ML20209B994
Text
do Ju 4[
M
/
! p ara 8
4 g 'i NUC(cfc o
W UNITED STATES A/
f /8 ',
g j;7 EAR REGULATORY COMMISSION p/ p.54MASHINGTON,D.C.20555 Y'
r, n;
yr
}
k pk
/
4' Ek4J983
.m
) -
MEMORANDUM FOR:
o ert u. nosnak. Ch3ef
--[ 3,(
p ~ h p.. 6 J
M Engineering Branch,'DE b
THRU:
Frank C. Cherny, Section Leader
, Mechanical Engineering Branch, DE pj Qp FROM:
Mark Hartzman Mechanical Engineering Branch, DE
SUBJECT:
TRIP REPORT - MEETING WITH ROBERT L. CLOUD AND ASSOCIATES ON 1/26/83 IN BERKELEY, CA
Purpose:
on The purpose of this meeting was to obtain clarification ned a number of items related to ITR #12 (DCl-IDVP-Piping) and ITR #17 (DCl-IDVP-Piping, Additional Samples).
In addition, a number of topics were also discussed to provide clarification on the overall IDVP.
The meeting was attended by myself and P. Bezier from BNL.
The following items were discussed:
1.
BNL ouestioned the appropriateness of the.Hosgri 20 Hz rigidity
~criterion for supports and equipment. Aside from the fact that it is in the DC design criteria, RLCA felt that it is a reasonable criterion since the amplification in the Hosgri spectra is essentially one (1) beyond 20 Hz in most spectra used for the piping analysis. (This is however not true for all problems).
ber 2.
The comparison of support loads calculated frem RLCA and PGE show significant differences in the values. The values calculated by RLCA usually exceed those by PGE by significant amounts, for example, as shown for one problem sample in Table II.
Furthermore they do not show any comparison with allowable loads or stresses.
We requested clarification on this issue. They stated that the design loads are those which were given to the support designers to design to an allowable stress.
Since some of the support loads calculated by RLCA exceed the PGE loads by a considerable amount, there is no assurance that they will not fail. Large bore supports are being addressed in a separate ITR.
The issue was also raised as to the presentation of the results.
RLCA calculated the deviation from the formula (verification-design)/ verification whereas we felt that they should have used the formula (verification. design)/ design. We believe that this is a
/
r
\\>
ywen}t#
-2 more appropriate approach. An example of our concern is shown in Tables I and II for a problem (110) taken from ITR #17 where the mode of presentation shows significantly different results. Their response was that their approach was presented to the Staff some time ago and that the staff sh.ould have objected to it then.
However, they agreed to use the form which we suggested. They also stated that large bore supports were being addressed separately in a different ITR so that the presentation of the results was not in itself of immediate significance. One of the purposes of these ITRs was to determine sources of significant differences between RCLA and PGE stress and support results and not specifically to determine if they satisfy the safety criteria.
For equipment nozzles the same remarks apply.
RLCA intends to perform a separate verification of the safety of these nozzles.
Future corrective action on nozzles and supports is also described in more detail in ITR #8.
3.
In problem 110 no spectra were available for the analyzed piping where they are attached to the reactor coolant loop. Apparently PGE used the RCL motions to determine the seismic stresses.
RLCA used the building spectra in addition to the anchor motion to calculate the pipe stresses and support loads.
In this case these exceeded by a considerable amount the allowable stress and the design support loads.
Although RLCA issued a type A error report we discussed'this to determine the procedure being used by PGE for reanalysis.
RLCA stated that they had not received the PGE results and therefore reserved comment then. The same issue of presentation of results was also discussed (ref. item 2 above).
,4 Other topics which were also discussed:
a.
Interpretation of response spectra curves at various elevations.
b.
t!o E01's associated with differences in spectra were specifically issued. We questioned this. They stated that these were usually combined with other E0I's on the same problem.
c.
Themethodsforcalculatingthemaximumacceleration(thesameas that used for calculating the forces),
d.
Boundary support condition used in the analysis of the RHR pump in problem 103. The question raised was whether pump lift off can occur. They stated that the weight of the pump is sufficient to keep it from lifting. We examined the calculation and it seems to be thorough.
e.
In problem 104 the natural frequency of the heat exchangers was determined as 9 Hz yet PGE assumed that it was rigid.
RCEA,even though they disagreed with the assumption, also assumed it to be rigid. The reason for this was that they were aware that there was a problem with the heat exchangers, but to be able to compare with s
....._.c
the PGE results they assumed the same condition as PGE, otherwise they could not have compared results. These heat exchangers are being verified separately.
f.
Criteria used for spacing the pipe masses around a support was stated as being the same as for the rest of the pipe. Support masses were not included in the models, g.
In problem 113 and 115 there were five separate analyses performed to represent five models of the steam generator bumper supports.
The results presented are an envelope of all five analyses.
W Mark Hartzman Mechanical Engineering Branch Division of Engineering cc:
R. Vollmer T. Novak D. Eisenhut J. Knight F. Cherny P..T. Kuo H. Polk P. Bezler, BNL M. Reich, ENL H. Schierling F. Schauer E. Sullivan
7-_,
J.
List of Attendees Ed Deninson RLCA R. Cloud RLCA C. Browne RLCA-Rob Foti TES Albert Leung TES S. Chim TES Peter Mason Bechtel Mike Tressler PGE.
Larry Shipley Bechtel 9
q
TA 6.2 t. C
.1.--
SEISMIC STRESS (psi)
D LCA PGandE Deviation
.' Node Node Veri fication Design l
Analysis
- Analysis, (Percentage)
.A B
0 615 460 52,903 3,156 94 +15 W /
0 B
620 462 48,492 2,947 94 8
B 625
- 463, 43,660 2,589 94 0
C 395 83 43,488 9,896 77 B
8 630 464 31,839 2,057 94 C
C 400 385 36,000 9,160-75 0
B 495 325 38,407 27,575 28 B
B 500 330 38,413 27,733 28 +39/
C C
410 385 30,356 9,160 70 C
C 455 397 35,422 11,864 67 C
C 450 395 35,297 10,328 71 8
C 425 386 34,244 12,560 63 8
C 315 104 24,112 1,632 93 B
C 415 386 25,937 12,560 52 8
C 320 107 23,393 1,538 g3 8
C 755 439 29,743 10,362 65 B
C 760 440 30,998 10,196 67 B
C 460 398 28,939 12,833 56 640 465 18,493 833B 95 t 'Z. I 2.0 %
8 ASeismic anchor movements included.
0
~
0ue to ZY seismic Due to XY seismic 0 Deviation = (Verification-Design)/ Verification x 100.
Values greater than 15% exceed acceptance criteria.
M Pe r e.e wh g e.
vet l u.x c.* l x LM d Mm (vc &&,un - b e vy)/De Gy
- l **-
Table 3-3 Comparison o f Stresses Veri fication Analysis / Design Analysis Piping Sample 110 3-15
~..._
.y l A 6 L-C I
,s<
Enveloped Seismic Loads C
Deviation Ve ri fi catio n Design Support Load A Number Direction Analysis.
Analysis (Percentage) 10-495L F
1.590 1,451 9
y
' ~
46-12R F
523 451 14
,x F
721 1,030 43 - MZ Z
46-7R F H.1.8 312 304 3
11-10SL F,x9,3 510 335 34 li 11-11SL F
1,707 164 90 4-Wl[s y
46-9V F
130 192 48 X
10-14SL F
520 530 2
7 10-785L F
162 287 77 X
46-10V F
268 234 13 X
F
(/
1.029 728 2 9
- U#.t Zs '\\
10-1SL F
384 X
10-2SL F
65 Z
46-11R F
637 179 72 e 19-Z y.
10-80SL F
585 1,147 96 - 40 "/,
7
~
537 80 10-795L F
299 X
^ Where F = Support force measured in pounds
.L!ndicates t' hat load direction is perpendicular to pipe Hindicates horizontal plane.
C Deviation = (Verification-Design)/ Verification X 100.
Values greater than 15% exceed acceptance criteria.
,,,,,,[.
t l
- e., e i l'1 V d d.. 9 Table 3-4 Comparison o f Support Loads Veri fication Analysis / Design Analysis Piping Sample 110 3 16
- -