ML20085G104
| ML20085G104 | |
| Person / Time | |
|---|---|
| Site: | Saxton File:GPU Nuclear icon.png |
| Issue date: | 04/01/1969 |
| From: | SAXTON NUCLEAR EXPERIMENTAL CORP. |
| To: | |
| Shared Package | |
| ML20083L048 | List:
|
| References | |
| FOIA-91-17 NUDOCS 9110240110 | |
| Download: ML20085G104 (16) | |
Text
_ . _ . _ . . _ _ _ _ . . . . . _ ~ . . _ _ . _ . _ _ _ _ - _ . . . . . . . _ _ . _ _ . _ _ . _ . _ . _ _ - _ _ _
- . . . , i t////M ,
i l
SUltiARY REPORT ON BUCKLING OF SAXTON CORE II l'UEL ASSD!BLIES AND PREVENTION OF' BUCKLING IN CORE III i
9 l
'.9110240110 910424 PDR FOIA DEKOK91-17 PDR ,
e StWiARY l
1 Buckling of the cons on sone Saxten Core 11 assemblies was observed after Core 1). operation. The extent of the buckling of the Core 11 cans and the design modifications used to prevent buckling in Core III are herein reported. .
The buckling of the central platonium assemblies was caus.J oy frictional forces between the grids and fuel rods arising from differential thermal expansion between the stainless steel assembly can and Zircaloy fuel cladding.
The buckling of the peripheral uranium dioxide fueled assemblies was caused ty thernal gradients across the assembly and was limitei to those assemblies with the Core II type grid design.
The major modifications to the loose lattice assemblies to prevent buckling during Core III operation consist oft (a) reduction of the grid f riction loads through resetting of grid springs; and, (b) stiffening of the can structure through the use of full length clips between can halves and replacement of six Circaloy water tubes by stainless steel water tubes with angle braces welded between the tubes and cans.
To prevent buckling in the load follow assemblies during Core %11 operation, the assemblies have been modified by reducing (Pa grid friction loads through resetting of grid springs and stiffening of i the can structure throught (a) replacement of fuel rods in two corner l locations by square stainless steel bars with angle braces veldad between l
the bars and cans; and, (b) angles velded to the inside of the can between fuel rods on the long sides of the can.
None of the buckled assemblies will be reused in Core III.
l l
l L
. m . . . . ~ ,
l t
l 1.0 St'y[1ARY,0{, L[C KLjyg 11 Eers.!!_flyten!yg_6ss33!!3b Buckling was observed in eight of the nine central plutonium assemblies. The buckling appeared to be of a randem nature with no apparent pattern or consistency. However, the four corner assemblies of the square pattern formed by 'he central nine assemblics appeared, in general, to have t'..; Jorst buckling.
The maximum lateral deficction of the buckles were estimated by visual inspection to be 0.06 to 0.08 inches.
1he center span between the sc ond and third grids of the plutonium assemblies experienced the worst buckling with the greatest f requency, the frequency and severity of buckling decreasing towards the end of the assembly. The dircui't k:f the buckling (toward or away f rom the fuel rods) appeared to be completely random and the severity of buckles independent of direction.
Rub marks, which were observed on several assemblies, could be attributed to handling or contact with spacer bars in the spent j
fuel cask d'uring shipment. However, in at least one case it is de' finitely concluded, based on the appearance of the marks, that I the rubbing occurred in the core and resulted from interferance with a control rod assta.bly.
The single plutonium assembly which did not exhibit buckling contained eighteen stainless steel clad rods. The effect of stainless steel clad fuel rods would be to reduce the friction load exerted by the Zircaloy rods and increase the effective strength of the can (the stainless rods being put into compression as well as the can during dif ferential thermal expansion).
The one plutonium assembly which contained eight stainless steel clad rods vv. found to have only minor buckling. By analysis, this number of stainless rods is insufficient to prevent buckling of the aseembly but the reinforcing effect evidently did reduce the extent of buckling in this assembly.
, s o
)
1.2 core,11,Uo2.A58eE.b11es Buckling was observed on three of the seven Core 11 design UO2 fuel assemblies. In this case, however, the buckling was minor in i nature and generally was restricted to the upper spans on the sides of the assemblies facing the center of the core. No buckling I was observed on any of the Core I design assemblies used in Ccre 11.
A detailed surcary of the buckling is given in Table 1 and Figures 1 and 2 show some typical buckles. Figure 3 shows a core cross section indicating buckled assemblies.
a I
e G
I l
l l
1
_m _
3-l_.__-.._.
_ _: _ u _ _.__ _ _ _.. _ .. _ _
- l
- . t I
r I
' TABLE 1 {
Core Buckling ,
Location Observed 4 Serial No. Tyge r SS Clad UO2 Rods ID No i
$03-1-7 503-1-19 SS Clad UO2 Rods 3r No SD No >
503-1-10 SS Clad UOy Rods SS Clad UO2 Rods 1C Yes Minor ;
$03-10-6 >
2B -No 503-10-2 SS Clad U03 Rods SS Clad 002 Rods 2F Yes Minor
$03-10-3 4B No
$03 10-4 SS Clad U02 R 40 :
$E No 503-10-5 SS Clad 007 Rods 2C - Yes t 503-12-2 Zr Clad Puo2-UO2 Rods Zr Clad Pu02-UO2 Rods 2D -Yes 503-12-5 Zr Clad PuO2 -UO2 Rods 2E Yes 503-12-4 3C No 503-12-3 Zr/SS Clad Pu02-UO2 Rods Zr Clad Pu0 7-UO 2 P de 3E Yes 503-12-6
'?
Zr Clad Pu02-UOy Rods 4C Yes 503-12-7 Zr/SS Clad Puoy-UO2 Rods 4D Yes
$03 1 Zr Clad Puo2 -UO2 R ds 4E Yes 303-12-8 503-7-1 SS Clad UOy Rods 1E .No SS Clad U02 R ds 3B No 503-2-3 SS Clad UO2 Rods 4F No 503-11-1 7
503-11-2 SS Clad UO2 Rods SC Yes Minor ;
Zr Clad-Puoy -UO2 R da 3D Yes ,
503-13-1 1
I +
l-I f
hs m-.s.-,,_ . . . . , . , . . . . . - . . - . - , - - - , , , . . - . , - . _ , - . ._;...-,--.. -
,._..,-s,~~.... ,---.,_.-.-_.,_..-,--c.,_.r%._. , , , - .
e 1 4
5
i e s * ** e M
. e t e .-
- ; w ; ;
i ,
r s s a s
, ,et,,e" .. ---....
l 7s,'
- ,,'.J .
/
- 2. .hS P--ss -- ~ ~ . ' N '\ ~
i r
%/ ', j A
i lf, m/,,
<- =
f, .41".
-,2<
,-r,-9jJ :.
c
- s. _
- 1 , N,
._ \ s>
- r ,,, , = ; -
..- , g. ,
0 {,yAI _, w,u f _m __,. ,i\
, , t
,i t i i s i 8 ,
/[ / $.s. . , - . . - i i
.a p;,,pe> ..sr
- c. .s o., i , ,,,.,
, -- - ..a z ,
I e *::.i .
( , v w- w .a se w g 1, L.-4.. g' }. . . 7.Q. 4 J. _. _q. ,
._. J
,o lK ,,,.,
\ \ ,,7 7-"-.
+ / * ,o .f ._.. ..y s
___=
,I ,i 1
.. /A./,c'./ . f, . -r)-u 1 ,
n ' ,
,i ,o i
s .
t' iR N ; ,e l u.L .,.4 r.'% ; ' ,,.; , ,
su,K-).M,t V
,f( 's /
V M .
M s.
i t 7. 3,/,
i
/
, ,/
?./' /
s ... .r.
, ), ,. ~ . .~. . ... .
_%t ,
%s - ,5**,
, m.. .. . .n m ... .
s.r ...... . . ..,
! FIGURE 3: Saxton Core Cross-Section Showing Buckled Fuel Assembly 1.ocations "B" l
. i 2.0 geygtgr,agggy;gg The following possible loading methods were examined as buckling modes for the assembliest A. Externally applied loads on the assembly due tot
- 1. Shipping and handling
- 2. Interference with reactor internals B. Loading generated internally to the assembly throught
- 1. Frictional effects during differential expansion
- 2. Temperature differentials across the assemblics 2.1 Externally, Applied, Loads, Evidence of buckling due to these types of loading would have been the collapse of the fuel assembly end spans between the nozzles-and end grids. The required end loading would also have had to have been of such magnitude that the top nozzle hold down springs would have collaps'ed. Examination of the assemblies showed no evidence of either of these conditions.
In addition', analysis showed that the conditions in the reactor during handling operations, which would be necessary to produce this magnitude of loading, could not be realistically predicted.
It wao concluded, therefore, that the buckling did not result ftom externally applied loads.
2.2 Inigggay,genggalgg,Lgaps 2.2.1 Plutonium,Assenglies (Zircaloy Cladding) Calculations show that the buckling of these assemblies occurred due to differential expansion on initial heatup. The buckling, however, would be of an elastic nature at' that point, disappearing on subsequent cool down, except for sone small amount of permanent set resulting from
i ! I I relaxatiot. due to irradiation. The large buckles developed by a ratchetting mechanism through a number of full
- temperature cycles of the core, the buckles growing by the ,
additional permanent set occurring with each cycle. f A full temperature cycle is from cold shutdown, through hot operating temperature back to cold shutdown conditions. i Examination of the reactor's thermal history, shows that only five such cycles had occurred prior to the mid-life detailed , observation of three fuel assemblies. Estimations of the pe rmanent set indicate that oniv small deformations would have been present at this stage; this is probably the reason that the buckling was not observed. Subsequent to this observation, eighteen additional cycles occurred which account for the large buck 1'es observed at the end of life. 2.2.2 E92,[9Slfd ,fssgg}}igs (Stainless Steel Cladding) Although the UOg assemblics which exhibited buckling were of the Core II design with six point contact grid support, the buckling in the assemblies is not attributed to In these assemblies, both the assembly frictivnal loading. can and fuel cladding are stainless steel. Therefore, any l frictional loading in the can caused by differential expansion between the rods and can at operating temperatures would be small and would result from t. ensile stresses in the can. In addition, buckling was only observed on the hot side of l l the cans and not randomly distributed around all sides as would be expected with axial f riction loads. It appears, instead, that the observed buckling in these assemblies resulted from a combination of thermal gradients ! across the assemblies and the resistance to bowing exerted j by the rod bundles in the grids used in the Core II design.
..--..-,---<-r,- -
- v. r ,m. ,.w- , ,--y.<, ,y, - ,,,m,w.,.-,,mv.,,-, ..ww..-r,yn.,
g v-, 3--er* ,~ ,, ~~y.--- _.p m .w .
-i
. A I
The six paint contact support used in the Core II grida provide an effective-built-in condition for the rods at each grid location and thus resist bowing of the assemblies through restraining moments on the rods. If thermal gradients i sufficient to produce bowing in an unrestrained condition were , j present in the Core 11 assemblies, the restraint offered by the grida could result in compressive buckling stresses on the i hot face of the assemblies. This would not be the case with Core I design assemblics where the grids provide a four point - support for the rods and little restraining moment. Examination of the core temperature distributions based on power distributions during Core II operation showed four assembly positions where thermal gradients would be sufficie.it I to cause buckling in Core II design assembifes and one locc:fon which was marginal. Of the four locations, one was occurled by a Core 1 assembly which exhibited no buckling. Two of the remaining locations were occupied by Core II assemblies which did exhibit b'uckling. The Core II assembly occupying the fourth location showed no buckling. In +.his case, the actual average temperature of coolant flowing through the assembly is in question. From instrumentation in the 3 x 3 test ' assembly (503-4-29) which was suspended in the 9 x 9 assembly at this location, coolant temperatures approximately 20*F below expected were indicated. Because of channeling effects through the 3 x 3-assembly, the indicated temperature would be basically the discharge f j temperature from the 3 x 3 assembly and would reflect the ef fect of deleted fuel rods in the assembly. however, the low temperatures in the 3 x 3.would also tend to reduce coolant temperatures in the 9 x 9 assembly and thus' reduce temperature gradients acrosn the assembly since these are_directly related to coolant temperature, Although the temperature effects car . be accurately predicted, it would appear that buckling did not occur because of reduced coolant temperatures. 9
- . - - . ~ . - - - . - - - . - . _ . - - . _ - . . - . - . - - - . - . - . . - _ - - - - - . _ - -
d r The last of the three Core II assemblies which exhibited i buckling was in the marginal location where the thermal gradients were not sufficiently high to predict buckling. ! The buckling in this case, however, was very minor and localized and could possibly have resulted from a local weakness in the can (a thin ligattent or out of flat condition). l i i f I l l . i I 9 l l l
- - 10
\
~.-,,,---,,_,_,,...,~,.-,--#~.-> -
. . - , ,-..,-,--A-w--- .,- . - ,.l.,,. ...- , ,,-.-_,..w,,,......~.,.rme,.~..-~.-,.v..-ec,..mm..-.. - - - - , , - , + - .,
+ -
- 3. 0 Ugpl[}p3}}gy,9{,[93{,j)),6)![Up[![!,Jg,[3fy[yJ, ppg [Ljyg The center f uel assemblies in Core III were of Core II design and I
contained Zircaloy water tubes and/or Zircoloy clad fuel. Therefore. j they would experience high friction loads due to differential expansion between the stainless steel assembly can and the Zircaloy rods and would be expected to buckle. To prevent buckling during Core III operation, the Core II assembly design required modification. There were two possible approaches to assembly modification to prevent I the buckling:
- 1. Increase the stiffness of che can sufficiently to withstand imposed loads.
- 2. Decrease the friction load by reducing the normal force applied to the fuel rods by the grid springs.
A combination of both approaches uas used. The spring contact force was reduced to the minimum which would not risk f retting of the fuel rods. However, this minimum contact force would still be sufficient to cause buckling. Therefore, the can was also strengthened. 31 h29sp,pa33fgg,3sepEb }}gs The fuel assembly has been strengthened by replacing six Zircaloy ( water tubes by six stainless steel ones and spot velding 0.028 inch thick angles between the can and the tubes. The one inch long clips previously used to fasten the two halves of the enclosure have been replaced by full length clips in the spans between grids. j A cross-section of a repaired loose lattice assembly is shown in j Figure 4. The grid springs have been reset to give a nominal 6.5 lbs contact ! force compared to the 15.5 lbs force previously used. The combined I l effect of the changes produces a safety factor of 1.5 between friction forces generated by the grids and the buckling strength of the cans. 1
. i :
$$(ET
,..,.C..H,, SHEET WESTINGHOUSE ELECTRIC CORPOR ATION f .-~.
I . f (' ! ( ( ", f' ? N)' (\.. / N,3 ) .../ '%_,) m e m m -. x l O O Os O O
':^:
p., O.
, s . _ , .
- m. ---
-3 ,
k, v v s, .
~
}
[A. . , ,
- .s , . .
3 (. - ). . , C/ s ( (, 3)
)
..' (d) (. ....
. ,, f.- m .s ,m .s ...
f ..
.3 s ., ..- .d N_ .....
(, ~' . ', ;^~1 l i w Q Q Q. s - ,, g FIGURE 4 Cross-Section of Repaired Loose Lattice Assembly i
~ 12 -
r 3.2 koad,,To}}ow Assepblies
, A slightly different method has been used for the load follow assemblies to.obtain the required strength. A solid square stainless steel bar, with two angles's' pot velded between the bar and the can, is used in place of two stainless steel clad fuel rods in opposite corners of the assembly. A full length angle 0.05 inch thick has been spot welded to the inside of the enclosure skin at the center of each long span between grids. Full length angle clips are also used between the ends of enclosure halves as was done with the loose lattice assemblies. A cross-section of a repaired load follow assembly is shown in Figure S.
The MAPI assembly to be used in the periphcry of the core will be similarly treated. The same buckling strength f attor has been achieved for these assemblics as in the loose lattice.
- 3. 3 Peripheral,p02,Assegblies Because of the large thermal gradients across these assemblies in Core Ill, only Core I type will be used.
An examination of the thermal gradients across the assemblies in the whole core shows that, although no buckling will occur, the worst gradient will produce a bow approximately 0.015 inch over the length of the assembly. This will always be of an elastic nature and will cause no interference problems. 3.4 Thergal,yydraulic, Considerations, The modifications used for both the loose lattice and load follow assemblics have not compromised the thermal-hydraulic performance in Core 111 operation. For both types of assemblies the minimum DNB ratio will not be below the current limit (1.30) specified in the operating license for the Saxton reactor. I I
(KETCH $NEET - _een non WESTlHGHOUSE ELECTRIC CORPOR ATION i I
'5DOO ; _
>o o v y vc o. v 9
~ ^
00000 O O *O b) r. L) '^ ^, ^.. )
^ ^
- D O c O O c C Om l
O O6 0 C O D 'D .? . 00 0 0 00 000 0 0 0 0 0 0 0 0 C, C O O D_ Q O O RE FIGUKE $1 Cross-Section of Repaired Load Follow Assembly 9
- _ _ , . . _ _ _ _ _ . . . _ _ _ _ . _ _ _ _ . - - . - - . _ _ . . _ . . _ . . . _..-.. . , _ _ . _ _ .__ . _ , - ~ . . . _ ,
. - _ ~ . - - . - - . - - . _ . . - _ . - . - - - - . - - - . - . ~ . . . . . _ . . - _ -
e se ,, g With the modifications discussed herein buckling is not anticipated ! in either the new or irradiated assemblies used in Core III. Assuming buckling did occur, however, up to 0.060 inch inward deflection of the can surfaces can be tolerated without danger of exceeding the minimum allowable DNBR. Deflections of this magnitude in the outward direction could also be tolerated without problems, i.t., binding of control rods would not occur. The fuel assemblies will be inspected at mid-life for any evidence of can buckling before reinsertion for continued reactor operation. In addition, the control rod scram times are normally checked every six months as a' check against gross buckling. 3.5 ,1ju elea r, Aspe c t s The modificat'sns proposed will have no effect on the operation of the load follow assemblies. However, the addition of stainless water tubes in the loose lattice assemblies will reduce the power output of these assemblies by approximately 3%. This may be alleviated slightly by increasing the nominal power output from 26 MWT te approximately 26.3 Indr without seriously affecting any thermal hydraulic or nuclear margins. 3.6 Health,gnd, Safety The modifications to Core Ill assemblies of Core 11 design will prevent buckling and present no hazard to the health and safety of the public. The assemblies will be given a detailed examination for buckling at mid core life. I
.u.._.____._._ ...~.,.... _ _ __.-- _ _ ._;. _ . _ ,_ ... _ _ _ ..,_._ ....__ _ ._}}