ML18046B160
| ML18046B160 | |
| Person / Time | |
|---|---|
| Site: | Palisades  |
| Issue date: | 11/30/1981 |
| From: | Fishman E EDS NUCLEAR, INC. |
| To: | |
| Shared Package | |
| ML18046B157 | List: |
| References | |
| 02-0660-1089, 2-660-1089, NUDOCS 8112220617 | |
| Download: ML18046B160 (52) | |
Text
-
1-I 02-0660-1089 Revision 0 Page 17 r,
3.0 PUMP .MODEL DJ!:SCRIPTION 3.1 General The computer analyses performed in evaluat-ing the Service Water Pumps were carried out on two distinct mathematical models.
The .1!.rst model (Figure 3 .1), utilized the EDS Computer Program "SUPERPIPE" and was a
- . "stick" model which represented the pump motor, discharge head, column and lower bowl assembly as a series of pipe (or beam) members (Reference 4).* Each component had appropriate c.ross sectional properties computed in hand calculations (References 6 and 7) which reflected the weight, and section modulus of the corresponding part of the pump. This "stick" was supported at the base plate elevation by an anchor with stiffnesses specified in each of the three translational and three rotational direc-tions. Additionally two lateral supports located at elevation 583'-2 3/4" were modelled with stiffness specified in their axial directions. Table 3.1 is a summary of the materials, cross sectional proper-ties, component lengths and lumped and distributed weights. A more detailed ex-planation of this model is given in the following sections.
The second ~odel (Figure 3.2) was developed using t"he computer program "ANSYS" which is maintained by Swanson Analysis Systems, Inc. This model was developed using quadi-lateral shell elements to represent the base plate support. *A portion of the pump column was incorporated into this model to allow for a more accurate transfer of load.
Further discussion of this model is given
( .
in Section 4.0.
~
I EDS Nuclear
_~::.::
02-0660-1089 Revision 0 Page 18 L
r: 3.2 Motor Unit r The motor which drives the Essential Service Water Pump is a General Electri.c vertical hollow shaft induction motor (Model SK6328XC185A) (Reference 8). The l overall length of the motor is 66.75". In
.the mathematical model the motor was repre-sented as a pipe with its outside diameter equal to 24.5" which corresponds
- l. to the outside diameter of the ~pper dis-charge head flange to which the motor II . attaches. This pipe member was 65.75" long. The motor mounting plate was also modelled as a pipe section, *£or the remain-ing l". The moment of inertia of the 65.75" pipe section was calculated from the value of static deflection at the motor center of gravity for the case when the motor is bolted to a rigid mass and considered a horizontal cantilever beam. This deflec-tion was given as .0~63 inches (Reference 8). Therefore, Where I = moment of inertia
! = distance fro~ fixed base to center of gravity E = modulus of elasticity W = total weight of motor
~ = static deflection The moment of inertia was calculated to be 151 in4
- Thus, with the outside diameter set at 24.5", the wall thickness was back-
["
calculated as .026" to obtain the proper moment of inertia.
EDS Nuclear
'W*ffl"MH'*Y 02-0660-1089 Revision 0 Page 19 L
[ .*
The entire weight of the motor unit, 3500 lb, was lumped at the location of its 1* center of gravity, 28" up. from the base.
The motor mounting plate was assumed iden-tical to the upper discharge head flange and was modelled as a pipe cross section I with an outside diameter equal to 24.5" and a wall thickness equal to 6.25". Since the entire motor weight was lumped at its I. center of gravity, the individual pipe components comprising the motor unit had no II weight per unit length.
!. 3.3 Discharge Head Assembly The discharge head assembly shown in Refer-ence 9 is bolted to the motor unit at its upper end and is bolted to both the upper most pump column flange and to the base
.plate at its lower end. The discharge head measures 37" in length. The top head flange comprises the upper inch. This flange was modelled with the identical cross-sectional properties of the motor mounting plate to which it is bolted. The flange member has no weight per unit length. Instead, one-half the total flange weight of 102 lb was lumped at each end of the member. The next 1.5" of the discharge head assembly (re-ferred to as the "upper" section) was modelled as a pipe with an outside diameter of 24" and a wall thickness of .441". This thickness was back-calculated from the section's moment of inertia value. The moment of inertia of this section was taken as the sum of the inertias for the 24" diameter, .375" thick outer discharge head shell and the four .375"x5" stiffening ribs.
EDS Nuclear
r.
I 02-0660-1089 Revision 0 Page 20 The section modelled next was 9.5" long (referred to as the "window" section), and represented the portiori of the discharge head containing the two 100 windows (cut-outs). The effect of the windows is to*
reduce the moments of inertia such that section properties about the two horizontal axes were no longer equal. Since the pipe member of "SUPERPIPE" requires that moments of inertia about the two horizontal axes be equal, a beam member was used for this 9.5" section. This beam member was inode.lled with a momept of inertia of 927 in4 about It the axis bisecting the windows, and of 2090 in4 about the perpendicular* axis* as cal-culated in Reference 7. Additionally a cross sectional area of 24.51 in2 and a torsional moment of inertia of l.132 in4 were calculated. !t is noted that the value of torsional stiffness is under-pred-icted by the inertia value used for this section due to the overprediction of the member's ability to warp (made by neglecting the stiffness of the adjacent members). This is not considered a signif-icant factor since the "window" section is above the point of input nozzle loads and will therefore experience no torsional loads.
- Shear areas for this member were taken as one-half the cross-sectional area.
The "upper" section and the "window" sec-I tion had a weight per unit length of 225
! lb/ft (calculated in Reference 10). The
! next 23.75" of discharge head (referred to as the "lower section) had an additional weight of 79.l lb/ft due to the contained fluid. Otherwise this "lower" section had the same member properties as the "upper" section of the discharge head. Proceeding in the downward vertical direction, the bottom head flange was modeled next in the EDS Nuclear
-1 .,
tr t
02-0660-1089 Revision O Page 21 l.
I mathematical model as a pipe member, l.25" in length, with its out~ide diameter equal f
I to 36" and its wall thickness equal to 10" (in accordance with Reference 11). A lumped weight was placed at each end of the flange in addition to the distributed fluid 1i weight.
The discharge head assembly was the only I- portion of the pump modelled, in part, in a
- i. direction other than vertical (global y axis). At a point 17.25" above the bas*e plate a "rigid member" was modelled in the horizontal direction for 12". This member brought the model to the outside *wall of the discharge head shell from the center-line axis. This member was made very stiff (O.D~ = 30", I.D. = 10") to most accurately transfer the input nozzle loads to the center line of the vertically oriented members. Modelled after this member was a 4.0625" horizontal pipe section having the properties of standard 16 11 carbon steel pipe. - At the extreme end of the horizon-tally oriented portion of the discharge head the discharge flange was modelled.
This flange was represented as a pipe of 23.50 11 outside diameter and 3.67" wall thickness in accordance with the dimension-al data of Reference 12 for a 16", 150 lb slip on flange. Because the weight of the horizontally oriented members was already fI I accounted for in the mathematical model I . along the pump's vertical centerline, the weight per unit length of these members was input as zero lb/ft.
L 3.4 Pump Column r
I The discharge column of the Essential Ser-vice Water Pump consists of a series of 16 11 , standard schedule carbon steel (ASTM SA-53 Grade B) pipe sections connected by EDS Nuclear
02~0660-1089 Revision 0 Page 22 r.
l.
l.38" thick flanges. All column flanges were modelled as pipe members with outside diameters of 21" and w~ll thicknesses of 3.262". This thickness was back-calculated*
from a moment of inertia value of 7391 -
in4 determined in Reference 7 where the stiffness contributions of the "spider" ribs and the 16", schedule 30 pipe were considered. A lumped weight of 32 lb was I* placed at the end of each column flange.
1* An additional 20 lb was coded at the center of a flange union to account for the bronze spider assembly *.
i
- 1. The upper most column flange is bolted to the bottom discharge head flange. Next in the mathematical model was coded a 2'-2.875" section of l~", standard schedule pipe.
One column flange was then modelled followed by six 5'-0" long sections of a flange, followed by 4'-9 1/4" of 16" pipe, followed by another column flange. The very last column flange is bolted to the upper end of the discharge nozzle in the lower bowl assembly to be described in the following section. In addition to the lumped weight at the flange ends a component distributed weight of 75.4 lb/ft and a fluid distributed weight of 79.l lb/ft was input for all pump column members identified as above the submerged water level. For members below this level an additional 87.1 lb/ft was coded to account for the effects of the external fluid (Section 2.2).
The pump column is restrained against lateral translation by two composite plate-angle members attached to the top of one of the column flanges at elevation 583'-2 3/4". The axial stiffness of the EDS Nuclear
02-0660-1089 Revision O Page .23 i
l
[. members which represented these supports was taken as 2.0 x 106 lb/in. The orientation of these member is shown in l' Reference 2.
3.5 Lower Bowl Assembly The Essential Service Water Pump is a two-stage pump which means that there are two impellers. Each impeller is housed in a cast iron enclosure referred to as an "in-termediate bowl".
I The discharge nozzle attached to the lower l
column flange is 1 1 -0 11 in length*. As was the ca~e for the entire lowe~ bowl assembly, the discharge nozzle was modelled with pipe cross sectional properties of 24" outside diameter and .785" wall thickness. This wall thickness was back-calculated from a moment of inertia value determined in Reference 7. The stiffness contribution of the internal shell section and ribs were taken.into consideration when: the moment of inertia was computed. A lumped* weight of 438 lb was placed at the centroid of the discharge nozzle taken as the center of the bearing indicated in Reference 11. In addition, a fluid weight per unit length of 166.2 lb/ft was coded for the member. This weight accounts for the weight of fluid running through the lower bowl assembly and also for the effects of the external fluid.
! The upper and lower intermediate bowls. were I. modelled next each for l'-6 1'4" with lumped weights of 794 lb placed at the centers of their bearings. cbded at the f.
( ends of each bowl was an additional lumped weight of 116 lbs which accouhted for the impellers.* Finally the suctipn nozzle was modelled for 10" with a lumped weight of I
l
- r EDS Nuclear i
(
l ¥5 j l
02-0660-1089 Revision 0 Page 24 I 242 lb placed at the center of the suction bearing. Protruding from the bottom of the
, . suction nozzle was a 6" section which was modelled consistent with the entire lower bowl assembly.
l i.
I
\.
EDS Nuclear
r 02-0660-1089 FIGURE 3.1 Revision O PUMP MATHEMATICAL MODEL Page 25 PUMP Ca>RC\ NATE L Axi::i y
! -rA x_J IJoeiH S.St.J 1* y z.iso l
16 Li WE?i
'° 40
~
ltD
.1'Zf
.7,t f
I 7C. Lx ~l.lTH 1'" *'~
lJJ4
.1io
... r--
ANcllot ">Uf'PDe:t L,ZVl!'L.
~~ ": ~.. . :*~ 100 .DS I x=-W:
z.su 110
' .II~
.llS ITO ISO 1 I
I LJ.TaA.1. <!.Ul'PPC'T' Ll!V'l!.L. I
~~1~4~M~W.~~!,.:!~LZY:::!_!e~L~-~,,~!~~o~~~::::'::===~==:t'~~~==:==:==:==:==:~~==!-.:.\\*11~~---------_J I~ J~
1 '"o 170, Zoe>, :to, 11.o ll01 z.10 1 140, Z10 1"101 'ZZIJ, ZSll, ZI" l - .II
.417 lDOO 11.lwT -~-- -- .-...
rj: I I. S&I
, 1.si1 1.n1
~
i*-m
. ~00 31.o
TABLE 3.1 ESW PUMP "STICK" MODEL TERMS:
w = lwnped weight We = component weight per unit length Wf = fluid weight per unit length I = moment of inertia about the X & Y axes L :::;
canponent length o.o. = outside diameter t = wall thickness c.1. = cast iron c.s. = carbon steel NODE COMPONENT w WC Wf I L o.D. t MATERIAL/DESIGNATION NO. NO. DESCRIPTION (lb} (lb/ft) Clb/ft) (in4) ft. (in) (in) 10 1 Upper Motor 0 0 151 3.229 24.5 0.026 C.S./ASTM A36 (assumed) 20 3500 l Lower Motor 0 0 151 2.25 24.5 0.026 c.s./ASTM A36 (assumed) 30 0 2 Mounting 0 0 16668 0.083 24.5 6.25 C.S./ASTM A36 (assumed)
Plate 40 51 3 Top Head .0 0 16668 0.083 24.5 6.25 c.s./ASTM A36 Flange It!
Ill lQ ro N
O'I
NODE NO.
- NO.
COMPONENT DESCRIPTION w
(lb)
WC (lb/ft)
Wf (lb/ft)
I (in4)
L ft.
O.D.
(in) t (in)
MATERIAL/DESIGNATION 50 51 4 Upper 225 0 2264 0.125 24 0.441 c.S./ASTM A53 Gr. B Cylinder 60 0 5 Window 225 0 See 0.192 N/A N/A c.S./ASTM A53 Gr. B Cylinder Note 1 70 0 6 Lower 225 79.1 2264. 1.979 24 o.441 c.s./ASTM A53 Gr. B Cylinder 80 151 7 Bottom Head 0 79.1 79231 0.104 36 10.0 C.S./ASTM A36 Flange 90 182 8 Column 75.4 79.l 7391 0.115 21 3.262 c.s./ASTM A36 Flange 100 32 9 Column 75.4 79.l 562 2.239 16 0.375 c.s./ASTM A53 ~r. B 110 32 10 Column 75.4 79.l 7391 0.115 21 3.262 c.s./ASTM A36 Flange to Ill 120 82 lQ ro 10 Column 75.4 79.l 7391 0.115 21 3.262 C.S./ASTM A36 N Flange -..J
,.--~
.. ~. ' :
NOOE COMPONENT w We Wf I L o.o. t MATERIAL/DESIGNATION NO. NO. DESCRIPTION {lbl UbLftl ( lbLft l (in4l ft. (inl (in) 130 32 11 Column 75.4 79.1 562 4.77 16 0.375 c.s./ASTM A53 Gr. B 140 32 12 Column 75.4 79.l 7391 0.115 21 3.262 C.S./ASTM A36 Flange 150 82 12 Column 75.4 79.1 7391 0.115 21 3.262 c.s./ASTM A36 Flange 160 32 13 Column 75.4 166.2 562 4.77 16 0.375 c.S./ASTM A53 Gr. B 170 32 14 Column 75.4 166.2 7391 0.115 21 3.262 c.s./ASTM A36 Flange 180 82 14 Column 75.4 166.2 7391 0.115 21 3.262 c.s./ASTM A36 Flange 190 32 15 Column 75.4 166.2 562 4. 77 16 0.375 c.s./ASTM A53 Gr. B 200 32 16 Column 75.4 166.2 7391 0.115 21 3.262 c.s./AS'l'M A36 lt1 Pl Flange lQ fl)
N CD
NOOE COMPONENT
---~
w We Wf I L o.D. t MATERIAL/DESIGNATION NO
- NO. DESCRIPTION (lb) (lbLft} ( lhLft} (in4} ft. (in) Un)
. 210 82 16 Column 75.4 166.2 7391 o.n5 21 3.262 c.s./ASTM A36 Flange 220 32 17 Column 75.4 166.2 562 4.77 16 0.375 c.S./ASTM A53 Gr. B 230 32 18 Column 75.4 166.2 7391 o.ns 21 3.262 c.s./ASTM A36 Flange 240 82 18 Column 75.4 166.2 7391 o.11s
- 2l 3.262 C.S./ASTM A36 Flange 2.50 32 19 Column 75.4 166.2 562 4.77 16 0.375 c.S./ASTM A53 Gr. B 260 32 20 Column 75.4 166.2 7391 o.n5 21 3.262 C.S./ASTM A36 Flange 270 82 20 Colwnn 75.4 166.2 7391 0.115 21 3.262 C.S./ASTM A36 .
Flange IU 280 32 Pl lQ (I) 21 Column 75.4 16.6. 2 562 4.77 16 0.375 c.s./ASTM A53 Gr. B N
\0
NODE COMPONENT w WC Wf I L o.o.; t MATERIAL/DESIGNATION NO. NO. DESCRIPTION (lb) (lb/ft) (lb/ft) (in4) ft. (in) (in) 290 32 22 Column 75.4 166.2 6892 0.115 21 2.975 C.S./ASTM A36 Flange 300 32 23 Discharge 0 166.2 3862 0.417 24 o.7e5 c.I./ASTM A48 Class 30 Bowl 310 438 23 Discharge 0 166.2 3862 0.593 24 0.105 C.I./ASTM A4B Class 30 Bowl 315 0 23 Intermediate 0 166.2 3862 0.594 24 o.1es c.I./ASTM A48 Class 30 Bowl 320 794 23 Intermediate 0 166.2 3862 0.937 24 0.105 C.I./ASTM A48 Class 30 Bowl 325 116 23 Intermediate 0 166.2 3862 o.5e4 24 o.7e5 C.I./ASTM A48 class 30 Bowl 330 794 23 Intermediate 0 166.2 3862 0.937 24 0.105 C.I./ASTM A48 Class 30 Bowl I'd Ill lQ m
w 0
.~-
~
NODE COMPONENT w WC Wf I L o.o. t MATERIAL/DESIGNATION NO. NO. DESCRIPTION (lb) (lb/ft) (lb/ft) (in4) ft. (in) (in) 335 116 23 Suction 0 166.2 3862 o.646 24 0.785 C.I./ASTM A48 Class 30 Bowl 340 242 23 Suction 0 166.2 3862 0.101 24 0.795 C.I./ASTM A48 Class 30*
Bowl 350 0 23 Suction 0 166.2 3862 o.5 24 0.795 C.I./ASTM A48 Class 30 Bowl 360 0 75 0 24 Rigid** 0 0 39270 l*O 30.0 10.0 c.s./ASTM A53 Gr. B Member (arbitrarily assumed) 75S 0 25 Straight** 0 0 562 0.339 16.0 0.375 c.S./ASTM A53 Gr. B Pipe 75F 0 26 Discharge 0 0 11623 0.12 23.5 3.67 c.s./AS'l'M A36 Flange 75N 0 ltj Ill lQ (I)
I-'
--**~- ..
- A rigid straight member assumed with the above cross sectional properties
- A section of elbow that protrudes from the discharge head is assumed to be a straight member with cross sectional properties equal to column pipe.
Additional Cross-Sectional Properties of Discharge Head "Window" Section NO'rE 1. SUPERPIPE computer analysis requires the following additional input for discharge head "window" section Cross-sectional area = 24.50 in2 Effective shear area = 12.25 in2 Torsional moment of inertia, Ix = 1.132 in4 ly = 927 in4 T Iz = 2090 in4 w
N
.u.
"l'il" Waj§i eet4 02-0660-1089 Revision 0 Page 33 L
r: 4.0 SOLE PLATE SUPPORT MODEL DESCRIPTION
, . 4.1 General In conjunction with the seismic qualifica-tiqn of the Essential Service Water Pumps, the stiffnesses of the supporting struc-tures have been investigated. The service water pump supports are located at the 583 1 -2 3/4" elevation and the 590 1 -7" elevation. The latter support consists of a base plate pinned at its four corners with a hole centrally located. *This support must react out both vertical,
! lateral, and rotational motions of the l pump. Due to its configuration a detailed analysis was required to determine the support stiffness.
Specifically, a finite element analysis was performed to accurately determine the stiffness characteristics for the base plate support configuration described below. The finite element approach used,
! employed a model representing one-half the l total configuration with the appropriate symmetry boundary conditions ~pplied. The following sections describe the geometry and the model, -the applied loadings with their corresponding boundary conditions, and the results obtained.
4.2 Geometry and Ma.thematical Model The overall support configuration modeled included the base plate, the pump discharge head assembly lower flange, and a portion I . of the lower pump column, Figure 4.1 *. The lower flange is included since it serves as the means by which the load from the motor unit and pump column is transmitted to the base plate. The lower pump column is in-cluded to facilitate the application of EDS Nuclear
we
.02-0660-1089 Revision 0 Page 34 L
L the various loadings onto the base plate while preventing local distortion at the pump column-lower flange juncture.
(' The base plate measures 40 11 square by 1 11 thick with a 26 11 diameter opening centrally located. The base plate is rigidly attached f to the concrete base mat by fo.ur (4) anchor bolts typically located 2 11 from each edge at the four corners. Located on top of the
- t. base plate is the lower flange measuring 36 11 o.o., 26 11 I.D., and l 1/4" thick. It is attached to the base plate by ei;ht (8) bolts equally spaced with a bolt ring diameter of 32 11
- The pump column, joined to the lower flange from below, measures 16" o.o. by 0.375 11 thick. A 2 11 length of
- I pipe was modeled. The material for all these components is carbon steel.
The mathematical model of the base plate support configuration used in the analysis was developed using ANSYS, a finite element computer program which is maintained by Swanson Analysis Systems, Inc. *The four-node quadrilat*ral shell element, STIF 63, is used in the analysis. Due to symmetry and the proper placement of boundary condi~
tions only one-half of the total conf igura-tion is modeled. The model consists of 256 elements and* 264 nodes. Figures 4.2 to 4.7 show the locations of the nodes and elements of the model. Detailed modeling of the bolts is not done, although, the effective restraint of the bolts are included in the model.
- r. 4.3 Loadings and Boundary Conditions I: Four (4) loadings have been considered I
EDS Nuclear (bending, compression - tension, torsion, and shear) in the determination of the
02-0660-1089 Revision 0 Page 35 L
stiffness characteristics. Due to symmetry, the bending and shear load cases account for two components of, stiffness each. The load case, type, direction, and magnitude of loading are listed in Table 4.1.
The nodal forces for each load case are tabulated in Reference 13. Load cases lA and 2A represent revisions to their initial load cases. After review of the original l analysis it became apparent that* addtional l*
boundary conditions were in order.
- The later revision more ~ccurately describes the probable support deflections.
The boundary conditions used in this analy-sis varied for the different loadings. In all cases the nodes at which the bolts joining the lower flange and base plate were located were coupled for translation.
Symmetry planes prohibit displacement per-pendicular to the plane. For load cases 1, lA, 2, 2A, and 4 the XZ plane is a symmetry plane. Asymmetry planes prohibit displace-ment in the plane. For load ca*se 3 the XZ plane is an asymmetry plane.
Additional boundary conditions were kept to a minimum thereby permitting maximum flexi-bility. For load case l/lA it was deemed necessary to account for the presence of the concrete base mat. To the left of node 122, inclusive of the base plate the direc-I
.! tion of loading is such that the base plate i would b~ placed in compression. Therefore the vertical translation of the base plate was helc;t fixed.
I EDS Nuclear
r 02-0660-1089 Revision 0 Page 36 t:
L The results from load cases l and 2 indi-cate that the lower flange and the base plate intersect. Since this is not physi-r cally possible, an additional constraint was imposed at the 26" diameter couplfng the translations of the corresponding nodes of the lower flange and base plate.
For load case 2 the direction of load was also changed since preliminary analysis I*. indicated that the base.plate support would never realize a net positive upward (ten-sion) load. The nodes at the 26" diameter of the base plate were also held fixed in the vertical direction following the same reasoning as described above concerning the bending loa'd case.
The boundary conditons for all load cases are included in Reference 13.
4.4 Stiffness Results The stiffness of the base plate support configuraton was determined (Reference 14) as the ratio of the applied loading to the respective maximum deflection of the lower flange inner diameter, neglecting any de-flection of the pump column. In the case of rotational stiffness, the lateral de-flection was resolved into a rotation about the pump column centerline. The computed stiffnesses are listed in Table 4.2.
r l
EDS Nuclear
l*
!M'& e w l 02-0660-1089
-L Revision 0 Page 37 I. Table 4.1 I" Load Case ~ Direction Magnitude l Bending Negative y 245437 in-lb l lA Bending Negative y 245437 in-lb L 2 Tension Positive z 24,000 lb 2A Co1Dpression Negative z 24,000 lb 3 Torsion Negative z 1$7,500 in-lb 4 Shear Positive x 24,000 lb I
I' l
l EDS Nuclear
11£:. . 02-0660-1089 Revision 0 Page 38 Table 4.2 Translational Stiffness (lb/in) Rotational Stiffness (in-lbLrad)
Kx = 2. 94 x 107 Kex = 3.54 x 107 Ky = 2.94 x 107 K6y = 3.54 x 107 Kz = 2.oa x 106 K 6z = 1.96 x 109.
I r
I r
EDS Nuclear
02-0660-1089 Revision o Page 39 LOWER FLANGE 2"
BASE PLATE II
+ . +
[,
~
f"
[. ~
20" r.
/
r
~
L A L _ 15 1/4" 16" Y~x 36" 40" PLAN VIEW I.
i 2"
r.
ZL x
- 1. 15 5/8" 26" PUMP COLUMN r
l Section "8,-e::,.'
FIGURE 4.1
PUMP DISCHARGE HEAD ASSEMBLY LOWER FLANGE - ELEMENTS Id !:ti 0 Ill (!) N
-0 ~ I
(!) t-'*0
.s:-. I'll
..... °'°'
0 0o
- S ,_. I FIGURE 4.2 oo 00 l.O
PUMP DISCHARGE HEAD ASSEMBLY LOWER FLANGE - NODES
"ti~
Ill Cl> N 0 lQ <: I 11> 1-'*0 (II O'I
~ ..... O'I FIGURE 4. 3 I-' 0 0
- J I I-'
0 Ooo
\0
BASE PLATE - ELEMENTS 14."1 IZ8 l"Z. 7 lz.&
121 18 FIGURE 4.4
I j-: (
BASE PLATE - NODES
'Z51
/'ZZ.
ZI&
'/.JO
3 z11.. ..__ _.__ _..'--__..__ _. ....___.._____.__ _. & -_ __,, ,.,~
l")I jt,(, 1.f.I l'Z6 14Z. '"7 FIGURE 4.5
LOWER PUMP COLUMN - ELEMENTS l~~ *21& '70 I 4'?' 11D 111 z~&
10)3 146 '~ Ile>
OJo e~
bf, 1,;
4Z
.f I 21 2.
211
'" 2D I LOWER PUMP COLUMN - NODES IO~ o Ill (I) N lQ <l I It> 1-'*0 (IJ
~I-'*°'
°'
~oo
- S I FIGURE 4.6 .......
0 00
~
- ASSEMBLED MATHEMATICAL MODEL - PLAN VIEW FIGURE 4.7
02-0660-1089 Revision 0 Page 46 I
5.0 ACCEPTANCE CRITERIA r AND RESULTS FOR THE
.5% DAMPING EVALUATION S.l Pump Column Examination of the SUPERPIPE computer out-f put (References 16, 19, 20 and 21) indi-cates that the maximum total stress in the Essential Service Water Pump is realized in the 16 11 , standard schedule pump column just above the lower lateral support. Total stress is considered the sum of the stress resulting from the Deadweight, Seismic and Thermal Nozzle loadings. While maximum stress levels do not occur at the same point along the pump in each load case, a conservative but still representative value of total stress can be calculated by summing individual maximum stresses due to each loading regardless of their location.
For the Service Water Pump .5% damping evaluation, this results in a total stress of 30024 psi (Reference 22).
The total stress in the Service Water Pump is limited in the Palisades FSAR to 110% of the minimum material yield stress. The specific FSAR eqution for Class I equipment pretaining to the SSE earthquake level is given below.
MOL + 2SL < 1.10 Where:
MOL = Maximum normal operating stress 2SL = Stress resulting from an SSE Earthquake.
Normal operating stress for the Service Water Pump is derived from Deadweight and Thermal Expansion nozzle loads.
1*
02-0660-1089 Revision O Page 47 r
r The pump column material is ASTM A53 Grade B
[ carbon steel with a minimum yield stress of 35000 psi (Reference 15). Substituting into the FSAR equation 1465 psi + 28559 psi < 1.10 (35000 psi) r 30024 psi < 38500 psi I .
I Thus column stresses are found to be acceptable and consistent with FSAR criteria.
5.2 Pump Shaft The shaft of the Essential Service Water Pump is stressed primarily due to bending during the SSE Earthquake event and to axial thrust induced by the rotating impellers. I I
I Each of these loadings were evaluated for I their impact on the shaft while loads due I to Oeadweight and Thermal Expansion of the I I
attached piping were consider~d negligible. I I
Maximum seismic stress in the shait was I I
found to occur in the region between the I base plate and the lower column supports i n I I
the .Si damping evaluation. Bending stres s was determined from the relative lateral deflections of the support bearings to be 5902 psi. To obtain a total shaft stress value, the axial stress due to impeller thrust was added. Axial force on the shaft can reach 6856 lbs inducing a stress of 1824 psi. Therefore the total shaft stress was found to be 7726 psi {Reference 26).
The shaft material, in the region of maxi-mum stress, is AISI C-1045 carbon steel with a yield stress of 86,700 psi. Since the maximum total shaft stress is only 9%
of the yield stress value, the shaft is considered loaded well within acceptable limits.
02-0660..:.1009 Revision 0 Page 48 I
5.3 Column Flanges The Essential Service Water Pump's column r flanges were evaluated based on the guide-lines of the 1980 ASME Boiler and Pressure Vessel Code as part of the overall investi-gation into the system's functionality f after an SSE earthquake event. Specifi-cally Code Subparagraph NC-3658.l in con-junction with Code Appendix XI and Appendix L were addressed.
The maximum bending moment acting at a column flanged connection due to an . SSE earthquake in addition to normai operating loads was found to be 152476 ft-lb in the
.5% damping pump evaluation. This moment occurred at the lower column support level and is equivalent to a flange design pres-sure of 1414 psi. With this design pressure the radial stress and tangential stress in the flange were computed as 31661 psi and 3626 psi respectively. The flange material, A36 carbon steel, has a minimum yield strength of 36000 psi. Therefore the radial and tangential stress levels in the flange are within the specified .allowable limits of the piant FSAR.
The maximum tensile force in any one flange bolt in the .5% damping evaluation was calculated to be 24092 lb (Reference 33) .
This exceeds the computed preload of 13800 lb. In fact the tensile load in five flange bolts was found to be greater than the applied bolt preload. This however, does not preclude the acceptability* of the connection. The maximum shear force in any one bolt was 11843 lb., taken as one-half the axial force in one of the lower lateral supports. Using the maximum shear stress Theory of Failure the flange bolts were found to be capable of withstanding the imposed loads (Reference 33).
02-0660-1089 Revision 0 Page 49 5.4 Column Vibration Below the lower lateral supports at eleva-tion 583'-2 3/4", the Essential Service Water Pump is unrestrained in the horizon-tal plane for its remaining 30'-S 7/8".
Therefore, large lateral deflections were I expected at the lower tip of the pump I column during an SSE Earthquake event. In the .5% damping evaluation, the maximum lower column horizontal translation was found to be S.86" for the SSE event in combination with normal operating loads (Reference 22). In discussions with the pump manufacturer, Layne and Bowler, Inc.,
it was determined that deflection of the bottom. end of the pump of as much as 6.00" was acceptable if the stresses in the column flanges and the bolting were within allowable limits (Reference 24).
5.5 Motor Vibration Above the base plate support at elevation 590'-7", the Essential Service Water Pump is unrestrained in the horizontal plane for the length of the Discharge Head and Motor Unit, some 8'-7 3/4". The Discharge Head Assembly is relatively stiff and lateral deflections for the SSE Earthquake event were lower than in the pump column, rea=h-ing only .134" in the .5% damping evalua-tion. However, the motor unit is sensitive to lateral translation and horizontal vibration during an SSE Earthquake is limited to .020" by the pump manufacturer.
Although the .134" lateral deflection
- realized at the top end of the motor during the SSE inertia analysis exceeds the 20 mil criteria, it was noted that .129" of this vibration is due to flexible rotation of the sole plate. Since only .005" of the vibration is due to relative bending between the top end of the motor and the sole plate, Layne and Bowler was able to confirm the acceptability of the .134" vibration (Reference 24).
r
- L 02-0660-1089 Revision 0 Page so r 5.6 Discharge Head to Sole Plate Bolts In addition to determining the stiffness of
[
the sole plate for use in the SUPERPIPE computer analysis, the ANSYS finite element model discussed in Section 4.0 was used to compute loads in the bolts connecting the sole plate to the Discharge Head Assembly.
This was accomplished by multiplying the bolt loads from the ANSYS analysis which were caused by a set of "unit loads", by the ratios of the actual sole plate support loads to these unit loads. Actual sole plate support loads were determined in the SUPERPIPE analysis of the Service Water Pump.
The resulting bolt loads for the .5% damp-ing evaluation were thus computed as + 8029 lb. shear and + 5330 lb. tension (Reference 34}. The corresponding shear stress in the bolt was found to be 18174 psi and the tensile stress was found to be 15959 psi.
Using an allowable stress specified in the Palisades FSAR for Class I equipment, the subject bolts are found to be inconsistent with the acceptance criteria in combined tension and shear.
5.7 Sole Plate The ANSYS model used to compute the stiff-ness of the sole plate to be input into the SUPERPIPE pump analysis was also used to compute sole plate stresses. Utilizing the SUPERPIPE analysis results, which gave the forces and moments acting on the base plate support, the stresses in the sole plate were back-calculated for bending, compres-sion, torsional and shear loads. Tension loads were never realized due to the rela-tive magnitudes of deadweight and seismic contributions. Upon adding absolute maxi-mum principal stresses due to the most severe compressive force, two shear forces,
~--~:;.:* ..
02-0660-1089 Revision 0 Page 51 a torsional moment and two bending moments, the total stress level found in the sole plate was 14726 psi (References 28 and 32). This stress value neglected the fact that maximum principal stresses do not occur at the same location for each of the six load applications.
For the sole plate material of ASTM A36 carbon steel the minimum yield stress is 36000 psi. Therefore, stresses in the sole plate for the .5% damping evaluation are of an acceptable magnitude.
In the course of determining the stiffness properties of the sole plate, the loads in the anchor bolts f ix~ng the sole plate to the concrete floor were also determined fer applied "unit loads". The analysis per-formed was elastic and as long as stresses remained in the elastic range, the result-ing bolt loads from the ANSYS output (Reference 32) could be multiplied by the ratio of the actual loads on the sole plat e to the "unit loads". The actual loads were those forces and moments on the base plate support determined in the SUPERPIPE analysis of the Service Water Pump and summarized in Reference 22.
The resulting anchor bolt design loads for the .5% damping evaluation were thus com-puted as + 10911 lb. in shear and + 2353 lb. in tension. The corresponding shear I.
and tensile bolt stresses were 18146 psi and 5094 psi respective l y. Using an FSAR I based allowable stress as specified for Class I equipment, the anchor bolts are found to be in agreement with the acceptance criteria for combined tension I and shear (Reference 28).
E OS Nuc lear
!Eliil~-~"'~4---*
02-0660-1089 Revision 0 Page 52 L 6.0 ACCEPTANCE CRITERIA AND RESULTS FOR THE 2% DAMPING EVALUATION 6.1 Pump Column The SUPERPIPE computer analyses (References 19, 20, 21, and 23) indicate that the maxi-mum total stress in the Essential Service Water Pump occurs in the 16", standard I
I schedule pump column just above the lower lateral support. Utilizing the same proce-dure for computing total stress in Section 5~1, a total stress value of 20440 psi was calculated for the 2% damping evaluation (Reference 25). This stress level is below the FSAR allowable of 110% yield which is l8500 psi (l.10 x 35000 psi) for the pump column material of ASTM A53 Grade B.
6.2 Pump Shaft The maximum seismic bending stre~s in the pump shaft was found to occur between the base plate and the lower column supports.
From the relative lateral deflections of the shaft bearings as discussed in Section 2.5, the maximum value of bending stress was found to be 4019 psi in the 2% damping evaluation. Adding to this the axial stress due to an impellar thrust of 6856 lb, the total shaft stres~ was calculated as 5843 psi (Reference 27).
The shaft material, in the region of maxi-mum stress, is AISI C-1045 Carbon Steel with a yield stress of 86,700 psi. There-fore, shaft stress levels are well within r acceptable limits.
i 6.3 Column Flanges , The maximum bending moment acting at a column flanged connection due to an SSE earthquake in addition to normal operating s was found to be 96636 ft-lb in the damping pump evaluation. This moment rred at the lower column support level and is equivalent to a flange design pres-sure of 896 psi. With this design pressure E OS Nuc lear
'02-0660-1089 Revision 0 Page 53 the radial stress and tangential stress in the flange were computed as 20063 psi and 2298 psi respectively * . The flange material, A36 carbon steel, has a minimum yield strength of 36000 psi. Therefore the radial and tangential stress levels in the flange are within the specified allowable limits of the plant FSAR. The maximum bolt tensile force was calculated to be 15269 lb. This load and the tensile load in two I adjacent . bolts were above the applied I*
I preload. The maximum bolt shear force was l
found to be 7978 lb. Using the Maximum Shear Stress Theory of Failure, the flange bolts are shown acceptable under the j\
applied loads. * \** .,\ c...u. **..rJ---"
~
. CY . <.;\
6.4 Column Vibration In the 2% damping evaluation, h~ __maximum lower column tip horizontal<:[ rati~ was found to be 3. 74" for the SSE -* eve-nt--'in combination with normal operating loads (Reference 25). Since the pump manufac-turer was able to confirm the acceptability of lower column deflections of as much as
- 6. 00" * (Reference 24), provided stresses in . j-the column flanges and the !>-9.~i~i::~----c\: , .1 __.-
- within allowable limits, <i:fhra~ions/Of \
- 3. 74" are considered acceptable _a _s _ well.
6.5 Motor Vibration The maximum lateral deflection of the top end of the motor unit was found to be .088" during the SSE Earthquake event. Although the .088" deflection in the horizontal plane exceeds the 20 mil limit established by the pump manufacturer, it was noted that
.083" of this vibration is due to flexible rotation of the sole plate. Since only
.005" of the vibration is due to relative bending between the top end of the motor and the sole plate, the pump manufacturer was able to confirm the acceptability of a vibration of this magnitude (Reference 24).
EDS Nuc lear
r
)limlii~~
02-0660-1089 Revision 0 Page 54 L
{.
6.6 Discharge Head to Sole Plate i- Bolts Following the procedure for determining forces in the bolts connecting the sole plate to the Discharge Head Assembly dis-i cussed in Section 5.6, design bolt loads were calculated in the 2% damping evalua-tion. The resulting bolt loads were com-puted as + 6572 lb. in shear and 4647 lb.
in tension. Bolt stresses were determined and compared to the acceptance criteria of the Palisades FSAR specified for Class I pieces of equipment. The shear stress of 14876 psi in combination with the tensile stress of 13914 psi was found to be within the allowable stress levels Reference 34).
6.7 Sole Plate Utilizing the SUPERPIPE analysis results for the 2% damping evaluation, the forces and moments acting on the base plate sup-port were multiplied by sole plate stress to applied sole plate load ratios deter-mined *from the ANSYS computer analysis of the sole plate described in Section 4.0.
Upon adding absolute maximum principal stresses due to the most severe compressive force, two shear forces, a torsional moment and two bending moments, the total stress level found in the sole place was computed as 12410 psi (Reference 29). This value of stress was conservatively determined by assuming that the maximum principal stress
[_ occured at the same location for all the applied loads.
The maximum sole plate stress is well below the yield stress which for the plate material of ASTM A36 Carbon Steel is 36000 psi. Therefore, the sole plate was found to be consistent with the acceptance criteria of the Palisades FSAR.
E DS Nuc lear
IBllm!i9ill!!ll~*-
02-0660-1089 Revision o Page 55 6.8 Sole Plate to Concrete Anchor Bolts Following the procedure for determining anchor bolt loads discussed in Section 5.8, the forces acting on the sole plate to concrete bolts were calculated in the 2%
damping pump evaluation. The resulting anchor bolt design loads were found to be +
8823 lb. in shear and + 2032 lb. in ten-r* sion. The corresponding shear and tensile bolt stresses were 14674 psi and 4399 psi respectively. Using an FSAR based allow-able stress as specified for Class I equip-ment, the anchor bolts are determined to be consistent with the acceptance criteria for combined tension and shear (Reference 29).
EDS Nuc lear
aii~Dl'a~"'f};Q_~:.~
02-0660-1089 Revision 0 Page 56 I*
7.0
SUMMARY
AND
- f. CONCLUSIONS The investigations of the structural and functional adequacy of the Essential Service Water Pumps di~cussed in this report were prompted by the need for a l documented safety evaluation of this component which was selected by the NRC's Senior Seismic Review Team as having potential seismic fragility. Following the philosophy of the Systematic Evaluation Program, the Safe Shutdown Earthquake was the only earthquake level considered in the I" pump evalutions. Structural damping ratios of both .5% and 2% were considered in the seismic inertia analysis because the Palisades FSAR does not specifically address the Service Water Pump in this regard. The acceptance criteria to which the pumps were evaluated were based on the Palisades FSAR requirements for Class I Systems and Equipment. Stresses due to normal operation and to the SSE Earthquake event were added and compared to allowable stress levels.
The .5% and 2% damping evaluations found stresses in the pump .column and pump shaft to be of acceptable magnitudes. In
\~ accordance with the manufacturer's requirements, shaft bearing loads and impeller clearances were not critical. The vibrations experienced at the upper and lower ends of the Service Water Pump during an SSE Earthquake were, however, critical.
In both the .5% and 2% damping evaluations, deflections at the top of the motor unit and at the bottom of the suction nozzle were within the limits established by the pump manufacturer for safe operation. The anchor bolts in the sole plate were loaded below allowable levels consistent with the plant FSAR acceptance criteria for loads in conjunction with the SSE Earthquake event.
E DS N clear
am---sma1___ liill5i~~~~~.u~~*
02-0660-1089 Revision 0 Page 57 Certain bolts joining the Discharge Head Assembly to the sole plate were found to be overstressed in the
- 5%. damping pump evalu-ation. All of these bolts were acceptable in the 2% damping evaluation due to a decrease in seismic acceleration values.
The sole plate itself was stressed within the allowables proposed in the plant FSAR.
Finally, investigations into the flanged column connections showed that some separa-tion of the flange faces may occur at the elevation of the lower column support.
This separation was found possible at only I one connection in both evalu*a tions. The extent to which bolt loads exceed thei~
l preload value was least in the 2% damping case. These connections are, however, strong enough to carry ~he imposed loads based on the allowable stress levels of the Palisades FSAR.
In conclusion, there are no modifications to the Essential Service Water Pump neces-sitated by the investigations of seismic adequacy discussed in this report. The loss of contact at one of the flang~d column connections will probably never occur in reality due to the additiondl force carrying capacity of bolts adjacent to those bolts having tensile loads in excess of their preload. If slight separa-tion occurs it would be only for the short duration of the SSE earthquake. Upon acceptance of the lower lateral supports, to be qualified with the design loads generated by analyses performed for this report, the Service Water Pump is con-sidered capable of maintaining its struc-tural integrity in the event of an SSE earthquake.
It may be important to note that the true structural damping of the system under investigation is thought to be greater than E DS Nuclear
~~~:;,. ;: ..
02-0660-1089 Revision 0 Page 58 2% of critical. The fact that thirty feet of pump column is submerged in water lends itself to this conclusion. A detailed examination into this area could lead to the justification of a higher damping .
value. This would in turn lead to *consid-erably lower inertia forces acting on the pump and consequently lower stresses and smaller pump deflections.
I l*
EDS Nucl ear
I. ..u ..
.. Y' H AA
-! 02-0660-1089 Revision O
_Page 59 r*
I" l a.a REFERENCES 1* 1. Bechtel Company Drawing No. C-221, Revision 10, "Intake Structure Miscel-laneous Details, 11 Location c-4 and F-4.
- 2. Bechtel Company Drawing No. C-226, 11 Revisio 1, 11 Intake Structure Plans,
.* Location B-7
- 1*
- 3. Palisades Nuclear Power Plant Final Safety Analysis Report, Appendix A, I 11 Design Basis for Structures, Systems 1* and Equipment (Except Containment Structure) for Palisades Plant. 11
- 4. C.K. McDonald, "Seismic Analysis of Long Column Vertical Pumps 11 , ASME I. Publication 74-NE-2.
- s. EDS Letter No. 0660-005-NY-Oll, dated June 18, 1981.
- 6. EDS Calculation 201, "Determination of Masses, 11 Rev. o. Job No. 0660-005-643
- 7. EDS Calculation 202, "Determination of Moments of Inertia," Rev. 0, Job No.
0660-005-643
- 8. General Electric Drawing No. 992C536, Rev. 3, (Bechtel Company Drawing No.
II. 950x9* Mll Sh. 9-3) Outline - Induction Motor
- 9. Bechtel Letter No. BEV-EDS-42, dated May 20, 1981. Transmittal of Pump De-tails, Bill of Materials, etc.
- 10. EDS Calculation 203, "Mathematical Model," Rev. 2, Job No. 0660-005-643 EDS Nuclear
..u ..
"CT M QAHW '*?RM 02-0660-1089 Revision 0 Page 60
- 11. EDS Letter No. 0660-005-NY-Oll, dated June 18, 1981. Miscellaneous Pump In-I formation
- 12. Tube Turns Catalog 411 "Welding Fitt-ings and Flanges", 1979.
I 13. EDS Calcculation 204, "Stiffness Calcu-lation - Loading Conditions", Rev. l, Job No. 0660-005-643.
- 14. EDS Calculation 205, "Stiffness Matrix Sole Plate Calculation", Rev. l, Job No. 0660-005-643.
- 15. 1980 ASME Boiler and Pressure Vessel Code,Section II, Part A and Section III, Subsection NC.
- 16. EDS Calculation 210, "Pump Stick Model Computer Analysis", Rev. l, Computer Run Sequence No. ACGYBMX dated 8/17/81, Job No. 0660-005~643.
- 17. EDS Letter No. 0660-005-NY-028, dated October 19, 1981. Design Input for ESW Pump Analysis.
- 18. EDS Calculation 215, "Stick Model-Seismic Analysis - Low Water Level,"
Rev. O, Computer Run Sequence No.
ACGYGUI, dated 10/20/81, Job No.
0660-005-643.
- 19. EDS Calculation 216, "Pump 7A - Stick Model - Nozle Loads", Rev. 0, Computer
[ . Run Sequence No. ACGYJQR, dated 10/22/81, Job No. 0660-005-643.
- 20. EDS Calculation 217, "Pump 7B - Stick Model - Nozle Loads", Rev. 0, Computer Run Sequence No. ACGYIUV, dated 10/22/81, Job No. 0660-005-643.
I. EDS Nuclear
02-0660-1089 Revision O Page 61
[.
I' 21. EDS calculation 218, "Pump 7C - Stick Model - Nozle Loads~*, Rev. O, Computer Run Sequence No. ACGYJRI, dated 10/22/81, Job No. 0660-005-643.
- 22. EDS calculation 213, "Computer Results
- .5% Damping", Rev. l, Job No.
0660-005-643.
- 23. EDS Calculation 220, "Pump Inertial Analysis - 2% Damping, Rev *. 0, Computer Run Sequence No. ACGYPCW, dated 10/20/81, Job No. 0660-005-643.
- 24. Layne and Bowler, Inc. Letter dated September ll, 1981 from Mr. Chi-Sheng Yang.
- 25. EDS Calculation 219, "Computer Results
-2% Damping", Rev. O, Job No.
0660-005-643.
- 26. EDS calculation 208, "Shaft Stresses -
.5% Damping", Rev. l, Job No.
0660-005-643.
- 27. EDS Calculation 222, "Shaft Stresses -
2% Damping", Rev. 0, Job No.
0660-005-643.
- 28. EDS Calculation 214, "Base Plate Stress and Anchor Bolt Load Calculation For The .5% Damping Evaluation", Rev. 1, r Job No. 0660-005-643.
- 29. EDS Calculation 223, "Base Plate Stress f and Anchor Bolt Load Calculation For The 2% Damping Evaluation", Rev. 0, Job No. 0660-005-643.
- 30. EDS Calculation 207, "Spectra Curves",
Rev. O, Job No. 0660-005-643.
I.
EDS Nuclear
[.
r 02-0660-1089' Revision O 1 Page 62
- 31. EDS calculation 221, "Spectra Curves -
2% Damping", Rev. O, Job No.
r 0660-005-643.
- 32. EDS Calculation 211, "Base Plate Model Computer Analysis", Rev. l, .computer I Run Sequence No.s ACIIZHG dated 8/13/81 and ACIIBAV dated 7/27/81, Job No.
0660-005-643.
- 33. EDS Calculation 209, "Flange Qualifica-tion, Rev. l, Job No. 0660-005-643.
- 34. EDS calculation 224, "Pump Discharge Head Bolt Loads and Stresses, Rev. O, Job No. 0660-005-643.
I.
1.
r:
EDS Nuclear
[ ..