ML20132C717
| ML20132C717 | |
| Person / Time | |
|---|---|
| Site: | Peach Bottom  |
| Issue date: | 01/31/1985 |
| From: | Khlafallah M, Lee H BECHTEL GROUP, INC. |
| To: | |
| Shared Package | |
| ML17150A278 | List: |
| References | |
| NUDOCS 8507300436 | |
| Download: ML20132C717 (55) | |
Text
r ENCLOSURE 2 i
O-TECHNICAL BASIS FOR THE USE OF ENERGY ABSORBERS AS SUPPORTS OF NUCLEAR POWER PLANT PIPING SYSTEMS BY M. Z. KHLAFALLAH ANO O H. M. LEE OF BECHTEL POWER CORPORATION SAN FRANCISCO, CALIFORNIA l
j l
r O- yBiF58a2k Bi36!ii';F January 1985
O CONTENTS Section Pm 1 INTRODUCTION 1 -1 1.1 Current Piping Support Practice 1 -1 1.2 Energy Absorbers as Pipe Supports 1-3 2 OPERATING PRINCIPLES OF PLATE-TYPE ENERGY ABSOROERS 2-1 2.1 Performance and Design Requirements 2-1 2.2 Energy Absorber Relational Characteristics 2-2
- 2.3 Operating Principles 2-4 2.4 Fatigue Design 2-5 3 SUMARY OF SHAXER TABLE TEST RESULTS 3-1 3.1 U-Loop Test 3-1 3.2 Space Frame Test 3-4 (
3.3 Scaled Spatial Piping System Test 3-5 3.4 Overall Experimental Work Conclusions 3-8 4 DESCRIPTION OF THEORETICAL AND CORRELATION STUO!ES 4-1 4.1 Computer Programs for Systems with Uncoupled 4-1 Nonlinear Degrees of Freedom 4.2 Step-by-Step Integration 4-2 4.3 Modified Iteration Algorithm 46 4.4 Implementation and Example 4-11 4.5 Correlation Studies 4-16 4.6 Sensitivity Study 4 26 i
5 DESCRIPTION OF LINERIZATION METHODOLOGY 5-1 5.1 Localized Equivalent Viscous Damping 51 5.2 Computation of Modal Damping Ratios 5-3 O
i
l D .
CONTENTS Section Paje 5.3 An Iterative Procedure 5-6 5.4 Verification of Linearization by Modal Damping 5-8 5.5 Conclusion 5-9 6 DESIGN PROCEDURE 6-1 Noma 11 ration of Energy Absorber Characteristic Curves 6-1 6.1 6.2 Design Methodologies 6-2 6.3 Design Response Spectra 6-5 6.4 Themal Analysis Considerations 6-7 7 ME101 COMPUTER PROGRAM 7-1 7.1 Description of ME101 7-1 7.2 Sumary of Verification Problems 7-2 8 FABRICATION AND TESTING . 8-1 8.1 Description of Energy Absorber Details 8-1 8.2 Testing 8-4 9 INSERVICE INSPECTION REQUIREMENTS 9-1 REFERENCES R-1 Sections 4 through 8 are proprietary.
O ii
- _ _ _ _ J
ILLUSTRATIONS j Figure Page 1.1 Energy Absorber Development Flow Chart 1-5 2.1 Concepts of Steel Plate Energy Absorbers 2-6 2.2 Deformation Model for X-Type Energy Absorber 2-7 Energy Absorber Characteristics 2-8 2.3 2.4 Typical Hysteresis Curve of an X-Shaped Energy 2-9 Absorber 3.1 Plan of Measurwnent points and Isometric of 3-10 U-Loop Piping System 3.2 Time-Histories, Without Absorber, Peak Table 3-11 kc. = 0.142 g 3.3 Time-Histories, with Absorber, No " Thermal" 3-12 I
Displacement, Peak Table kc. = 1.32 g 3.4 Time-Histories, with Absorber, + and -1.0-Inch 3-13
" Thermal" Displacement, Peak Table kc. - 1.32 g 3.5 Hysteresis Curve of 1/8-In:h and 2xl/8-Inch 3-15 Energy Absorbers Obtained from Shaker Table Test O of U-Loop 3.6 Test Model of Hope Creek Core Spray, Piping 3-16 System - IV 3.7 Extreme Values of Pipe Strains, Mcelerations of 3-17 the Valve Operator, and Corresponding Snaker Table Response for Increesing Eerthquake Intensities 3.8 Maximum Reaction Forces Between Pipe and F.ame at 3-18 Restraint Devices and Rigid Rod Connections 3.9 Fourier Spectra of Snubbers and Energy Absorbers 3-19 3.10 Hysteresis Loops of Srubbers and Energy Absorbers ,3-20 Subjected to the Same Earthquake 3.11 Damping Ratio of Syste:n IV with Different Support 3-21 Devices in Position and Corresponding First
> Natural Frequencies 111 i
O i i
l l
l
ILLUSTRATIONS Ra,e O Fi ure 3-22 3.12 Force-Displacement. Behavior of a Snubber and a 2-Inch x 1/8-Inch Energy Absorber from Separate Tests Using the Same Input Table Motion 4-28 4.1 Local Coordinate System of Truss Element 4-28 4.2 Graphical Solution of Equation (4-45) for the Inelastic Truss Element Under Prescribed Load 4-29 4.3 A Simple One Degree of Freedom Nonlinear System for Program Verification 4-29 4.4 Response Time-Histories and Hysteresis Curves Computed with New Algorithm and with ANSRII of the One Degree of Freedom System 4-30 4.5 A Three Degree of Freedom Nonlinear System for Program Verification 4-30 4.6 Response Time-Histories and Hysteresis Curves Computed with New Algorithm and with ANSRII of the Three Degrees of Freedom System 4-3i Computed Time-History from wew A19erithm aad from O 4.7 ANSRII of Piping Sys-IV (Hope Creek Three-
. Dimensional Model) 4-32 4.8 Input Table Acceleration Time-History and Its Response Spectrum for U-Loop Without Energy Absorber 4-33 4.9 Response Time-Histories of Relative Pipe Corner Acceleration and Displacement and Fourier Spectrum of Displacement Response of U-Loop 4-34 4.10 Computer Simulated Response Time-Histories with Parameters from Raw Measurements 4-35 4.11 Response Time-Histories Affected by Low Young's Modulus and High Young's Modulus 4-36 4.12 Effect of Frequency and Modeling on RMSDR of Pipe Corne'r Displacement 4-37 4.13 Response Time-Histories with Overestimated Damping Factors iv O
ILLUSTRATIONS Figure Page
~
O 4.14 Effect of Damping Factors ao and a1 on RMSDR of Pipe Corner Acceleration and Displacement Responses 4-38 l 4.15 Response Time-Histories with Best Correlations 4-39 with Experiment 4.15a Triangular Energy Absorbers 4-39a 4.16 Computed Response Time-Histories and Hysteresis 4-40 j Curves Compared with Experimental Results of the 1/8-Inch Energy Absorber
- 4.17 Computed Response Time-Histories and Hysteresis 4-41 l Curves Compared with Experimental Results of the 2x1/8-Inch Energy Absorber 4.18 Response with Properly Adjusted Young's Modulus 4-42 and Hysteresis Loop of 1/8-Inch Energy Absorber 1
4.19 Computed Response Time-Histories with Identical 4-43 !
Parameters Except the Stiffness is Proportional to the Damping Factor
! 4. 20a Theoretical Hysteresis Loops Compared with an 4-44 A Experimental One for Varying Parameter r with
, U a = 3.35 4.20b Area per Cycle Corresponding to Figure 4.20a 4-44 l
- 4. 20c Hysteresis Loops Over a Time Period of Four 4-45 Seconds for Different r Corresponding to Figure 4.20a 4.20d Dissipated Encrgy Corresponding to Figure 4.20c 4-45
- 4. 21 a Theoretical Hysteresis Loops compared with an 4-46 Experimental One for Varying Parameter a ,
r=7
- 4. 21 b Area per Cycle Corresponding to Figure 4.21a 4-46 4.21c Hysteresis Loops Over a Time Period of Four 4-47 Seconds for Different a Corresponding to Figure 4.21a 4.21d Dissipated Energy Corresponding to Figure 4.21c 4-47 Y
O
ILLUSTRATIONS Figure P_ag 4.22 Sensitivity to Changes in the Absorber Rysteresis 4-48 Loop 4.23 Sensitivity to Changes in System Damping and 4-49 Hysteresis Loop 4.24 Sensitivity to Changes in System Flexibility and 4-50 Hysteresis Loop 4.25 Typical Response Time-Histories Computed with 4-51 Properly Adjusted Parameters Compared with Experi-mental Results.
5.1 Response Time-Histories with Local Linearization 5-10 Compared with Experiments 5.2 Hysteresis Loop with Constant S-Value Compared 5-11 with Experimental One 5.3 Typical S-Curve Derived from Jenning's Equation 5-11 with r = 7, 2 = 3.35 Compared with the Estimated S-Curve of the 2xl/8-Inch Absorber 5.4 Relationship Between Damping Coefficient E , and 5-12 Frequency e for Rayleigh Damping Based on Different Modes Compared to Calculated Modal Damping Coefficients 5.5 Linear Time-Histories with Overall System Damping 5-13 Factors Calculated with Modes 1 and 5 Compared to Nonlinear Results 5.6 Distribution of MODS /PKLY Ratios of Pipe Sys-IV 5-14 5.7 Design Configuration of Pipe Sys-Y 5-15 5.8 Design Configuration of Pipe Sys-VI 5-16 5.9 Distribution of MODS /PKLV Ratios of Pipe Sys-IV, 5-17 V, VI, and Their Combination Typical Computed Modal Damping Ratios of Pipe 5-18 5.10 Sys-Y l
6.1 Normalized Hysteresis Curve and S-Curve 6-8 ,
vi f
O
ILLUSTRATIONS Figure Page 6.2 Response Factor vs Damping Ratio at Control 6-9 Frequencies 8.1 Energy Absorber Details 8-7 8.2 Fatigue Design Curve 8-8
< 8.3 Overall Test Setup 8-9 8.4 Six-Inch Assembly Under Test 8-10 8.5 Test Assemblies and Fixtures 8-11 8.6 Load vs Displacement for 4xl/4-Inch II Specimen 8-12 at _+0.8 Inches Displacement - 1st Cycles 8.7 Load vs Displacement for 4x1/4-Inch II Specimen 8-13 at +0.8 Inches Displacement - 100th Cycle 8.8 Impact Test Setup 8-14 8.9 Load and Displacement Time-History Data 4xl/4-Inch 8-15 II Specimen 550-lb at 10-Inch Height 8.10 Load and Displacement Time-History Data 4xl/4-Inch 8-16 O II Specimen 550-lb at 15-Inch Height 8.11 Load and Displacement Time-History Data 4xl/4-Inch 8-17 II Specimen 680-lb at 20-Inch Height vii O
l
TABLES Page Table .
7.1 Features Checklist for Verification Problems 74 8.1 Standard Energy Absorber Data 8-18 8.2 Fatigue Test Specimens 8-19 8.3 Test Frequencies 8-20 9.1 Section XI Support Examination Categories 9-4 O
viii O
Section 1 INTRODUCTION 1.1 Current Piping Support Practice Currently, Seismic Category I piping systems in nuclear power plants are designed for nomal, seismic, and other transient loads using the following types of pipe supports:
o Spring-type supports, including variable and constant-effort types, are used as vertical dead weight supports at locations where essentially free thermal expansion is necessary. The springs in these supports are coil type, have low stiffness values, and are designed to perform elastically throughout their travel range.
Spring-type supports have a negligible effect on piping response during seismic events. As such, they are not nomally accounted
! O for in the seismic analysis computer modeis of piping systems.
o Rigid-type supports, including rods, struts, stanchions, frames, etc., are designed to restrain the piping system under all loading conditions. They essentially prevent the pipe from moving in the restrained direction and are designed to remain elastic under all j
l specified piping load combinations.
i l
I o Snubbers, including hydraulic and mechanical types, are used as i seismic and dynamic supports only. They are either mechanically or hydraulically activated. They offer no support against gravity, j allow essentially free movement under thermal expansion and con-
' traction of piping systems, but, under rapid dynamic piping move-ments, they activate and function as a rigid strut. Snubbers are designed to remain elastic under specified load combinations.
O 1 -1
l With the exception of springs, pipe supports currently in use are Q designed with the intent of maintaining their rigidity against speci- i fled loads while remaini'ng within the material yield stress. Seismic analyses of piping systems employ linear elastic mode superposition l methods in which rigid supports and snubbers are modeled as linear elements. Although gaps and other sources of nonlinearities exist in rigid support and snubber designs, modeling these supports as linear elements has been considered adequate in light of the conservatism inherent in the overall design process. The damping values used in the seismic modal analysis are based on Regulatory Guide 1.61 for recent plants, or the values contained in the Final Safety Analysis Reports (FSARs) of operating plants.
The escalation of seismic criteria has led to a continuing increase in the number of snubbers and the overall number of pipe supports over the last few generations of nuclear plants. As a consequence, recent plants contain stiffer systems than old operating plants. Our operating exper-feace base iadicates that riexibie syste=> Perron# 11 durias seis ic O events when compared to rigidly constrained systems. The increase in the number of snubbers, the rise in the number of snubber operating problems, and the decrease in system flexibility has led to numerous industry programs to investigate alternatives. Two important areas covered in these investigations are damping and flexible system design.
The recomendations by the Pressure Vessel Research Comittee (PVRC) on alternative generic damping values, which were included in Code Case N-411 of the ASME Section III Code, are one example of damping.
The alternative damping. values in Code Case N-411 were developed from l
available test data of piping systems typically supported with springs, rigid supports, and snubbers (1). References (2), (3), and (4) establish the fact that general system damping values are strongly dependent on the types and quantities of pipe supports used as well as the magnitude of the actual response involved. Fluids in hydraulic snubbers, small gaps in mechanical snubbers and frame supports, and O
1-2
friction forces in springs and bearing supports are the energy dissi-Q pating mechanisms to which the observed damping values are attributed.
Due to complexity and no'nunifomity of these mechanisms, it is not feasible to predict system damping values analytically, thus test data was used to make generic recommendations.
The energy absorber designed in accordance with Code Case N-420 and des-cribed in this enclosure is an alternative that enhances both damping and system flexibility.
1.2 Energy Absorbers as Pipe Supports The energy absorber, as presented in this enclosure, is a new pipe support concept that allows large amounts of energy dissipation under simple, well-defined, repeatable, and reliable conditions. Accurate analytical estimates of damping values associated with energy absorbers are possible. When used as seismic supports of piping systems, energy
() absorbers will provide the following basic benefits:
o Add significant amounts of damping that can be reliably calculated for each significant mode. The amount of damping added for signif-icant modes of vibration can be much higher than the values recom-mended in Code Case N-411.
o Reduce the systen frequencies, which will result in more flexible designs, especially for systems designed for high seismic loads.
o Virtually eliminate the need for snubbers, thus enhancing the system's reliability, o Significantly enhance the system's capability to accomodate higher than design earthquake loads or other dynamic events within pipe and support allowable limits.
1-3 e
Energy absorbers are simple, flexible supports designed to undergo controlled and predictable yielding under dynamic displacement. The hysteretic action of thi energy absorber plates results in added damping to the piping system. Cumulative fatigue effects can be evaluated according to the requirements of Code Case N-420.
Figure 1.1 shows the major activities that led to the development of this new pipe support concept. Each of the steps on the chart is shown in the sequence where it occurred. The successful completion of each step served as a benchmark for the succeeding step. Therefore, the energy absorber development and its analytical solution methods followed a course that could be verified by proven experimental data.
O O l I
l-4
..,n - - _ . - , , _ , , , _ _ _ _ . - - - _ _ _ _ , _ _ . . . - . _ . _ _ , . _ _ _ . _ _ _ _ _ _ _ . , _ _ _ . . _ - . . - - _ . _ . _ , _ . , . - , , _ _ _ , _ _ , , -
I
[ EXPERIMENTAL STUDIES '
O / N l ENERGY ABSORBER DESIGN l , l SHAKING TABLE TESTS ON PIPING SYSTEMS l
/ " \ /\ )
l FATIGUE l l MATERIAL l l GEOMETRY l lTHERMAll l SEISMIC l i
ANALYTICAL AND THEORETICAL STUDIES l
/ \
l SENSITIVITY H CORRELATION STUDY l-
/ NN l HYSTERESIS l l DAMPING l l SYSTEM FLEXIBILITY l O
UNEARIZATION
/ .
N l LOCAL DAMPING l l GLOBAL DAMPING l l MODAL DAMPING l
\/ \ /
l TIME HISTORIESl l MODE SUPERPOSITION l
\ /
ICOMPARISON WITH NONLINEAR TIME HISTORYl l
DESIGN PROCEDURES l ANALYSIS l [ DEVICE l f
Figure 1.1 Energy Absorber Development Flow Chart 1-5 1
Section 2 Q
OPERATING PRINCIPLES OF PLATE-TYPE ENERGY ABSORBERS l
2.1 Performance and Design Requirements Energy absorbers are made of simple, specially shaped steel plates designed to:
o Exhibit a smooth, well-defined force-displacement relationship.
o Possess moderate, finite, and well-defined stiffness and peak forces to be used in the thermal expansion analysis of piping systems.
o Yield uniformly over the maximum possible volume of the plate's material.
O o Have a well-defined hysteretic behavior under cyclic displace-ment. The hysteresis loops should only be a function of the current strain status and should be independent of strain rate.
o Exhibit negligible stiffness degradation.
o Achieve maximum cyclic energy absorption capability per unit volume of the material, per cycle, over as large a range of strain as I allowed by fatigue design considerations.
o Possess a fatigue endurance capability sufficient to maintain the plate's integrity under cyclic design and service load conditions.
o Be insensitive to environmental conditions, such as temperature and l radiation.
f Be or s'=aie coastruct$oa so as to eahaace reitabiitty aad eit='#-
O o ate maintenance and functional testing.
2-1
.-l ___
&,-. L_, ~. A mn ~ +--r u --. a e, "u.
During the past few years, several energy absorption concepts have been investigated (5), (6), leading to the conclusion that only the simplest Q energy absorbers made of ordinary ductile steel plates fulfill all of the above requirteents. Ffgure 2.1 illustrates two simple concepts for steel plate energy absorters that are ideal for satisfying all design requi ements and allow for reliable calculation of their damping effects.
Concepts 1 and 2 can be used interchangeably because they possess similar characteristics and hysterttic perfomance and provide the same damping effect to the piping system. Concept 2, employing x-shaped plates, has been selected by and used in the Bechtel energy absorter development program. This concept was selected because its end connec-tions art simpler than those of Concept 1. To facilitate construct-ibility and installation, to allow for pipe movements in the unrestrain-ed direction, and to increase reliability, multiple x-shaped plates wert incorporated in the design shown in Concept 2(a). This concept allows the use of simple pin connection at the pipe end. A detailed descrip-tion of these energy absorbers is provided in Section 8.
O Energy Absorber Relational Characteristics 2.2 Figure 2.2 depicts the configuration of a one-dimensional x-shaped energy absorter plate. It is readily apparent that this configuration provides a large volume of unifom plasticity when yielded, and that its hysteretic characteristics can be easily detemined as a function of its dimensions. Hysteretic behavior obtained from simple bending tests on x-shaped plates has correlated very closely with that derived for an ideal x-shaped beam.
1 For an ideal x-shaped beam (plate), the force-displacement relationship in the elastic region is:
i 3
=
Et b (2-1 )
2 12a t
O 2-2
C) corresponding displacement, and E is the material Young's modulus. As shown in Figure 2.2, a, b, and t are plate dimensions.
The relationship between the displacement d at strain eis 2 (2-2) d= 2c(a /t)
Af ter the plate has yielded, the force-displacement relationship becomes:
F,_, , 3 ,
1 (2-3) 2 F, 2(c/cy) where Fy and c y are the yield force and yield strain respectively and c is the strain.
Equation (2-3) indicates that the ultimate load for an x-shaped plate Fu approaches 1.5 times its yield load Fy .
O Equations (2-1) througn (2-3) also show that:
o The yield force Fj and the ultimate force Fuare proportional l to t and b/a but independent of the plate's length (for a given ratio of b/a),
3 2 o The initial elastic stiffness is proportional to t ,1/a ,
and b/a. 1 2
o The yield displacementy d is proportional to a and 1/t.
The above relationships are shown graphically in the force-displacement
< diagrams in Figure 2.3. The fom of the hysteresis curves at any strain value, based on the ideal x-shaped beam relationships, is indicated by the graphs of Figure 2.3. Figure 2.4 shows a typical hysteresis curve f rom a test specimen. The close correlation observed in the figures has been demons'trated by subsequent analysis.
1 2-3
1 The simple e6ergy absorber characteristics and their interrtlationships i shown in Figure 2.3 illustrate the capability to size energy absorbers to suit a wide variety of applications.
I 2.3 Operating Principles Energy absorbers designed according to the requirements of Section 2.1 and as described in Section 2.2 are simple flexible supports that interact with the supported piping as described in the following paragraphs ,
a) Themal Expansion Energy absorbers will accommodate piping themal expansion by flexing. Their finite, moderate, and well-defined elastic stiff-ness and peak forces allow accurate and reliable themal expansion analysis of the piping systens. Piping systems with energy absorb-ers are analyzed to satisfy the applicable code stress limits,
< nozzle load allowables, and other applicable criteria as required.
b) Seismic and Dynamic Loads Under seismic and dynamic excitation, the cyclic displacement of the piping will result in a hysteretic behavior of the energy abso rte rs. The hysteretic behavior will cause an increase in the effective damping of the system dynamic response, hence it will prcvide the desired control. An increase in the system's response will result in an inenrase in the effective damping, thus converg-ing on a stable level of system response. The design and analyti-cal pmcess allows complete control of damping values, strain levels in the energy absorbers, and other pertinent system response parameters. This has been proven in shaker table testing, and the test results have been correlated to theoretical and analytical methods with a high degree of accuracy as shown in the subsequent
- sections.
2-4
--,--,e- -w-- - - - - ~ . -----n_-e- - - - - - - - - ---..,,-.._,g .,g , ,.---- ,.m ,,- - - - - ---gr,- . - - , -- - , , - , - , . - ,
c) Seismic Anchor Movements (SAM)
Energy absorbers afe considered as flexible elements. Since the energy absorbers are more flexible than a locked snubber and become even more flexible as they undergo yielding during a seismic event, the system response due to SAM will be lower than for a correspond-ing system supported with snubbers or rigid supports.
An important feature of energy absorber design is that loads from ther-mal expansion, seismic and dynamic loads, and seismic anchor movement are not additive. Once an energy absorber has yielded, no increase in load beyond its load-displacement characteristic is possible. Al so ,
tests have demonstrated that thermal expansion loads in energy absorbers undergo a shakedown at the onset of yielding under dynamic excitation.
This is further discussed in Section 3. .
2.4 Fatigue Design O Energy absorbers are designed for cumulative fatigue effects resulting from all cyclic thermal, seismic, and dynamic loads in accordance with the requirements of Code Case N-420. Load rate testing of prototypical energy absorbers to develop fatigue design curves is discussed in Section 8.
O 2-5
4 l
i 1
i PROFILE DEFORMATION STRESS DISTRIBUTION
-a m-
[
m i
l Nonworkable l due to local
{
yielding only.
I J e _.
l l
1fy+
Concept 1 Workable J l i
- fy 4 O -- >
2 Concept 2 Workable l
I /
l 1
Concept 2a :. .. .Jorkable E _ __ __
i i
Figure 2.1 Con:ects of Steel Plate Energy Absorbers l 2-6 I
i
li l lll 'l O
j a
t=
i i
n 5 i
i . _
i . _
i w I ,
.il l I " I l J
C a q
> 1 J 1 o d 1 q
s 1
1
- v N1 1 A 6
a a a u
3 a
d 4 -
X J
j o
/
O i i
i i
/
1
/
1 t
p a
o i
i 1 w i / O u i 1 o
/ _
i i
i
/
/ ne i / u i /
/ u i / j o
i /
i /
a i 0 i < /
e / 4 i / 2 i , / '
i 8
/ 2
_ i /
i / a -
i / a i / n _
i / 6 i / t i / j i /
i /
i l :
i )
i . /
i /
i /
i /
i i
B u/
" a O ^s
- ! E .l l
- l ,!i !!l l\ l
l t.54 "
i l
O 2.25F y
- 3.375K Fu .
L '
Fy -
K e.25Fy 8.I25K 8.5L F 6 7d, d, 2d y Influence of the Thickness t of an Idealized X-Type Absorber to Stiffness K, Yield Load F y, and Ultimate Load F u
Iu Fy -
8.5e o I.5.
4K K e.44K t t t 8.25d y dy 2.25d, Influence of the Length a of an Idealized X-Type Absorber to Stiffness K, Yield Load F , and Ultimate Load F I.5F u I.Sh/s
- 1. 5F, -
F, ble Fy - 1.5 e.5F u 8.5F y -
K 8.5b/a 8.5K I
O Y Influence of the Angle (b/a) of an Idealized X-Type Absorber to Stiffness K, Yield Load F , and Ultimate Load F u
Figure 2.3 Energy Absorber Characteristics 2-8
_.-.---.___m.- _ . - _ . _
O -
4.00 _
l 2.as .
E e.se S vD l
m a -2.se .
-4.00 .
-e.68 0.00 e.68 1.28
-t.28 DISPLACEMENT CINCHES)
LOAD VS DISPLACEMENT FOR 8'X3/8*-II SPECIMEN AT
+-l.2 INCHES DISPLACEMENT-- 200TH CYCLE Figure 2.4 Typical Hysteresis Curve of an X-Shaped Energy Absorber t
Section 3 SU W RY OF SHAKER TABLE TEST RESULTS Extensive experimental studies of plate-type energy absorters have been conducted at the Earthquake Engineering Researt:h Center (EERC) of the University of California, Berkeley. In these studies, various plate geometries and materials of construction were investigated, resulting in the development of triangular and x-shaped energy absorber concepts.
Shapes similar to these have been previously utilized in some civil /
structural applications of energy absorption. Fatigue testing of sample plates, fabricated from mild AISI 1020 steels, demonstrated sufficient fatigue endurance at high strain levels (up to 2.5%). Th us , the feasibility of using these energy absorbers as seismic supports of piping systems without ever replacing them was established. A series of shaker table tests at EERC was then conducted to demonstrate the concept and obtain response data for subsequent use in analytical correlation and parametric studies.
The behavior of piping systems with energy absorbers, under seismic loadings, was studied in three separate and extensive series of shaker table tests at EERC. The results of the tests are briefly sumarized in the following paragraphs. Special emphasis is given to the first and third test due to the volume and significance of the test data obtained.
I 3.1 U-Loop Test I
a) Test Objectives and Setup
! The first extensive series of shaker table tests was conducted,
- under the sponsorship of the Department of Energy, on a single i plane U-loop piping configuration using triangular energy absorters. A complete description of this test is contained in O
i 3-1
Reference (7). The objective of this test was to stu@ the general concept of controlling (ynamic piping response with energy absorb-Q ers and the effects of combined seismic and themal loadings.
Although triangular piates were used, the conclusions drawn from this test art equally applicable to x-shaped plates.
Figure 3.1 shows the configuration of the U-loop used. It consisted of a 3.5-inch. 0.D. pipe fixed at the boundaries with concentrated masses added at each elbow and filled with water. One energy absorber in the horizontal direction and two in the vertical direction were used in various test runs, singly or in combina-tion. Over 80 test runs were perfomed on the loop with or with-out energy absorbers, simulating both cold and hot piping con-ditions. Parametric studies on energy absorter sizes were per-fo med. The table motion used was the TAFT earthquake. Peak acceleration intensities were varied between runs from 0.125g to 1.499 Pipe stresses, displacements, accelerations, energy absorter forces, and input table motions were recorded.
O Time-History Response of Unrestrained System b)
Figure 3.2 shows the measured unrestrained pipe displacement response to an earthquake of 0.142g intensity and a duration of 20 seconds. It indicates high response under the lowest input motion, c) Time-History Response of Cold-Restrained System Figure 3.3 shows the measured pipe response at the same location as in Figurt 3.2 with one horizontal energy absorter installed on the U-loop. The test run was for a TAFT earthquake of 1.32g intens-ity. The effectiveness of the energy absorter in controlling the system's response is evident.
O 3-2
d) Time-History Response of Hot-Restrained System Q I Figure 3.4 shows th'emeasured system response under the same test conditions as described in Section 3.1(c) but with simulated thernal expansion conditions. The response is shown for two conditions, one which thermally deflected the energy absorber in one direction and the other which deflected it.in the opposite direction. The results indicate that under dynamic excitation, a classical shakedown of the thermal forces in the energy absorbers occurs in one to three dynamic displacement cycles. The dynamic response of the piping system is identical under cold and hot test conditions.
e) Conclusions Drawn from U-Loop Test As is readily seen, the U-loop configuration was essentially a one degree-of-freedom system when excited in either the horizontal or the vertical directions. Therefore, important system parameters O such > a tur 1 freaueacies aad d matas coefficieat ere fouad accurately and the effects of energy absorbers were easily iden-ti fied. This fact was later utilized in correlating the test results with computer solutions, verifying nonlinear programs, and gaining a clear understanding of the characteristics of energy absorbers. Several important conclusions can be drawn from this test:
o Energy absorbers are able to dissipate a significant amount of energy as can be seen from the hysteresis curves of Figure 3.5. Thus, the dynamic response of the pipe is considerably reduced as compared to the response of a free pipe, while the force transmitted from the pipe to the support structure was limited to the yield force of the energy absorbers.
O 3-3 p
- - . = _ __ _ _- . -- -
1 o The damping coefficient of the free pipe was measured around V 25 of critical. When energy absorbers were used, the system damping increa' sed significantly, in some cases to values well above 20% of critical.
o Themal expansion or contraction effects are not additive to the dynamic system response.
o Performance characteristics of energy absorbers, measured from the shaker table tests, were verified to have a constant strain-stress relationship, thus correlating with models using the perfect kinematic hardening rule (8), (9).
3.2 Space Frame Test a) Test Objectives and Setup To examine the behavior of x-shaped energy absorbers in a general O uncoupied three-dimensioaai system, a spatiai piping system supported by a three-story steel frame was mounted on the shaker table. A detailed description of this test is contained in Reference (10). The frame was 18 feet high,12 x 6 feet in plan I
dimensions, and weighed approximately 26.6 kips. Several config-urations were tested, such as the frane without piping, the frame with the piping rigidly attached, and the frame with the piping supported by energy absorbers. TAFT earthquake records were used for all configurations. The results showed the expected reliable perfomance of the energy absorbers and provided conclusions similar to those gained from the U-loop test. The following additional advantages were revealed from this interactive piping-structure experiment:
O 3-4 D
b) ConcTusions Drawn from Space Frame Test
, o The fundamental frequency of the piping system supported with energy absorbers was lower than that for a rigidly attached configuration, o With energy absorbers, responses at higher frequencies were lower, overall system stresses were lower, and spectral peaks were broadened indicating that high damping was introduced by the hysteretic action of the energy absorbers.
o Energy absorbers close to valves with eccentricities were highly effective in reducing pipe stresses and valve acceler-ations.
3.3 Scaled Spatial Piping System Test a) Test Objectives and Setup O
This test was sponsored by the Electric Power Research Institute (EPRI). Bechtel provided advisory consultation, provided the system configuration, and facilitated the acquisition of some test material. In this experiment, an extensive series of tests (over 110 shaker table runs) were performed. A full description of the experimental setup and primary test results can be found in Refer-ence (11). In this experiment, a half-scale model of a section of the core spray piping system design from the Hope Creek nuclear plant (Figure 3.6) was constructed inside a rigid frame and mounted on the shaker table. The snubbers were installed and tested and were then replaced with energy absorbers, which were installed and tested. The applicable Hope Creek operating basis earthquake (OBE)
, and safe shutdown earthquake (SSE) response spectra curves and multiples thereof, properly modified to account for the scaling effect, were used as input motions to the shaker table. Despite the modeling considerations, however, the model was viewed as a O $=aiier prototvae aaa testea accora$ asis. Tae purpose of tais test was to:
3-5
o Apply the findings from the previous two tests to a realistic three-dimensional system.
o Evaluate the feasibility of replacing snubbers with appro-priately sized x-shaped energy absorbers on a one-to-one basis at the same locations.
o Obtain comparative system response data with snubbers and energy absorber setups.
o Develop benchmark data for analytical and correlation studies.
b) Stress and Valve Accelerations Figure 3.7 compares naximum stresses and valve acceleration in the piping at increasing intensities of earthquake. It is observed that at design earthquake intensity the two responses are g approximately equal. At higher than design earthquake intensities, V however, the responses with energy absorbers are lower than with snubbers, c) Support Loads The results here are predictable (Figure 3.8). With increasing ground motion intensity, there is no significant increase of force in energy absorbers after they have reached their yield values, whereas the snubber forces increase. This is a significant advan-tage because the maximum forces transmitted by energy absorbers to the building are smaller and are clearly defined. The rigid rod forces are sinflar in magnitude in both installations.
O 3-6
d) Fourier. Spectrum of Pipe Acceleration Response O Another way of comparing the snubber and energy absorter system responses is via the fast Fourier transfonnation (FFT) of pipe accelerations Figure 3.9. The snubber system response has a random characteristic with significant energy over the whole frequency spectrum (attributed to the impacting phenomenon associated with snubber activation). The energy absorber system response is more associated with the natural frequencies of the system, thus is more predictable.
e) Load-Displacement Response Figure 3.10 compares a snubber and its corresponding energy absorber hesteresis loops from similar tests. The snubbers displayed uneven, irregular, and multiple-impacting character-istics. The energy absorters displayed smooth and predictable characteristics. The areas within the loops represent energy absorted and are a measure of damping.
f) Damping The comparison of damping values is of grtat interest. Figure 3.11 shows the free decay curves of the piping system without snubbers or energy absorbers, with snubbers, and with 4-inch long and 2-inch long energy absorbers. It is observed that system danping without snubbers or energy absorbers is about 1.2% of critical. With snubbers it is 5.75. With energy absorbers it is 5.6% for the 4-inch size and 7.9% for the 2-inch size. The direct correlation of used energy absorbers to damping is evident. Given an install-ation in which optimization of energy absorbers is made by design, higher damping values will result. It is important to note that subsequent analytical work in which the hysteretic characteristics O
3-7
. _ _ _ .. ._ .. . _ . _ .- . . _ _ _ - . J
\
\
of the installed energy absorbers were used, produced damping O v.iues that are virtuaiis ideaticai to the measured vaiues. 4 direct correlation therefort exists between experimental data and analytical methods to calculate the damping associated with the use !
of energy absorters. l g) Conclusions Drawn frwm Scaled Spatial Piping System Test As was the case in previous tests, energy absorbers were demon-strated to be an effective means of supporting piping systems for earthquake loads. In addition, an important aspect of this test was that snubbers wert actually installed and tested befort being replaced with energy absorbers. Af ter the latter were tested, a direct comparison was made of the perfonnance of each.
The results indicated a more clearly defined behavior on the part of energy absorbers than the snubbers. The hysteresis curves of energy absorbers showed smooth and well-defined characteristics, O simiiar to thm auf ned f,em st. tic dispiacement tests. Snuaen, on the other hand, showed a behavior that is more difficult to predict or reproduce. See Figure 3.12 for comparative plots. The shakedown of themal fortes in the energy absorters was observed in the same manner as in the U-loop test. Throughout the test series, the energy absorbers were nounted as a direct one-to-one replace-ment of snubbers without optimization of location. Even with straight replacement, the experinental data showed advantages of energy absorbers over snubbers.
3.4 Overall Experimental Work Conclusions The summary of the experimental wort presented in the preceding sections has conclusively proven the viability of energy absorbers as supports for piping systems. The experimental wort demonstrated the following advantages of energy absorbers:
O 3-8
s ,
O c , t
. \ 1 o They provide some degree of decoup1f ng of seismici:1put to the pfping. 1 c s 3
o The) provfor as good or better contro1'of system seismic response as snubbers',J ,
x o They are more advantageous' than snubbers at large intensity excitations"due to the significant increase in their damping effect.
o The results' are predictable and repeatable. , Some. eurgy absorbers were subjected to more than 32 earthquake tests of varying intens-ities wit'i no degradation in retponse repeatability or character-istics.
O l'
, \
\
\ ..
o They possets sufficient fatigue endurance to{susta11 design condi-
'l ,
tions for the plant life. s C
y .
\g v' N o They are simple, require yo maintenance, andfare easily replaceable should sucll 'a need arise. .
.s o They can be ussd to replice snubbers in existing plants with a resulting increase in flexibilty and system damping,
', s y
1 \
\ \ ~
q
, {
\,
0s
(' s s4 I
. l
's i k
n o -
35 l
%. l r
) *%
O
\ [
\ "
/
$ f I
' N
/ s.
/ # N g [
4 AfIs-
- g. If F I ,
O , .0-2 PIPE LOOP 3 elbow p.
& SasulaTED v4Lvt ,
p 5 DeDIGT A850RK4 6 NeeGEO Sup*0RT 7 Npe0E0 COceECTs0N N00 P
/ / P >^.'
!s-r,., ,
4
<. 's Figure 3.1 Plan of Measurement Points and Isometric of U-Loop Piping System 3-10
O g ,, ,
090480.06 I
.V. _- - a m miy n n rv i , - a_m _ . , , . r
.r ,
i iy iy rv i
-4 . 9 -
CH-9 W120NTE TAGLE DISPLACBElff 1.0 -
e .
O-a i ir
^
y . ' ' / r4 v[
~
- 7 'i-V 1 ^ i-[- 6 ^ i
-d- a
,I l l
-1.0 . l CH-2 Moa!ZONTE TASLE MTION fI ! l O -
w I I i,t i
_,m_, aaA l 1 I ) l \ , b i
i v v 4 ' y rg -v U-
-(
i (Ij 1 f 1 I I I j j l
-e.5 _
l CH-20 WINAL PIPE DISPLACDENT ' l 'l 25.0 3.0 sec III I l l ,l I i
i !!
II I ' {j 1l
< yi i I6' 1
- i
- 4
{I Figure 3.2 Time-Histories. Without Absorber, Peak Table Acc. = 0.142 g 3-11 l
O 890480.85 L4.9 .
e nA Ahb O_
m n. y. , , , . . .
w y v y ~. %
' ~
TAKE DISPLACDENT CH-O las!ZWITAL ,
1.9 .
. ..., k h h h .h A h ; M IIA k. f.\ kt A . n 4. .k a A m.e e .
i ' T 7 6 ' ig - g'ti W y4 l' ( 3ylj' W {V V W' y y N IV'
-1.9 _
CH-2 HORIZONTAL TA8LE ACCELERAT!GI e.s -
O .
kl A AJi is i A ! 0I dh k ma-BL ski
),l i m a
, ,p- 1 o g v vp - p qr,rm rror " ,
~
CH-29 IW12 INTAL PIPE DISPLACDENT uos.e _
k 5 lh ih!I8 l- J l A AA m __sa .ns@ela a i n
' IP'T !)y 7.iTj ]Nff r '
jHI lI J TI' W' "
w.e _ 23.c CH-32 ABSORBER F0PCE etc Figure 3.3 Time-Histories, with Absorber, No " Thermal" Displacement, Q Peak Table Acc. = 1.32 g 3-12 O
090480.09 Q g,, ,
= l
! T.
^
s
/\, t/\ , i l , i b b, W u /\', yn gAvy tt r y M 12GITAL f TAKE DISPLACDENT
' ~
.e .
a .k 'k !hh ,IA > ihk k.b kA A k.ka'4.damM.m 18VVY.(1WlWPVyy7 I V' VI) 8 4.
W j l'y
} 'g ;' j'
, ' ' ' , ' VVi ~
W 12DIITAL tan g acert m T!WL CH-2 e.5 _
O i i- ,,,,,,,,,,
- ~
CH-20 l W .O - W!!DIITAL PIPE DisPLACDElff
..L !! >.11 !bil 6,A 1 IJ L l! .d 3.
li >:NI d k Illl1 A gg egyyywrp
._,.. Lei y
i l l
, , . . , ,.c WN.e -
D' 3.0 ese Figure 3.4a Time-Histories, with Absorber, +1.0-Inch " Thermal" Disolacement, Peak Table Acc. = 1.32 g 3-13
O 890480.1O u.e .
- f
^ b A i nA /\ A A /\ _
z
.y. , .
- i. ;. . w. y v y ~ , v
-44.0 -
CH-O IWi!!DNTAL ) TADLE DISPLACDENT 1.0 .
o _.
.*,,..b N Ab I A h M , ' A b, .6 kg A [i. .h e 4. .l p .. n, e W 'i i"j' y d pl(i W [v V PCy p ' M' yVT 8
' ' ' 6 " M ~ } ' gip '
ISE!!DNTAL TW ActIIDAT!W
-1.0 .
'"~'
g e. . -
qv M4WMe i 2%
Z,
, , . .-.c M .m A .
i,, . . . . . . . . . . . . . .
-0.5 _
CH-28 ICE!!!3 ITAL P!PE DISPLAcDENT uec.e .
l
- s Ao A .5 \\
, 1 . .
'l l'
' ~ " ~ P'Tg"y }l [I 'l}"'l
-400.e _
CH-32 ABSORBER FORCE g,e orc Figure 3.4b Time-Histories, with Absorber, -1.0-Inch " Thermal" Displacement, Peak Table Acc. = 1.32 g O 3-14
,. -- ,- i i i ,
M.=: IM.83 -
100.-h ,
i<
l' l
- I N.-,-
f -
v D
200.- - ; ; ; ; ; ;-
-1.58 -1.es .54 8.M 8.58 1.M l.58 DIST 29 i e i i <
l586H.N 300.-
298.- F } i IM.-F C e.
g 7
-lN.-F
- 298.- F l H
- 300. - -
l l l l 1.00 .54 8.M 0.58 1. M DIST 29 Figure 3.5 Hysteresis Curve of 1/8-Inch (Top) and 2xl/8-Inch Energy Absorbers Obtained from Shaker Table Test of U-Loop 3-15
! Rod Restraint #'
I-O- Snubber / Energy Absorber g
3 ,
W Spring s2 g
O Acceleration A Displacement C3 Strain Gases Jzr Load Cell b
9 e'l l'
63 oY
@ 59 4.
'J Top of Frame e*
h 57 4 55 hlJ<gSe N a 0 -
- u 3l% a b
54 pC A
h@
52 i A M 36 37 39 3s
\
Shaking Table Figure 3.6 Test flodel of lope Creek Core Spray, Piping System - IV l
l
- O l 3-16
O O O EmsI/In3 e.e -
0 M e " ~ ~~' -
0 o
A e.E -
f 9.4 - 0 -
S P~~* # A
,/
F*/*,.
T g,3 .
R #
/ '
A "
a[ -
I a.2 -
,/
N m' -
e.s -
g.,
E.e 0 0
- q le3
- ,E 4e , l a
N -e ,
/
ta A . *~ /
' X 3.0 ,# -
Y / ,/
p W A /
C 2.s y' C
~'H - y g.3 p *A * $ ~... ~
~ ~ "
mee m s:- Q Y '~ H e.E 1.E 2.8 e.e e.E I.9 1.E 2.s e.e 9.5 f.e f.E 2.9 9e t.e INTDeSITY FACTOR OF 08S INTENSITY FACTolt OF SSE INTENSITY FACToft OF S$$
o Sw H Hect. Leble occ. 333 . g..g.a 3.r. gg g ,n g, gy, a V vert. table eso. Des . Seet.J OpweL ine emele cerL%
En. ehe. 2 In.
o En. m . 4 in.
ssc - s.r. w t h cert w .
Figure 3.7 Extreme Values of Pipe Strains Accelerations of the Valve Operator,
' and Corresponding Shaker Table Response for Increasing Earthquake Intensities
O O O I '"3 -
Note 1 Note l s.e ses.s -
\- ~
/
.e K r --a
- ,.A ,
,F ,# ,,
ap . -
/
y des.e -
/
/ W e ,a ./
e ,* *
/
C E , fee.e -
I * .
- 7 a 2
~'
w n -
1 o
y-CIb.) ,
- ,, s. .
/
< ,e n , "
K -
,/
- ^
,f p.'
,/ er ~ , .
i 'f , ase.e j g e ./ 4
$ W E , ..
2.0 e.e S.S 1.0 9.5 2.0 e.e S.S l.8 1.5 2.e ee e.S 3.0 1.5 IIsitwsITY FACTOR or ces ItsTttegITY FACTee CF SSE igettps11Y FACTOst or SSS o W SSS = Se. led Sof. hideva Eethe,also
- s. re th,,ues.
a En. e. 2 In. 0.S = Sc.I.J Opw.llas e
- e to. 4. . 4 la. ssc - s.t. h t4 E -th.,=S..
Note 1: Flattening of snubber forces denotes local yielding in snubber support steel.
Figure 3.8 Maximum Reaction Forces Between Pipe and Frame at Restraint Devices (Top CH 53) and Rigid Rod Connections (Bottom CH 52)
~
O .
y ,
,g
- 3
- s
.A .i . m
\
- R
[-
R 2 - - Wl; 2 -
- W .
.9 9 g 2 :
2 a o
- .9 .
9 5
- m u
2
- .9
- .9
- :2 5 3
- n .
4 ,
u g
m
.g d
5
- 4 .
u B B R 3_ 8_ c
- m Ai Awov v Aussy g O
>=
q ., -
O g
.g C
g ,
b 6 .
> <. a <. g I jM -l R
m E
- m .n.
.n.
e ;
u 92
=9 ;
- g. t.
<g_ ,
~
.g -
j.g-
- T f a E
[ :
'g
- ~s i l~s
, [ * @
ci
.e f ',.
< c ;
e ,
. L 3
, , , . . e 4
gs 4 -
C ;
- e. e. 6 e. e. e. e. e. .-
- g' 3 3 3 I RS2 535 RR! - -
l 88 Aw l ei Ams, l
O 3-19
lf )
, l
=
O
' ; ? - ; l - -
,e s . .
s e.
. m.
2
- e i.
2 e e .
e e ,
- e . m - )
, s. s m m
e }
e i
t o
i s t
. , e. , - o o
e s s. B
,e a e s a (
s.
eT8 .
t s r
s e i
,e l e
- e. o e 0 1
b
- s. r i e .
e o
- s. .
. s b
a m .
A
, e. .
s y
. . o g
- . . s. r
. - e e .
n 1
- e .
. E
- - - s. ._ .~ - - - - -
d
~
- . . . . . . n e e e s a
- ' m8 e s s , s e s e
, m o e 2 , - s i m- )
, g l
- - - p oe 3 - Tk
( a u
O sq rh et _
e b r _
7 e 2 ba
. uE _
. e n n . r. S e
_ 2 e
s f a m
=
n
- e. mo -
s oS w .
e u ;
e se
- o .
- i. ph
. e ot
. . o 5
. L o 2 t
. 2 T
. ,se s id se
- S
- e. s
. I e o i et D .
. rc
- . ee tj
. sb s
e yu e HS
- n e .
i.
i
- s. ,
. 0 1
. 3 e
e e q2 r
, . . . . u n g mne s
o
- en ,
e e e
, e e i
u o es
- e. a 2 i t e
, s 2 f a ,
O yg J ;l: ;ll l fl!j j i*
0.87
8 3 ggggg,33 ,
~
f ( - $.7 %
S A ^f; . . _ - - suussEn 8 - 36 o
- l I ,
18.88 gag . 33, gigt 33 33'"
s.m -- i a i i igne,g -
~
-8\*)1I
( = 7.9 %
l l ii I AA..--_ 2 33fg. 8 - 47.5 %
l l
' l' i' VV"~~' - ENERGY f- i5 Hz ggg(ggi - ABSORBER
.,z . -
is.se oi . an. oist as i i e.fi i i
,b,,,,,,3
\\
( = 5.6 %
bd^^ - $*xl/8' 8 - 35.2 %
VVVV",- _-- - --
O 1 i ENERGY ABSORBER f - 6.65 Hz 4h kk , , , . i to.es . m. oist a
- ,hn& .6.. ..!.f.f !,Ill),lly!D , ///.
NO DEVICE 8 - . %
llhlIIllIId""" - f = 6.15 Hz n.= oi. .. .i., a n.a i
Figure 3.11 Damping Ratio of System IV with Different Support Devices in Position and Corresponding First Natural Frecuencies 3-21
. .- ___-____ = _- _ __.
O O O j 4gg,. .
119382.10 .
Snubber _ 150302.10 -
~
300.- '
~
200.- '
100 ~'
$ ~
q ,
e
=
s g
0-
_jgg,__ 2*xt/8" y '
- u. Energy y Absorber \ -
-200.- -
-300.- '-
i l -400.- -
.200 .100 0.000 9.100 0.200 1-DIST 21 ll Figure 3.12 Force-Displacement Behavior of a Snubber and a 2-Inch x 1/8-Inch Energy Absorber from Separate Tests Using the Same Input Table Motion
i I
Section 9 O
INSERVICE INSPECTION REQUIREMENTS 1
When nuclear pipe supports are constructed and installed to the rules of the ASME nuclear codes, the inservice inspection (ISI) of Section XI of the code applies. Therefore, energy absorbers constructed and installed to the rules of Code Case N-420 are subject to Section XI rules. While energy absorbers are not explicitly mentioned in Section XI requirements for ISI, the intent of the existing rules are applicable.
Table IWF-2500-1 of Section XI (enclosed as Table 9.1) specifies the exam-ination categories for all support types. With the exception of snubbers and I variable and constant springs, the required examination method is the visual, i
VT-3 method described in Paragraph IWA-2213: 6 I "IWA-2213 Visual Examination VT-3
"(a) The VT-3 visual examination shall be conducted to detemine the O generai mechanicai and structurai conditions of components and their supports, such as the presence of loose parts, debris, or abnomal corrosion products, wear, erosion, corrosion, and the loss of integrity at bolted or welded connections.
"(b) The VT-3 visual examination may require, as applicable to detemine structural integrity, the measurement of clearances, detection ofphysical displacement, structural adequacy of supporting elements, I connections between load carrying structural members, and tightness of bolting.
"(c) For component supports and component interiors, the visual examination may be perfomed remotely with or without optical aids to verify the structural integrity of the component."
i 9-1
! O i
, _._. _ , _ . _ , _ . , ~ - , . . . . . . - , . , , _ - - - - , - - - , - - - -
Visual VT-3 examination rules apply to the energy absorbers described in this enclosure. To facilitate performance of a VT-3 examination on energy Q absorbers without removal or disassembly, the openings and scratch plates described in Section 8 w re provided.
A visual VT-3 examination on energy absorbers will provide the level of assurance intended by the ISI requirements for the following reasons:
o Energy absorbers are merely a simple type of pipe support designed to strain and fatigue considerations. They are not_ springs. They are not snubbers.
Energy absorbers do not contain activation mechanisms, internal moving parts, or fluids. They contain simple bolted connections. Therefore, from an ISI point of view, they should be viewed as being similar to rigid-type component standard supports.
o The basic failure mode of an energy absorber is fatigue cracking of an absorbias a'at - co at' rat'9ue crac'$as occur' t'r ta' cic'ic "
O of the material corresponding to the imposed strains is consumed. When energy absorbers are applied to piping systems, they are designed to sat-isfy design conditions and fatigue requirements without replacement.
Their fatigue design will be based on fatigue design curves developed from testing of prototypical specimens per the requirements of Code Case N-420.
Fatigue cracking, when it begins, is usually surface-type cracking originating close to plate edges and is very visible and easily identifiable by the simplest visual examination (the naked eye). Based on prototypical testing, a visible surface crack takes a relatively large I number of cycles before it results in complete failure of the plate (ranging from the low 10s at the highest strain levels tested to the mid-l 9-2 O
100s at low strain levels). Formation of a visible crack does not degrade the function of the plate until the crack has significantly propagated and O resulted in separation of a large percentage (25% or more) of the plate's cross section. Therefore, p1'ates will not fail catastro- phically in fatigue, rather, if failure were to occur, it would be over a reasonably large number of cycles. One earthquake event does not contain enough cycles to cause a surface crack to propagate significantly and cause degradation of a plate's function. Themal expansion cycles occur over a long span of time.
Energy absorbers are typically designed to sustain low strain levels under themal expansion conditions. The number of themal cycles assumed in the design is usually very conservative. Thus, a crack propagation would require 100 or more themal cycles. One inspection interval contains much less than 100 themal expansion cycles.
o Perhaps most significantly, the design of energy absorbers discussed in Section 8 incorporates multiple plates. It requires two plates or more to function, and most applications will contain three or more plates.
Based on fatigue testing, fatigue failure is basically limited to one plate in an assembly of plates. This is expected because random, natural distribution of crack initiating factors will favor one plate to take the lead in crack fomation and propagation. Loss of one plate in a com-plete assembly will not completely incapacitate the energy absorber. This further reduces the significance of a fatigue failure of one plate.
o Energy absorber designs additionally incorporate two aids aimed at simplifying and enhancing the visual ISI examinations. The first is windows cut out of the sides of the boxes to facilitate viewing of the plates and bolts. The second is the scratch plate and markers at the pin location, which will provide physical evidence of pipe movements and their magnitudes. This physical evidence of pipe movement can be directly correlated to design conditions and fatigue endurance.
9-3 9
o 1 'l l I
} l.
-l I l- I
.lI ._c
_s_s ,.
_c
- - it it ii ii .9 i : : :: ri: ::: s p- n:ii it tii nii .: '
X 3
$l* I j e
e s
e s
e s
e s
ols l l ! !! ! j
'l- l - _ _ _
=
5 1e ,
a e
i
- o I -l - 1 1 1 1 E i ;-
2 N
gj j 3
j 1; .1 h
j I I i
l!"d,j I ,
l 11i 1h1 i 1I J h h h m O 9-4
e o , 1 11 1 I! I i 1-I a a 3 3 y es e9 -9 -9
- !! !! 11 11 e
22R 22 222 222 3
a Iff 56 465 666 1_1_1_ 1_ 1_ t_ i_ l_ f_i_t_
b$ $ 5 55 1 1 T
5 p
I
> . I I I I
-a 5
=
s
- e. - 3 - - - .
Ow g I A , $
4 i
i8 5_ 5_ 5_ 5_
zij I
u!. II I* 11.
I '
16II' I
ilf a
i j
i
} I 13 jg =I E
=
I
=
I
=
0 9-5
1 l
0 1 ~l l 1 l l l Il 1 I s ss s s i
i
- ,-S e ,e ,e
-9 -9
,e
-9
,e
-9 e
I
- !! !! 11 11 !!
Ia eee ee eee eee ee:
R 349 4: 144 44
,, 1 I_I_I_ II__ I_II_
_ I_I_I_ I_I_
=
55 g s ? ? ? ? i 5 8 l-G1 ,
ll
. 1, 1, 1, 1, 1, s
- s. g; e Wg $ . m W. 3 . . - - -
30 5 I I I 1 I I $
! : : : :$ a w :$ $ $ $
i w
diIi '
1 ,11 i y j Ifgi'lll'111>I n.
lyii l p- i, I4 !I a 11 e
I y= Ill !
. .. . . e se bb * * *
- 9-6 D
r
~
REFERENCES
- 1. Welding Research Council', " Technical Position on Damping Values for Piping--Interim Sumary Report," Bulletin 300, December 1984.
- 2. Shibata, H., et al. "A Study on Damping Characteristics of Piping Systems in Nuclear Power Plants," Institute of Industrial Science, University of Tokyo, undated.
- 3. Ware, A.G. , " Damping Test Results for Straight Sections of 3" and 8" on Pressurized Pipe," NUREG/CR-3722, April 1984.
- 4. ANCO Engineers, Inc., " Laboratory Studies: Dynamic Response of Prototypical Piping Systems," NUREG/CR-3893, August 1984.
- 5. Spencer, P.N., V.F. Zackay, and E.R. Parker, "The Design of Steel Energy Absorbing Restrainers and Their Incorporation into Nuclear Power Plants for Enhanced Safety; Volume 1--Summary Report," University of California, Berkeley, UCB/EERC-79/07, February 1979.
- 6. Earthquake Engineering Research Center, "The Design of Steel Energy Absorbing Restrainers and Their Incorporation into Nuclear Power Plants for Enhanced Safety:
a) Kelly, J.M., and M.S. Skinner, Volume 2 " Development and Testing of Restraints for Nuclear Piping Systems," UCB/EERC-80/21, June 1980 b) Kelly, J.M., and M.S. Skinner, Volume 4 "A Review of Energy Absorb-ing Devices," UCB/EERC-79/10,1979.
c) Owen, W. S. (MIT) et al, S. J. Hartman (UCB), et al, " Volume 3 -
Evaluation of Comercial Steels," UC8/EERC-79/09, February 1979.
- 7. Stiemer, S.F., and W.G. Godden, " Shaking Table Tests of Piping Systems with Energy-Absorbing Restrainers," University of California, Berkeley, O UCB/EERC-80/33, Septeater 1980.
R-1 1
- 8. Hill, R..' The Mathematical Theory of Plasticity, Clarendon Press,1950.
- 9. Prager, W., An Irrtroduction to Plasticity, Addison-Wesley,1959.
- 10. Stiemer, S.F., W.G. Godden, and J.M. Kelly, " Experimental Behavior of a Spatial Piping System with Steel Energy Absorbers Subjected to a Simulated Differential Seismic Input," University of California, Berkeley, UC8/EERC-81/09, July 1981.
- 11. Schneider S., H.M. Lee, and W.G. Godden, " Behavior of a Piping System Under Seismic Excitation," University of California, Berkeley, UC8/EERC-82/03, May 1982.
- 12. Mondkar, D.P., and G.H. Powell, "ANSR-II, Analysis of Nonlinear Structural Response User's Manual," University of California, Berkeley, UC8/EERC-79/17, July 1979.
- 13. H&K, Inc., "ABAQUS-ND: A Finite Element Code for Nonlinear Dynamics,"
O Electric Power Research Institute, NP-1552-CCM, February 1981.
- 14. Bathe, K.J. , E.L. Wilson, and R.H. Iding, "NONSAP: A Structural Analysis Program for Dynamic Response of Nonlinear Systems," University of California, Berkeley, UC/SESM-74/3, February 1974
- 15. Clough, R.W., and J. Penzien, Dynamics of Structures, McGraw-Hill,1975.
- 16. Stricklin, J. A., and W.E. Maisler, "Fomulations and Solution Procedures for Nonlinear Structural Analysis," Computers and Structures, Vol. 7, 125-136, 1977.
- 17. Hilber, H.M., T.J.R. Hughes, and R.L. Taylor, " Collocation, Dissipation, and Overshoot for Time Integration Schemes in Structural Dynamics,"
Earthquake Engineering and Structural Dynamics, Vol. 6,99-117, 1978.
l l
R-2 l t
- 18. Mondkar V.P., and S.H. Powell, " Finite Element Analysis of Nonlinear O Static and orn-ic Responses, internationai Journai for a-ricai Methods in Engineering..Vol . 11, 499-520, 1977.
18a. Newmark, N.M., "A Method of Computation for Structural Dynamics," Proc.
ASCE, Vol . 85, No. EM3, 67-94,1959.
- 19. Jennings, P.C., " Response of Simple Yielding Structures to Earthquake Excitation," C.I.T. Report 63-60, June 1963.
- 20. Caughey, T.K., "Sinusoidal Excitation of a System with Bilinear Hysteresis," Journal of Applied Mechanics, Trans. ASME, Vol. 27, Series E, No. 4, 640-643, December 1960.
21 . Jennings, P.C., " Equivalent Viscous Damping for Yielding Structures,"
Journal of the Engr. Mech. Division, Proc. ASCE, Vol. 94, No. EM1, 103-116, February 1968.
O 22. iwan, W.O., Estimatin, ineiastic Res,onse Spectra from Eiastic S,ectra.-
Earthquake Engr. and Struc. Dynamics, Vol. 8, 375-388, 1980.
- 23. Grossmayer, R.L., and W.D. Iwan, "A Linearization Scheme for Hysteretic Systems Subjected to Randon Excitation," Earthquake Engr. and Strue.
Dynamics, Vol . 9, 171-185, 1981.
- 24. Traill-Nash, R.W., " Modal Methods in the Dynamics of Systems with Non-Classical Damping," Earthquake Engr. and Struc. Dynamics, Vol. 9, 153-169, 1981.
- 25. Tansirikongkol, V., and A. Pecknold, " Approximate Modal Analysis of Bilinear MDF Systems," Journal of the Engr. Mech. Division, ASCE, Vol.
106 No. EN2, 361-375, April 1980.
- 26. U.S. Atomic Energy Comission, " Combination of Modes and Spatial Components in Seismic Response Analysis," Regulatory Guide 1.92, December O 1974.
R-3
- 27. Gasparini, D., and E.H. Vanmarcke, " Simulated Earthquake Motions Compared O. .i n prescrised Response Spectra, M.r.T., oePt. or Civii En9 1neerino, R76-4, January 1976. -
- 28. Nigam, N.C., and P.C. Jennings, "SPECEQ/SPECUQ, Digital Calculation of Response Spectra from Strong-Motion Earthquake Records," A Computer Program Distributed by NISEE/ Computer Applications, Earthquake Engineering Research Lab., C.I.T., June 1968.
- 29. Bechtel Power Corporaton, "ME101 Benchmark Problem: RHR Head Spray SK-M-1550, Rev. M," San Francisco, Project 08031, Problem 1-10-09, December 1982.
- 30. Bezier, P., M. Harti:t.an, and M. Reich, " Piping Benchmark Problems:
Dynamic Analysis, Uniforn Support Motion Response Spectrum Method,"
Brookhaven National Lab./NRC, NUREG/CR-1677, Vol.1, August 1980.
31 . Timoshenko, S., and N. Goodier, Theory of Elasticity, McGraw-Hill, 2nd O ed. 1961.
- 32. Fung, Y.C., Foundations of Solid Mechanics, Prentice-Hall,1969.
- 33. U.S. Atomic Energy Commission, " Damping Values for Seismic Design of Nuclear Power Plants," Regulatory Guide 1.61, October 1973.
- 34. Tsai, N.C., " Spectral Peak Ratio Curve", Bechtel Power Corporation, File No. J820,821, February 19, 1975.
- 35. Hill, R., The Mathematical Theory of Plasticity, Clarendon Press,1950.
- 36. Stephen, R.M., " Cyclic Loading of Energy Absorber Restraints," Structural Engineering Laboratory, Report 84-2, July 1984 O
R-4
.