ML20196C544

Alternate Seismic Criteria & Methodologies for Fort Calhoun Station Vol 1 Criteria & Methodologies

ML20196C544
Person / Time
Site:	Fort Calhoun
Issue date:	07/31/1988
From:	OMAHA PUBLIC POWER DISTRICT
To:
Shared Package
ML20196C542	List: ML20196C538 ML20196C544 ML20196C571
References
NUDOCS 8812080036
Download: ML20196C544 (60)

Text

_ _ _ _ _ _ _ _ _ _ .

ALTERNATE SEISMIC CRITERIA & METH000LOGIES FOR FORT CALHOUN 5TATION VOLUME 2 JUSTIFICATIONS OF CRITERIA & METHODOLOGIES Prepared for U.S. Nuclear Regulatory Comission Prepared by Omaha Public Power District i

July, 1988

$$k20ggggpg{QQjg3 p PDC

TABLE OF CONTENTS VOLUME 2: JUSTIFICATIONS OF CRITERIA & HETH000LOGIES PAGI TABLE OF CONTENTS

1.0 INTRODUCTION

1 2.0 PIPING AND PIPE SUPPORTS 3 2.1 Stress-Strain Correlation and Strain Limit 3 2.2 Yield Strength Increase 11 2.3 Hanufacturer's Allowable loads 13 2.4 Static Load Rating 14 2.5 Random Vibration for Combinations of Modes and Levels 15 2.6 PVRC Damping 16 2.7 CQC Method for Combination of Modes 18 2.8 Support Ductility 19 3.0 ELECTRICAL RACERAYS 26 3.1 Allowable Spans for Conduits and Cable Trays 26 3.2 Cyclic Load Rating 27 3.3 Damping 28 4.0 HVAC 30 4.1 Qualification by Test 30 4.2 Qualification by Experience Data 31 4.3 Damping 32 5.0 EXPANSION ANCHOR BOLTS 33 5.1 Factors of Safety 33 5.2 Tension / Shear Interaction 34 5.3 Edge Olstance and Spacing Requirements 35 5.4 Concrete Topping 36

6.0 REFERENCES

37 l

1.0 INTRODUCTION

This document provides a set of alternate seismic criteria and methodologies for Fort Calhoun Station Unit 1. These alternate criteria and methodologies do not replace the current Fort Calhoun Unit 1. USAR criteria and evaluation methodologies. These are retained as the principal design basis for the existing structures, systems and components. The alternate criteria and methodologies provide an alternative basis for future new designs, modifications and reanalyses of piping and pipe supports, electric raceways, HVAC and anchor bolts. In conjunction with the alternate criteria, revised seismic response spectra have been developed for the Reactor Building, Auxiliary Building and Intake Structure using current site-specific seismic hazard evaluation and soil-structure interaction methodologies.

Fort Calhoun, Unit 1, was designed in the late.60's and early 70's, using the design techniques of that period. Since that time, there have been considerable advancements in seismic design criteria and analysis methodologies, particularly in the last five years. In general, these advanced criteria and methodologies show that previous approaches overpredict l

the seismic response and underpredic+ the seismic capacity of structures, I systems and components. This has resulted in the design of complex and rigid systems which actually reduce maintainability and reliability during normal plant operations. Recent seismic testing and experience data collection programs sponsored by the NRC, utility groups and EPRI further demonstrate that the piping, raceway and HVAC systems installed in power and industrial facilities are very unlikely to fail during an earthquake, even if they have been designed only to carry dead loads.

To reduce overconservatisms in seismic analysis and design, the most effective approach is to develop a consistent set of alternate criteria and methodologies which evaluate both seismic demand and capacity aspects.

Seismic demand can be reduced by using acceptable state-of-the-art seismic load generation and analysis techniques. Seismic capacity estimates can be increased by using results of extensive industry-sponsored research and test programs.

The alternate seismic criteria and methodologies are not part of any specific OPPD new design or modification effort. It is not intended to apply them wholesale in reevaluation of piping, electric raceway and HVAC systems to remove supports. The intent of the subaittal is to license a set of alternate criteria and methodologies which can be used (in addition to thote accepted in the USAR) for future new designs, modifications and reanalyses. By requesting timely approval from the NRC, both the NRC and OPPD will benefit from a review and approval period in which issues can be discussed separately from any particular new design or modification effort.

This document contains two volumes. Volume 1 discusses the alternete seismic criteria and methodologies and Volume 2 discusses the justifications of these criteria and methodologies. In Volume 1, Section 2 defines the scope of this document: Section 3 discusses in detail the alternatt seismic criteria for piping (large bore, small bore and instrument tubing), pipe supports (standard

component type supports and linear type supports), electrical raceways (conduits, cable trays, raceway supports and raceway-to-support connections),

HVAC (ducts, supports and miscellaneous hardware), and anchor bolts used for piping, raceway and HVAC systems; Section 4 discusses in detail the alternate seismic methodologies for load generation and analysis techniques for the systems involved. In Volume 2. justifications are provided for all criteria and methodology items which are exceptions to the applicable codes, standards, and/or current USAR approaches.

l l

l l l

\

2.0 PIPING AND PIPE SUPPORTS The criteria for evaluation of piping and pipe supports are based primarily on the ASHE B & PV Code (1). This section provides justification for the cases in which additional or relaxed criteria have been used.

2.1 Stress-Strain Corvelation and Strain Limit The piping strain criteria, as stated in Volume 1, Subsection 3.1.1.3, provide an alternate approach to the ASME B & PV Code Class 2/3 stress evaluation for the Faulted (Level 0) condition.

The basis of the ASME B & PV Code, Appendix F, stress limit for the Faulted condition is that the piping will remain structurally integral and that the pressure boundary will remain intact. The Fort Calhoun alternate criteria are also intended to ensure piping integrity, and therefore augment the ASHE B &

PV Code, Appendix F, criteria. "Piping integrity" is defined herein as piping thich maintains structural integrity and shows no significant decrease in rated flow canacity, i The stress-strain correlation approach was reviewed by the NRC and approved l for its use in the San Onofre Nuclear Generating Station - Unit 1 Long Term l Service Program (4),

2.1.1 Strain Limits Piping integrity can be shown by the establishment of limits on material strains to ensure an acceptable limitation of deformation and to provide a suitable margin to rupture. The strain limits established for the Fort Calhoun alternate criteria for piping evaluation are:

(tiIpercentforcarbonsteel

$ 2 percent for stainless steal where et - Hasimum piping membrane-plus-bending strain based on an elastic stress to inelastic strain correlation, as discussed in Volume 1.

As stated in Volume 1, Subsection 3.1.1.3, generic application limits for both stainless steel and carbon steel piping are required to ensure applicability of the stress-strain correlation. In cases where stainless steel strain exceeds 1 percent, the following additional checks are made:

(1) Compressive wrinkling (local buckling)

(2) Low-cycle fatigue (3) Increased pipe reactions and displacements 2.1.2 btress-Strain Correlation

' To calculate the piping strains up to the limits specified requires inelastic l or nonlinear analysis methods. However, performing nonlinear analysis is expensive and time-consuming. Therefore, an approach which allows the use of l

standard linear elastic anal d s techniques and conservatively converts the

elastica 11y-calculated stresses to strains is desirable. The conversion used is as follows:

For carbon steel: (t-Kgf For stainless steel: (t-K 3 2.0 E where it - tbximum piping membrane-plus-bending strain ce- E' istically-calculated stress for pressure, gravity and s lstIc loadings, based on stress intensification factot approch per AsHE B & PV Code NCIND-3652.2 (psi).

E - Young's .cdulus, from Apper.oix I of the ASHE B & PV Code (psi).

KS

- Strain correlation factor, equi,1 to the inverse of the material parameter n as defined in ASHE B & PV Code, Table NB-3228.3(b)-1, 2.1.3 Justifications Justifications for the stress-strain correlation and strain limits were presented in submittals to the NRC (2) (3) for Southern California Edison's SONGS-1 Long Term Service (LTS) Seismic Reevaluation Program. Acceptance conditions made in the NRC Safety Evaluation Reports for the LTS Strain

• Criteria (4) are incorporated into the alternate criteria for Fort Calhoun.

In general, the stress-strain correlation is based on the fatigue evaluation procedure of the ASHE B & PV Code and is verified by comparison with test results, particularly the Greenstreet elbow test results (5). Both bases are briefly described below. Justification for restrictions, when placed on the use of strain criteria above the 1 percent limit are also briefly described in this subsection. /

A. BSEV Code Fatious Evaluation ____Proadun Based on a review of the alternatives to determine strains from elastica 11y-calculated stresses, an approach similar to the fatigue evaluation procedures of the ASHE B & PV Code (1) is relected. This approach is chosen because the ASME B & PV Code fatigue evaluation is a strain-based methodology *. Also, the ASHE B & PV Code fatigue evaluation contains a procedure for the simplified elastic-plastic evaluation of piping components. By reviewing the background behind these procedures, it is concluded that their philosophy and approach provide a method for conservatively calculating strains from elastica 11y-calculated stresses.

The ASHE B & PV Code simplified elastic-plastic methodology was developed to account for the effects of the strain concentration phenomenon which occurs Note

• The ASHE B & PV Code design fatigue curves are developed from strain-controlled fatigue data and then converted to representative ~

stresses, which are not the actual stresses applied but have the advantage cf being directly comparable to stresses calculated on the ,

assumption of elastic behavior.

~. - --- --

f when stresses are greater than the yleid stress. This is shown in Figure a 2.1-1. When the stress exceeds the yield stress, the actual strain, is the elasticallyt.

[ exceeds the elastically-calculated strain (0,).

calculated stress divided by the Young's modulus (,a ,'7 o

e . The ASHE B & PV Code defines the strain concentration to measure the differences between the elastically-calculated strain and the actual strain beyond the yield point.

As shown from Figure 2.1-2a, the strain concentration factor is constant and equal to elastic stress concentration factor between Points A and B, when the material behavior is perfectly elastic. The strain concentration factor h-e increases steadily above yield stresses until a maximum value is reached 1 (Point C in Figure 2.1-2a). The strain concentration decreases after this as the deflection increases.

The ASHE B & PV Code idealizes the above material behavior by introducing the strain concentration factor Ke (ASME B & PV Code NB-3228.3). As an example, the strain concentration factor for stainless steel is shown by the curve in Figure 2.1-2b. The maximum strain concentration is defined by the inverse of thematerialparametern,andtheslopeofgegtrainconcentrationinthe transition region (between 1.0 and 1/n) by n /n(m-1). Figure 2.1-3 shows F

the relationship between the strain concentration and stress for the ASHE B &

PV Code and several tests on 304 Stainless Steel.

The K, factor developed for the ASHE B & PV Code fatigue evaluation correlates the elastically-calculated stresses to actual strains for use in the ASME B & PV Code fatigue curves. For the Fort Calhoun Alternate Criteria, the maximum K factor will be conservatively applied to determine the strain correlation fa,ctor K5 as follows:.

2 K, -

K, -

f B. Comoarison with Greanstreet Elbow \*st Rg3ylti D11criotion of Tes11 The experimental work performed by Greenstreet (5) determined load-deflection and load-strain responses for sixteen 6-inch nominal diameter commercial carbon steel pipe elbows and four 6-inch stainless steel elbows **. The material for carbon steel elbows is ASTH A-106, Grade B and the material for stainless steel elbows is ASTH A-312, Type 304L. These material properties closely match those used on Fort Calhoun piping. Each specimen was loaded with an external static force of sufficient eagnitude to produce predominantly plastic response. The influences of elbow bend radius (long radius and short radius) and wall thickness (Sch. 40 and Sch. 80), as well as the effect of internal pressure (five specimens were loaded with internal pressure) were studied. The load-deflection curves and load-strain curves obtained in the Note ** Elbows are of particular importance because they are often the most flexible components in a piping system. Elbows are forced to accommodate displacements arising from various loadings including earthquake. Therefore, the integrity of elbows usually governs the integrity and functionality (flow area) of a piping system.

)

N x - _ ______

tests were limited by the test appar.itus. As a result, the test results used in the comparison do not represent the maximum capacity of the elbows.

Greenstreet plotted ten load-strain curves, all without internal pressure. A detailed breakdown of these tests is summarized below:

Fig. Number in loadino Condition Greenstreat

_Recort Soecimens Descrio. Material (2) tigment(l) Pressure 22 PE-1 Sch.40-LR C.S. in-plane (+) No 23 PE-2 Sch.40-LR C.S in-plane (-) No 24 PE-3 Sch.40-LR C.S. out-of-plane No 25 PE.9 Sch.80-LR C.S. in-plane (-) No 26 PE- ; Sch.40-SR C.S. in-plane (-) No 33 PE-i9 Sch.40-SR C.S. in-plane (-) No 34 PE-20 Sch.80-SR C.S. in-plane (-) No ,

2' PE IS Sch.40-LR S.S. in-plane (-) No 21 PE-17 Sch.40-SR S.S. in-plane (-) No l

.2 P'.-18 Sch.80-LR S.S. in-plane (-) No l

Note: (1) A positive in-plane moment causes the elbow to open; a negative in-plane moment causes the elbow to close.

(2) C.S.: Carbon Steel ASTM A-106, Grade B S.S.: Stainless Steel ASTH A-312, Type 304L. ,

inten ulation of Comparison The SCE SONGS-1 LTS Report (2), Appendix A, shows load-strain curves reported by Greenstreet together with load-strain curves calculated by stress-strain correlation methodology. A factor of 2.0 is multiplied to KS a e/E for stainless steel material in order to conservatively derive a satisfactory comparison with test results. In reviewing these load-strain curves, the following observations are made:

(1) The stress-strain correlation methodology will overpredict the strains, when compared with the Greenstreet test results. The overprediction is in the range of 25 percent to 100 percent for nine out of ten load-strain curves (excluding specimen PE-19).

(2) For specimen PE-19, test-recorded strains are approximately 10 percent t l higher than the calcula ted strains in the vicinity of the test cut-off i region, 1

-. ________________________________~

s f.y yn . (, .

, * <N ..

g .

y E (3) With ex' m

m ., it appears that the calculated strains and hA 4 ust-re m - :r:. is will converge at higher strain levels.

m The If, percen: c.at.sance for specimen PE-19 (observation 2) and the potential .'.h @

cony hgence at higher strain levels for other specimens (observation 3) are E not t o ncern because of the very conservative comparison made. Hajor q ?'y' ;

Te,41 a conservaUsms in the comparison are discussed as follows: 4 .e.y j (1) The calcelated strains based on stress-strain correlation methodology are - 'M 4.h i" intended tu predict membrane-plus-bending strains (i.e., strain averaged through the ull thickness plus strains at the surface due to an ktdf

p. d

(

equivalent liner distribution of strain through the wall thickness). As stated in the Grevistreet test report (5), the test load-strain curves QS -

were plotted for loc ' ions with strain gages mounted at or near the .. .n "maximum" or "peak" strains on elbow surfaces. ,

Figure 2.1-4 shows an elbow cietch section under in-plane bending.

In-plane bending was applied to C 1 tests with load-strain curves plotted, except PE-3. Strain gages ct 90', 270' and the near vicinity S er

,h ]

(O' is at extrados and 180' at intradoO would pick up predominent d.M -

circumferential strains because the axis ciussing 90' and 270* is a major 'A 3.*$

F axis of ovality for the deformed cross-section, as well as a neutral axis under in-plane bending. Strain gages at O', 180' and the near vicinity

.%Q

/4 .,

[ would pick up both axial and circumferential strains, with the axial T,Y J' strain at its maximum. Figure 2.1-5 shows the maximum axial and the . . r. '

maximum circumferential strain distribution through the wall thickness g'*

^

under in-plane bending. The axial strain is more or less linearly p{ p distributed through the wall thickness and is therefore a 3 n ..

membrane-plus-bending type strain. However, the circumferential strain , Q 1.,

contains local or peak effect (local strains at any point). This is g'I . i particularly evident for the elbow's inside surface at or near the major q, i.fl "

r axis of the crotch section. Greenstreet plotted the circumferential $ , . .i f strain versus angular location (from O' to 180') as function of load for 1 ' l. . '

r specimens PE-18 and PE-19 (Figures 29 and 30 in (51). Near the 90' 4.;'. ,

location, where the maximum circumferential strain appears, the 1 test-recorded strain on the inside surface is much higher than the -;., '

test-recorded strain on the outside surface. Since the tests were j ' l performed without internal pressure, the membrane part of the <N. l+ ,':

circumferential strain should be small. This indicates that .'he . = .'

~

test-recorded circumferential strain is not linearly distributed through .

the wall thickness and therefore contains local or peak effects, particularly on the inside surface.

g4-A review of the Greenstreet test load-strain curves indicates that the ... .' .

gage locations for the maximum strains are, for most cases, at or near .

=

90' and 270'. For Specimen PE-19, the load-strain curve was plotted by ib Greenstreet for the maximum strains located on inside surface at the 90'  ? - e E location (strain gage number A+2 as shown in Figure 2.1-4). At this . ,.,.

location, the local strain effect is at its maximum. '

~k ' }.y$

As a conclusion, it is censervative to compare the calculated membrane-plus-bending strains with test-recorded maximum strains which g contain local or peak effect. . f

~

.s

~

K g-6h9 5

$a t

(2) All load-strain curves reported by Greenstreet are from tests where the elbows were not pressurized. In actual cases, elbows are typically pressurized and the internal pressuro will improve the integrity and stability of elbows. This is because by reducing the elbow flexibility with its internal pressure, the measured strains, both axial and circumferential, would be reduced.

(3) The test load-strain curves beyond the test cut-off points were extrapolated by straight lines which are tangent to the test cut-off regions as shown by dotted lines on several load-strain curves. These extrapolated straight lines are relatively flat. In actual cases, the test load-strain curves should be skewed upward beyond the test cut-off points due to the strain hardening phenomenon.

(4) The loads in the Greenstreet tests were applied statically. Studies (6) 0 'v that the margin against failure of piping systems is significantly greater for dynnmte loads, such as an earthquake, than for static loads when piping responses are held to the same allowable level. Also, in establishing the allowable strains, the material overstrength and strain rate effects under dynamic loads were not credited.

Therefore the calculated strains based on stress-strain correlation methodology are conservative, when compared with experimentally determined strains.

C. Apolication Limits Application of the strain criteria is limited generically for both stainless and carbon steel material, and specifically for stainless steel when it is strained beyond I percent. The application limits are designed to address any concerns expressed as part of the SONGS-1 LTS criteria review by the NRC (4),

orassuggestedbyNUREG 1061 (7) (8). Specifically, application limits for the following top.cs are discussed to provide clarification of intent:

(1) Comoressive HrInkling. To avoid compressive wrinkling failure, the calculated strain for stainless steel pip 6 will be limited by the lower of two percent and 0.2t/R, where t and R are pipe nominal wall thickness and mean radius, respectively. The 0.2t/R check is recomended in NUREG 1061, Volume 2 (7) as a simple and conservative means of preventing compressive wrinkling failure in straight pipe.

(2) Loy_Cyrle Fatiguti To avoid low-cycle fatigue, the elastically-calculated stress 0.75tH/Z due to SSE loading will be limited by a simplified fatigue check for stainless steel pipe exceeding one percent strain. This fatigue check is based on Harkl's correlations on moment-loading fatigue test (9).

The allowable usage factor for this fatigue evaluation (U,) will be assessed on a case-by-case basis, depending on the operating conditions of the particular piping system under review. The value of Va 15 highly dependent on the cumulative usage factor from other cyclic loading, (i.e., thermal expansion cycling in conjunction with thermal transient cycling and other dynamic load cycling). As a maximum, it will be limited to 1/3. as stated in Volume 1. Subsection 3.1.1.3.

(3) Eg gssive Deformation: NUREG 1061, Volume 2 (7) recommends the limit of fifteen percent reduction in pipe cross-section flow area to avoid excessive pipe deformation. The Greenstreet tests (5) illustrated that the maximum test strain level, as limited by the test apparatus (less than I percent), would produce a maximum ovality of 9 to 15 percent. A 15 percent ovality corresponds to less than 1 percent of flow area reduction. These tests indicate that the flow area reduction at the strain levels presented in this document would be much less than the 15 percent flow area reduction limit. Therefore, the excessive deformation check is readily satisfied.

(4) Mcuracy of Elastic Pining Reggonse Analysis: Elastic piping response analysis is adequate when the calculated stress exceeds the Code limits because:

(A) Piping coeponent and system tests, as summarized in (11) (piping component tests by Imazu, Greenstreet, Teidogucht ant.i oiping system tests by EPRI/ANCO) show that elastically correlated piping stresses to test results were significantly higher than ASME B & PV Code allowables (typically 2-4 times level D) and ther was no indication of phe failure or gross structural instability. Recent testing by EPRI/h.'C Research Project 1543-15. "Piping and Fitting Dynamic Reliability Program" (10), indicates that the loading-to-failure for .

piping components is much greater than the code allowable values ut least fifteen times Code Level D stress allowables). For the stress-strain correlation approach, the exceedance from the ASME B &

PV Code level D limit of 2.4Sh is much less than these values.

Further, the exceedance frore. the ASME B & PV Code Level D limit is for limited piping systems and at isolated locations. Therefore, it can be safely concluded that, at the stress levels corresponding to the strain limits as presented in this document, there should only be local piping yielding. The piping system as a whole remains essentially elastic.

(B) Elastic piping response analysis will predict conservative piping boundary loads, even when the elastically-calculated stresses exceed the ASME B & PV Code limit. Campbell, et al. (6), recomended that the piping stresses be factored down to reduce the excessive conservatism in the piping design practice, if the elastic pt;1ng response analysis by current method is performed. The boundary loads are not to be factored. This approach is equivalent to increasing the stress limit while using the same .

elastically-calculated boundary loads in the qualification of supports, equipment, valves, penetrations as recomended in this report.

(C) Piping displacements would not be expected to be a concern under seismic loading because the con, rolling factor is usually thermal .,

loading. Further, at the stress levels allowed by the stress-strain correlation method large displacements are prohibited by the limits '

placed on excessive deformation of piping components.

l l

l

To address concerns.that pipe reactions and displacements may be greater than t%se estimated by the linear elastic analysis (13), the elastically calculated pipe reactions and displacements are increased by a maximum factor of 2.7, for stathless steel pipe exceeding one percent strain.

This factor (the inelastic response factor X from Volume ) Subsection 3.1.1.3) is based on the following two concepts :

(1) Increase the displacements by a factor of 4.0. This is the ratio of displacements between a fixed-fixed beam with either a hinge in the middle or hinges at both ends and the same beam without a hinge for mid-point static loading.

(ii) Reduce the displacements by a factor of 1.5. This is the ratio of the dynamic margin against failure to the static margin against failure as discussed in (6).

(5) P_lattir_lensile Ins.tability: To avoid plastic tensile instability, only ductile pipe materials will be considered for the stress-strain I

correlation application. Ductile materials can withstand tensile strain l much higher than 2 percent. At the tensile (or ultimate) strength, mild l

carbon steel (0.25 carbon) exhibits approximately 20 percent strain and I

stainless steel (Type 304) exhibits approximately 60 percent strain (11). The maximum design tensile strain limits recommended in (7) are on i

the order of 2 percent to 5 percent. Therefore, the plastic tensile instability check is readily satisfied.

2.2 Yield Strength Increase In applying the ASME B & PV Code requirements for each service level, a departure from the code will be taken on a case-by-case basis to increase the material yield strength (S ) yspecified in the A.SHE B & PV Code.

The material yield strength specified in the ASME B & PV Code is the minimum yield strength. It may be increased by 20 percent for Level B, C and D conditions to accour.t for material overstrength and strain rate effects. This increase factor is listed in Volume 1. Subsection 3.1.3.2. It should be noted that the increase is to account for inherent conservatism in the code-specified yis1d strength only and it is not a redefinition of the ASME B

& PV Code allowables. As shown below, the increase over code minimum could be as high as 30 percent (1.18 x 1.10), although :redit would be taken only 20 percent.

The yield strength increase (30 percent) was reviewed by the NRC and approved for its use in the San Onofre Nuclear Generating Station - Unit 1 Long Term Service Program (4).

A. Yield S*ungth Increase to Account _for the Material OverstreA2th Reference (14] documents the steel strength test data for A36 structural steel commonly used for all major structures at the V.C. Summer Nuclear Station, Unit 1. From these test data, the average yield strength of A36 structural steel is demonstrated to be 21 percent higher than the minimum code-specified yield strength with a low coefficient of variation on the samples (0.09).

l In addition, the report by Smith, et al. (15), states that the measured yield strength of over 60,000 specimens of mild steel was found to be, on the average, 18 percent greater than the ASHE B & PV Code-specified yield strength.

B. 1111d Strenoth la ngtse to Actnunt for the Strain Rate Effetti The strain response of ductile materials depends upon the loading rate.

Review of the literature shows that extensive data on the influence of rate of strain on the yield strength of mild steel are available (16] (17] (18). The yield strength versus strain rate curves from these references show that there is a significant increase in yield strength with increased strain rates.

For pipe supports and supporting structural steel members, the seismic loading is a result of piping excitation. Piping fundamental frequencies are usually in the range of 2-10 Hz. A typical pipa support reaction time history is shown in Figure 2.2-1. This shows the predominant response at about 6 Hz and a higher node response superteposed at about 12 Hz. This figure demonstrates that the load rate (proportional to strain rate) is significant even at high load levels. A sarple calculation is given below for this support's reaction time history to estimate the strain rate.

Minimum frequency of 'osd = f j- 6 Hz Young's Modulus E - 30 x 103 g3g Haximum stress in critical re-ber = a - 30 ksi (assumed near yield)

Time required for the stress in member to increase is from

- t 0 to a . T, is E

I T - 1 - 0.042 sec.

4f)

Therefore, the strain rate is 0 ksi - 2.4 x id" in/in/sec.

3 T ET 30 x 10 x 0.042 Figure 2.2-2 shows the effect of strain rate on ytsid strength of mild steel (163. The strain rate obtained for the sample calculation is significant. At this strain rate, the increase in yield strength over the static yield '

strength is about 16 percent.

This analysis demonstrates that for typical piping system 'requencies, the increase in yield strength as a result of strain rate ;;f ets is granter than i 10 percent, f

W.

1

. 'q-

\ '

h e

4 Mmm. .- N . h umt' -

2.3 Manufacturer's Allowable Loads In cases where manufacturers have recommended allowable loads for Normal, Upset, Emergency and Faulted conditions, these allowable loads will be used. However, many components are provided with allowable loads that are applicable only to normal working conditions. For these components, allowable loads will be developed for Upset, Emergency and Faulted conditions based on allowable load increase factors which are given in Volume 1, Subsections 3.1.3.1, 3.2.3.3 and 3.3.3.1. These increase factors are based on the allowable stress increase factors permitted by the ASME B & PV Code for the corresponding service levels.

A cinimum factor of safety of 1.5 is required, based on the maximum reduction from test ultimate load (0.67) specified by ASHE B & PV Code Appen.11x F-1370.

2.4 Static Load Rating In cases where tests are performed to determine the capacity of catalog components, the factor of safety shall be used to determine the component's allowable load for each service level. These factors of safety are provided in Volume 1, Subsections 3.1.3.1, 3.2.3.3, 3.3.1.2, 3.3.2.2 and 3.3.3.3.

The factors of safety for determining allowable loads based on static tests are based on the ASME B & PV Code load rating procedure for linear type supports, given in ASME B & PV Code NF-3262 (for Normal, Upset and Emergency conditions) and Appendix F (for Faulted condition). The ASME B & PV Code factors of safety for Normal. Upset and Emergency conditions are based on the ratio of allowable stress to minimum tensile stress. This approach was simplified by the following two assumptions: (1) the allowable stress for Normal and Upset conditions is 60 percent of the minimum yield stress, (2) the c.inimum tensile stress is 1.7 times the minimum yield stress. The resulting factor of safety, 2.83 (for Normal and Upset conditions), was rounded up to 3.0. The factor of safety for the Emergency condition is 3.0 divided by 1.33, the ASHE B & PV Code allowable str6ss increase for the Emergene.y condition, or 2.25. The factor of safety for the Faulted condition,1.5, is equivalent to the maximum reduction from the test ultimate load specified by ASME B & PV Code Appendix F-1370 (0.67). Note that this method gives ratios of allowable loads for Normal, Upset, Emergency and Faulted conditions which are compatible with those given by the allowable stress increase factors allowed by the ASME B & PV Code.

The procedure used to establish the test ultimate load is based on both the ASME B & PV Code and the AISI Manual. The recommendation that the test ultimate load be based on the average of three tests, provided no individual test deviates from the mean by more than i 10 percent, is consistent with AISI F1(a) (19). The 10 percent reduction for less than three tests is based on the ASME B & PV Code NF-3261 specification for load ratings based on a single test.

i l

l i

2.5 Randon Vibration for Combination of Modes and Levels Long piping systems are generally connected to different elevations or to areas of the structure which have different dynamic behavior. Therefore the l dynamic input (e.g., floor response spectra) will be different at different 1 piping supports. In current practice, when the effect of different ins t motions is considered, the responses due to the motions at each support (level) are combined by the absolute sumation or the Squars-Root-of-the-Sum of-Square (SRSS). The use of these combination methods does not consider the phasing that could exist between the support cotions and thus it is either too conservative or can not be justified. The accuracy of the total response obtained by these methods can not be assessed in general.

To accurately calculate the total dynamic response for the case of piping subjected to multiple level inputs, it is necessary to consider the phasing between the motions at the piping supports and also the phasing between the modal responses. A method based on random vibrations principles (20) [21]

that properly accounts for phasing will be used, as stated in Volume 1, Subsection 4.2.1.1. The modal-level combination rule used in this method can be written as:

R d

- ( Rh R P g )l/2 d

R - Maximum dynamic response

- Maximum dynamic response of mode i due to the Rfk excitation applied at support level k d

R - Maximum dynamic response of mode j due to the 3j excitation applied at support level 1 P Correlation coefficient for modes i and j and ikjl - sJpport levels k and 1 The correlation coefficients P ijkl vary between -1 and 1. Their value is 1 for modes responding in phase and for support input motions also in chase.

Their value is -1 for modes responding out of phase but supports moving in phase or vice versa. If the modes are uncorrelated (e.g., modes with very different frequencies) and the support motions are also uncorrelated then the cor.* elation coefficient will be zero and the formula reduces to SRSS combination of modes and levels.

This methodology gives results which are consistent with time history analysis results, being less conservative and more reliable than the multiple level input and the envelope input methods commonly use in practice.

A similar formula is applied to combine the pseudostatic responses due to Seismic Anchor Motion (SAM) [21).

i 3

2.6 PVRC Damping The modal damping values provided by Regulatory Guide 1.61 [22] are used for response spectra, equivalent static coefficient or time history dynamic analyses. As a less conservative and more realistic alternative, damping values provided by ASME B & PV Code Case N-411-1 [23] [24] (PVRC Damping) are acceptable for use on all response spectra and equivalent static coefficient analyses of piping systems, as stated in Volume 1. Subsections 4.2.1 and 4.2.2.

The use of two different sets of damping values is restricted by:

(1) Analysis of any specific piping :ystem shall consistently use one or the other of the two different sets of damping values.

(2) Damping values provided by ASME B & PV Code Case N-411 are not acceptable for use on piping systems which use energy absorbing supports (covered by ASME B & PV Code Case N-420 (25]) or those in which stress corrosion cracking has been identified.

Justification for the use of PVRC Damping '. provided by two sources: (1) review of the restrictions imposed by the Regulatory Guide 1.84, Revision 24

[26), and (2) NUREG 1061 [7] [8] recommendations.

2.6.1 Regulatory Guide 1.84 Restrictions Regulatory Guide 1.84, Revision 24, provides five restrictions on the use of ASHE B & PV Code Case N-411. These are restated below with a discussion on how each of these restrictions is met for the piping systems at Fort Calhoun:

(1) Damping values are limited te analyzing piping response for seismic and other dynamic loads filtered through building structures, provided response mode frequenties are limited to 33 Hz and below. Use of the damping values is to be complete and consistent.

Resoonse As stated in Volume 1 of this report, the methodology is limited to dynamic seismic analysis. The methodology requires that tne ASME B & PV Code Case N-411 damping values be applied consistently for the analysis of any specific piping system.

(2) Damping values specified may be used only in those analyses in which current seismic spectra and procedures have been employed.

Resoonse The damping values are used in response spectra or equivalent static coefficient analyses for spectra developed in accordance with the methodology specified in Volume 1, Section 4.1, and analysis techniques employed in Volume 1, Section 4.2. Both spectra development and use are based on current industry practice.

(3) When used for reconciliation work or for support optimization of existing designs, the effects of increased motion on existing clearances and on line mounted equipment should be checked.

Resoonse It is current industry practice to review effects of increased motion for reanalysis of existing piping systems. ,

(4) This ASME B & PV Code case is not appropriate for analyzing the dynamic response of piping systems using supports designed to dissipate energy by yielding (i.e., ASME B & PV Code Case N-420).

Resoonse The use of ASHE B & PV Code Case N-420 with ASME 8 & PV Code Case N-411 is specifically excluded.

(5) This ASME B & PV Code Case is not applicable to piping in which stress corrosion cracking has occurred unless a case-specific evaleation is made and is reviewed by the NRC staff.

Resconse The use of this Code Case for piping identified to have stress corrosion cracking is specifically excluded.

2.6.2 NUREG 1061 Recomunendations NUREG 1061, Volume 2, Subsection 2.2 [7], reviews the PVRC recomended interim position on damping values against the values of Regulatory Guide 1.61. In NUREG 1061 Volume 5, Subsection 5.2.2 [8], the Piping Review Committee sumarizes its recomendations by stating "The Committee believes that the ASME/PVRC recommendations are more realistic and should be implemented immediately."

2.7 CQC Method for Combination of Modes Three dimensional piping systems generally possess natural frequencies which are closely spaced. This phenomenon means that the responses of the corresponding modes will tend to be in-phase. Thus, to accurately calculate the total response, it is necessary to use a modal combination rule which is able to consider the phasing between modal responses. The modal combination rules described in Regulatory Guide 1.92 [29) either do not account for phasing between modal responses or do so in a simplified way (Grouping, fen Percent, Double Sum) which does not always give accurate results and in general tends to overestimate the total response.

A modal combination method which properly accounts for the phasing of the modal responses is the Complete Quadratic Combination (CQC) method. This method will be used as stated in Volume 1, Subsection 4.2.1.1. References (30) [31) [32] give a complete description of the method.

The CQC rule can be expressed as follows:

R -( I E Rj Pjj Rj)1/2 ij where R - Combined response Pjj - Correlation coefficient between mcdes i and j Rj, Rj - Modal responses for modes i and j (including direction sign for each mode) 1 The coefficients Pjj vary between zero and one. They will be one for modes which respond in phase and they will be zero for uncorrelated modes (e.g.,

modes with very different frequencies).

This modal combination rule has been shown to give accurate and less conservative results than those calculated with the combination rules in R.G.

1.92. The CQC method was reviewed by the NRC and approved for its use in the San Onofre Nuclear Generating Station - Unit 1 Long Term Service Program [4).

2.8 Support Ductility In cases where the elastically calculated tension and bending stresses of structural steel members exceed the stress allowables for the Faulted condition, a maximum ductility of 3 (u 5.3) will be used on a case-by-case basis to evaluate the inelastic behavior of the steel structures to maintain their design function (Volume 1. Subsection 3.1.3.2). Numerous observations of the actual performance of steel structures subjected to seismic motions have shown that structures are capable of absorbing and dissipating a considerr amount of energy when strained in inelastic response beyond their elastic imit. On the other hand, an earthquake is capable of inputting only a limited amount of energy into a structure. Unless corrected for inelastic response capability, a linear elastic response analysis is incapable of acctunting for the inelastic energy absorption capacity of the structure.

In order to account for inelastic energy absorption capacity, NUREG-0098 (34),

Table 3, recommends a design ductility range of 2 to 3 for structures or components that can deform inelastically to a moderate extent without unacceptable loss of function. This is applicable to piping lincar type supports since A-36 structural steel commonly used for supports is a ductile material. Reference (343 further states that the upper limit of the ductility range would be adequately conservative.

The use of a maximum ductility of 3 was reviewed by the NRC and approved for its use in the San Onofre Nuclear Generating Station - Unit 1 Long Term Service Program (4).

STRESS

/: ELASTICALLY ASSUMED STRESS-STRAIN CURVE a, - . -

1 I ACTUAL l ST R E S S-ST R AIN I CURVE g, -.

3 -

I l

I l

\ SIMPLIFIED l l STRESS-STRAIN i l CURVE I I l l  :

I I I I l l  ;

I I

(, si STRAIN 0, = ELASTICALLY CALCULATED STRESS

(, EL ASTIC ALLY CALCULATED STR AIN

=

(, e TOTAL OR ACTUAL STRAIN o, = ACTUAL STRESS

' Figure 2.1-1 Elastic Strain vs. ".ctual Strain z

9 C w

ct H

2 w

B z A o C -

o I

z E l I

5 e I i

1 EL ASTIC ALLY-C ALCUL ATED STRESS, Sn 3Sm

a. Actual Case l

SLOPE l-n 3 ,

3.3 n(m-l) 1 Ke l 1 , I l

l I/n l l l I

I I l l I

1 a i l 1 1.7 2 3 EL ASTIC ALLY-C ALCUL ATED STRESS, Sn i 3Sm i

b. Code Idealization for Stainless Steel Figure 2.1-2 ASME B & PV Code Strain vs. Stress Relation P

4.0 2

(--------.

E I J y 3.0 j ,

I . /

O j

! /

W '

E B& PV SEC 1119 l f

< < p I

/

E a ANSI 2.0_

J

< 1 /' B 31.7

1. * (1969)

$ 0 j f

>- /

o >-

a i s

< a l /

/

< p o

~

fn

1. 0 ----- M

.J w

0.0 0.0 1.0 2.0 3.0 EL ASTIC ALLY-C ALCUL ATED STRE SS, sn 3Sm Figure 2.1-3 Code vs. Test Stress-Strain Relation l

l l

i EXTRADOS 00 i

,a

--~,

/

/ \

./ g I ~ \

270  ! ~

j 90

\ j

\ /

\ /

/

\ /

%s%* . ,,,,,,,,,"s STR AIN G AGif A+2 ELBOW SH APE (OV ALITY) ]

UNDER IN-PL ANE BENDING FOR PE-19 180 INTR AD O S Figure 2.1-4 Elbow Crotch Section Under In-Plane Bending

, Em , ' b'  ;

O.S. =g 7

il MAXIMUM AXIAL STRAIN if AT O' OR 180 :f

?!

I

I

I l.S. I

'D

f rCL  ; (b  ;

0.S. /j f/

//

MAXIMUM CIRCUMFERENTIAL [7 STR AIN AT 90 OR 270 l

d '/

/ l.S.

=  :

'm (b  ;

2

'D .  : ,

LEGEND:

(m - MEMBR ANE STR AlH (b - BENDING STRAIN (p - LOCAL OR PEAK STR AIN O.S.- OUTSIDE SURF ACE l.S.- INSIDE SURF ACE Figure 2.1-5 Maximum Strain Distribution Through the Hall Thickness Under In-Plane Bending 5.. J '

l' g i

s rI l I h I\ 1 l l ,li

~

'  ?"

\ ) \

)

<t

(\ \; )

... c. ... ,,,.g,,,,

' Figure 2.2.1 - Typical Support Reaction Time History 100 90 8'

E x 70 60 p

tn /

N /

b 40 "**

30 ~

I 20 l

10 --

I Cro gos iga gga gge ,ge  ; to gor go s

! AVERAGE RATE OF STRA!N PER. SEC.

l Figure 2.2.2 - Relationship Between Material Yield and Strain Rate for Mild Steel (16)

L

i 3.0 ELECTRICAL RACEMAYS The criteria for evaluation of electrical raceways are primarily based on application of the two relevant design codes: for hot rolled steel, the ASME B & PV Code [1]; for cold formed steel, the AISI Hanual (19). This section provides justification for the cases in which additional or relaxed criteria have beers used to supplement these standards.

3.1 Allowable Spans for Conduits and Cable Trays Tne criteria for establishing allowable spans for conduits and cable trays which are given in Volume 1 Subsections 3.2.1 and 3.2.2, are intended to verify that normal industry standards and manufacturers' recommendations are employed for the design of the conduits and cable trays themselves. The design of raceway systems using normal industry standards and manufacturer's recommendations does not require rigorous evaluation of seismic loads. This approach is justified by industry research which shows that the supports and support anchoracas, not the conduit and cable tray units, are the critical components for racevay seismic performance.

The seismic performance of electrical raceway systems has been extensively investigated [35). In the course of this investigation, inspections were made of a wide variety of electrical raceway systems installed in industrial and power facilities which have experienced earthquakes with intensities of ground motion up to 0.6g. Most of the systems which were inspected were designed only to support dead loads; some were strictly field-run, and others were severely overloaded. In general, raceway systems were found to perform well under seismic loading. Several hundred raceway systems were investigated at 23 facilities. Of these, there was no recorded failure of any of the cable tray or conduit spans themselves from inertia loads. The single instance of conduit span failure was sttributed to large anchorage movements caused by ground failure and building settlement. In addition, there was no recorded loss of function of any raceway system, even those which failed structurally.

The report [35) concludes:

"The integrity of the primary (overhead) vertical support appears to be the major requirement to ensure structural integrity and the available test and experience data show that design for gravity loads is adequate to insure structural integrity for inertia loads from earthquakes larger than typical SSEs."

Based on this investigation, the alternate criteria require only conformance to normal industrial standards for the conduits and cable tray spans. This requirement, in addition to the rigorous seismic evaluation of conduit and cable tray supports, is sufficient to ensure the safety of electrical raceway systems.

3.2 Cyclic Load Rating The cyclic load rating procedure given in Volume 1, Subsection 3.2.3.3, uses a cumulative fatigue damage criterion which is based on IEEE 344-1987 [36.1 It requires consideration of the cumulative fatigue damage which would be caused by 5 OBE events followed by one SSE event. Support components are qualified if this cumulative fatigue usage is less than unity.

The procedure for establishing a safe-life fatigue curve from test data is based on the ASME B & PV Code Case N-420 [25] procedure for the development of design fatigue curves for linear energy absorbing supports. The safe-life fatigue curve will be established by a regression analysis of low-cycle fatigue test results which follows ASTM E/39-80 (37), with the factors of safety recommended by ASME B & PV Code Case N-420. This procedure is equivalent to the ASME B & PV Code Case N-420 with the following two assumptions: a) for low-cycle fatigue, the logarithm of cycles to failure is linearly related to the logarithm of the applied load amplitude; b) the fatigue life is Icg-normally distributed with constant variance in the range of interest.

I

3.3 Damping Damping values used for analysis of raceways systems are given in Volume 1, Subsection 4.3.1. The damping values given for analysis of conduit systems correspond to those specified in R.G.1.61 (22) for analysis of bolted steel structures. In addition, these daniping values are the same as those used for conduits at many other nuclear plants.

The damping values for cable trays are based on results from two extensive full scale shake table test programs (38) (39). The key results from these programs, which covered a wide variety of support and system configurations, are as follows:

(1) Cable tray seismic response exhibited significant nonlinearities due to movement of cables within trays. This effect may be accounted for, in linear analyses, by the use of effective damping values which are much larger than those typically used for analysis of other similar structures.

(2) Incrtased input motion intensity caused increased damping due to cable movement.

(3) The e'fective damping values displayed in the tests ranged from 5 to 50 percet!t.

Figure 3.3-1 shows results for laterally restrained and unrestrained rod-hung cable trays with three different cable fill levels from (38).

The results corresponding to the highest intensity input motions (which were still generally much less than the seismic capacity of the systems) demonstrate effective damping levels consistently above 15%. Other cable tray support cunfigurations exhibited somewhat higher levels of damping.

Therefore, the use of 15% damping for the SSE is justified.

i W

j a - 50 lb/ft CABLE FILL, RESTRAINED A - 25 lb/ft CABLE FILL, RESTRAINED e - 10 lb/ft CABLE FILL, RESTRAINED o - 50 lb'It CABLE FILL, UNRESTRAINED A - 25 lb'ft CABLE FILL, UNRESTRAINED l:  ;

A 4

.30 a "

A ea ,e -

me se N a m

.20 d af e ,

a 2

b C a "a

.10 l

m e e o*"

O E

o O o.1 0.2 0.3 0.4 0.6 0.6 0.7 0.8 0.9 1o AVERACE PEAK RESPONSE ACCELERATION (g)

Figure 3.3-1 Effective System Damping as a Function of Average Peak Response (38) 4.0 HVAC 4.1 Qualification by Test As given in Volume 1, Subsections 3.3.1, 3.3.2 and 3.3.3, the qualification by test of HVAC ducts, duct supports or any miscellaneous hardware is acceptable. Both the AISI Manual [19), and ASME B & PV Code, Subsection NF

[1], specify testing as an acceptable method for the determining component load capacities.

The justification for the factors of safety applied to static test results indicated can be found in Section 2.4 of this volume.

The industry has generated a large volume of test data for the seismic qualification of numerous components. The NRC in NUREG-1030 [40) has addressed the general methodology of correlating existing seismic qualification tests with equipment currently installed in operating nuclear power plants. The report concludes that to be able to use existing test data, the level of excitation at which functionality must still exist must be established.

Test data are available in the industry for seismic qualification of ducts, duct supports, and much of the miscellaneous hardware. Test have been performed to obtain duct, duct support and hardware capacities to generalized force and moment loading as well as full scale shake table tests of duct / support systems loaded by synthetic time histories for plant specific seismic qualification.

The attributes addressed in Subsection 3.3.1.2 will ensure that the application being considered has an equal or greator capacity than the tested component. In addition, the loading considered far the test specimen shall be compared with the loading anticipated for the cc.nponent to be qualified. this will result in a conservative evaluation which meets the intent of h0 REG-1030.

4.2 Qualification by Experience Data The approach of using experience data on the performance of equipment found in industrial and conventional power plants that have been subjected to actual earthquakes of large magnitude is currently practiced in the industry (40),

[41). As given in Volume 1, Subsection 3.3.1, 3.3.2 and 3.3.3 the same approach is valid for the qualification of HVAC components where applicable experience data are available.

In general, the scismic capacity of equipment installed ',n nuclear power plants meets, and in most cases exceeds, the capacity of equipment installed in conventional power plants and industrial facilities. This is also true for HVAC installations. Many of these conventional facilities have experienced strong motion earthquakes without experiencing damage. Therefore, this information can potentially be used as an alternative to a rigorous seismic qualification. This general approach has been endorsed by the NRC in NOREG-1030 [40), and NUREG-1211 [41).

In order to meet the intent of NUREG-1030, the seismic ruggedness of the HVAC component must be compared with that for which experience data are available.  ;

The attributes addressed in Volume 1. Subsection 3.3.1.2, ensure that the l application being considered has an equal or greater ruggedness than the experience data base installation. In addition. the seismic motion intensity from the experience data installation must envelop the design motion intensity for Fort Calhoun.

I l

l t l

4.3 Damping As given in Volume 1, Subsection 4.4.1, the damping factors to be used for the seismic analysis of cold-formed duct / support systems are 4% and 7% for OBE and SSE respectively. These values are in accordance with Regulatory Guide 1.61

[22) for bolted steel structures. Most HVAC lateral connections are either bolted or friction joints, and support-to-duct connections are either bolted or friction. Since the friction type joints will result in even higher damping, the values from Regulatory Guide 1.61 are conservative. Applicable test data for dynamic test of full scale duct / support systems (42] have shown that typical damping values of 7% and 10% for companion angle (bolted) and pocket lock (friction) type construction respectively. The test results are applicable to installations at Fort Calhoun. The lower range (bolted) damping factor will be used for all SSE evaluations. A value of 4% will be conservatively used for OBE evaluations.

5.0 EXPANSION ANCHOR BOLTS The criteria for evaluation of concrete expansion anchor bolts are based on current industry research and practice, as well as on the results of special studies undertaken to address specific expansion anchor bolt issues at Fort Calhoun.

5.1 Factors of Safety i The basic fact (ts of safety which will be used for expansion anchor bolt evaluations, 4 for non-shell bolts and 5 for shell bolts, as given in Volume 1, Subsection 3.4.2, are consistent both with manufacturers' recommendations for working loads (for example, see (43]) and with the factors of safety

?,pecified in IEB 79-02 (44) for safety-related piping subject to SSE loads.

On a case-by-case basis, reduced factors of safety of 2.5 for non-shell anchors and 4 for shell anchors will be used. These reduced factors of safety are based on recent industry research which has shown that "a factor of safety louer than that traditionally used in design could be allowed if proper installation could be verified ... either by demonstrating that proper quality assurance and control procedures were followed during the original ccnstruction or by conducting an in-situ inspection of expansion bolts" [33).

Although factors of safety as low as 2.0 for non-shell enchors is justifiable, a factor of safety of 2.5 is recommended to account for realistic (median centered) in-structure spectra generation approach [27). The additional restrictions and inspection criteria listed in Volume 1, Subsection 3.4.2, for the use of the reduced factors of safety are intended to mest this requirement for verification of proper installation.

i c

l r

5.2 Tension / Shear Interaction As given in Volume 1, Subsection 3.4.3, combined tensile and shear loads will be evaluated by a two-part interaction equation. This interaction relationship is recommended by EPRI (33] as providing the best fit to a large number of tension / shear tests.

1 i

l

)

l

5.3 Edge Distance and Spacing Requirements The edge distance requirements given in Volume 1, Subsection 3.4.4, allow the full anchor bolt tensile capacity to be used for bolts further than five bolt diameters from a free edge, and full shear capacity for bolts further than ten bolt diameters from a free edge. Between five and ten bolt diameters, an additional limitation is placed on the shear capacity. Most concrete expansion anchor bolt manufacturers specify that the full anchor capacity may be used at an edge distance of five bolt diameters. However, recent tests have indicated that this requirement may not be sufficiently conservative for evaluation of shear loads. The additional formula uses the projected area of a half cone on the concrete edge surface to approximate the effective tensile capacity of the concrete resisting the shear load, based on the approach of (12).

Spacing requirements given in Volume 1, Subsection 3.4.5, conform to both menufacturers' recommendations and to the EPRI recommendations given in (33).

For some multiple-anchor configurations, these requirements may be overly ccnservative (e.g., those with deeply embedded wedge-type anhcor bolts for which the ultimate tensile capacity is limited by wedge slip failure, not concrete core failure). Alternative spacing requirements may be developed in these cases.

l 1

b d

5.4 Concrete Topping

Additional limitations and inspection requirements, as given in Volume 1, Subsection 3.4.6, are applied to anchor bolts installed in concrete topping.

These requirements were developed from a study of anchor bolt installations and tests at Fort Calhoun (28).

i l

O i

I j

i i

1

)

)

6.0 REFERENCES

[1] ASME Boiler and Pressure Vessel Code,Section III, 1980 Edition with Addenda thror,h Sumer 1981.

[2] SCE Document "SONGS-1 LTS Seismic Reevaluation Program, Technical basis for Stress-Strain Correlation", prepared for the NRC, dated 1 January, 1986.

(3) SCE Document "Supplement to the Topical Report Entitled SONGS-1 LTS Seismic Reevaluation Program Technical Basis for Stress-Strain Correlation", prepared for the NRC, dated March 7, 1986.

[4] NRC's Safety Evaluation Report, "Safety Evaluation by the Office of Nuclear Reactor Regulation Relating to the Long-Term Service Seismic Reevaluation Program, Southern California Edison Company, San Diego Gas and Electric Company, San Onofre Nuclear Generating Station, Unit No. 1. Docket No. 50-206, "provided by NRC letter to Kenneth P. Baskin (SCE) from Thomas M. Novak (NRR), dhted July 11, 1986.

[5] Greenstreet, W.L., "Experimental Study of Plastic Response of Pipe Elbows," ORNL/NUREG 24, February 1978.

[6] Campbell, R. D., Kennedy, R. P., and Thrasher, R. D., "Development of Dynamic Stress Criteria for Design of Nuclear Pipt'ig System,"

SMA 17401.01, prepared for PVRC by Structural Mechanics Associates.

Inc., March, 1983.

[7] NUREG 1061, Volumes 2 "Report of the U.S. Nuclear Regulatory Commission Piping Review Committee Evaluation of Seismic Designs -

A Review of Seismic Design Requirements for Nuclaar Power Piping",

l April 1985.

[8] NUREG-1061 Volume 5 "Report of the U.S. Nuclear Regulatory Commission Piping Review Committee: Evaluation of Seismic Designs, Summary-Piping Review Committee Conclusions and Recomendations,"

April 1985.

[9] Mark 1, A.R.C., "Fatigue Tests of Piping Components", Trans. ASME, Vol . 74, pp. 287-303,1952.

[10] EPRI/NRC Research Project 1543-15. "Piping and Fitting Dynamic Reliability Program". Interim Results Presented at the EPRI Workshop on Piping Intengrity, dated October 7 and 8, 1986.

1 (11) SCE Document "SONGS-1 LTS Seismic Reevaluation Program, Technical Basis for Piping Strain Limits and Development of Linear, Elastic Analysis Methodology," Prepared for the NRC dated May 31, 1985.

[12) ACI Standard 349-80, "Code Requirements for Nuclear Safety Related Structures."

(13] E.C. Rodabough Associates Letter from E.C. Rodabough to Dr. Long Shieh (Lawrence Livermore National Laboratory), "San Onofre Unit 1  !

Review of Licensee's Strain Based Me$hodology for Seismic Evaluation of Large Bore Pipe," dated February 13, 1986.

l (14] Applicant's Additional Testimony to the ASLB, South Carolina Electric and Gas Company, Docket No. 50/395 OL, December 18, 1981. t (15] Smith, P.O., Maslenikov, 0.R., and Bumpus, S.E., "LLL/ DOR Seismic Conservatism Program: Investigations in the Seismic Design of Nuclear Power Plants," UCRL-52716.

(16) Manjoine, M.J., "Influence of Rate of Strain and Temperature on Yield Stresses of Hild Steel," J. of Applied Mechanics, Volume 11.

ASME Trans., Volume 66, pages A211-A218, 1944.

[17] Timoshenko, S., "Strength of Materials, Part II," Third Edition, Van Nostrand, Princeton, New Jersey, 1956. l (18] Beedle, L.S., and Tall, L., "Basic Column Strength", Proceedings of ASCE, Volume 86 (ST-7), July 1980. [

(19) American Iron and Steel Institute (AISI), "Cold-Formed Steel Design Manual," 1986 Edition. '

i (20) Asfura, A., and Der Kiureghian, A., "Floor Response Spectrum Method I

for Seismic Analysis of Multiply Supported Secondary Systems,"

Earthquake Engineering and Structural Dynamics, Vol. 14, 245-265, 1986. r i  !

(21] Asfura, A. "A New Combination Rule for Seismic Analysis of Piping r Systems", Proceeding of the 1985 Pressure Vessels and P1;ing Conf t.rence, PVP-Vol . 98-3, New Orleans, Louisiana, June 1985.

[22] U.S. NRC Regulatory Guide 1.61, "Damping Values for Seismic Design of Nuclear Power Plants," dated October 1973. ,

(23] ASME ASHE B & PV Code Case N-411-1, "Alternative Damping Values for [

Response Spectra Analysis of Class 1. 2, 3 Piping Section III, Division 1," approved February 20, 1986.

l l

(24] MRC Bulletin #300, "Technical Position on Damping Values for l l

Piping-Interim Summary Report," December 1984, (25] ASHE ASME B & PV Code Case N-420. "Linear Energy Absorbing Supports l for Subsection NF, Class 1, 2 and 3 Construction,Section III, '

Division 1," approved February 14, 1985.

(26) U.S. NRC Regulatory Guide 1.84, "Design and Fabrication Code Case i i Acceptability, ASHE Section IV, Division 1", revision 24, June 1986.

I  !

4

)

l

• r _-. -. .__ __ _-__ _ ,. . - , m ,__. - ,_ , - .-,. ____,_

[27] Seismic Qualification Utilities Group SSRAP Raport, "Use of Seismic Experience Data to Show Ruggedness of Equipment in Nuclear Power Plants". Draft June 29, 1987.

[28] Impe11 Report No. 01-1390-1648, "Criteria for Anchors Installed in Concrete Toppings at Fort Calhoun," Revision 0, January 1988.

[29] U.S. Nuclear Regulatory Commission. Regulatory Guide 1.92 Combining Modal Responses and Spat 189 (Taponents in Seismic ,

Response Analysis, dated February 19 1 (30) Der Kiureghian, A., "Structural Ruponse to Stationary Excitatica "

Journal of The Engineering Mech 6nic Division ASCE, Vol. 106, No.

EM6, December 1980, pp. 1195-1213.

[31] M11 son, E.L., Der Kiureghlan A., and Bayo. E.P., 'A Replacement for the SRSS Method in Seismir Analysis," Earthquake Engineering and Structural Dynamics, Vol. 9, 187-194, 1981. ,

[32] Der Kiureghian, A., "A Response Spectrum Method for Random Vibration Analysis of MOF Systems," Earthquake Engineering and Structural Dynamics, Vol. 9, 419-435, 1981.

[33] Electric Power Research Institute Report No. NP-5226, "Seismic Verification of Nuclear Plant Equipment Anchorage," May 1987.

[34] NUREG/CR-0098. "Development of Criteria for Seismic Review of Selected Nuclear Power Plants," N.H. Newmark, N.J. Hall, May 1978.

[35) EQE Incorporated, "The Performance of Cable Tray and Conduit Systems in Actual and Simulated Seismic Motion," Draft B November 1 1986.

[36] Institute of Electrical and Electronics Engineers (IEEE), Standard 344-1987, "IEEE Recommended Practices for Seismic Qualification of Class 1E Equipment for Nuclear Power Generating Stations."

(37: ASTM E739-80, "Standard Practice for Statistical Analysis of Linear or Linearized Stress-Life (S-N) and Strain-Life (E-N) Fatigue Data."

(38] URS/Blume Report No. 8050, "Analytical Techniques, Models and l Seismic Evaluation of Electrical Raceway Systems", Rev. O, August 26, 1983.

[39) Reimer, G.S., and P.H. Koss, Development of Analysis and Design Techniques from Dynamic Testing of Electrical Raceway Support Systems, Revision 5 Technical Report for Cable Tray and Conduit Raceway Test Program, Bechtel Poner Corporation, San Francisco, California, July 1979.

i (40] NUREG-1030, Report of the U.S. NRC Office of Nuclear Reactor Regulation: "Seismic Qualification of Equipment in Operating Nuclear Power Plants - Unresolved Safety I.: sue A-46", February 1987.

[41] NUREG-1211, Report of th9 U.S. NRC Office of Nuc1 Gor P.eactor Regulation: "Regulatory Analysis for Resolution of Unresolved i Safety Issue A-46, Seismic Qualification of Equiprant in Operating Plants," February 1987.

[42] Report MA2-79-1, "Summary Report for HVAC Duct Seismic Qualification and Verification /Imprevement Program." Tennessee Valley Authority, Civil Engineering Dranch, June 16. 1969.

[43] Hilti Fastening Systems, Inc., "Anchorays and fastener Design Manual," 1985.

[44] USNRC IEB 79-02, "Pipe Support Base Plate Designs Using Concrete Expansion Anchor Bolts," Rev. 2, November 8, 1979.

i i

l i

I