ML20234F085

Revised Summary Evaluation of Reactor Pressure Vessel/ Neutron Shield Tank Seismic Response & Amplified Response Spectra Generation

ML20234F085
Person / Time
Site:	Yankee Rowe
Issue date:	12/30/1987
From:	CYGNA ENERGY SERVICES
To:
Shared Package
ML20234F069	List: ML20234F066 ML20234F085
References
NUDOCS 8801110321
Download: ML20234F085 (40)

Text

SUh151ARY EVALUATION OF REACTOR PRESSURE VESSEL / NEUTRON SHIELD TANK SEIShilC RESPONSE AND ARS GENERATION FOR YANKEE ATOMIC ELECTRIC COMPANY l

l Decemtwr 30, 1987 l

11ll!!!!!!!Il1ll11111111111111 $$$11kpOCK321B$$10029 0

ppR '

l P

'j

}

k; 1

ISSUE

SUMMARY

Perform a stability analysis of the Reactor Pressure Vessel (RPVyNeutron Shield Tank (NST) to demonstrate that the RPV/NST will not uplift or slide under the combined dead and NRC seismic spectra loads.

Generate the amplified response spectra (ARS) at significant Reactor Support Structure j (RSS) locations and at the Main Coolant Loop (MCL) nozzles. ,

i l

l l

l llllll111111111!Il111111111111

REFERENCES:

• 1. " Seismic Reevaluation and Retrofit Criteria," DC-1, Rev. 4, Cygna, April 1987.

• 2. "RSS Structural Analysis Report," Rev. 3, Cygna,1983.

• 3. "NRC Presentation, Resolution of Pending SEP Issues, YNPS," Volume 5, Cygna, August 1986.

4. "PRA: A Compater Program for Pipe Rupture Analysis," Version 3.0, Cygna.

5. Newmark, N.M and Hall, WJ., " Development of Criteria for Seismic Review of Selected Nuclear Power Plants," NUREG/CR-0098,1978.

6. USNRC Standard Review Plan, Section 3.71, NUREG-0800, Rev.1,1981.

7. "PCI Manual for Structural Design of Architectural Precast Concrete," Publication ,

No. MNI 121-77, Prestressed Concrete Institute,1977,

8. Cygna Calculation 83033/26/F, Rev. O.

9. Cygna Calculation 86064/3/F, Rev. O.

10. Cygna Calculation 87150/1/F, Rev. O.

11. "INSPEC, A Computer Program for Calculating Spectra", Version 3.0, Cygna,1987.

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

American Concrete Institute, April 1981.

13. " Digitized Psuedo Spectral Acceleration Data for SEP Plants", Memorandum for D.M. Crutchfield, Chief of SEP Branch, from H. Levin, Division of Engineering, USNRC, September 17, 1980

14. Chang Chen, " Definition of Statistically Independent Time Histories", Journal of the Structural Division, ASCE, Vol.101, No. ST2, February 1975.

• These documents have been docketed i

ll111111llltllll111lll11111111 YAN\87150\Sumeval2.rpt i

RESPONSE

1. INTRODUCTION 1.1 Objective The objective of this study is to demonstrate that, under the combined dead and NRC spectra loads, (1) the RPV will . not lift at either the RPV/NST-RSR* connection (upper gap) or the NST-RSR/RSS interface, and (2) it will not slide at the RSR and RSS interface.

1.2 Description of Structure As shown in Figure 1.1, the Reactor Pressure Vessel (RPV) is supported by the Reactor Support Ring (RSR) through 28 lugs. The RSR is the integral top support of the NST. The NST-RSR is in turn supported by the Reactor Support Structure (RSS).

As shown in Figure 1.2, the connection between the RPV and RSR is locked with 28 4"-diameter pins oriented radially. This arrangement allows the RPV to expand in the radical direction. If it can be demonstrated that the RPV will not uplift from the RSR, these pins can prevent the movement of the RPV along the tangential direction. More specifically, the RPV cannot slide along the RPV-RSR interface if there is no uplift. .

13 Potential For Uplift / Sliding Because of the 28 pins, the RPV cannot- slide against the RSR. The RSR, however, is supported by a concrete corbel which is part of the RSS and is unrestrained horizontally and vertically upward.

The sliding of the RSR can only be prevented by the friction force between the steel contact surface and the concrete corbel.

1.4 Previous Analysis The methodologies used in the previous analysis have been reviewed by the NRC staff and their consultants in 1986 during the Systematic Evaluation  !

Program (SEP) [ Reference 3}. I i

" Reactor Support Ring (RSR) is the integral top support of the Neutron Shield Tank (NST).

lll1lllll11111111111ll11111llt YAN\87150\Sumeval2.rpt

l 1

The synthetic time histories used in the ~ previous analyses .are very conservative, as shown in Figures 13 and 1.4. The spectrum of _ the synthetic time-history is generally 15% to 30% higher than the target spectrum. Therefore, these analyses were repeated using less conservative synthetic time-histories which better fit the target spectrum. )

1 I

1.5 New Analysis The procedures used in the new analysis are as follows: j (a) Generate synthetic time-histories to fit the NRC spectrum with 10%

damping.

(b) Perform a nonlinear time-history analysis of the RSS.

(c) Perform a linear time-history analysis of. the RPV and demonstrate that the RPV will neither uplift nor slide.

l (d) Generate ARS. j l i l

l l

l.

i I

)

l l

l 111lllllll111111111111111lll11 YAN\87150\Sumeval2.rpt

s  !

• .".O,,.C sm :".,:".M . ~ " . ...,

. J'* % - d *MY. 1.C , , , , , , , ,

- L'1 - a=--*=-==

(RPV) . .

-[REACTORPRESSUREVESSEL t

.RSR

~

ar -

-.4.' . ge

. ; ., J. , , . '. . ? ,' .;

, p 3 0 .[

MAIN' COOLANT *M -

. y .{

[- LOOP (MCL) ,INNERj [ '(

,* $'wt.

. NST 'F,jt ".

J .'" }p -

' ;' . ' ~. . WALL i i

.$ MCL3 ./ .

-.0"' . das ". 16 ,

j

• *{ DE T AI L. D -

. F r :. _ A k_%ss ,. dt I < '.!, ) REACTOR **^*'*"* OUTER 1

, L_'__n' V 4,/

g _,u J SUPPORT RING (RSR) NST j

, ", .1 ,s w <-

, ({ -.m RPV 8

i A r-- -

BLDG '

y)C 1' -$- , g h h ',

Y 4 <. 's!

pk' J '

n'sa fTN

?

n .c.

/

!*=== Q .

dj . 1 T. 1, , ,

)

!,, . ... . i  ?, l  ; l V

.L 4 1

, ,,; -g,E' .

d ,

II" ~

y NEUTRON E.IIELD .g . I I.

,3 TANK (NST) -5 puw a m t' ms. w.. .e

/ ~

- 9, .

i T , ,

• 4

)

e  ! 'l ,

y l

)I.<. e--i=J

. i i i MI. M b- g kgibg Ji 2). .* N

~

.'. 'u* c'

.-  % anc N:.s

.y.......,,- e i .'.q.~._

p - - .., j-mM N es, 6 e

.. .og

. 1 i . *iFET10N AL ELEVATION g.. . .L.OWER SHIELD TANE l

• . un \

l Figure 1.1 Reactor Pressure Vessel (RPV) and Neutron Shield Tank (NST) 4 YAN\87150\Sumeval2.rpt I lifilillllllllllilllllililllll i

l

_ _ - _ _ J

1. - w 17 6 -}=t l

I'

.. , .g .

m_4 O fl* I;I J

@g , lb:t =i l1

• f

. _~

.I _

v -l J+ l il '

2 H~- 1 ;

J i.s__2

.,!.i,,-

l t s

M-Il .\

lj f(.f j f , lj,l  :

a i n

'd s

{l. l,1--**

E g:  ;

1_ _sf, I

~

n . '4s . si -

i,

, I at t.y g ..

+

g.

w\8f -

i . 3 .. ,

. fr$ [1 {lf% N'y.(jj T.

y l.-

(S  ;

/',p 1 ( l;l. l:

l I .

y.

• 1!'

N

.?

's f.rs 3p'!Is I

/ N - E \d &pt g

A.t' . . ' 4> .

s. \ . 32 c _. 18 -;,:

d k\_ % A7.3, \ r )' -- .

.j

\ -

g; gl-- --[,l ei y ,- . +

2

&. + '

.l. y/ L.j)l

^

\

l _

L_

/ ,., \. .. s, ,

"y,,.j

y. -  ! -. ..- .g ..

\s%:/./

/, , l[ , a,

.\ ' si l$ l M* hNQ' Y s 7c.f . jl - ../ .

. i j[!!

f, 1 3

{}

si j Figure 1.2 Reactor Support Detail i

l 6

A[tj g ,

YAN\87150\Sumeval2.rpt ll111111ll1111lllll1ll11ll!I11 f 1

m

e.6e l l

MilflCIAL IIM MilleRI l towlerius unc settlaun '

elIN LANrlNs Mlle e e.e7 e.so . . . - . .

l  :

G e.no . . . . .

n E / \ i E / \

5 / \

e. so . - . . . . . . . . . . . . . .

...j. .

W / s e

z / N N b,e.no f'

" p i

/ NRC SPtCTRUM e,,e . . . . . - . .

e.i e.: e.s e., e.s e.6 s.e a.e e.e e. s. 6. no. re. so. go. so. 6e. iso.

FRE00ENCT IHZI

• Figure 13 Response Spectrum of the Synthetic Time-History Used in Previous Analysis YAN\87150\Sumeval2.rpt lll1!!11111111111111111111!I11

_ _ _ _ _ _ - ~

i l

M. ..Wla.!#.4.!!=

.ai. .. .n.

SYNTHETIC TIME-HIS"ORY%

h.- / ' %

g .. . s-

- \

/

E ,,, / \ -

U /

N

/ N N / N c

e .. -f m

/ IIRC 5pectrum

.. . . ... .. .I

• ".., .:. ... .l. ..\..i ' ' i.. ... . . ' . i i i ' ' i.. d. d. d. d. d. '=.

FPICOUENCT IHZ)

I Figure 1.4 Response Spectrum of the Synthetic Time-History Used in Previous Analysis YAN\87150\Sumeval2.rpt lIlll!Illilllillllilllllilllll 1

l l

d

2. SYNTHETIC TIME-HISTORY GENERATION Two statistically independent ;p sthetic time-histories were generated to fit the NRC spectrum with 10% crita damping ratio. The 10% damping NRC spectrum was converted from the 590 A iping NRC spectrum using the scaling relationship provided by Reference 13 as L k vs:

PSAx% = PSA5% x 10 (CTx(new damping (x) .05))

where PSA is spectral acceleration and CT si a period dependent coefficient as defined below.

Period CT 0.04 0.05 0.065 -0.290 0.08 -0.600 0.10 -0.904 0.13 -1.270 0.20 -1.700 0.30 -1.990 0.40 -1.950 0.75 -1.810 1.0 -1.960 2.0 -1.600 Statistically Insignificant Coefficient, Use 5% PSA Value The generated acceleration time-histories are shown in Figures 2.1 and 2.2. The correlation coefficient between these two time-histories is 0.058 which demonstrates their independence. This correlation coefficient was calculated per Reference 14.

]

The time-histories were generated in accordance with SRP 3.7.1 [ Reference 6}

SRP 3.7.1 specifies that no more than five points of the spectrum obtained from the synthetic time-history should fall below the design response spectrum. Those five points should be no more than 10% below the design response spectrum.

The comparisons of the design and synthetic spectra are shown in Figures 23 and 2.4. As shown in these figures, the spectra of the generated time-histories have 4 no more than five points below the design spectrum, and these points are not I more than 3% below the design spectrum. Therefore, the generated time-histories satisfied SRP 3.7.1.

i I

T 1111llllll111lllllllll111lll11 YAN\87150\Sumeval2.rpt )

VRL

.2000

.1600 e .1200 z l' .< [ il I r g .0B00 j i

). ,

j[ ,

J

,i ,' 1 i i y

t

.0'100

{I i j.) i

{ j Il

}

! (3 t } I l d A l l, ) ' ll l I I LI II l'  !

I A 2 E

.0400 l'fq' jf{ j ((}'I Y/

d

.0800

)

1

.1200

. 1600

.2000 0 2.00 't.00 b.00 B. 0D 10.0 1.00 3.00 5.00 7.00 9.00 TIME NGC1 10% DAMr!NO CPCCTRUM - FINAL i

Figure 2.1 Synthetic Time-History to be Applied in Horizontal Direction YAN\87150\Surcieval2.rpt n ,

O YAL

.2400

.2000 tn .1600 z

$ .3200 g Ed i h I

.0000  ;

la 0400 W 1  !

l V l

AA .

) i i , /

.0400 1( r;

.0800 i

.1200

.1600 0 2.00 4.00 b.00 0.00 10.0 1.00 3.00 5.00 7.00 9.00 TIME NRC3 10% DAMP!NO SPECTRUM - FINAL 1

Figure 2.2 Synthetic Time-History to be Applied in Vertical Direction YAN\87150\Sumeval2.rpt l lll11111!Illll111111ll!!!lllll IL

11 NRCl 107. ORMPING $PECTRUM

, PGA = 0.190G 21 SPECTRUM GENERATED FROM ACCELERATION TIME HISTORT DURATION = s fl. 000 ORMPING = 0.100 1.00 0.90 -

0.80 -

0.70 cs 5 O.60 E

$ 0.50 d

8 a:

0.90 SPECTRUM OF SYNTHETIC TIME-HISTORY 0.30 -

0.20 -

NRC SITE SPECIFIC SPECTRUM 0.10

-0.00 ' . .

o. . _. _

PER]CO IN SEC.

Figure 23 Comparison of Horizontal Spectra YAN\87150\Sumeval2.rpt 11llllltlll1111111!!!I1ll11lll

_L

11 NRC3 10*./DAMPING SPECTRUM PDA = 0.190G

2) SPECTRUM GENERATED FROM ACCELERATION TIME H]STORT DURATION = 10.000 DAMPJNG = 0.100 1.00 - .

0.90 .

0.00 -

0.70 e

5 0.b0 -

E

$ 0.50 .

d O

• SPECTRUM OF 0.40 SYNTHETIC TIME-HISTORY -

j 0.30 0.20 NRC SITE SPECIFIC _ .

SPECTRUM 0.10

-0.00 ,

8 a e s l

• o 8

e o a

o 9 8 J J

J PER100 IN SEC.

)

Figure 2.4 Comparison of Vertical Spectra I

13 4[ ;, , YAN\87150\Sumeval2.rpt lllllllll111111!!!!I1lllllllll l

1 k

3. NONLINEAR TIME-HISTORY ANALYSIS OF RSS 3.1 Introduction A nonlinear time-history analysis of the RSS was performed using the PRA computer code [ Reference 4} Relative displacement time-histories at significant RSS locations were output.

As will be briefly discussed in Section 3.2, the RSS was. originally analyzed i using linear and nonlinear models (References 2, 8] to demonstrate that the RSS is adequate under Yankee Composite Spectra (YCS) and NRC spectra loads. Both these analyses and PRA computer code have been reviewed and accepted by the NRC staff and their consultants during SEP.

The nonlinear analysis was repeated using the new synthetic time-histories and slightly modified model. The differences between the existing SEP nonlinear model and the new model will be discussed in Section 33.

32 Original Imear and nonlinear analyses of the RSS The RSS was originally ar.salyzed using the linear model (Figure 31) for the YCS (7% damping) and NRC Spectra (10% damping) loads. However, the results of the linear analysis (using response spectrum method)-indicate that the RSS columns yield under the NRC Spectra load. The adequacy of the RSS had to be assessed by the extent of the nonlinearity such as the total number of the inelastic excursions and the amount of the plastic hinge rotation. Consequently, the nonlinear analyses using the nonlinear RSS model (Figure 3.2) and the very conservative synthetic time-histories (Figures 13 and 1.4) were performed. The original linear and nonlinear analyses are briefly discussed below.

As shown in Figure 3.1, the linear model is three dimensional. The column area was modeled using the transformed area calculated as follows. {

Transformed Area = Ac + As (Es/Ec) where Ac is the concrete area, As is the steel area, l Es is the modulus of Elasticity for steel and i Ec is the modulus of Elasticity for concrete.

Under a NRC 10% damping spectra load, the maximum displacement at the top of interior column is 133". To study the effect of damping ratios, the original nonlinear analyses were performed using 7% and 10% damping. As shown in Figure 3.2, the original nonlinear model is two-dimensional. The column area was modeled using the gross area (Ag = Ac + As). In addition, the modulus of elasticity of the columns was adjusted to fit the initial slope of the moment-curvature diagram of the cracked column section.

14 A M ,j, YAN\87150\Sumeval2.rpt ll1ll11111ll1111111111111111ll

i

. \

)

Consequently, the vertical frequency of the nonlinear model is lower than l that of the linear model. However, since the nonlinearity of the RSS is l primarily controlled by its horizontal stiffness and capacity, the discrepancy in the vertical stiffness will not affect the nonlinear horizontal responses such as the number of the inelastic excursions and the amount of the plastic hinge rotation. Therefore, the original nonlinear analyses were accepted by the NRC staff and their consultants during SEP for the evaluation of the )

RSS. The horizontal ARS for piping and equipment analyses were generated using the results of the nonlinear analyses. However, the vertical ARS were generated using the linear model which has the realistic vertical frequency.

1 At the top of interior column (Node 30), the maximum lateral displacement is ]

1.41" and 1.54" for the 7% and 10% damping analyses, respectively. The l maximum plastic hinge rotation is 0.00434 and 0.00481 radians for the 7% and 10% damping analyses, respectively. Therefore, the nonlinear response was not reduced by the increase in the damping ratio.

33 New RSS Nonlinear Analysis Since it was demonstrated by the SEP analyses that the RSS will respond inelastically under a NRC spectra load, the linear analysis need not be repeated. The RSS model (Figure 3.2) used in the new analysis was nearly identical to the previous SEP nonlinear model except its vertical stiffness was adjusted to be consistent with that of the SEP linear model as follows. i (a) The column area of the new model was increased to:

)

Column Area = Transformed Area x (Ec/ Reduced E used in the original model)

(b) The upper part of the RSS was modeled with the rigid beams. In thg original nonlinear model, these rigid beams have an area of 100 ft-which is not large enough to represent the rigd characteristics. In the new model, this area was increased to 1000 ft . This modification has '

minor effect on the vertical stiffness of the model. J l

Since the RSS yields under the combined dead and NRC spectra loads, a 10% )

critical damping ratio was used. The justifications for using 10% damping are as follows.

(a) The 10% damping was selected based on the recommendation of NUREG/CR-0098 (Reference 5) which allows a 7% to 10% damping for the reinforced concrete structure with stress at or just below yield point. Reference 5 states that "the lower levels of the pair of dampings given for each item are considered to be nearly lower bounds, and are therefore highly conservative; the upper level are considered to be average or slightly above average values, and probably are the values that should be used in design when moderately conservative 15 A [ ;, ,, YAN\87150\Sumeval2.rpt lllll1lll1ll111111lll11ll11ll1 m

, a estimates are made of the other parameters entering into the design criteria".

(b) From the previous analyses, the effect of damping is not significant in reducing the nonlinear response of the RSS.

(c) . The results of the new nonlinear analysis indicate that the nonlinear effect is insignificant. The maximum horizontal displacement at Node 30 obtained = from the nonlinear analysis (1.25") is approximately the same as that obtained from the linear response spectrum analysis (133"). The maximum plastic hinge rotation of the column is small i (0.0037 radians). Therefore the energy dissipated through the plastic hinge rotation is small. Compared with the original nonlinear analysis,-

the new analysis produces noticeable reduction in the maximum lateral 3 displacement and plastic hinge rotation. These reductions are j primarily due to the better fit of the synthetic time-histories used in j the new analysis.  !

(d) Although there are few inelastic excursions, many excursions have the maximum stress near the yield point which justify the use of 10%

damping.

The model PRAthecomputer damping. code The auses and a viscous damping coeffeients matrix=[C]d were selecte a[M]so+that

[K) the' to first -

)

two modes of the RSS have 10% damping (Figure 33). '

The synthetic time-history shown in Figure 2.2 was generated to fit the NRC spectrum. In accordance with Reference 1, this time history was scaled to 2/3 to be applied in the vertical direction.

Since the RSS model is two-dimensional, the horizontal ground accelerations can only be input in one direction. To consider the effect due to the input in the orthogonal horizontal direction, the horizontal accelerations as shown in Figure 2.1 were multiplied by 1.1. This 1.1 factor has been reviewed and accepted by the staff during SEP.

3.4 Absolute Acceleration Time-Histories i

The ground displacement time-histories were combined with the relative displacement time-histories generated in the ' nonlinear analysis to develop the absolute displacement time-histories at the RSS locations.

The displacement time-histories were double differentiated to develop the acceleration time-histories as shown in Figures 3.4 and 3.5. The acceleration time-histories shown in these figures are for Node 32 which represents the elevation of the RSR.

YAN\87150\Sumeval2.rpt 14lllllll ' lllllll!!!llllllll i

.l

)

i l

i 114 i RIGID BEAM 113 I 101 102 103 "201 12 I 7 EXTERIOR COLUMN lil i 1 < 202

^

-203 L _.

08 104 ,, 2 o204 "205 106 A 4 a 5 '206 INTERIOR COLUMN n'

i Figure 3.1 RSS Linear Model YAN\87150\Sumeval2.rpt

1 l

I l

l l

i I

I o 222.u7' .3 , 3, _

l bn RIGID BEAM TK 33 -

Et 32-

)

i RIGID 31-BEAM g39.966' NODE #7--

\ 14 21

)

30 g 27.637' 29 INTERIOR EXTERIOR 7

COLUMN COLUMN %

i i

D *O 7FfT mb/gm/ 757f

""b'" l Figure 3.2 RSS Nonlinear Model i

l 18 i g4[ 3, ,

YAN\87150\Sumeval2.rpt l 1111ll11111ll1111111111llll111

-)

1 1

l m  !

g y 12

.i 10 a e /

x /

6 l r J

" 4 )

,a J

• 2 E

I i i i 1 2 4 .6 6 to - 8 FREQUENCY (Hz) 1 1

Figure 33 Input Damping Ratio for RSS Nonlinear Time-History Analysis 19  !

4[ , , , YAN\87150\Sumeval2.rpt llllllllll11111ll1ll11ll1111ll L __ _ _ _ _ _ _ _ _ _ _ _ . . . _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ _

vnt 100 8

g 80 1

~

60 8 r\

- n 0

'10 1 g ['

A '

y Pil !l t}

U b

10 j {

-20 '

/ 1 1 1

-90 v k

-60

-80 j

-100

.0100 2.00 4.00 5.99 7.99 9.98 1.01 3.00 9.99 6.99 8.98 TIME PBD DIDP & ACCEt T-H AT 32X 4 32Y 1

Figure 3.4 Absolute Horizontal Acceleration Time History at Node 32 of RSS Model I

20 l A[e , 3 YANt87150\Sumeval2.rpt 3

lll11111111!!!l1111111llllllll ,

- --------_-_---__lb

vnt 62.5

^

N l g 50.0 m

%37.5 5O 25.0 I

~ l-J Imi .s

/

/

I I

L f 7 d  ; i lf I f I'y' u g , 3 q 6 i r l equ ,

i ,,

fI jl I

l l

-12.5 - -

i 1 q

l I l

-25.0 q i

-37.5

-50.0

-62.5

.0100 2.00 4.00 5.99 7.99 9.98 1.01 3.00 9.99 6.99 B.98 TIME ADO DIGP & ACCEL T-il AT 32X 4 32Y t

Figure 3.5 Absolute Vertical Acceleration Time History at Node 32 of RSS Model

____m

{eg'M lfj YAN\87150\Sumeval2.rpt 111111111ll11111111111lltll111 l

l

h

4. ' INEAR TIME-HISTORY ANALYSIS OF RPV/NST L ,

4.1 Introduction A linear time-history analysis was performed for the RPV/NST to demonstrate that uplifting and sliding do not occur under the combined ,

dead and NRC spectra loads.

The RPV/NST model (Figure 4.1) used in the previous analysis [ References 3, 9] has been reviewed by the NRC staff and their consultants during SEP.

4.2 Model Changes The previous model used nonlinear gap elements. The gap element is equivalent to a spring which has rigid stiffness under compression but has zero stiffness under tension. Since it was demonstrated in the original SEP analysis that the gaps were not opened, these gap elements were replaced by rigid springs in the new RPV/NST model (Figure 4.2) so that linear analysis can be performed. Note that the rigid springs replacing the lower gap elements were deleted becasue connecting a node through a rigid spring to the ground is equivalent to connecting this node directly to the ground.

The structural damping was modeled using a damping matrix [C} = a[M) +

[K) to model the damping. The a and coefficients were selected so that 3% damping can be produced at the fundamental frequency (9.5 Hz) of the RPV model and at 40 Hz as shown in Figure 43.

i 43 Results The vertical force time-histories of the springs representing the upper gaps (between the RPV lugs and the RSR) are shown in Figures 4.4 and 4.5.

Since these two springs are always in compression, the upper gaps will not open.

The time-histories of the vertical reactions representing the lower gaps (between the RSR and the RSS) are shown in Figures 4.6 and 4.7. The reactions are always upward, demonstrating that the lower gaps will not open.

The required friction coefficient to prevent the RPV/NST from sliding was calculated as the total horizontal sliding force (reactions at Nodes 14 and 15, Figures 4.8 and 4.9) divided by the total compression at the lower gap.

The time-history of this coefficient is plotted in Figure 4.10. From this figure, the maximum required friction coefficient is 035, which is less than 0.4 and 0.55 recommended by PCI and ACI, respectively, [ References 7,12]

for the friction coefficient between steel and concrete. Therefore, the RPV will not slide under the combined dead and NRC spectra loads.

22 i y;,, YAN\87150sSumeval2.rpt lll!I1lllllll11111llll11111111 i O

1 l

1

\

The maximum concrete bearing pressure on the corbel supporting the RSR is I 039 ksi which is smaller than the all wable bearing pressure of 0.7 fc = 2.8 7

ksi.

l M -

' 23

_l YAN\87150\Sumeval2.rpt

. 3

'N I.D a i Il!I111lll11lll1111lll11111111

, /- ~t .l 3

, s .

t g

( s I i GFD(!tlD K E#O

• YMCL. - >

x GMCL g .3-,[f MCL h ,

1 \ h GROUND GROUND RSS (MCL Support) ,

g ,

j

~

XMCL ,

BEAM (Tan.ID) n BEAM (RPV)

MRPV [ f h '

t BEAM (RIGID) j .

Y i

/

@/ XRSR GRSR  % RSR

'YRSR- %g

+

fi///

le l

l s e & riitt GROUND (RSS LEDGE) g g GROUND I I , h, '

BEAM (NSIr p p P3AM (NST)

1. /

M "NST (inne r)

NST (outer) h LUMPED MASS \

UbE h = NODE NUMBER Y = TRANSNATIONAL SPRING

= RIGID REGIOth = ROTATIONAL SPRIFG

"""= BEAM ELEMENT Figure 4J Original (SEP) Nonlineer RPV/NST Model with Gap Elemen'.s p m.--- --

24

[eltilfj g

YAN\87150\Sumeval2.rpt Ill;ililllllillllllllllllllll!

l

,1 .

- n- r

)

's -)

l D, ,

4 . ,.

3

+

/  ; - '

3 4

/

\

't . ) j '

.i + 2

. , . 5

, 1 n<

i'  ;; 2 ELEMENT NUMBER (TYP.)

.\ '

' l MCbMASS u u.

I j , /. ( Ky (MCL)

'( .

RIGID a

l N2 ,[16 4' /

~ "

/ /"

ADDITIONAL MCL E9,(MCL)3 1 L9 i HORIZONTAL MASS I

f ( h' 3 i I, 4

. . /x K (LUG) v <

-a I-71.25" 3 -

/

4 ~ 13 11 h

• 4 RIGID (UPPER GAP)

%I .' 10 ef;,cp,5 12

$ Rib 1D BEAM -

RPV BEAM n'

m

/ Ke (RSR) }6 5 RPV MASS Kx (RSR) l y (RSR)

RPV SUPPORT RING (RSR)

1. .g

_._ _ _s _ _ _ = 15 0

i I

7/g 7 O

107" l l12l @ l

(

OUTER NST BEAM ~ INNER NST BEAM OUTER NST MASS INNER NST MASS 9 '8-NODE NUMBER (TYP.)

i Figure 4.2 New Linear RPV/NST Model YAN\87150\Sumeval2.rpt lililllllillllllllllllllllill!

I 1 ______ ___ 2

j m

Y6 L

Q C4 6

g3 g A - 7 E2 a

hI b

I I I t i 10 20 50 40 50 l TRE Q UENC Y ( Hz )

I Figure 43 Input Dumping Ratio for RPV Time-History Analysis i

26 '

gT[P; , YAN\87150\Sumeval2.rpt 3

lilillllllllllilillilillllllli

i l

I MIPS l, 1

-390

-360 l l

-300

)

I l

-400 I

-420 j l j t

l l

i i 9 l I

(hlj 1

!ij lll l

}D l

' -990 g ,

UGFV ih l 11  ! n

-960 p i a , e my G I l

!' li lj I

-480 I

> l l

1

-500 l 1

r 1 -520 1

-590 j 0 2.00 4.00 6.00 8.00 10.0 1.00 3.00 5.00 7.00 9.00 TIME RPV STABILITY ANALYSIS (MODEL 2D9 LINEAR)

Figure 4.4 Vertical Force of Spring Element 8 Representing Upper Gap 27

,4 [ t=j , 3 YAN\87150\Sumeval2.rpt lilllilllllllllllilillllllllli

- - - - _ _ _L

I Ii

]

i KIPS i

-Swa 1

-360

-380 l

-400 '

l 9

-420 i i

l i I

!g ,

!h. I '

k 440 ,

([l'i' .i d i  !

3 UGFV i n It i I l I f \ < I ,

-%D

($ ]' '

]

,f

! i '

!I

-980 j t 1

-500 -

i

-520 l -510 0 2.00 4.00 6.00 8.00 10.0 1.00 3.00 5.00 7.00 9.00 TIFE APv STABILITY ANALYSIS IMSDEL 209 LINEARI 1  !

Figure 4.5 Vertical Force of Spring Element 9 Representing Upper Gap ,

Nbb b

• lilillililllllllllllllllllllil L_

MIPS 9

675 650 625 600 ,

i 575  ;

j l lI' }'f

.A)I

, , 1 550 - r ,!: ll!  ! I l' ,

i t fl j:

i

. a l f i l'I isi, Fy y

e!

L i i

,j S 525 , i , i W1 ,,

\ "

\

[\

i' i l 500 i 475 s 450 425 0 2.00 4.00 6.00 8.00 10.0 1.00 3.00 5.00 7.00 9.00 TIME RPV STABILITY ANALYSIS IMODEL 209 LINEAR) l Figure 4.6 Vertical Reaction at Node 14 Representing Lower Gap 29 y;,, YAN\87150\Sumeval2.rpt lll1lll111lll1ll!I111111111111

. _ _ _ _ - _ _ - - _ _ _ _ _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _l J

l e

K!PS 675 650 625 l

600  ;

I I

! l 't

{ j 575 l i

f , kJ

. A L II i . O I 1 ll kl  ! h. ! bI

! l l f l)'  ! i FY I  ! i f

] , t I l l l

525 , ,,

I b

1 500 j 475 s 450 425 1 0 2.00 4.00 b.00 8.00 10.0 1.00 3.00 5.00 7.00 9.00 TIME I

PPV STABILITY ANALYSIS (McDEL 209 LINEAN l

1 i

Figure 4.7 Vertical Reaction at Node 15 Representing Lower Gap 30 g[ ;, 3 YAN\87150\Sumeval2.rpt 111lll1lll111llll11111ll111111 l U

KIPS e e+

200 l

160 p l

120

. l s I h\ '

h l 0

f 0

.A J n' $

1 a

l r "

0 y [

-40 k

l

-80 0 y

-120 s j

-160 f

-200 0 2.00 4.00 6.00 8.00 10.0 1.00 3.00 5.00 7.00 9.00 TIME RPV STABILITY ANALTSIS (MSDEL 2D9 LINEARI Figure 4.8 Horizontal Reaction at Node 14 YAN\87150\Sumeval2.rpt ll1111lll11111lllllllll1811111

_ _ _ _ _ - _ - _ - _ - x

K1P5 200 l

160 ,

l 120 80 40 h\ h b

( f )

f 0

"vi v' v g

-90 l7 I

-80 l

l

-120

-160 f

I

-200 O 2.00 4.00 6.00 8.00 10.0 1.00 3.00 5.00 7.00 9.00 TIME I

RPY STABILITY ANALYSIS (MODEL 209 LINE AR) 3 l

Figure 4.9 Horizontal Reaction at Node 15 32 tT[ ; 6 3 1 YAN\87150\Sumeval2.rpt llllll1111llllll111111llllltll L

MU

.4000

.3200 t

.2400 l

k

.1600 l

f

.0800 A k k f i I h A "'

'V ' Y it

.0800 g

l

{

1. 4

.1600

.2400 t

I

.3200

.4000 0 2.00  % 00 6.00 B. 00 10.0 1.00 3.00 5.00 7.00 9.00 TIME I

RPV STAB]LITY ANALYSIS (MODEL 2D9 LINEAR) 1 Figure 4.10 Re-quired Friction Coefficient to Prevent RPV Sliding YAN\87150\Sumeval2.rpt llllllllllI1111111111lllll11ll

\

5. ARS GENERATION ARS were generated at the RSS and Main Coolant Loop (MCL) nozzle locations using the time-histories calculated in Sections 3 and 4 Main Coolant Loop using '

INSPEC Computer Code, Version 3.0 [ Reference 11} The earlier version (2.0) of I INSPEC has been reviewed and accepted by NRC staff and their consultants. The only difference between Versions 2.0 and 3.0 is an option has been added to Version 3.0 to allow it to generate the spectrum with PVRC damping.

The ARS were generated for the analyses of the piping systems and equipment supported by the RSS or MCL Some of the ARS are shown in Figures 5.1 through 5.4.

In accordance with USNRC Regulatory Guide 1.122, the ARS used for piping and  !

equipment analyses were broadened by 15% (broadened ARS are not shown).

l J

l I

i J

34

! llllllill111111llll1llllllll11 YAN\87150\Sumeval2.rpt

\

s

1. 20 . . . . _ . _ _ . . _ _ _ , . . . _ . . _ . . _ . . . . . . , , . . _ _ _ . _ . _ . . _._

P$$ NeuLlufRA MODEL 102 DAMP RPy Listan m00tL u!TH St DANP NMC $PICf AA SEllNIC LORD A55 EL.1079'-8' Hom!!0NTAL SR$

DanPING N-vil 1.00 "

I C

g e.eo ._ - . .

E E  :

E

$ e..o U t W  ;

a F E E y e. ,o g . . _... __ .. . . . . . . . _ _

c h \- wwsaw_ ~

e.no '. - _ . . _ _ _ . _ _ . _ _ . __. _

t

* oIi ' o$ e'.i e'. c'.s e'.i ' ' ' i ', 's ',' 's - i ' ' ' i o --W 4- e a * ' 'ioe FREQUENCY (CPS) f CTG ER T S RV CE Figure 5.1 Horizontal ARS at RSS Elev.1079'-8" (Node 32) i 35 gM. 3 YAN\87150\Sumeval2.rpt 1llll111llll111111111111111111 i

4 0.59 . . ._ . _ _._ . _ _ - - . . . - -- . . - - - - - - -

RSS NONLINtan n00EL 102 DRMP RPu LINE AR M00tL Mifn St DAMP NRC SPECTRA $[l$Mic LOA 0

, ASS EL. 3079'-8' VE AtlCAL ARS

. DAMPluG es.gil 0.15 g o.n . _ _ - . _ _ _ _ _ . . . _ . _ . _ . . . _ . ._

E t E  :

"i o.2, ; _ _ . . _ _ _ . . . _ _ . . _ . . _ . _ _ _ - . . _ . - - _ _ ~_

b W

d E

g o.i.E _ _ _ _ _ ._ .._ . . _ . . _ _ _ _ . . . _ _ . _ .

o.oo {

b I

o. i c'. ~o '. s o'.i e'so'.W' ' ' i ' 'e 'a ' ', 's '. ' ' ' ' i o s A ' A e a ' ' 'ioe -

FREQUENCY (CPS)

PROGRAM INSPEC CTCNR ENERGT SERVICES Figure 5.2 Vertical ARS at RSS Elev.1079'-8" (Node 32) 36 g[ ; , , YAN\87150\Sumeval2.rpt lilllllhlllllllllllllllllllli L__ _ .. . . . . .. - - . .

1 l

l l

l

(

l . 20 ,. _ . _ . . .

--- - - - - - ~ - - - - - ~ - ~ ~ ~ - -

..,-.---..-.-..f'-~-- RSS NONilNERR MODEL tot DAMP hPV Lluten MODEL ml1H St DAMP mRC SPECTRA Sil5MIC LORO I'N * -- - - - - - _ . . . _,..,,_ __ ,

E o *0 : - _ _ _ . - . _ . _ _ . _ . _ _ _ _ , _ _ , _ ,_

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

g E

5 u

' 68 F' - -

~ .____ _ _ _

u d

5 i g*ap- . _ - _ _ . . _ _ . _ _ t. - - - . . . - . .

sn w.-e~

o.e6 : ._ __ _

~ -

0.00 '

~ 0 ] 0,so's' ' ' y a , , ,,

i a s 0*I 02 0. 9 # ' '

20 33 50 '60 joo FREQUENCY ICPS)

PROGRAN INSPEC ,

CTCNR ENERGs sgnygcg5 l

Figure 53 Horizontal ARS at MCL Nozzle l l

l 37 I

' lJL A YAN\87150\Sumeval2.rpt lillllilllllllllllllilllllllli I

l i

i I

1 i

i i

I t . 20 . . . ._. . - - . _ _ _ . . . _ . ,. . . _ . . _ _ _ . . _ . _ . . . , , _ . . _ . . _ _ _ . _ _ .

RSS WONLINEAR M00fL 80% OmnP hPV L int an MODE L u t tM 3 DeMP WRC SPECf RA SEtSMit LORD

, RPV NCt N0ffLt VERflCAL Amt

. DanelmG u.gt) g e.eo t.

E i

= -

E .

w .

J 0. 60 . . _ . _ , . - _ _ . . . . _ _ _ _ _ _ . . .. . , . . _ . . _ .

U .

e  :

5' i 5 t g e.nop...___....___.-.____.._._ _ . . ._. . . _ _ . _ _ . . . _ _ .

,n .

. s L %_  !

0.20 . . _ . - . _ - _ , _ _ . . _ . _

i vt ,

e . co,h . ._ . g , j, ,.y., ,3. ,,, ,, _, . , .

FREQUENCY (CPS)

PROGRAM INSPEC CTCNR ENERGT SERVICES j

Figure 5.4 Vertical ARS at MCL Nozzle u )

g [ ;, i_ YAN\87150\Sumeval2.rpt lilillllillllllillllllllllllli

i

}

', - ~ ;~ .

CONCLUSION: -

The RPV/NST will neither uplift nor slide under theTombined dead and GRC seismic spectra loads, Therefore, the linear RPV analysis and the ARS generated using the .:

time-histories generated by this analysis are valid and the ARS are acceptable for use in ,

the piping analysis. j 1

39 g { ;, , , YAN\87150\Sumeval2.rpt llllllll1111lllllll1llll111111 I

l