ML20214E521

Assessment of Structural Seismic Design Adequacy of Yankee Nuclear Power Station

ML20214E521
Person / Time
Site:	Yankee Rowe
Issue date:	01/31/1985
From:	Tsai N, Wong W, Yang M NCT ENGINEERING, INC.
To:
Shared Package
ML20214E515	List: ML20214E521
References
NUDOCS 8603260240
Download: ML20214E521 (36)

Text

.__ - __ . . . . _ . . .. _. . . -- . _ . _ _ ._-.._

t

. Enclosure 4 Attachment 2 ASSESSMENT OF STRUCTURAL SEISMIC DESIGN ADEQUACY OF YANKEE NUCLEAR POWER STATION N. C. Tsai M. S. Yang l

W. L. Wong l

J Published January 1985 -

i- . .

NCT Engineering, Inc.

Lafayette, California 94549 Prepared for EG&G Idaho, Inc.

Under Subcontract No. C84-110324

. and the U.S. Department of Energy

Under Contract No. DE-AC07-76ID01570 ii

-- -ca:; <

. c:

~'5- '

8603260240 860314 PDR ADOCK 05000029 P PDR vw mn -tw

o EXECUTIVE

SUMMARY

'This Technical Letter Report presents the findings from our review of the licensee's seismic reevaluation of the structures at the Yankee Nuclear Power Station.

Seven reports were re@iewed. They include the reports for the seismic reevaluation

, criteria, vapor container, reactor support structure, turbine building, diesel generator building, primary auxiliary building, and fire water tank. The evaluations are sufficient except for certain outstanding items that require clarification or additional evaluation.

The licensee properly identified certain current structural deficiencies in the turbine

, building, diesel generator building and primary auxiliary buildings, and proposed the proper strengthening modifications. Our findings are summarized briefly in Table 1.

I

( .

CONTENTS Page No.

EXECUTIVE

SUMMARY

I t

INTRODUCTION 1

-REVIEW CRITERIA 2

SUMMARY

OF REVIEW FINDINGS 3

t Seismic Reevaluation Criteria 3 Vapor Container 4

Reactor Support Structure 6

Turbine Building 10 Diesel Generator Building 12 Primary Auxiliary Building 15 Fire Water Tank 18 1

CONCLUSIONS 21 REFERENCES 23 APPENDIX TABLES

1. Summary of Review Results 1

INTRODUCTION

(

The Yankee Nuclear Power Station was among the eleven older nuclear plants included in the Systematic Evaluation Program (SEP) that was initiated in 1977 by the Nuclear Regulatory Commission (NRC). Seismic design adequacy under the Safe i

Shutdown Earthquake (SSE) condition was one of the safety topics identified in the SEP. NRC subsequently initiated the Integrated Plant Safety Assessment (IPSA) program upon the conclusion of the SEP. It identified, among others, all seismic related issues that were not satisfactorily resolved by the licensees during the SEP. The licensee of

(

Yankee nuclear plant then submitted additional information to NRC to address those outstanding seismic design issues. This information included their selmic reevaluation results on eight structure-related subjects, as listed below:

i (1)

Seismic reevaluation criteria (Reference 1)

(2) Vapor container (Reference 2)

(3) Reactor support structure (Reference 3)

(4) Turbine building (Reference 4)

(5) Diesel generator building (Reference 5)

(6) Primary auxiliary building (Reference 6)

(7)

Fire water tank (Reference 7)

(8) Main steam feedwater piping support structure EG&G Idaho, Inc. is contracted with the NRC to provide technical assistance in reviewing and performing independent assessments of the seismic design adequacy of the Yankee nuclear plant. Specifically, EG&G Idaho will assist NRC in confirming the results of previous SEP evaluations and resolving those outstanding issues identified in the IPSA report.

NCT Engineering provides technical assistance to EG&G Idaho by reviewing 11eensee's reevaluation of the seismic design adequacy of the Yankee plant in the first seven structure-related areas listed previously, i.e., from the reevaluation criteria report to the fire water tank report (References 1 to 7). The main steam feedwater piping support structure report need not be reviewed because, according our judgment, the subject structure is so light compared to the feedwater piping that I it may be regarded as a part of the pipe support system.

The bases and results of our review of the seven remaining structure-related areas are presented in the following, and the conclusions are listed in Table 1.

I

i .

REVIEW CRITERIA t

Since the Yankee nuclear plant was not designed to current codes, standards, and NRC requirements such as the Standard Review Plant (SRP) and applicable Regulatory Guides, we performed more realistic assessments of the seismic capacity

(

of the facility and took into account the ' conservatism associated with the original analysis methods and design criteria. For instance, we typically accepted licensse's use of the as-built material strength instead of code specified design strength in their seismic reanalysis. As another example, depending upon the importance to the safe i

function of the plant, limited and local inelastic responses might be acceptable. In general, our review was based on a set of criteria and guidelines that were developed for the SEP plants. They are described in the following four documents:

(1) NUREG/CR-0098, " Development of Criteria for Seismic Review of Selected Nuclear Power Plants," by N. M. Newmark and W. J. Hall, May,1978.

(2) "SEP Guidelines for Soll-Structure Interaction Review," by SEP Senior Seismic Review Team, December 8,1980.

(3)

U.S. Nuclear Regulatory Commission, " Systematic Evaluation Program Position RE:

Consideration of Inelastic Response Using NRC NUREG/CR-0098 Ductility Factor Approach," June,1982.

(4) U.S.

Nuclear Regulatory Commission, "SEP Topic III-6, Seismic Design Considerations, Staff Guidelines for Seismic Evaluation Criteria for the SEP Group II Plants, Rev.1," September 1982.

For cases that are not speelfically covered by the above criteria, the following SRP sections and Regulatory Guides were used as the basis for our review:

(1) Standard Review Plan, Sections 2.5, 3.7 and 3.8.

(2) Regulatory Guides 1.29, 1.60, 1.61, 1.92, 1.100, and 1.122.

In the event that the licensee's reanalysis deviated from the aforementioned review criteria and guidelines, we reviewed, based on our experience and best engineering 2

=_ -

t .

judgment, the justifications presented by the licensee. In this report we will identify

, such differences and provide our recommendation on their acceptability. This is to recognize that plant specific deviations on a case-by-case basis may be necessary and be found acceptable so long as they reasonably meet the intents of the SEP review guidelines.

(

SUMMARY

OF REVIEW FINDINGS The findings from our review of the seven seismic evaluation reports are separately

g summarized below.

Seismic Reevaluation Criteria I,

The general criteria and basis for the licensee's seismic reevaluation and/or retrofit are depicted in the criteria report (Reference 1). We did not review those portions of the report that are~ related to the seismic evaluation of pipings and pipe supports because they are not within our current scope of review.

The criteria covered the seismic input, soil-structure interaction, damping and energy absorption, loadings and load combinations, performance criteria, etc. They appear adequate except for the following:

(1)

The report is incomplete because it only specified the Site Specific Spectrum (SSS) as seismic input and the associated criteria for analysis and evaluation.

The Yankee Composite Spectrum (YCS) was not specified, neither were the associated analysis and performance criteria. NUREG-0825 (Reference 8) required that the structure be evaluated for both the SSS and YCS spectrum input if the structure exceeded the linearly elastic limit when subject to the SSS spectrum, in which case the structure is required to remain within the elastic limit subject to the YCS spectrum and to remain functional subject to the SSS spectrum. The report did not address this requirement, either.

(2)

The criteria for combining the effects from three earthquake components and for generation of amplified (in-structure) response spectrum were not discussed in the report.

i 3

i .

(3)

De generalized assumption that soil-structure interaction is negligible need not

, be valid for every structure.

The above represent deficiencies of the etiteria report itself, and are not necessarily applicable to the individual structure evaluation reports. Derefore, we will

(

assess, again, the sufficiency and consistency of the criteria that were actually adopted by the licensee in their evaluation of the individual structures when we review the structure evaluation results.

Vapor Container

] Licensee's evaluation of the vapor container (VC) was summarized in Reference

2. The findings from our review are as follows, Loadings and Performance Criteria Both the YCS and SSS spectra were accounted for in the analysis. LOCA loads were also included. The load combinations and performance criteria appear sufficient.

I The buckling stress criteria for both the shell and straight members also appear adequate.

We have concern with three items. They arei j

(1)

The allowable shear stresses for the steel columns and tie beams were not

, explicitly specified. In the subsequent evaluations, no evaluation results for shear stress in the steel columns and tie beams were available, either.

- (2) - A bearing capaelty of 20 ksf was specified here for the soll underneath all footings, which is inconsistent with the 10.6 ks'f capacity as specified previously in Section 5.4.1 of the criteria report. The source for the 20 ksf capacity is a .

Reference 14 in the VC _ report which, to our understanding, is an in-house document of the licensee and has not been reviewed by the NRC.

j (3) Pipe anchor loads at VC penetrations were accounted for in the analysis to evaluate the local shell stresses. De validity of the pipe anchor loads from the four main steam (MS) lines and four boiler feedwater discharge (BFD) lines, however, is questionable because it is our understanding that the evaluation of 4

l

( .

these pipes was completed only recently, subsequent to the publication of the g VC report.

Analysis Methodology

(

The modeling of the vapor container by finite elements is sufficient, which included soll springs at the footings to account for the soil-structure interaction. Pipe anchor loads at the vapor container penetrations were -accounted for in the analysis model to determine the local stresses developed around the penetrations. It appears that both the YCS and SSS spectrum seismic analyses used a modal damping of 5% of critical, which is sufficient in view of the fact that the structure essentially responded like a one-degree-of-freedom system with anchor bolt foundation connections.

Evaluation Results Licensee's evaluations were performed for the shell, shell penetrations, steel columns, tie beam, diagonal tie rods, base plates and anchor bolts for the columns, reinforced concrete pedestals and footings, and soll pressure. They concluded that the vapor container is capable of withstanding the LOCA load case and the two seismic (SSS and YCS) load cases.

They found local yielding would occur at some footing pedestals under both the YCS and SSS loads. In addition, yielding of some column base plate anchor bolts would occur. They concluded that such local yielding would not interfere with the safety function of the vapor container. We concur with the licensee's 4

conclusions and their justifications for the acceptance of the local yielding of the foundation, provided that the following concerns be resolved:

.(1) Results of evaluation of the shear stress in the support columns and tie beams were not available, and they should be provided.

(2)

Inconsistency in soll bearing capacity between the vapor container report here (20 ksf) and the criteria report (10.6 ksf) should be resolved. The 20 ksf value is based on a Reference 14 quoted by the vapor contain report which, to our understanding, has not been reviewed by the NRC.

(3)

Evaluation results for the elevises and turnbuckles on the diagonal tie rods, which was also a concern from the previous SEP review (Reference 10), were not 5

i

( ,

available.

They should be provided unless the " connections" as referred to by

. the licensee in Section V.I of the vapor container report, "The original design of the connections of the supporting columns and tie rods developed the full strength of the members. As a result,' the connections are considered acceptable.",

included the turnbuckles and clevises; besides, this statement was only for the g

YCS load case, and no conclusion could be found for the SSS load case.

(4)

The design of anchor bolts embedded in concrete is usually governed by the pullout capacity and not by the yield strength of the bolt material. Therefore, the column base plate anchor bolts should be evaluated against pullout failure unless the pullout espacity een be shown to exceed the yield strength of the bolts.

(5)

Sufficiency of the clearance between the reactor support structure columns and the VC at the shell penetrations should be evaluated, to justify the assumption that the two structures are uncoupled from each other in the seismic analysis.

The same concern applies to the clearance between the VC shell and the radioactive pipe tunnel ~of the primary auxiliary building (see Drawing 9699-FC-47B).

Amplified Response Spectra For the YCS load case, amplified response spectra of 2%, 3%, 5% and 7% at five locations were generated. No corresponding amplified response spectra were generated for the SSS load etce. This appears to have deviated from the requirement of NUREG-0825.

All concerns with the evaluation of the VC are summarized in Table 1.

Reactor Support Structure Licensee's evaluation of the reactor support structure was summarized in the reactor support structure analysis report, Reference 3. The findings of our review are as follows.

6 l

(. .

Loadings and Performance Criteria

(

Both the YCS and SSS loads were considered in the licensee's evaluations, and the structural damping adopted in the linear elastle analysis was 7% and 10%,

respectively. These damping values appear justifiable according to the evaluation results

( to be discussed later.

Like the evaluation of the vapor container foundation, the licensee used a bearing capacity of 20 ksf for the soil underneath the reactor support structure foundation,

t which is inconsistent with the 10.6 ksf as specified in the criteria report, Reference 1.

The source of the 20 ksf bearing capacity was just a transmittal from the licensee to CYGNA and, therefore, its acceptability to NRC must be established.

The material properties and performance criteria appear sufficient. The licensee performed nonlinear analysis for the SSS loads because stresses developed in the exterior columns and foundations exceeded the yield capacities, to verify that the structure will function under the SSS loads. The criteria for evaluating the bond stress and dowel embedment at the upper column connection with the concrete structure also appear sufficient.

More detailed discussions will be given later when we summarize our review of the evaluation results.

Analysis Methodology A 3D model was used for the linear seismic analysis for both the YCS and SSS loads.

For the SSS load case, nonlinear time history analysis was performed to verify the functionality of the exterior columns because their yield capacity was exceeded based on the linear analysis; for the nonlinear time history analysis, a simplified 2D model was used. In addition, finite element models were used for detailed nonlinear and linear static analyses of, respectively, the ring foundation and the collars of the exterior support columns. Soll-structure interaction was not taken into account in the analysis model but, according to the reevaluation criteria report, Reference 1, the frequency change due to soil-structure interaction for the reactor support structure was about 10% and may be neglected. We conclude that the models and methodology used in both the linear and nonlinear analysis are sufficient except for two concerns:

i 7

1

l t .

(1)

To evaluate the sufficiency of the nonlinear time history analysis methodology

, l we need a brief description of both the analytical basis and the status of quality assurance compliance of the CYGNA computer code PRA.

(2)

In the static finite element analysis of the steel collars, as illustrated in Fig.

, B.4a to B.4c, we. need a clarification on how and where the loads from the seismic analysis model were applied to'the finite element model of the collar.

Input motions to the nonlinear time history analysis for the SSS case included 4

both the synthetic motions having a 10-second duration and the El Centro records having a 30-second iuration of significant motion. We conclude that the time history input motions are sufficient from the response spectrum and duration points of view.

Evaluation Results The evaluation of the foundation soll is, as pointed out earlier, contingent upon the validity of the soll bearing capacity of 20 ksf. The mat foundation appears structurally adequate. De ring foundation would yield under the SSS loads, but would remain functional according to the nonlinear static analysis. We need a clarification, however, on the definition of " yield capacity" of the ring foundation in comparison to the corresponding ultimate capacity.

. The steel collars had been installed by the IIcensee to strengthen the exterior column-to-foundation connections, to resolve a concern identified in the previous SEP review of the Yankee plant in 1982 (Reference 10). Evaluation of the steel collars indicated that they were adequate to withstand the YCS and SSS loads. De licensee, however, did not evaluate the sufficiency of the rock bolts against pullout from the concrete foundation.

De columns were found adequate for the YCS loads, but would yield flexurally under the SSS loads. Ductile behavior of the columns can be expected, however, because of the relatively low shear stress developed in the columns. De nonlinear time history analysis was intended to verify sufficient ductility being available to

( maintain integrity at the structure-column connections. In the previous NRC review during 1982, four concerns were specifically identified regarding the sufficiency of the 8

- - - _ - . . . - , . - . _ _ _ . - - . - . - - _r. , _ _ _ _ - __ - , _ _ - , . , , , _ . , _ _ , , -- _ .._-_-_ __.-._,__..__.. __-

t .

1 I

dowels at the structure-column connections and the stud bolts on the inner face of the columns under the SSS loads (Reference 10):

t I'

(1)

Dowel Embedment - Concerns were identified regarding the development length, i

the bond stress determination and the strem-strain relationship of the #14 dowels, b

(2) Stud Bolts - 1he previous concern was regarding the shearing capacity of the welded heade:1 studs on the inner face of the steel columns to transfer the loads i developed from the dowels at the structural-column connections.

The licensee responded to these concerns in the reactor support rtructure report.  !

l Our consultant, Mr. Boris Bresler of Wiss, Janney, Elstner Associates reviewed the  :

j licensee's responses. Details of his review and conclusions are summarized in a letter, attached as an appendix to this Technical Letter Report. Bresler concluded that the i dowel embedment length and welded stud bolts would be. sufficient to resist the SSS

{

loads.

His "best estimate" calculations indicated that the plastic rotation capacity is of the same order of magnitude as the largest calculated demand. We judge that this would not prevent the structure from functioning safely under the SSS loads. There is, however, one puzzling result related to the evaluation of the P-Delta effect of the interior columns. That is, in Table E-3 the maximum displacement at top of the -

Interior column for the YCS case (0.543') exceeded that for the SSS case (0.111'), whleh contradicts with the fact that the YCS represents a smaller earthquake than the SSS does. In addition it is not clear why there are two rows of identical results presented I

in Table E-3.

! In summary, we conclude that the reactor support structure would withstand the

! ~  !

YCS loads elasticaHy and remain functional for the SSS loads provided that all the concerns identified previously are resolved. Table 1 lists all our concerns with the j evaluation of the reactor support structure.

l Amplified Response Spectrum '

The amplified response spectra were generated for both the YCS and SSS load l cases. The licensee performed a sensitivity study of the generated spectra to different synthetic ground motions for the YCS load case. 1 hey also compared the SSS case i

9

- ~ -- - - . .. - .- - -. - _ . . ..

s '

amplified spectrum from one of the non-linear time history analyses to that from the linear seismic analysis. The study showed that the spectra to be used for subsystem evaluations are adequate, and we concur with the IIcensee's conclusiens.

Turbine Building

(

Licensee's evaluation of the turbine building was summarized in Reference 4.

The results of our review are as follows.

i l, Loadings and Performance Criteria

• Both the YCS and SSS seismic loads were included in the structural evaluations. ,

I The performance criteria were consistent with those previously speelfied in the criteria report.

!, The bearing capacity adopted for the soll underneath the footings was 10.6 ksf, which is consistent with the value specified in the criteria report.

Analysis Methodology i

A 7% modal damping was used for both thi YCS and SSS load cases. Dree-dimensional seismic models were used for the lateral analysis of both the turbine 1 building and turbine pedestal. The modeling and analysis methodology appear adequate

except for two concerns. First, the analysis model shown in Fig. B-8 indicates that the turbine building provides structural support to the office building in only the EW direction. his is inconsistent with the statement on p.11 of the turbine building report that the turbine building supports the office building in both horizontal directions, and hence a clarification is necessary. Secondly, for the YCS case the 7% damping ,

appears sufficient for the evaluation of the turbine building, which is a braced steel i frame, but appears high for the reinforced concrete turbine pedestal According to licensee's evaluation results, the stress induced in the turbine pedestal for the YCS load case was about only 20% of yield and we believe a damping of less than 7% should i

be used. Such deficiency, we judge, would be immaterial because the use of a lower damping, say 5%, for the YCS load case would not jeopardize the structural Integrity of the pedestal. For the generation of the YCS case amplified response spectrum at the pedestal, however, this defielency could be significant and needs to be addressed by the IIcensee.

10

Evaluation Results

(

The licensee's evaluation appears sufficient except for two concerns. First, p.15 of the turbine building report states that the sum of the modal masses from the first 31 modes of the turbine building model exceeded 90% of the total mass. The sum of

( 2 the modal mass was actually about 26 k-sec /ft for either horizontal direction (Table D.1) which is less than 10% of the total building mass of about 330 k-sec2/ft (Table E.1), and clarification is required. Secondly, the result of evaluation of the unreinforced concrete masonry walls was not available although the evaluation criteria was presented

(

in the report. These masonry walls would probably fall under the seismic condition and, if so, the potential impact on the safe function of all nearby Category I items must be addressed.

Structural Deficiency and Strengthening Modifications As a result of their evaluation, the licensee identified the following deficiencies in the turbine building.

(1) Overstress and hence cracking would occur under both YCS and SSS loads in the concrete shield wall along Grid Line J because the wall was under-reinforced.

(2) No structural connections were provided in the current design for the transfer of seismic loads from the floor slabs into the structural frames and walls.

(3) One diagonal brace (in Grid Line 5) would buckle for the YCS load case, and nine braces (three in Grid Line 5, two in Grid Line G, and four in Grid Line C) would buckle for the SSS load case.

(4) Five brace to beam-column connections would not be capable to resist the SSS load case.

(5) Under the SSS loads, eight footings would be uplifted and the 10.6 ksf soll bearing capacity would be exceeded under four footings by less than 17%.

11 y _.

_q __ , - - , , , . , _ _ - - - . , - - _ - - - - _ _ , . . , , - , ~ - -

O '

The licensee suggested the following strengthening modifications:

(1) Install additional braces and shear walls, to redistribute the seismic lateral forces and hence. relieve overstress in the existing braces, connections, and shield wall along Grid Line J. Such strengthening, according to the lleensee, will also g alleviate the footing uplifting problem under the SSS loads.

(2) Provide pcsitive structural connections for shear transfer from concrete floor slabs to lateral load resisting system.

f While installing additional braces would strengthen the seismic resistance of the structure, it is not immediately clear to us such measure would necessarily also lessen the effect of seismic uplift at the column footings.

Amplified Response Spectrum Amplified response spectra were generated at two locations in the turbine building and one location on the turbine pedestal. As pointed out before, we believe that the amplified response spectrum for the YCS load case at the turbine pedestal should be generated using a damping of lower than 7%. Otherwise, the generation of the amplified spectrum appears sufficient.

Table 1 summarizes the previously identified concerns with the evaluation of the turbine building and pedestal Diesel Generator Building Licensee's evaluation results of the diesel generator building was summarized in Reference 5. The findings from our review are as follows.

Loadings and Performance Criteria The licensee considered both the SSS and YCS load cases. The performance criteria for the steel frames, the diagonal and K-braces, and the reinforced blockwalls of the annex building appear adequate.

12

t ,

Analysis Methodology s

t The 7% modal damping used in the analysis of both the diesel generator building

~

{

and the annex is sufficient. Licensee's evaluation was based on the assumption that f horizontal X-braces will be installed in the roof of the diesel generator building to i( justify its' rigid diaphragm behavior, and that more braces are introduced in the nitrogen and accumulator towers. Such strengthening modifications were implemented in the analysis models. Tlie analysis models appear sufficient except for the following concerns:

, (1)

The effect of soll-structure interaction was not addressed by the 11eensee,'and a justification is required for ignoring such effect. ,

l l

(2)

In Fig. B-8, the properties of the three beam elements in the vertical model of the diesel generator building are not available, and they should be provided for review.

I I

(3)

In the vertical analysis, clarification is required on how the coupling between the diesel generator building and the nitrogen tower and accumulator tower was accounted for. It appears from the evaluation report that the building was analyzed separately from the two towers.

(4)

The licensee stated on p. 7 of the evaluation report that structural coupling between the diesel generator building and primary auxiliary building was assumed to be in the vertical direction. It appears from Detail B of Fig. A-4 that the two buildings are actually coupled in the horizontal (Y) direction, and not the vertleal dirce'tlon, and clarification is required. In addition, if the two buildings are indeed coupled in only the vertical direction clarification is required on how the coupling was accounted for in the vertical model of the diesel generator building shown in Fig. B-8.

(5) There appears to be inconsistency in the locations of the knee braces between the design drawing (Fig. A-4) and the model (Fig. B-4). For example, the knee brace at the upper left corner of the center bay in Section S1, Fig. A-4, appears to have been erroneously located at the upper left corner of the side bay of the seismic model. Verification of the consistency in the locations of all knee 1

braces is therefore required between the model and existing design.

13

.g g-(

Evaluation Results e

The licensee presented only the results of evaluation of the SSS load ease although the YCS load case was mentioned in the licensee's evaluation criteria. The results for the SSS loads indicated the diesel generator building would be adequate provided that the aforementioned strengthening modifications to the roof and the vertical b' racings in the nitrogen tower and accumulator building are implemented.

We have some concerns with the evaluation results, as follows:

4 (1) ne story seismic drift index of 0.016 as shown on p.10 of the evaluation report is not, as stated by the licensee, less than the UBC allowable drift index of 0.005. Clarification is therefore required.

(2)

According to the results shown in Table E-1, seismic uplift of certain column bases is possible under the SSS loads but was not addressed by the licensee.

Herefore, evaluation of the column bases against seismic uplift is required.

(3)

Evaluation results for the knee braces were not presented in the evaluation report, and they should be provided for review.

(4)

No evaluation results for the reinforced blockwalls in the anner were presented in the report although the pertinent performance criteria were specified.

All concerns identified previously for the evaluation of the diesel generator building are also summarized in Table 1.

Structural Defielency and Strengthening Modifications i The licensee identified the diesel generator building roof diaphragm and the lateral load resistance capacity in the nitrogen tower and accumulator building ~ to be the deficiencies.

They planned to Install horizontal X-braces in the building roof, and vertleal braces in the nitrogen tower and accumulator building.

1 14

Amplified Response Spectrum

(

No amplified response spectrum was generated, and a review becomes inapplicable.

Primary Auxiliary Building

(

Licensee's evaluation results for the primary auxiliary building were summarized in Reference 6. 'Ihe conclusions of our review are as follows.

! Loading and Performance Criteria Both the YCS and SSS loads were included in licensee's evaluation. 7% damping was used in both load cases. As pointed out later, when discussing the evaluation results, the stresses developed in the reinforced concrete, steel columns and X-braces were so low that the use of 7% damping for the YCS loads was not justifiable. We believe that a 3% to 5% damping would be more reasonable for the YCS loads. In addition, the criteria for combining the effects from the three earthquake components is not available.

The allowables for the steel frames, diagonal braces, steel beam to concrete wall connections, radioactive pipe tunnel, and unreinforced concrete block walls are adequate. Reference to the 1979 UBC for the ultimate shear strength ~of reinforced concrete walls is, however, inconsistent with all other evaluation reports and the criteria report, and justification is required.

Analysis Methodology

'Ihree-dimensional stick model was used to represent the primary auxiliary building I and the radioactive pipe tunnel. In the horizontal analysis, equivalent columns were included in the model to represent the reinforced concrete walls. A dummy subdiaphragm was also introduced at the first floor levd in the horizontal model, in order to properly.

model the connection of the pipe tunnel support column to the first level,

.i Response spect-um analysis was performed for the structure using the 7% damping YCS and SSS spectrum as ground motion. Soil-structure interaction was not represented in the analysis models. Our comments on the analysis model are as follows: ,

15 l

( ,. .

4 (1) As pointed out earlier, the use of the 7% damping YChi. spectrum was not

< . justifiable because of the very low stress induced in the structure.

.(2) Using simple calculations, we judged that neglecting the effect of soil-structure interaction in the horizontal analysis appears sufficient, but 'may not be so for I

( the vertical analysis.' Therefore, justification for neglecting the soil-structure.

interaction in the vertical analysis is needed.

l (3) The validity of representing a wall by several equivalent columns interconnected ir by rigid links is questionable. For example, the equivalency in moment of inertia between the wall and. three equivalent columns is not immediately clear unless the flexural rigidity of the wall is so high that the flexural deformation becomes

! negligible compared to the shear deformation.

s 2

(4) The property of the dummy subdiaphragm at the first level of the horizontal model was not available. Neither could the dummy subdiaphragm be located from the figures illustrating the horizontal model.

b (5) Clarify whether those structural nodes constrained by the in-plane rigid diaphragm assumption are unnecessarily restrained for any out-of-plane degrees of freedom.

(6) For' the vertical model shown in Fig. B-5 the radioactive pipe tunnel appears

, . uncoupled from the primary auxiliary building at the R.C. roof (i.e., mass D1).

! This appears to contradict the actual design, and clarification is required.

The above comments,"we believe, need not materially change the lleensee's conclusions about the seismic adequacy of the building, but some of them would affect  ;

l the adequacy of the amplified response spectra.

• 3 Evaluation Results The results indicated that both the primary auxiliary building and the radioactive pipe tunnel were adequate for both the YCS and SSS loads except for one single diagonal brace and some steel beam to concrete wall connections. The unreinforced concrete p block walls were overstressed and were expected to fall at the early stage of an earthquake. This justified the licensee's neglecting the stiffness of the block walls in 16 1

the seismic analysis model. The generally very low stress in the structure subject to

i the YCS loads suggest that the slidity of using 7% damping in this case is questionable.

As pointed out earlier, we believe that a 3% to 5% damping would be more appropriate.

Using the lower damping value for the YCS load case, however, is not anticipated to significantly change the licensee's current conclusions on the structural evaluation. Our i concerns with the evaluation of the structure are as follows:

(1)

Verification is needed of the accuracy of the very large shear stress of 130 psi in Table E-1 for the R.C. wall located at Level 2 and along Column Line No. 5.

t (2) Tables E-4 and E-10 should be completed by providing the axial areas and allowable compressive stresses that are currently absent from the tables for five columns at Level 2.

(3) According to Table E-10, the column located at intersection of Column Lines 8 and Fb would be subjected to seismic uplift. Sufficiency of the foundation and base connection for this column must therefore be evaluated against seismic uplift.

All concerns identified previously are summarized in Table 1.

4 Structural Deficiency and Strengthening Modification The licensee's evaluation identified two structural deficiencies, and recommended the following strengthening measures:

(1) The single diagonal brace located in Column Line 8 was severely overstressed.

The licensee recommended the installation of 'an additional diagonal brace to form an inverse "V" with the overstressed brace.

(2) ' The bolts at the steel beam to concrete wall connections located at the intersection of Column Lines 5.5 and G,6 and Fb, and 6 and Ec were overstressed in shear. They need to be strengthened.

17

t o

Amplified Response Spectrum l

Amplified response spectra were generated at two locations at Elevation 1039'-

6". Both YCS and SSS load cases were included in the spectrum generation.

Our concerns on the sufficiency of the amplified response spectra are given below, r

(1) Amplified response spectra for the YCS load case should be generated based on a structural damping lower than 7%. This is because the structure was only lowly stressed under the YCS loads.

(2) The validity of the vertical response spectra is questionable unless the vertical soil-structure interaction effect can be shown to be negligible.

Fire Water Tank Licensee's evaluation of the fire water tank was summarized in the fire water tank analysis report, Reference 7. The finding from our review are as follows.

Loadings and Performance Criteria Only the SSS spectrum was considered in the fire tank evaluation. The YCS load case was not evaluated and we judge this to be sufficient because, in accordance with our assessment of the analysis results, the tank would remain within elastic limits subject to the SSS load.

The licensee stated that. the stress developed in the steel shell was to be evaluated against Section 3 and Appendix E of the American Petroleum Institute (API) Standard 650, which was used also on the original design of the tank shell. According to the fire water tank report, API Standard 650 gave the allowable compressive stress to be 6.60 ksi considering possible initial nonlinearity and fabricattor. Imperfection in the shell. In the actual evaluation, however, the licensee used instead the full theoretical value of 24.1 ksi for an ideally perfect and linearly elastic cy!!ndrical shell. We believe this ideal buckling capacity to be unrealistic. Even Reference 9 recommends that a maximum of only about 40% of the theoretical value should be used, which is about 9.6 ksL 18

( , .

Other than the buckling stress criteria, the criteria used by the lleensee in the

, evaluation of the fire water tank appear sufficient.

Analysis Methodology Housner's simplified representation of the hydrodynamic response of the water with a vibrating (sloshing) mas and a rigid man was adopted by the licensee. %e tank shell was represented.by a fixed-base stick model (for dynamic analysis) and a fixed-base finite element model (for static shell stres analysis), and soil-structure

, interaction effect was not considered. To estimate the significance of soil-structure interaction effect, we performed some simple calculations asuming the tank to be a rigid structure. The resulting lateral, rocking and vertical frequencies were about 25, 28 and 20 Hz, respectively, which are much higher than the rigid base frequency of-the tank and suggest the effect of soil-structure interaction would indeed be negligible.

The modeling and analysis appear sufficient except, for two concerns:

(1) Modeling of the shell was based on a 5-ring construction in which the bottom ring was 0.35" thick and the remaining four rings were 3/16" thick. According to the design drawing, however, the shell was a 4-fing construction with the same overall tank height of 32'.

(2) The stress distribution shown in Fig. E-1 appears questionable because it is very different from what would be calculated from the beam theory. D erefore, description is needed on how the hydrodynamic loads derived from the simple Housner model were applied to the static analysis of the finite element model that led to the results shown in Fig. E-1.

Evaluation Results The licensee calculated a maximum compressive stress of 9.32 ksi in the shell which exceeded the allowable of 6.60 ksi as specified by API Standard 650 but was well within the theoretical buckling stress of 24.1 ksi, and did not consider the shell overstressed. Maximum tension in three anchor bolts due to the SSS loads exceeded the allowable by about 30%, but the licensee considered the tension region to be very limited and redistribution of stres would take place, with which we concur. Soil pressure under the ring foundation was 7.7 ksi, which was within the allowable. The 19

,t ,

licensee thus concluded that the fire tank shell and foundation 'are adequate for the

SSS load case and, hence, the YCSload case although the YCS load case was not analyzed.

, We have several comments on the licensee's evaluation results, as follows.

(1) As pointed out earlier, a sufficient evaluation of the buck $ng potential of the tank shell requires both the use of an acceptable buckling stress criteria and the clarification of the methodology of load application to the finite element model that generated the stress distribution shown in Fig. E-1.

i (2) The tank base anchor bolts should be evaluated also for possible pullout failure and shearing failure unless these two failure modes can be shown less critical than tension yielding of the bolts themselves.

(3) Licensee's evaluation was incomplete. Dere was no evaluation of the local

. integrity at the pipe-to-tank connection due to the pipe anchor loads. D ere was no evaluation of the bearing capacity and hence settlement potential of the compacted backfill beneath the tank bottom.

! (4) Amplitude of the water sloshing should be calculated and checked against the l free board, to ensure that water would not impact the tank roof and cause damage. 2 f

Other than the above comments, which are summarized in Table 1, the licensee's evaluations app ar acceptable for the SSS loads. 'Ihe results indicate an evaluation for the YCS loaas would not be necessary.

4 Structural Deficiency and Strengthening Modifications i

Based upon the licensee's current evaluations, no structural deficiency was

. identified for the SSS loads and hence no strengthening measures are presently required.

! This conclusion, of course, is contingent won licensee's satisfactory resolution of all s

the concerns identified in Table 1.

Amplified Response Spectrum l No amplified response spectrum was generated, and our review is not required.

I 20

,,-.w.,,,,, -- w.,-- - - ,- -m-,,...,.s a- ..n - - -,-wn ,w, , -- - -,---s- - - -

rw, ..-w., ----, ,- -- w-,-,.. - - - - , - , , - , . -.-- .-- a --,m._.,,---,,---a, _ , ,-

CONCLUSIONS We reviewed the licensee's seismic reevaluation criteria report, five structure evaluation reports, and fire water tank report. The five structures include the vapor container, reactor support structure, turbine building, diesel generator building, and

, primary auxiliary building. Our conclusions are summarized in the following.

Seismic Reevaluation Criteria Report

<. We reviewed only those portlens of the report related to the scope of our review although the report also includes evaluation criteria for piping and other systems. The criteria report is in general sufficient except for the following deficiencies:

(1) The report is incomplete in that it was not updated to specify the YCS load case, the elastic requirement under the YCS loads and functionality requirement under the SSS loads as set forth by NUREG-0825, the combination of earthquake component effects, and the generation of amplified response spectrum.

(2) We do not concur with the general assumption that soil-structure interaction effect is negligible even though the licensee has demonstrated this to be true for the vapor container and reactor support structure. It must be evaluated on the individual structure basis.

The above represent deficiencies of the criteria report itself, and are not necessarily applicable to the actual structural evaluations. Therefore, we also reviewed the criteria actually adopted for each Individual structure.

Structural Eve.luation Findings from our review of the licensee's seismic evaluation of the five structures and fire water tank are summarized in Table 1. In general, the evaluations are adequate except for certain items that require clarification or additional evaluation. Current structural deficiencies were identified and proper strengthening modifications were proposed by the licensee for the turbine building, diesel generator building, and primary l

l 21

( ,

i auxiliary building. We believe that after the outstanding concerns are resolved and the strengthening modifications implemented, the Yankee Nuclear Plant structures would be sufficient to withstand both the YCS and SSS loads.

Amplified Response Spectrum t

As shown in Table 1, the only deficiency in the response spectrum generation includes:

( (1) Vapor container - Response spectrum for the SSS case is not available.

(2) Turbine Pedestal - A structural damping lower than 7% should be used for the spectrum generation in the YCS case.

(3) Primary auxiliary building - Vertical soil-structure interaction should be taken into account. In addition, a structural damping lower than 7% should be used for the YCS load case.

t 22

. - . - . . . _. -,=__ _-. . - - .- ._

1< .

REFERENCES

! 1. Cygna Energy Servlees, Seismic Reevaluation And Retrofit Criteria for Yankee Nuclear Poweri :"* n, Rowe, Massachusetts, Doc. No. DC-1, Rev. 2, August 1982.

i n 2. Cygna Energy Servlees, Vapor Container Structure Yankee Nuclear Power Station, 1-l Structural Analysis Report, Report No. EY-YR-80023-5, Rev. 3, April 1984.

d

3. Cygna Energy Services, Seismic Analysis of Reactor Support' Structure Yankee j Nuclear Power Station, Rowe, Massachusetts, Report No. EY-YR-80023-6, Rev. 3, March 1983.

. 4. Cygna Energy Services, Turbine Building. Yankee Nuclear Power Station, Structural Analysis Report. Report No. EY-YR-80023-9, Rev.1, December 1982.

4

5. Cygna Energy Servlees, Diesel Generator Building & Annex, Yankee Nuclear Power Station, Structural Analysis Report. Report No. EY-YR-80023-8, Rev. 2,

• January 1983.

6. Cygna Energy Servie~es, Seismic Analysis of Primary Auxiliary Building and i

Radioactive Pipe Tunnel, Yankee Nuclear Power Station Rowe, Massachusetts, Report No. EY-YR-80023-7, Rev. 2, January 1983.

i

{ 7. Cygna Energy Services, Fire Water Tank, Yankee Nuclear Power Station Rowe, i

Massachusetts. Report No. EY-YR-80023-15, Rev. O, March 1983.

8. Integrated Plant Safety Assessment' of Yankee Nuclear Power Station at Rowe, Massachusetts -NUREG-0825, February 1983.

9. R. J. Roark and W. C. Young, Formulas for Stress and Strain, 5th Edition, McGraw-Hill Book Company.

10. Attachment No. 2 to the letter from D. M. Crutchfield, NRC, to J. A. Kay, YAEC, dated February 1,1983.

23 4

- - - . ..,..,.-,.-,y...rm...,..myy_.,_____-,,_, _ , _ _ , ,_____,_.._.,,m,,,, , , _ - - - , _ , . ,__m._,, _ _ _ _ _ . - , . - _ . . . _ , , , ,

Table 1: Summary of review results t Structure Evaluation Analysis Evaluation Deficiency & Amplified Criteria Methodology Results Strengthening Response Modifications Spectrum Y

(1),(2) Acceptable (1),(3),(4) None (7)

( C w (5),(6)

Reactor Support (2) (8),(9) (5),(10),(11) None Acceptable Structure (12) (13) (14),(1F! (16) (12)

~

Diesel Generator Acceptable (17) to (21) (15),(22),(23 '

Yes Not Building (24) Required Primary Auxiliary ' (12),(25), (27) to (30) (15),(31), Yes (12),(26)

Building (26) (32),(33)

Fire Tank (34) (35),(36) (36) to (38) None Not e

Required Notes:

(1) Provide the allowables and evaluation results for shear stress in the steel columns and tie beams of the VC.

(2) Establish the acceptability of the 20 ksf soll bearing capacity which is inconsistent with the 10.6 ksf bearing capacity specified in the criteria report.

(3) Verify the validity of the pipe anchor loads from the four main stream (MS) and four boiler feedwater discharge (BFD) lines that were applied to the VC analysis for evaluating local shell stresses at the pipe nozzles, because it is our understanding that the VC report was published prior to the completion of reevaluation of the MS and BFD lines.

(4) Provide evaluation results for the elevises and turnbuckles on the diagonal tie rods for both the YCS and SSS cases.

(5) Provide evaluation of the column base anchor bolts against pullout from the concrete foundations unless the pullout capacity can be shown to exceed the yield capacity of the bolts. ,

(6) Provide evaluation of the sufficiency of the clearance between the VC and reactor support structure columns at the shell penetrations, to justify the assumption that the VC is seismically independent of the reactor support structure, ,

24

l '

t

] .

Table 1 (continued)

Provide also evaluation of the clearance between VC and the radioactive pipe

, tunnel of the primary auxiliary building.

i (7) The SSS load case amplified response spectrum should be generated.

i ' (8) Provide information on the analytical basis and the status of quality assurance

) compliance of the CYGNA computer code PRA which was used in the nonlinear l seismic analysis of the reactor support structure.

l 4

(9) Clarify how and where the seismie load was applied to the finite element model of the collars as shown in Fig. B-4a to B-4e of the reactor support structure report.

(10) Verify the accuracy of the 0.543' (YCS case) and 0.111' (SSS case) maximum 1

displacements at the top of interior column because we anticipate YCS case j displacement should be smaller than that of the SSS case.

), -

(11) Clarify the definition of " yield" of the ring beam foundation in relation to the j

ultimate strength.

4 (12) The low stress induced in the turbine pedestal and primary auxiliary building did not substantiate the use of 7% damping for the concrete for the YCS case, and a lower damping value should be used for the generation of amplified response j, spectrum.

(13) In the turbine building report, the analysis model in Fig. B-8 appears to indicate  !

l that the turbine building provides structural support to the office building in only the EW direction while on p.11 the licensee stated that the office building j is supported in both horizontal directions by the turbine building. Clarifiction '

i of this inconsistency is required.

(14) On p.15 of the turbine building report the licensee stated that the sum of the '

modal masses from the first 31 modes of the seismic model exceeded 90% of '

mass. Table D.1 shows the sum of the modal masses to be

the ttotal a buildp/ft 26 k-sec which is less than 10% of the total bunding mass - of 33 sec /ft as listed in Table E.1.

(15) Evaluation results are required for the unreinforced concrete masonry walls, and the effect of the anticipated failure of these walls during earthquakes on other i

Category 1 items should be addressed.

(16) While the recommended installation of additional braces would strengthen the seismic resistance of the structure, it may not necessaruy also lessen the effect i

of seismic uplift at the column bases. Please clarify.

(17) '!he effect of soil-etructure interaction was not taken into account in the analysis, and it should be addressed.

l (18) Provide properties of the three beam elements in the vertleal model shown in i Fig. B-8 of the diesel generator building report.

1 25 1

l

4 7- .

4^

l Table 1 (continued)

(19) Clarify how the coupling between the diesel generator building and nitrogen tower

~

and accumulator tank tower was accounted for in the vertical analysis because it appears that from the analysis models they were analyzed separately.

i I

(20) Clarify the direction (vertical or Y-) of structural coupling between the diesel generator building and primary auxiliary building. In addition, if the two buildings

! are coupled in the vertical direction, as the licensee so stated, clarify how such coupling was accounted for in the vertical analysis of the diesel generator building.

(21) Verify the consistency in the locations of the knee braces between the design j, drawing (Fig. A-4) and analysis model (Fig. B-4).

i (22) Verify the accuracy of the story seismic drift index of 0.016 (p.10 of evaluation. i i report) because it exceeded the UBC allowable of 0.005 but the licensee stated 4

the opposite.

. (23) Provide evaluation for seismic glift of certain column bases for the SSS load  !

case because Table E-1 Indicates the uplift potential existed.

~(

(24) Provide evaluation results for the knee braces in the diesel generat'or building i j and the reinforced blockwalls in the annex.

l (25) Clarify why the 1979 UBC was used as the criteria for evaluating the reinforced

! concrete walls in the primary auxiliary building. UBC was neither specified in '

the criteria report nor used in the evaluation of other Category I structures.

1

-(26) Effect of soil-structure interaction in the vertical analysis should be addressed  ;

j for the primary auxiliary building evaluation. '

1 I (27) Validity of representing a wa!! by several equivalent columns in the horizontal $

c model of the primary auxiliary building should be clarified. ,

! (28) Property and location of the " dummy subdiaphragm" at the first level of the J

horizontal analysis model of the primary auxiliary building should be identified.

(29) Clarify whether the structural nodes constrained by the in-plane rigid diaphragm assumption are unnecessaruy restrained for any outef-plane degrees of freedom.

I (30) Clarify why in the vertical model shown in Fig. B-5 of the primary auxiliary

building report the radioactive pipe tunnel appears to be uncoupled from the

building at the R.C. roof.

i (31) Verify the acuracy of the 130 psi shear stress shown in Table E-1 for the R.C.  !

wall located at Level 2 and along Column Line No. 5 in the primary auxiliary l i

building.

e l (32) Complete Tables E-4 and E-10 of the auxiliary building report by providing the i

axial areas and allowable compressive stresses that are missing from the Tables for five columns at Level 2.

1 (33) Provide evaluation of seismic uplift for the auxiliary building column loce'ed at intersection of Column Lines 8 and Pb because Table E-10 indicated ths uplift i

potential.

26

Table 1 (continued) s (34) Justify the validity of the 24.1 ksi buckling stress criteria for the evaluation of the fire tank shell.

(35) Reconcile the discrepancy between the 4-ring construction and 5-ring analysis

, model for the fire tank wall.

(36) Provide descriptions on how the hydrodynamic loadings derived from the simple Housner model were applied to the finite element model which led to the stress distribution shown in Fig. E-1.

, (37) Provide evaluation results for the tank base anchor bolts against potential pullout from the ring beam foundation and shearing failures unless these two failure modes can be shown to be less critical than yielding of the bolts themseDes.

(38) Provide evaluation results for (a) local shell integrity at the pipe-tank connecticn due to the pipe anchor load, (b) bearing capacity and settlement potential of

, the compacted backfill beneath the tank bottom, and (c) sufficiency of the free board at top of tank wall vs. the maximum amplitude of water sloshing so that the tank roof would not be impacted by the sloshing water during earthquakes.

(

l 27 i 1

APPENDIX The appendix presents Mr. Brester's conclusions from his review of the dowel embedment and welded stud bolts at the exterior column to structure connections of the reactor support structure. It was prepared in the form of a letter report to the NCT Engineering, as attached here, f

i l

A-1 l

Wiss, Janney, Elstner Associates, Inc. gg WE 2200PCMELLstatET.5U11Et2S co suttmo ano afstanc ( cattas

. t ut avy.LLE. Ca s.4061636 .

e4tSe s2t ract gr j

di$.a. October 18, 1984

• .*Efd'*

:::"a.-"

(

..c ,

! M*.'A"3

• t g.,c,,,,

,, Mr. N. C. Tsai l i Es'o.,'

NCT Engineering Inc.

U E'i Lac.na P. O. Box 1937

, , , _ (3650 Mt. Diablo Blvd.)

1* iO.C% 1.afayette, CA 94549 i M.','

Nh l NT4L' Re: Review of RSS Column Joints Yankee Nuclear Power Station y @',, WJE Job No: 840727

• " 'l *L'."

Dear Mr. Tsai:

• UTi

$h

" ; .'f**,,,,, ,, .

At your request, we have reviewed portions of the March, 1983 (CYCNA Report No. EY-YR-80023-6) Seismic Analysis of Yankee Nuclear

• ' , O Power Station Reactor Support Structure (RSS) dealing with dowel

' ' y",,,

ot embedment and stud bolt connections at the column tops.

8 l' U"o'.,..

.* ! L",'. DESCRIPTION OF THE C01EIGt CONNECTIONS e 9 S'iCsCse Six exterior and two interior columns support the cylindrical concrete reactor structure weighing approximate'y 26,000 kips. The exterior concrete columns, 7.0 feet in diameter, a : proximately 40 feet high, and enclosed by a steel shell 3/8 inches thic> are connected to the reactor structure at the top of the upper cylin'rical structure through 44 No.14 vertical steel bars. The embedment lengths of these devel bars at the base of the upper structure range from 30 to 42 inches. The 44 bars are (presumed to be) arranged in one ring on a 5-foot, 8-inch diameter circle which results in an average clear spacing between the bars of about 3-1/8 inches. The interior concrete columns, 7 feet, 6 inches in diameter, about 28 feet high, and enclosed by a steel shell 1/2-inch thick, are connected at the top to the upper cylindrical structure through 64 No. 14 vertical steel bars. The embedment lengths of these dowel bars at the base of the upper structure range from 31 inches to 38 inches. The bars are (presumed to be) arranged in one ring 75 inches in diameter which results in an average clear distance between the bars of about 1-7/8 inches. If the bars are arranged in two rings, the spacing of the bars in the inner ring.

would be about 4 inches.

A-2 CHICAGO SAN FR ANCISCO PRINCETON HONOLULU DENVtA

Wiss, Janney, Elstncr Associates, Inc.

NCT Engineering Inc. October 18, 1984 Mr. N. C.' Tsai Page 2 a

Steel headed studs welded to the exterior shell are embedded in the concrete columns below the construction joint at the top of the

~

, column. The 7/8-inch diameter stud bolts are estimated to be 5 inches long and are spaced 8 inches apart vertically and circumferential1y.

There are 330 bolts in the exterior column and 500 in the interior column.

REYLEW OF DOWEL EMBEDENT ANALYSIS The dowel abilitity to develop yield strength and undergo sufficient inelastic deformation is essential in order to allow development of a plastic hinge with adequate rotations without pull-out as assumed in the RSS nonlinear analysis. Assessment of this performance requires identification of the stress-strain diagram for the No. 14 steel dowels and evaluation of bond stresses between the dowel and surrounding concrete and the adequacy of the anchorage length.

The CYGNA report identifies the reinforcing steel as No. 14 deformed bars conforming to ASTM A15-54T. This specification covers three grades of steel - structural, intermediate, and hard grade - with minimum yield point values of 33, 40, and 50 ksi respectively. Unless the grade uf the bars is identified, it is difficult to establish the range of the yield point values: under the general catageory of A15-

$4T bars, these values can range from 33 to 60 kai. As suming intermediate grade, the yield point values can range from 40 to 50 kei.

Furthermore, the CYGNA report assumes a bilinear stress-strain diagram for the No. 14 steel bars. The actual stress-strain diagram for intermediate grade steel is likely to have a yield point above the minimum value of 40 ksi, and if the bars comply with the specification, the stress-strain diagram must have a flat plateau (plastic zone) before commencement of strain hardening. This characteristic derives from the requirement in the specification [Sec. 6(b)] that the yield point shall be determined by the drop of the beam or halt in the gage of the testing uachine. Also, the ultimate strain of the No.14 bars is likely to be less than 0.2, probably of the order of 0.16.

The value of the average bond strength assumed in the CYCNA report is p = 0.7 kai. Using this value and the available anchorage lengths of 30 and 30.75 inches, the maximum tensile stress before pull-out (or excessive slip) have been calculated (see Eq.1 in the report) as 51 and 53 ksi respectively. Thus, while the existing anchorage does not meet the 1979 UBC requirement (Section 2612(f)], it is adequate to develop tension stresses in the devels in the range of 40-50 ksi.

A-3

F - - -

Wiss, Janney, Elstner Associates, Inc.

1 i

NCT Engineering Inc. October 18, 1984 Mr. N. C. Tsai Page 3

(

The proposed evaluation of the plastic rotations 1 capacities (Page 18 of CYGNA report) presents three basic problems: assumption of the bilinear stress-strain diagram, assumption of unfiors bond stress cistribution along the anchorage length, and selection of an appropriate value of the yield point stress fy .

4 The distribution of the bond stress along the bar changes as the

' load increases. At-low load levels in the steel reinforcement, high '

bond stresses occur in the sone adjacent to the main transverse crack, which in the case of the RSS, is likely to develop at the top of the column (at the end of the steel shell). . With , increasing load, some local slip occurs in the zone adjacent to this crack and the peak bend stress shif ts away from the crack. At yield stress level for steel' i

with an ideal plastic plateau in the stress-strain diagram, the bond  !

stress reduces to practica11'y sero as there can be no bond stresses in the zone where the force in the steel bar remains constant, equal to f A y b (yield stress x area of the bar). With further elongation in  !

the bar, low bond stresses are generated in the strain hardening range, and the peak bond stresses are shifted further away from the initial transverse crack. As the plastic hinge develops, the distribution of the bond stresses along the anchorage length of the. tension dowel are far from uniform. Therefore, use of equations of the type given on Page 18 of the CYGNA report (Eqs. (2) through (5)) is questionable.

It should be noted that the assumption of uniform bond stress

' distribution for the purpose of evaluating length of anchorage required to prevent pull-out is acceptable. The same assumption, however, may not be satisfactory for evaluating rotation capacity, as the strain' 3

distribution and elongation of the bar are highly sensitive to values l

of u(x) and E sh*

The difficulties of determining c oax defined by Eq. (3) of the report can be easily demonstrated bf examining the inelastic strain contribution (the second ters of the right hand side of the equation).

In the ideally plastic range of the stress-strain diagram, Esh = 0 and p= 0, and consequently, the inelastic strain is 0/0, or indeterminate.

Therefore, the validity of rotation capacities given in Table E-16 are questionable.

4 Another uncertainty in evaluating the plastic rotation capacity is the magnitude of the yield stress yf. The general effect of the l

variation in f using Eqs. (2)ythru on C can be seen from the following table [obtained (5) in a manner similar to that used in obtaining

{ values in Table E-16).

1 a

i A-4

Wiss, Janney, Elstner Associates, Inc.

NCT Engineering Inc. O,ctober 18, 1984 Mr. N. C. Tsai Page 4 t-f., kai 40 50 la, in 30 30 4, 10,. a 36* 4*

for interior Column, Db = 75 inches It can be seen that with fy = 50 kai, the conservatism in the calculated value of the plastic hinge rotation capacity disappears, as the calculated capacity 4x10 radians is just about 10% over the maximum calculated demand value of 3.6x10~4 radians. In view of the uncertainties in Eqs. (2) through (5) and the uncertainty of the yield

' stress f y, it may be concluded that the calculated rotation capacity is of the same order of magnitude as the calculated rotation demand.

A more significant measure of performance is the relatively small value of slip required to meet the plastic rotation demand, ds calculated on Page 18 of the CYGNA report: (90)(3.63 x 10-4) = 0.033 inches.

REVIEW OF STUD BOLT ANALYSIS The transfer of the maximum tension forces from the dowels to the composite concrete column (steel shell filled with concrete) is accomplished through the headed steel studs welded to the exterior shell.

For the exterior column shell 4 inches in diameter, there are 33 bolts at 8 inches spacing around the circumference, and therefore,10 rows are required to accommodate the 330 bolts called for. The 44 No.

14 dowel bars are spaced at a distance of 4.85 inches (on centers),

and therefore, on the average (44/33) = 1.33 times the devel tension must be transmitted through one longitudinal row of bolts (10 bolts).

Using a value of 1.1 f y= 55 kai (in lieu of CYCNA's 44 kei), the average shear per stud bolt is 16.5 kips. For the interior column shell, 90 inches in diameter, there are 35 bolts at 8 inches spacing around the circumference , and therefore, 14 rows at 8 inches are requried to accommodate the 500 bolts called for. The 64 No.14 dowel bars are spaced at 3.63 inches (on centers), and therefore, on the average (64/35) = 1.83 times the dowel tension must be transmitted through one longitudinal row of bolts (14 bolts). Again, using 1.1 fy '

= 55 ksi, the average shear per stud bolt is 16.2 kips. i 1

A-5

Wiss, Janney, Elstner Associates, Inc.

NCT Engineering Inc. October 18, 1984 Mr. N. C. Tsai Page 5

(

Both of these values are within acceptable limits as the 1979 UBC allowable horizontal shear load for 7/8-inch headed stud (3.5 inches long) in 4000 psi concrete is 18.0 kips - under service load conditions.

Therefore, it appears that the load transfer through the stud bolts is satisfactory.

CONCIESIONS This review has been limited to the evaluation of the dowel embedment and welded stud bolt connections at the top of columns supporting the RSS. Within this limited scope, we have come to the following conclusions.

1. Although some uncertainty exists about the yield point value of th. embedded No. 14 dowels, it is reasonable to assume a value for f =

y 50 kei, the upper limit for intermediate grade A15-54T reinforcing steel bars.

2. Using these material characteristics, the embedment length of the No.14 steel dowels is sufficient to develop these yield stresses without pull-out or excessive slip in the connection.

3. The calculated plastic rotation capacity cannot be determined precisely in the absence of test results on bond stress distribution along the length of anchorage with large (No. 14) reinforcing steel bars arranged .with relatively close spacing.

However, based on "best estimate" calculations, it appears that the rotation capacity is of the same order of magnitude as the largest calculated demand (per CYGNA report).

4. The load transfer from the No.14 steel dowels through the welded studs to the outer steel shell of the column results in a calculated load per bolt, well within the 1979 UBC limits.

?

Very truly yours, '

s Wise, Janney, Elstner Associates , Inc. '

B. Bresler Principal BB:1js A-6

...