a consultant to DE&S. Mr. Richard Kibling represented PCI during the tests. The data acquisition equipment was operated by Mr. i Michael Kelly of DE&S, who is also a qualified test engineer. Mr. Kenneth Brittain and Mr.

Brady Hightower of EPRI were responsible for operation of the crane and loading devices.

1 TR-ECCS-GEN-011, REV. 0 18

l 1 i l

l TABLE 4.1 Weights of Covers, Bolts Nuts, and Additional Mass on Rods Durine a Tpical Test Series I

Cover Weight (Ib) 84 Added Weight (Ib)-lateral test 70 (all on the rod at the disk end)

Added Weight (Ib)-axial test 80 (40 to each rod)

Weight studs and nuts on cover (Ib, 6 approx.)

TABLE 4.2 Typical Accelerometer to FFT Annivrer Data Acanisition Calibration (Strainer #1)

Lateral Axial Pre-Test Post test-in. Pre-Test Post test-in-water water Accelerometer #1 (North end)

Serial Number: 7371 Expected Output Frequency Oiz) 159.2 159.2 159.2 159.2 Actual Output Frequency (Hz) 160 159.5 159.5 159.5 Accelerometer #2 (South end)

Serial Number: 7370 Expected Output Frequency Oiz) 159.2 159.2 159.2 159.2 Actual Output Frequency 01z) 159 159.5 159.5 160 l

l l

I l

L l TR-ECCS-GEN-011, REV. 0 19

'i L- - - - _ - - _ - _ - - _ - - - - - - - - - - - - - _ - - - - - - -

TABLE 4.3 Planned Test Matrix For Strainer In Air And Water Lateral or Axial Tests Clean Strainer Strainer Target Approx. Spring No. of Weight 10 % Status Deflection Load (Ib) Stiffness Tests (ib) (in) (each Spring)

(Ib/in) 2600 In air 1/2 300 260 3 minimum (up to 10) 2600 In water 3/4 400 260 3 minimum (up to 10) 1500 In air 1/2 150 150 3 minimum (up to 10) l 1500 In water 3/4 225 150 3 minimum (up to 10) 1100 In air 1/2 100 112 3 minimum (up to 10) 1100 In water 3/4 175 112 3 minimum (up to 10) l i

l TR-ECCS-GEN-011, REV. 0 20 L

t _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ - - _ -

TABLE 4.4 1 Typical Test Senes j Lateral Axial Test Spring Comments Test Spring Comments Deflection (in) Deflection (in) l-Al- 1 4-Al 1 1-A2 1 4-A2 1 1-A3 1 4-A3 1

]

l-A4  % 4-A4 1 1

1-A5 1 l 1-A6 1 l l-W1 1 4-W1  %

I l-W2 1-1/4 4-W2 1 l-W3 1-1/4 4-W3 1-1/4 Note 2 1-W4 1-3/8 Note 1 4-W4  % )

1-W5 1-1/4 4-W5 1 l-W6 1-1/4 4-W6 1+ Note 3 1-W7 1-1/4 4-W7 1+ Note 3 1-W8 1-1/4 Notes: 1. Slight overload (acceleration above the FFT preset range) on Channel A

2. Lead block (at Tank wall) fractured during loading, but test was still recorded Slightly above 1" applied f r

l l

l l

l l

l

)

1 TR-ECCS-GEN-011, REV. 0 21

5.0' -RESULTS 5.1. Approach to Results Interpretation The calculated hydrodynamic mass of the strainer is based on the ratio of the frequencies for free-vibration response of the oscillating body. In the arrangement for support of the drainer in these quick-release free-vibration tests, the strainer essentially responds as a lumped mass single-degree-of-freedom rigid body oscillating on the suspension springs.

Th:: oscillations are principally in the vertical direction.

A'i example of typical acceleration records of motion at both ends of the strainer is shown in Figure 5.1. The duration of the acu:leration record is 40 seconds although the oscillation has died out in the first 5 seconds. The peak-to-peak acceleration magnitude is listed for each channel (e.g. Y: 0.274g Channel B). ~ As seen from the acceleration records, the free-vibration response is essentially rigid body decaying oscillatory motion in the vertical direction u.ith the two ends basically in phase (heave mode).

The power spectrum of each acceleration record, as shown for the typical case in Figure 5.2, was obtained to determine the frequency content of the motion. The linear spectmm ranges in frequency from 0 to 10 Hz. The dominant response frequency (of the heave mode) is readily spnarent and is noted as shown; 1.425 Hz for Channel A (top plot) and the same for Chanac. B (bottom plot). The frequency resolution was typically 0.025 Hz but, as noted in Section 4, for one or two tests data was acquired for a duration of 8 seconds and the frequency resolution was then 0.125 Hz. The data from these 8-second readings is not used for frequency calculations, since its resolution is so much lower.

5.2. Free Vibrations in Air (Lateral and Axial Tests) l-l The results for the " quick-release" free-vibration testa in air are presented in Appendices EI A, E2A, and E3 A (lateral) and Appendices E4A, ESA, E6A, E7A, and E8A (axial) of Ref.18. The top pair of plots are the time history records for vertical acceleration at the -

two ends of the strainer and the lower pair of plots are the corresponding power spectra of the acceleration responses. The dominant frequency of vibration (vertical heave) at both

, ends (Channel A and B) for a typical series of tests is indicated in Table 5.1 and 5.2. The L frequency resolution was 0.025 Hz for all tests. The tables also show the total weight used in the test, i.e., strainer, cover, weights, and miscellaneous items such as rods, clevises, nuts and bolts.

TR-ECCS-GEN-011, REV. 0 22

In certain tests, a second frequency peak in the power spectrum (the pitch frequency) is evident at a slightly higher frequency for the lateral test. However, the amplitude of this mode is only about 20% of the dominant heave mode, indicating that the motion is essentially vertical translation.

The maximum peak-to-peak' acceleration for each test in air was typically between 0.2g ud 0.3g laterally, and between 0.3g and 0.4g axially. The decay rate in the lateral and axial test acceleration time-history records is an indication of moderate damping in air.

5.3. Free Vibrations in Water (Lateral and Axial) 5.3.1. Test Results The results for the " quick-release" free-vibration tests in water are presented in Appendices ElW, E2W, and E3W (lateral) and Appendices E4W, ESW, E6W, E7W, and E8W (axial) of Ref.18. For each test, the top pair of plots are the time history records for vertical acceleration at the two ends of the strainer and the lower pair of plots are the corresponding power spectra of the responses. The dominant frequency of vibration for each of the tests is indicated in Tables 5.1 and 5.2. The frequency resolution was 0.025 Hz for all tests. The vibration frequencies are very consistent between tots, as evident from the tables.

A second frequency peak in the power spectra (the pitch frequency) is usually evident at a slightly higher frequency, but again, the amplitude of the power spectrure for this frequency is small.

I The maximum peak-to-peak acceleration for each lateral test in water ranged from 0.17g to 0.31g. The upper end of this range is about equal to the peak-to-peak acceleration range in air, although the frequency ofvibration is lower. This reflects the fact that the initial displacements in water were larger than those in air.

Note that for the strainer data shown in axial test 4-W3, the lead block at the tank wall fractured just as the full tension was applied, acthg as its own " quick-release". The data

' is, therefore, included in the table.

l For the axial tests, the vibration frequencies are also very consistent between tests, as evident from Table 5.2. The maximum peak-to-peak acceleration for each test in water L

ranged from 0.16g to 0.24g. This range is less than the corresponding peak-to-peak acceleration ranges in air, reflecting the lower frequency of vibration. The initial i

' displacements were about the same in air and water. 1 The high damping in water is evident from the rapid decay in the osci!!ating acceleration

- records for both the lateral and axial tests. I L

I TR-ECCS-GEN-011, REV. 0 23 L_ _ __ _-- _ .

5.3.2 Test Validity Blevins (Reference 10) recommends the following be checked for vibrating bodies in fluid:

L(2' xf)/c << 1 L/D < 1 :

2xfD 2/v > 10,000 where:

L = harmonic displacement amplitude f = cyclic frequency of the vibrating body in the fluid I

c- - speed of sound in the fluid

? = L4700 ft/sec (Reference 11, Page 3-50)

D? = characteristic dimension of the body v = kinematic viscosity of the fluid

= 2 1.31 x 10'8 ft /sec (Reference 9)

Others have suggested that the Reynolds number (Re = VD/v, where V is 'the velocity) 4 should be greater than 10.

Using was 0.17gthe (65.7 data in/secfor at aafrequency typ)ical strainer of 1.19 Hz (lateral) as an and example, the minimum 0.16g (61.8 in/sec') at the pe same frequency (axial). Assuming harmonic motion, the resulting displacement amplitudes, L, are:

L (lateral)= 0.59"

& (axial).. = 0.55" For this strainer, D == 45" (outside diameter of the test specimen)

Thus, the associated parameter values are:

Lateral Test - Axial Test Criterion L(2xf)/c 8 x 10-' 7 x 10 << 1 L/D 0.01 0.01 <1 ~

2 2xfD /v 8 x 10' 8 x 10' > 10' Re 10.5 x 10' 9.8 x 10' > 10' Thus, the recommendations ofBlevins, and the suggested checks by others, have been satisfied. Results for the other strainers are similar.

STR-ECCS-GEN-011, REV, 0 24 i Y____-______________-_____-______-___

.g -

i 5.4. Effective Hydrodynamic Mass Coefficients (Lateral and Axial)

The natural frequency of a simple spring-mass system is:

f= [l/2x]*V(K/M) '

where:

K = spring stiffness M = Mass = W/g W = Weight in a Ig field -

i g = acceleration due to gravity Since 1), the mass of the vibrating body is inversely proportional to the square of the oscillation frequency (for an essentially single-degree-of-freedom system on simple springs), and 2), the system stiffness is the same in air and water, then M.;,/M = W.;,/ (M =*g) = [f ,/f.w]2 where:

M.a = mass during the test in air = strainer + hardware weight = W.w/g M = mass during the test in water

= (strainer + hardware weight +added water weight)/g f = average frequency in water f.w = average frequency in air Thus, solving for M , we find:

M = W.a /(g*(f / f.w}2) l Hence, the weight of water " participating" in the vibration is:

W6ya = (M *g - W.w) = W.w'({f.w / f }'- 1)

The volume displaced by strainer envelope is the selected reference volume, Va, for lateral l

vibration The reference volume for axial vibration is a sphere with the characteristic diameter of the disks,.

The added mass coefficient (and C,. in this case) is the ratio of the weight of

" participating" water, W6ya, to that associated with the reference volume, Va:

l C,. = Wayu / yVa From Reference 9, the density of water (y) at 55'F (the test temperature) is 62.4 lb/ff. 1 TR-ECCS-GEN-011, REV. 0 25 L

The tables below provide the pertinent information for the lateral and axial tests, using the total weights, i.e., including the miscellaneous weight of the rods, clevises, and nuts and L - bolts. The data for C. are shown for the average value of the in-air and in-water frequencies, as well as C. calculated from the mean-in-air plus 2cr(standard deviation) frequency and the mean-in-water minus 2cr frequency.

l Lateral Tests Data Set 1 2 3 4 OAIJODow 1.0 1.4 0.9 2.0 CT/ODai,e - 0.44 0.67 0.31 0.55 C.(mean +/- 20) 0.44 0.52 0.57 0.54 1.

Note that for the fourth strainer, there were twice the normal number of radial stiffeners in each disk. The test values of C. are well below the C. of 2.0 typically used for an impervious cylinder immersed in water and subject to lateral vibration, even when a 2a variation on frequency is included.

( Axial Tests Data Set 1 2 3 4 5 Core tube end one solid, one one solid, one both solid one solid, one both plate perforated perforated perforated perforated conditions C.(mean +/- 0.44 1.00 1.25 1.09 0.57 l:

2a) l These compare with a C. of 1.6 or higher, with a significantly different (and larger) reference volume, typically used for a cylinder immersed in water and subject to axial I

vibration (UD ~ 2).

The derived value of the effective hydrodynamic mass coefficient is an equivalent value for the perforated strainer and represents the combined hydrodynamic effects of the

" enclosed" and " entrained" or "added" mass. For the perforated body, there is no totally l: ' enclosed fluid mass and the " effective" mass is dependent on the " porosity" or openness of the strainer surface and core plate. For the strainers tested, the perforated plate had hole diameters of either 3/32" or 1/8" with open areas of 33 or 40 percent, respectively.

L li

' TR-ECCS-GEN-011, REV. 0 26 L-- - - - - - -. - _ - - - - - . - - - _ - - - - - _ - - - - - - - - - - - - - - _ _ - - . - - - - - - - - - - - - - - - - _ - - _ - - _ - _ - _ - - - - _

Table 5.1 Typical Ouick-Release Lateral Test Results - Vertical Oscillation Frequency Strainer Tested weight = 1534 lb In-Air In-Water Test Freq. At B Freq. At A Test Freq. At B Freq. At A l

(S. End) (Hz) (N. End) (Hz) (S. End) (Hz) (N. End) (Hz) 1-Al 1.4250 1.4250 1-W1 1.2500 1.2000 1-A2 1.4250 1.4250 1-W2 1.2000 1.1750 1-A3 1.4250 1.4250 1-W3 1.2000 1.1750 1-A4 1.4250 1.4250 1-W4 1.2000 1.1750 1-A5 1.4250 1.4250 1-W5 1.2250 1.1750 1-A6 1.4250 1.4250 1-W6 1.2000 1.1750 1-W7 1.2000 1.1750 1-W8 1.2000 1.1750 Table 5.2 Ouick-Release Axial Test Results - Vertical Oscillation Frequency Strainer Tested weight = 1544 lb In-Air In-Water Test Freq. At B Freq. At A Test Freq. At B Freq. At A (S. End) (Hz) (N. End) (Hz) (S. End) (Hz) (N. End) (Hz) 4-Al 1.4250 1.4000 4-W1 1.2250 1.2000 4-A2 1.4000 1.4000 4-W2 1.2000 1.1750 4-A3 1.4250 1.4000 4-W3 1.1750 1.1500 4-A4 1.4250 1.4000 4-W4 1.2000 1.2000 4-W5 1.2000 1.1750 4-W6 1.2000 1.1750 4-W7 1.2000 1.1750 Note: The frequency resolution in all tests was 0.025 Hz TR-ECCS-GEN-011, REV. 0 27

FIGURE 5.1 Typical Lateral Test In Air (Acceleration Vs Time)

A=7371 M EHL B=737B S E G 2/25/90 MVERAGE 10Hz A:AC/6.2V B:AC/G.2V INST 0/E DUAL ik 50 SU"

.295 .

MASE MEP E+0 .

BL i

. . RC. 0 l REAL lP- EU SET G

n i

ei .

Cn A

) . [.'. . ., .

EU= G

E+0 1V=

1 TIME A LIN 40SEC .1926+1 1 3: .234 SEc B: .26?E+0 G l

.199 Ch E EU= G E+0

( R[AL ,9lf..i .9930+0 V ,

'llP : UNIT

. . . . . . . . . . . , .. .....J..... X:H:

.199 3 .

Y:EU E+0 TIME E Llh 405EC 'C0H or f ELNK E: .312 SEC 5: .274E+2 G UNIT SET 22:16 UNIT X l UNIT Y ICh A EUICh B EUI I i l EXIT l

l.

l l

TR-ECCS-GEN-011, REV. 0 28 L_______________________-.___________________

FIGURE 5.2 l

Typical Lateral Test In Air (Power Spectrum) l y

A=7371 H EHL B=7370 S Dtb 2/25/99 AVEP.hGE 10Hz A:AC/8.2V B:AC/B.2V INST 0/8 DUAL ik SP SUM

MASS MEM

.103 E-i . ..... ..; .. . ..; .. . . ..; .. . . . . . . . . . .

g,: 1, G . . . . . . . ........ .. .

..j.. . . ... ...

Ch A

. . . . ..j.. . . .. i . . ..i.. ..i .. . ..i.. . . i .. .. EU= G e

ivs O PWR SP A LIN 10Hz .1026+1 3: 1.425Hz =: .737E-2 6 fCh 9

.993 E-2 . . . . . . . . . . . . . ... . .

EUs G iV=

gg . . ....

.9930+B.

G . . , . .. ; . . .

UNIT a ..

x:Hz y.EU

~

0 - COH BLNK

~

0 FHR SF E LIN 10Mz 0FF

: 1.425Hz =': .745E-2 G UNIT SET 22:17 UNil X l UNIT Y Ich 6 Eulch B EUI i l l EXIT TR-ECCS-GEN-011, REV. 0 29

6.0 APPLICATION OF TEST RESULTS The measured frequencies of vibration (from spectral analysis of acceleration records) for all

. tests show very little variation from test to test within a series, statistical analyses of the test l data have been performed. The mean, x., and standard deviation, o, of the test frequencies

' for each series was performed in Reference 18. A somewhat pessimistic measure of the statistical variation in the hydrodynamic mass participating in the response, C.(2o), is obtained by using the in-air mean + 2o and the in-water mean - 2 frequencies. These coefficients are shown in Table 6.1 together with those, C.(avg), based on mean frequencies.

Mean + 2ais standard practice for reliability work in the nuclear industry. IEEE-323 (Reference 12) recommends that a 10% margin be applied to test data when comparing it to L application requirements. Thus, the recommended effective hydrodynamic mass coefficients for design assessment, C.(des), related to the relevant reference volumes, Va, are 1.1 times the 2avalues, as shown in Table 6.1.

6.1. Effect of Geometric and Structural Design Parameters on C.

The recommended approach and selected parameter values (effective hydrodynamic mass coefficients and associated reference volumes) for the lateral and axial loads are based on the tests for the strainers identified in the Reference 4 drawings. The recommendations should be applicable to the other strainers that are of similar design and installed configuration.

An investigation of the effects of the various strainer physical parameters on the magnitude of the hydrodynamic mass (and coefficient) was undertaken. Essentially the hydrodynamic mass

' is dependent on factors that impede the free-stream flow. With the typical perforated plate used in the PCI stacked-disk strainers (open areas of 33% or 40%), the flow is relatively _

unobstructed. The openness of the core tube is typically much less and can be expected to .

impede flow more (providing greater hydrodynamic mass). The hydrodynamic mass will be particularly affected by the area of solid plate (projected normal to the' flow) in the strainer.

>, Thus the number and nature of radial and circumferential stiffeners in the disks can cause an increase in hydrodynamic mass.

The variation of these structural details for the strainers tested in the lateral direction cause counterbalancing effects in the effective hydrodynamic mass, masking trends from any single parameter variation. For example, comparing lateral C. results in Table 6.1 for the second

and third rows (which both had 3/32" perforated plate), the smaller IJD for the third strainer

. (and smaller core tube to disk diameter ratio) would suggest a lower hydrodynamic mass.

However, the larger number of disk radial stiffeners in the third strainer appears to counterbalance this effect. The range of results are reflective of the range of dimensions and differing structural details in the tested strainers.

TR-ECCS-GEN-011, REV. 0 30

For tests in the strainer axial direction, the area of the strainer flanges and solid cover plates has a strong influence on the hydrodynamic mass. Thus, the low value of C. for row 1 is reflective of a small diameter of the flange cover plate relative to the disk diameter and the high C. value for row 3 is reflective of the solid covers on each end. Rows 4 and 5 show the effect of replacing the solid cover plate on one end with 1/8" perforated plate, and it approximately halved the hydrodynamic mass. The disk gap width is also likely to affect the hydrodynamic mass in the axial direction with a narrow gap causing relatively higher values.

Again, due to the counterbalancing effects, a quantitative assessment of separate parameters is not simple.

If the strainers are longer than those tested, then it is possible that end effects will not be as pronounced. For example, in_ one installation, strainers are installed on a pipe only a short distance from the Torus shell. Thus, while the fluid is able to flow easily around one end, it is blocked from flow around the other by the Torus shell. Thus, the added mass might be higher for this case.

Classical practice (for example, Reference 10, Table 14-1, Item 1) states that the added mass for an impervious cylinder vibrating in a fluid is equal to the displaced volume of the fluid.

For water-filled piping in water, this means that the effective hydrodynamic mass coefficient (C.) for an infinite cylinder is 2.0: a value of 1.0 for the contained water and a value of.1.0 for the entrained water. Stated another way, C. = coefficient of" contained mass" (C.) + coefficient of"added mass" (C.)

The results from the dynamic testing show that the effective hydrodynamic mass coefficient (C.) for the strainer, with flow around the two ends, is ~ 0.5, based on the displaced volume.

' A value of C. less than 1.0 indicates that even the fluid " contained" in the strainer is not fully participating in the vibration.

L Thus, in order to extrapolate the test data to longer strainer modules, there are two possible j

cases which bound the conditions

either C is zero and all the mass is "added" or C. is zero and all the mass is " contained".

If all the mass is contained (C. = 0), then length should have no impact on C.. The fluid is not

" flowing" around the ends, but flowing out of the strainer. Since flow occurs only through the

_ . cans and the gaps between them, and the cans are uniform, the length cannot have an impact on the coefficient.

l.

l If all the mass is added (C. = 0), then length may have an impact on C.. The American l- Bureau of Shipping (Reference 13) provides the following length-to-breadth correction factor, k, to be applied to C. for lateral flow situations with solid cylindrical members:

I

! .TR-ECCS-GEN-011, REV. 0 31 i

L- - - - _ _ _ _ _ _ _ - - - - - _ . -- - - - - - - - - - - - - - - - - - - - - - --------__ ______ _____________________________

j . _.

k = (UB)2/ (1 + (UB)')

where, L = length of the member B = breadth the member projects to the flow (cylinder diameter)

For the strainers tested, taking B as the OD of the disks and L as the effective length (active, overall or in-between, depending on spool and flange diameters relative to the disk diameter) would give a conservative estimate of the factors if the strainers were impervious. The "end effect" for a perforated body is significantly less than the effect for a solid body because the flow is, primarily, through the perforations and is not obstructed in the same way as for the solid. This is convincingly confirmed by the lateral test results (see Table 6.1). The aspect ratios range from 0.8 to 2.0 implying a correction factor range of 2.05 (2.56/1.25) whereas the C,. range is only 1.3, suggesting the correction factor is overstated by a factor of approximately three.

For hydrodynamic loads in the axial direction, minor correction for length may be warranted.

Reference 10, Table 14-2, Item 6 for rectangular solids suggests that the added mass is .

relatively constant for an aspect ratio (UD) between 0.6 and 2.8_. At an aspect ratio of 3.6, the implied increase in added mass over that for an aspect ratio of 1.0 or 2.0 is approximately 13 percent.

6.2. Application to Other Sizes Lateral Vibration - Having noted above that the C, are not simple combinations of par; emers, wumn attempt to see if some consistency exists. For lateral vibration, let us use the, data in the nrst, second, and fourth sets of Table 6.1. These all have 6 radial stiffeners per disc, and thus do not have the extra " resistance" found in the strainer represented in data set

3. As stated above, we believe that C, should be a weak function of OAUODaa (since most of the flow is through the strainer) and perhaps a function of CT/ ODaa. Reviewing the data, it appears that the equation below provides a reasonable fit: j C., = 0.6*(OAUOD44)*"*(CT/ ODaa)*"

l

. 1 The table below shows the pertinent data:

Actual C., OAUODaa CT/ ODaa Predicted C.,

0.48- 1 0.44 0.49 0.57 1.4 0.67 0.59 0.59 2 0.55 0.61 l

l TR-ECCS-GEN-011, REV. 0 32 L_=_ _ = _ = _ . _:__-___

i l

l l l i

The predicted values agree well with the actual values and the expression above can be used  ;

for similar designs (6 radial stiffeners) and parameters. Care must be taken, however, in severely extrapolating the results to significantly longer modules, since the OAIJODaa effect would then be overstated. For example, if one had a stacked module with ,

an OAL/ODaa of 8, which is about the largest value seen in the designs to date, and a l CT/ ODaa on the order of 0.75, the resulting C. would be on the order of 1. The authors l strongly believe that a value above 1 is not possible with these designs.

1 For designs such as that represented by the data in set 3, which have many radial stiffeners due i to their eccentric arrangement, it is expected that the above (OAIJOD4a)(25*(CT/ Odaa)a25 j correction holds, with the only variation being the constant coefficient. For similar designs, l the constant can be replaced by 0.9. This would result in a predicted value for the strainer represented in data set 3 of C.(predicted) = 0.9'(0.31 *0.933)(25 = 0.66 vs the design value of 0.63 Again, care should be taken in extrapolating this value to severe lengths, since the primary

. reason for the higher C. for the data in set 3 is the resistance through the disks, which the length will not affect. A value of C. greater than 1 should not be used.

Axial Vibration - Axial correlation is more difficult than the lateral, since the length has less of an impact, but " solid" end portions have more. One item which is clear is that the changing a 32" perforated plate to a 32" diameter solid cover plate added 680 lbs of effective mass.

Postulating that all this is due to a sphere with a 32" diameter, we back calculate a C..

8 C. (solid end plate) = [680/1.3333*x* 16 )*1728/62.4 = 1.1 Thus, any solid plate will be assigned a C. of 1.1, with an effective volume equal to the

' surrounding sphere. If there are two solid end plates, as in one of the tests, two spheres will be postulated.

The amount of effective mass for the remainder is more difficult. It is obvious from the test of the strainer with perforated plate on both ends that even a strainer without a solid cover has an effective mass associated with it. Some of that mass is "added" because the end plates are partially solid. . Some of the mass is " contained" by the disks and core tube, and the mass in j the core tube may not be very effective. To estimate the effect, let us conservatively use the lateral reference volume minus the volume associated with any solid ends. Since the strainer represented in data set 5 has no solid ends, this is simply 26.5 ft', and the C. is C. (strainer) = 760/(26.5*62.4) = 0.45 L

The effective mass for any strainer is then the sum of the volumes of the solid end plates times 1.1 plus (the lateral volume minus the sum of the volumes of the solid end plates) times 0.45.

The following table shows how the predicted and design values agree.

I TR-ECCS-GEN-011, REV. 0 33 4

Actual Design V,(lat) V V V strainer Predicted Act/

Effective Mass (ft') Cover 1 Cover 2 (ft') Mass Pred (lb) (ft') (ft') (Ib) 844 27.2 6.30 0.00 20.90 1019 0.83 968 20.5 9.93 0.00 10.57 978 0.99 1440 26.5 9.93 0.00 16.57 1147 1.26 760 26.5 0.00 0.00 26.50 744 1.02 1670 41.95 9.93 6.30 25.72 1836 0.91 As shown above, the agreement is not bad, although it does overpredict the results for the longer strainers. Thus, to develop effective axial masses for other size strainers, a conservative estimate would be Effective Mass = 62.4*[1.l*(Sum of the Spherical Volumes for any solid end plates) +

0.45*{V,(lateral)- (Sum of the Spherical Volumes for any solid end plates)}]

The authors believe this becomes very conservative as the length increases, but in the esence of better data the above is proposed for use.

6.3. Comparison to Values Used in Design The structural evaluations of the strainers described in this report were typically completed before these tests were performed. In completing the structural evaluations, conservative estimates of the effective hydrodynamic mass (and associated volume) were made. In particular, it was postulated that length to diameter ratio had a dominant effect on the effective hydrodynamic mass. As shown in the data correlation above, the length to diameter effect on lateral effective hydrodynamic mass is small, on the order of the fourth root. The axial effective hydrodynamic mass does not appear to be a function of the length to diameter ratio.

In order to show how conservative the structural evaluations were, the effective hydrodynamic mass used in the evaluation for the one of the strainers (ref.19) is compared to the effective hydrodynamic mass predicted using the above tests. The results shown are per strainer, and show how conservative the structural evaluations were.

1 OAL = 148" ODa = 36" CT = 24" V a(lateral), one strainer = 20.47 ft' Effective Lateral Mass (predicted) =

0.6 *(148/36)a25(24/36)" 25 *20.47 *62.4 = 1000 lb Effective Lateral Mass (Ref.19) = (page 21 of Ref.19, module 6) = 1934 lb Effective Axial Mass (Predicted) = 0.45*20.47*62.4 = 575 lb Effective Axial Mass (Ref.19) = (page 18 of Ref.19, module 6) = 1895 lb l

TR-ECCS-GEN-011, REV. 0 34 I

L- -------

Table 6.1 Strainer Hydrodynamic Added Mass CoefHelents '

Lateral 2 Data Set OAUODaa CT/ODaa Va(ft') C.(2o) C.(des)

I 1.0 0.44 27.2 0.44 0.48 2 1.4 0.67 20.5 0.52 0.57 3 0.9 0.31 20.1 0.57 0.63 4 2.0 0.55 42.0 0.54 0.59 Axial' OAUODaa CT/ODaa Va (ft') C.(2o) C.(des) 1 1.0 0.44 27.6 0.44 0.49 2 1.4 0.67 14.1 1.00 1.10 3 2.0 0.55 19.4 1.25 1.37 4 1.3 0.6 19.4 1.09 1.19 5 1.3 0.6 19.4 0.57 0.63 Notes:

1. Related to reference volume, Va

2. Va(lateral)is displacement volume ofstrainer envelope

3. Va(axial) is volume of sphere with strainer disk diameter l

1 TR-ECCS-GEN-011, REV. 0 35 l

l

p ,

D t

7.0 CONCLUSION

S A test program was undertaken to generate the data required to develop empirically-based values for the hydrodynamic added mass coefficients (and resulting effective hydrodynamic l mass coefficients, C,.) for use in qualification calculations related to installation of replacement ECCS suction strainers in BWR suppression pools. Hydrodynamic loads directed both l

laterally (flow normal to the longitudinal axis of the strainer) and axially (flow along the longitudinal axis of the strainer) were ofinterest.

' Results were based on tests using several PCI stacked-disk stramers with different geometric

, arrangements. Recommended values of the added mass coefficient (and effective L hydrodynamic mass coefficient, C,.),' appropriate for design assessment with lateral l - hydrodynamic loads, based on the reference volume equal to the displacement of the strainer i envelope, ranged from 0.48 to 0.63. The comparable added mass coefficient for axial i hydrodynamic loads, based on the reference volume associated with a sphere of diameter l~ equal to the strainer disk, ranged from 0.49 to 1.371 These values include a pessimistic assessment of the experimental statistical variation and a 10 percent margin. The lateral values are significantly less than those typically used in current analyses; a range.from 1.25 to

- 2.0 for lateral loads. The axial values result in hydrodynamic masses much less than those l 'used in current evaluations, and show that the reduction may be as large as a factor of 2 or L higher, depending on end plate conditions.

i The most significant factor in variation of the mass coefficient for axial hydrodynamic loading is the ratio of the area of the solid end plate to the total projected end area of the strainer.

! The low values are associated with perforated plate on the flange or a solid flange area less than 40% of the projected area.

The recommended effective hydrodynamic mass coefficients for both lateral and axial loads are appropriate for perforated stacked-disk stminers similar in design and geometric -

d arrangement to those tested. The parameter ranges included disk diameters from 36" to 45",

~

aspect ratios from 0.8 to 2, disk widths and gaps from approximately 2" to 4", perforated plate with 3/32" and 1/8" holes (33% and 40% open area respectively) and' perforated core tube diameters from 14" to 24". Strainers outside this relatively wide range of design parameters, particularly if they incorporate significant areas of solid plate, should be assessed on a case by case basis. A very conservative assessment for different length effects (higher aspect ratios) can be made following the guidance suggested in Section 6.0 of this report.

l p

l ..

TR-ECCS-GEN-011, REV. 0 36 t l 1

__ ______________ L

L L

t

8.0 REFERENCES

l' \

l- 1. NRC Bulletin %-03, " Potential Plugging ofEmergency Core Cooling Suction Strainers by L Debris in Boiling-Water Reactors," May 6,1996.

2. Duke Engineering & Services, Inc., Report No. TR-ECCS-GEN-05 " Supplement I to Hydrodynamic Inertial Mass Testing ofECCS Suction Strainers - Free Vibration Data Analysis", Rev 1, 7/14/97.  ;

t .

3. Duke Engineering & Services, Inc., Document No. TR-ECCS-GEN-007, " Strainer l Generic Dynamic Test Plan," Rev. O.

4. Performance Contracting, Inc., Engineered Systems Division, Strainer Drawings:

a) See Reference 4a in Reference 18.

b) See Reference 4b in Reference 18.

c) See Reference 4c in Reference 18.

d) See Reference 4d in Reference 18.

5. Duke Engineering & Seivices, Inc., Drawing No.DR-ECCS-GEN-001, " Concentric 11 Disk Non-Safety Related Strainer Test Fixture Design," Rev 0.

i' 6. Duke Engineering & Services, Inc., Report No. TR-ECCS-GEN-01 " Hydrodynamic i I

Inertial Mass Testing ofECCS Suction Strainers," Revision 2. l

7. Grinnell Pipe Hangers Catalog PH-97, Grinnell Supply Sales Company.

8. See Reference 8 in Reference 18.

9 John P. Comstock, Editor," Principles ofNaval Architecture," Society ofNaval Architects and Marine Engineers,1967.

10. Robert D. Blevins, " Formulas for Natural Frequency and Mode Shape," Van Nostrand 1 Reinhold Company,1979.

11. Theodore Baumeister and Lionel S. Marks, " Standard Handbook for Mechanical  !

Engineers," Seventh Edition; McGraw-Hill Book Company.

12. IEEE Standard 323-1974, " Qualifying Class IE Equipment for Nuclear Power Generating Stations".

13. " Rules for Building and Classing Mobile Offshore Drilling Units," American Bureau of i Shipping,1980.

14. Duke Engineering & Services, Inc., Document No. TR-ECCS-GEN-008, " Strainer Generic. Dynamic Test Procedure," Rev. O.

15. See Reference 15 in Reference 18.

TR-ECCS-GEN-011, REV 0 37 a

1 l

L l 16. See Reference 16 in Reference 18.

17. See Reference 17 in Reference 18.

18. Duke Engineering & Services, Inc., Document No. TR-ECCS-GEN-010, " Strainer Test Report (Dynamic)," Rev. O.

19. See Reference 19 in Reference 18.

l l

t i

i i

l.

( .

1

(

!- J l

l TR-ECCS-GEN-011, REV. 0 38 i

(- _ . - _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

_