ML20100M213
| ML20100M213 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 11/28/1984 |
| From: | Oldag D CYGNA ENERGY SERVICES |
| To: | Ellis J Citizens Association for Sound Energy |
| References | |
| 84042.03, NUDOCS 8412120240 | |
| Download: ML20100M213 (87) | |
Text
{{#Wiki_filter:< saevers 101 California Street. Suite 1000, San Francisco, CA 941115894 415!397-5600 November 28, 1984 84042.033 Mrs. Juanita Ellis President, CASE 1426 S. Polk Dallas, Texas 75224
Subject:
Responses to Cygna Questions from the Independent Assessment Program Reviews ' Comanche Peak Steam Electric Station Independent Assessment Program - Phase 3 Texas Utilities Generating Company Job No. 84042
Dear Mrs. Ellis:
Enclosed please find copies of responses to questions.from the various disci-plines associated with Phase 3 of Cygna's Independent Assessment Program. Feel free to call if you have any questions or wish to discuss the enclosed documents. Very truly yours,
/
D. C. Oldag / Administrative Asst's ant ajb Enclosures cc: Mr. S. Treby (NRC), w/ attachments Mr. S.- Burwell (NRC), w/ attachments Mr. D. Wade (TUGCO), w/o attachments Ms. J. van Amerongen (TUGC0/EBASCO), w/o attachments Mr. D. Pigott (Orrick, Herrington & Sutcliffe), w/o attachments
> 1 1
5 0$ O E E oS$$$505 A PDR -4 ( a**
- San Francisco Boston San Diego Chicago FhcNand
$ 't*
'Ev7 C Z Mrs. Juanita Ellis November 28, 1984 84042.033 Page 1 of 1 Attachments
- 1. L.M. Popplewell (TUGCO) letter to N.H. Williams (Cygna), CPPA-41,195, "Cygna Potential Finding Report (Fisher Valves)," October 2,1984.
- 2. L.M. Popplewell (TUGCO) letter to N.H. Williams (Cygna), " Telephone Conversation of October 30, 1984 between L. Weingart, J. Burgess and J.
van Amerongen," October 30, 1984.
- 3. L.M. Popplewell (1UGCO) letter to N.H. Williams (Cygna), "Cygna Review Questions, Reference Cygna telecon dated 10/23/84, Double Trunnion Support MS-1-004-005-C72K," November 1, 1984.
- 4. J.B. George (TUGCO) letter to N.H. Williams (Cygna), " Cinched U-Bolt Testing & Analysis Program, Additional Information," November 1, 1984.
- 5. L.M. Popplewell (TUGCO) letter to N.H. Williams (Cygna), " Telephone Conversation of October 30, 1984 between J. Minichiello (Cygna) and J.
Finneran (TUGCO)," November 8, 1984.
- 6. J.B. George (TUGCO) letter to N.H. Williams (Cygna), " Cinched U-Bolt Testing & Analyses Program, Additional Information," November 16, 1984.
1 a .. . - - TEXAS UTILITIES GENERATING COMPANY
- r. o. sox ies: on tw mosz. texas 20
' f
& N CPPA-41,195 [ Ld2/1
-(> [Mk bh October 2, 1984
, 4 CYGNA Energy Services 4 '
'gg(
101 California Street J* t S r sco, Califomia 94111 h. b M Attention: Ms. Nancy Williams 70//UQ
"~
h Project Manager L,CYGNA < C m ANCHE PEAK STEAM ELIW6 M STATION Rfoyo^ CYGNA POTENTIAL FINI ING REPORT (FISHER VALVI $r.sLocare:
/O /t 1/ Siu REF: CPPA-39, 7gfcno,: g Sc/ '
Gentlemen: MLE: #*I# M" C C4-ThefollowingissubmittedinresponsetoSYd$f[MNE'r " Report. This report involved several main steam relief valves where the actual "as-built" loading conditions were not properly qualified or con-fimed acceptable. The main steam relief valves (Tag No. PV-2325 through PV-2328) were subjected to an operability test. They passed the test and no pipe support rework was required. As previously mentioned in referenced correspondence, a review of Fisher active and passive valves with similar support configurations was conducted for potential impact. This review identified the following valves which also required qualification / acceptance to the "as-tuilt" loading. Active: HV-2185 through W-2188 FV-2193 through FV-2196 W-2397 through HV-2400 HV-2401A6B through HV-2404ASB W-4165 through HV-4176 HV-4178 and HV-4179 HV-7311 and HV-7312 Passive: TV-4691 through TV-4694 HV-5384 and 1-7800 1-8034 and 1FC-7812 1-7155 and 1HCV-014 The "as-built" loading /information for these valves was transmitted to Fisher and have subsequently been qualified. The only physical modifica-tion required as a result of this qualification was the replacement of five (5) 9%-1/4 snubbers with SM-1/2 snubbers. Although the 9%-1/4 snubbers were adequate for the "as-built" loading, Fisher requested that S6M-1/2's be installed in order to use an existing qualification report thereby expediting closure of the issue. A spit IssissN s ##* 7K24 m f *Tf Liff km Ef. Af'1Nf f' t'e ptf #*,4N F L.
l.
... -~
CPPA-41,195 . CYGNA Energy Services Page 2. October 2, 1984 , This issue has been addressed in accordance with Deficiency Review Report (DRR)-054 which identified the concern as a potentially reportable item.
, Our subsequent disposition to Significant Deficiency Analysis Report (SDF.k) CP-84-16 indicates no adverse safety conditions would result if the item IL.:.ined undetected and the issue is not reportable under the provisions of 10CFR50.55(e).
Please contact this office in the event additional infonnation can be provided. Very truly yours, TEXAS UTILITIES GENERATING C04 PAW w L. M. Po plewell Project Ingineering Manager DIP /Jd/RPB/cp cc: ARMS D. H. Wade J. J. Van Amerongen File SDAR-CP-84-16 l l s u-
$ O 4L:
J. m n, M TEXAS UTILITIES GENERATING COMi%NY . kJ P. O. BOX 1002 G1 EN ROSE. TEXAS 76043 b* / st E'/oya/pf-CYGNA I Octob'r 30, 1984 -- M
---.EVQ Cf '.L DATE REC'D/LOGGI,pg __ f// g /
Loc no.: _ gy g
#IM' Cygna Energy Services 1 al rnia Street B /./ A . g CROSS REF. FILE _ A. / h _ g San Francisco, Califorr.ia 94111 v"-
Attn: Ms. Nancy Williams, Project Manager REF: Telephone Conversation of October 30, 1984 between L. Weingart, J. Burgess, and J. Van Amerongen
Dear Ms. Williams:
Attached please find the information requested to close out the Phase III open issue on the Fisher Main Steam Relief Valve. If there are any further questions or comments, please contact Ms. Jeanne J. Van Amerongen (Extension 500). Very truly yours, n L. M. Poppl well Project Engineering Manager LMP/JVA/bh cc: L. Popplewell l D.H. Wade R.E. Ballard J. Finneran J. Burgess l J. Van Amerongen i 4,,,,,........ ,,u : ,,,,,,,-es... we ... ..$i .
e W
. . - . - _ - . . . . . . ~ , . . . , . .-. ...... . .
. ( -
?-.
TEXAS UTILITIES GENERATING COMPANY P. o. Box leIE
- C1.E38 ROSE. TEXAS 18ee8 TSG-6494 September 19, 1984 CCL ma -
P. O. Box 12728 -
. . .-- . I Research Triangle Park, NC 27709 .
Attention: Mr. Stephen Lehrman , COMANCHE PEAK STEAM ELECTRIC STATION EQUIPMENT QUALIFICATION DOCUMENTATION SPECIFICATION 2323-MS-78 P. O. NO. CPF-11167-5, SUPP. 7 Gentlemen: By copy of this letter, we enclose to the vendor stamped " APPROVED - CHECKED FOR GDTERAL COMPLIANCE WITH PLANS AND SPECIFICATIONS. THIS DOES NOT RELIEVE THE VENDOR FROM RESPONSIBILITY FOR THE CORRECTNESS OF HIS WORK OR FOR FULFILLING THE OBLIGATIONS OF HIS CONTRACT WITH TEEAS UTILITIES GENERATING COMPANY", the following document: SEISMIC QUALIFICATION REPORT OF MAIN STEAM RELIEF VALVE FOR COMANCHE PEAK STEAM ELECTRIC STATION
. REPORT NO. A-655-84 '
Dated September 14, 1984 Flease refer to the above "TSC" number in all transmittals resulting from this letter. Very truly yours, MPeter
. ns Supervising Engineer TUGt0 Nuclear Engineering FBS CLW: RMacb i
Attachment cc ARMS (IL) L. Barnes (1L IA) FMG-M-997 E. Q. File (IL,1A)
- 3. F. Jones (1L) FMG J. Burgess (IL)
D.Headrick(IL) , r. . ,.,,,v ..r rm. : reure, u.u v ,, .,== r
, . , _ oo, w SEISMIC QUALIFICAilON REPORT. .
of MAIN STEAM RELIEF VALVE 3 ,. , , , ,
^--
g,7 COMANCHE PEAK STEAM ELECTRIC STATidN TEXAS ' UTILITIES GENERATING COMPANY Ov"Edr'2EsIEs"N"A*.'vj" - Report Date : September 14,1984
. CCL Report Number : A-655-84 CCL Project Number :84-1813.12
) m.. a TUGCO P.O. Number:
r O amovio nern as ae= _ CPF-11167-S Supplement 7 O not_aer. oven arvier a=o ersueurr s t von oaa coueuance wnw B y *~r (([ [. [ d
" ,,'l, ^"*,,,1 *" f,. ," *o^."
~
Blaok
,. . .... ... . ss,",' o, '*0,s ,.o,'s'i." In s
etn rutratuae fut oeuaafsons or His
- . . - ""{.
- g. . . . . . . . ' ",
7 4" I)iut as GENEAaftNG Co. D *
. \
c eked By . 5
. , ep
; -@e@yhrman, y ,o.......,,,,.
P.E.
,,,,,,;,,,0'p 0
CARO:h,
^@ N:
Approved By b- '
<- i C. Gary Higfies,.Nfi .3 ) j CORPORATE C TNG AND 5 DEVELOPMENT MPANY,LT i
b,ffo i d4 u ,,
,'9",,,$.D,
,.o$$,
Approved By TEXAS UTILITES GENERATING COMPANY Prepared By CORPORATE CONSULTING AND DEVELOPMENT COMPANY, LTD. P.O. Box 12728 - - - - - - - Research Triangle Park, North Carolina, 27709-9998 for TEXAS UTILITIES GENERATING COMPANY
i ( 1
. . \, \.
TABLE OF CONTENTS ,.,. ,,,
' ~ ~
- 1. INTRODUCTION.. ..........................'.................... 1
- 2. TEST CONFIGURATION........................................... 3
- 3. TEST RESULTS................................................. 8
- 4. CONCLUSIONS.................................................. 10
- 5. REFERENCES...............................'..................... 11 APPENDIX A INSTRUMENTATION SHEET APPENDIX B TEST ENGINEER'S CHECKLIST FOR VALVE TAG NO. 2-PV-2327 APPENDIX C TEST PROCEDURE 1813.12-1, REY. 1
- APPENDIX D TEST PROCEDURE- 1813.12-1, REV.1. SUPPLEMENT NO.1 APPENDIX E PHOTOGRAPHS APPENDIX F TEST MONITOR LOG APPENDIX G SEISMIC LOAD
SUMMARY
AND SNUBBER AXIS ORIENTATION 6 0 4 O q
q_ i 10
- 4. CONCLUSIONS
.-+'.......
The valve has been shown by the testing described in this report to be of sufficient structural integrity to withstand the loads postulated for the valve during the SSE. Neither the valve's pressure retention nor the valve's operability was impaired by the application of static seismic loads. No changes in the performance of the valve were detected during or after the static seismic operability tests.
/
6
o s *
', g Fe'Controis infomosonal, Inc, 20t C Center Street .
., MeAJtown, lowe 50154 Rione 515/754 3011 0 -
'MGM- 968 FISHER Fisher Controls August 29, 1984 Texas Utilities Generating Company P. O. Box 1002 Glen Rose, 21 76043 Attention: Manager - TUSI Nuclear Engineering
Subject:
Submittal of Fisher Controla Seismic Analysis and Supplementary Qualification Report
Reference:
Comanche Peak Steam Electric Station, Units 1 & 2
- Main Steam Relief Valves TUGCO P.O. No. CP-0078 and CPF-12049-S Fisher Representative Order No's 1-63500-A.-B and j 007D-LF93114 Fisher Project No's NES 170 and LSC 833C910 Fisher Qualification Projects 76NC02 and 84QN89 Gentlemen:
Please find enclosed for your information Fisher Controls Seismic
- Analysis and Supplementary Qualification Report FQP-5A-1 Supplement A, Revision A, dated August 22, 1984, as requested by the referenced purchase order,
)
&carely
)
q ,
= _
Larry C. nsack l Project ger Tisher Controls international, Inc. LCBanf Enclosure - 71 sher Controls Seismic Analysis and Supplementary Qual. Report, TQP-5A-1, Supplement A, Revision A, dated 8/22/84. (3 copies) ect Texas Utilities Generating Company i Attn: Dave Headrick Vinson Supply Company (Dallas) Aten Dick Jacobson
. i NtJ . O/ O
** Tazas Utilities Gene. ..:ing Company ) e3,/h,V3 -
')QQ i
Auguse 29, 1984 Page 2 m'- M ec: Fisher Controls . Aten: Woody Dickinson John Dresser 1-63500 Base Order (NHS 170) 7D-LF93114 (8330910) Desk i e D
.s e 4
G I J J
, , fLc . H-D 8Y TEXAS UTILITIES GENERATING COMi%NY -
P. O. BOX 1002 Ct.EN ROSE, TEXAS 76043 ( db
. . tLLkJuuh November 1, 1984 4 -
h, bt_ dOhX 2 lCYGNA _ 74041 P / Cygna Energy Services JOS 1:3
- kOM 101 California Street M Suite 1000 San Francisco, California 94111 DATE REC'D/ LOGGED:
Loc I:0.: j'gg! g RILE: E-/ / I-Attn: Ms. Nancy Williams, Projeci gagebF. FILE
- 2. / N7/- [A /r7[ ]
COMANCHE PEAK STEAM tLtt lM AL .41 AllON U CYGNA REVIEW QUESTIONS REF: CYGNA Telecon Dated 10/23/84, Double Trunnion Support Ms-1-004-005-C72K
Dear Ms. Williams:
Attached is TUGCO's response,to the above reference Phase 3 follow-up question. If there are any further questions or comments, please contact Ms. Jeanne J. Van Amerongen (Extension 500). Very truly yours, L. M. Popplewell Project Engineering Manager LMP/JVA/bh cc: L. Popplewell D.H. Wade R.E. Ballard J. Finneran D. Rencher J. Van Amerongen 4 ,.,,,-...v..r ,r u ree.,,,, ,,...i.,,,...... u -
~
CYGNA QUESTION:
, AS part of CYGNA's preparatory review for tte Phase 3 double trunnion open issue, CYGNA has found that one support, MS 004-005-C72K, may need further' review by TUGC0 for the weld between this supportthewasstanchion modeled in (item the pipe 8)stand the oad (item Since 13).
as analysis as two supports (i.e., as a moment moment carried by this weld. restraint),"there is a theoretical As shown on page 19 of the 8/17/84 calculation, more weld may be required to transmit this moment than presently exists. CYGNA requested TUGC0 to clarify. _TUGC0 RESPONSE: The original design of this support relied on the weld between items 8 and 13 to transmit the entire tension or compression load. 186901 This weld was designed for a translational force of lbs., and no moment. The revised analysis to include the moment restraining effect of trapeze supports was issued on 5/12/83 (ref. GTN-65560). The problem had previously been qualified without modeling these moment restraining supports in. stress was within allowable limits.In both analyses all pipe However, loads on this support increased due to the restraint of rotation. Calcula-tion by NPSI showed the weld in question was no longer adequate for the increased loads. Hence, two U-bolts were added to accomodate the increased loads. In the new support configuration, the two cinched U-bolts aid in transferring the moment into a force couple which acts down the axis of each snubber. 100% of the moment effect. The weld is not required to transfer attachments are designed for the increased loads analysis. from the latest Therefore, the weld between items 8 and 13 is not required to transmit this moment directly, and the connection is adequately designed. Even if the weld and cinched U-bolts were incapable of transferring the moment, the result would be less load either way.on the snubbers and pipe stresses are acceptable T i i
- . . . _ _ - - - . . - - ~' ~ ~ ~ ~
- = - - - - - . . . - . . . - a.
=
TEXAS UTILITIES GENERAT, P. o. Box 1002 G1.EN ROSE. TEXAS 'N043 COMPANY '[/ r-
'khect November 1, 1984 g
Ms. N. H. Williams i Project Manager i D CYGNA Energy Services 101 California Street, Suite 1000 San Francisco, California 94111-5894 [l/j4/[ (( p COMANCHE PEAK STEAM ELECTRIC STATION Independent Assessment Program Phase 3 Cinched U-Bolt Testing & Analysis Program Additional Information REF: (a) N. H. Williams (CYGNA) letter to J. B. George (TUCCO)
"U-Bolt Cinching Testing / Analysis Program Dhase 3 open Items", 84042.015, dated August 23, 1984 (b) Transcript of " Discussion Between CYGNA Energy Services and EBASCO Services, Inc.", dated September 13, 1984 (c) R. C. Iotti (EBASCO) letter to N. Williams (CYGNA)
" Additional Information As Follow-up To Meetings of 9/13/84", 3-1-17(6.2) ETCY-1, dated September 18, 1984 i
(d) N. H. Williams (CYGNA) letter to J. B. George (TUGCO)
" Status of Cinched U-Bolt Testing and Analysis Program", (
84042.018 dated October 1, 1984 l
Dear Ms. Williams:
Reference (a) contained CYGNA's questions on the TUGC0 cinched U-bolt testing and analysis program. All of the questions raised in that reference were discussed during the meetings of September 13, 1984 (Reference b), and several of the questions were resolved as a consequence of that meeting and the addi-tional information provided to CYGNA via Reference (c). Reference (d) requests additional clarification and/or information on TUGCO's stated position on those CYGNA comments / questions which have not been fully resolved. Accordingly, we are providing in Attachment 1 the information you request as answers to the questions posed in Reference,-(4),- --
! . . . m , eg l**.. sl 6 r7((f 6
) J% :n : hC(/1
' JAn: IiE.'D/LOCCDs // 7 M PC ::J. : k
;:L:: S . /.'/ S r . 4 f no:s nEP. PItE 3./ Sir. #4 M G u,,, ,,,,,x . ,, nu ,. , m m ,%,.
a n,< . . .m a .- (f
- ', 2.
Ms. N. H. Williams Page 2-November 1, 1984 We trust this will provide all of the information necessary for you to complete this portion of the Phase 3 Assessment. Please call if there are any questions. Very truly yours, TEXAS IlITIES GENERATING COMPANY lY
./
,/ bl
. George
.Vice-President / Project General Manager JBG/I.MP/JCF/RCI/gh -
cc: D. Wade (TUCCO J. Van Amerongen (EBASC0/TUCCO) R. Iotti (EBASCO) J. Finner.in (TUGCO)
J .' , .. _ , - n.
..', =u ATTACIDENT 1 TUGC0 ANSWERS TO CYGNA'S QUEST 10NS' 0F OETTER 84042.018, 10/1/84
- 1. Q. CYGNA Question 2-(transcript page 27)
-Re: -Classification of Preload TUGC0 has classified the pipe stress due to preload as primary in the first alternative and increased the allowable primary stress. In the second alternative, TUGCO classifies only the membrane portion of the preload stress as primary, and then neglects the bending (or membrane) portion in the primary plus secondary evaluation. TUGCO's basis for this is that the stress is non-cyclic in nature (Robert C. Iotti and J..C. Finneran Affidavit Regarding Cinching of U-Bolts, pp. 47-and 67).
In the first alte- .itive, CYGNA does not find sufficient justification for the use of 3Sa for primary stress limits. In the second alterna-tive, CYGNA does not find sufficient justification to neglect preload as part of the secondary range. In effect, while loads such as dead-weight or settlement are non-cyclic in nature, they are compared to the appropriate Code allowables. CYGNA believes that the total stress, due to all contributions at a point, should be considered in the evalua-tion. Therefore, what is the effect of considering preload as a cyclic ~ load? A. We are surprised that CYGNA still considers the issue of classification of preload an open question, since no questions were asked at the con-clusion of our stated position in this regard (see transcript page 29). To further amplify the explanation provided during the meetings (trans-cript at pp 27,28) we are providing below the results of a sample fatigue analysis,-conducted in accordance with Appendix XIII, Article 1153, Shakedown Analysis, for the 4-inch Sch 160 pipe. The alternating stress, Salt, is given by S = Ke Sn alt where Sn is the peak stress which equals 64.16 KSI when the effect of preload is included, and Ke is the simplified elastic / plastic damage factor. The latter is given by the equation Ke = 1 + ((1-n)/n(m-1)] ((Sn/3Sm)-1) where for the material in question, m=1.7, n.0.3, and 3Sa=50.52/ksi Ke = 1.9
OE %4 g uggg gg q GW/ i
'. ' " iEBA5CO.5ERVICE3 90 CORPORATE 3
==v-- = - :
m w,J, .... #e/2.1/sv -
.ospv.
g r , #-
^*'*
- l .,.uco.m
- esic -
cuam- Wk passac+ f = "_- . _ , Ia Oak avasect yu m.--
. it.; [L )6 # M2 T i a d S'- J "_ U-&DCT O M ~~ - -
. wha & Ma%g
^
~
g . i
, I-t
~
" ~ ~
-, r 4
1
#so 4ff 3
! . 317
.. r \
j . .
'\
\' .s
- .').
. , _\ . . ..
.< Kats
~
\
/ .
s e 1 I IS) i
- e .
o.
) .
......,...,....,,...,...,,,:,............,,....,,...;,;.,;,c.c.;,,,,,;....;.,.:...w,.......,,,,,_,,,,,,,,,,,,,,,;,;...,....,,,,,,,,,_
.ESA5CO.5EftVICE5 tNCORPORATE3 E.edt'h K h 4 .. ,. I o / 1 J s y ,,,, - ..
g
..... ..... - ;,, , -n.
i
. . . . -vo s.e enosner [-w h.
.. . s mL L so' w st.4 .e. U- eter e C..
t't< 4. M V . . a %.. .& k s.d&H w w k4 F't.9
; ,, in i .
+c _
,F l
- s. . . .
- j. . . s i
N. . l , N o tit
;. a
.) .
^
N l JEBASCO SERVICE 5 INCORPORATED d'f,/#3 l whC2:uoeh .... to/11/sy - g . . ..
- cane. - ..... _
,, , , _ .seg.
,) -
2 u . T o s.c .
.,,,,,. .c mA $ __ L_
...or w F*h -
A e c a st.J e.h, u.-- s e t r m . . 4 ,
- pf4 .d ase*F du. 2 " M
.%ha .d tL<T.y au1
. 1 l
m - set .
= - - - .
i 2, ano.
- -- sis m
..d,,%1I
/,e 550
/
./
/ .
n
/ . 4 =
% ') l
/
T - - l }
-usso l
i 0 l l l J- . 1
-----.--,-ene--o---n,,n,-w-r--w,..,,-.,..-----nne,,w,,,- -,.-,.,,w-. -
~ - , , , , , _
. , . i EBASCO SERVICES MCORPORATED bd LM1~h.. y
.lol1dsy . . . . .. - _
.. , , , _ = =n-
.\ cusse 'TO G-c
......,c ~ A h k
..m. w 4 .'h io" ue= st A Art u sect 4 2 -
-A
~
M kkk S k 6
.dn'1 Ay .2.6i
# O g
_J l 9
= .3*F0 e
/
/. .
. 4 k$
j
/ -
7 Er3513 sJ.
~
O o
ha ~.
._..._..,.....-.."e .--
'==***N-- --
'~5-- E
~
s
. .. . . .. . + s
.4: -.. ;.
y
- ,.5-
\
s
~'
;. -.} -
- .[N. . . .
i.
. - ,g
-s -
.v.
g =' . o
.. . . g : -
4 - ,\ - M) -
.=
.ih 3" * -r c'
8- . , s* * '
.- Q { .,
, q Q- .
* . * . .~
, : i
&.' ~
;L }, .
e
. ;+p .9
~ - --
1., :
.M h
-(
'y N i j: j : -;.
ii i. ;
;.j 2-j
~
,i
;., 'j i f.
q)g .i
- 4 i
I i j sP
** . A a i 1
~
- p .Q e m .e . 1 1..
~
n 5;; ~ i
@gs$,-RQ h
. . *. Q~w
' }Mf 8Uti .
- g.:gs gre:n
,,. 4;.. H;. :%g .
i : Qyg n.. y. a, g.
$ gr n i
4 Fi 4 .M 19; i N MSegh p tg N-a at,
- q. a
$*JSjh ,,,/for ea s,'c 1 4l:k k:hE d .dh, h: M i .i g4n ;te,.: j.
% .g m a s ;c -
g,. jh sd c i
. cst c;r- s /to- 1 , o
- 9. .~ .
0- N ia ni'#Rww.q;;;N g
- - - 1. I ' ,'
~
*. - a . rlrgg d 4.c w *o ; g :j .e HEW g% Li u g n gWa*-i ;j "o , !rffe. l- ; ywW w% %W 1 i g ; ;
- i. N ia a.34 this j[d m ges,
..u a. .~4 ,
t ~ 3,1 :
' of l C N g tvW C Q.R.9p'i ri, y w.m a, Mm ew y er, i
.s- ,
og cla o y jQ Lfg NN
;_O, A.
1 :jn i:. -! ! ', . i f.t o g
.t !- .
M 3. M$ i.R $s: % [ie g. h .= k;} Ncv.en'd Mi .q g bi-3
- i. :-
5 .cs.- s,; , . y4.! - 1 -w%. gi.Q $. *. I- ie. e 6-9 .a g. ; T;y d . -. hwi p- Q .h'.k:, O {
.[ 3
. f'
- j%fy }
-1 L .4 f t g
!3%I .;. p
. m
- ..E lr '- 1 ~
l , s, l .~g g il .j[I 1 , ,. f' fI3 1
'(7 8 j h.
y J ~I '.~j,Q i i J ol h
L . a_._.__ x _ . . . _ . . _ . . . _. -
..s The conditions which permit usage of the procedure above are all satisfied. Namely.
(a) The stress range excluding thermal bending' stresses is within 3Sm. As previously discussed with CYGNA, this stress range is approximately 35'ksi and hence within 3Sm. L(b) The temperature is well below 800 F. (c) The material has a ratio of.specified minimum yield strength to specified minimum tensile strength of 0.8. (d) The maximum allowable thermal stress condition of Appendix XIV, Article 1410 is satisfied since in this instance y'~= 175 kai for x = 0.16. The alternating stress S dt is then equal to 60.95 ksi. For this alternating stress, the allowable number of cycles is ap-proximately 8000 (see Code Figure 1-9). CYGNA provides no indication of how many cycles of preload should be considered, and we have already stated our position that preload should not be considered a load that is cyclic in nature. Intany case, even if we consider that the 200 cycles, used in our Affidavit at p. 71,. are still applicable to the case with preload, the incremental usage factor is. U, = 8 = 0.025
- which again indicates that the integrity of the pressure boundary, 4
based on fatigue considerations, would not be significantly affected l -by the localized U-Bolt effects, regardless of whether the preload is considered as a cyclic or a non-cyclic load. , Finally, we question the choice'of words used by CYGNA in regard to l the first alternative chosen by Applicants to assess the acceptability [ of stresses in the pipes. l. L We'do not understand what CYGNA intends when it states that "it doesn't l find sufficient justification for the use of 3Sm for primary stress L -limits." Applicants have clearly stated that this is one of the alternatives used in lieu of direct guidance by the Code and that.the use of a 3Sm limit is prompted both from inference from the Code (see footnote 23 of Affidavit at p.54) and the fact that the preload stress
'has some of the characteristics of a secondary stress,'with a prepon-l derant portion of thestress being due to a bending component (Affi-davit at 53). We do not understand what additional justification
- ' CYGNA would want. Dr. Bjorkman himself has stated that he would l
i- consider 3Sm to be the proper allowable for cinching alone and also
- cinching plus thermal (Tr. at 12996). If CYGNA has -justification for not using 3Sm, then CYGNA should state such justification. Otherwise, we'are at a loss for replying to a question which we do not understand.
l
.a
- 2. Q. CYGNA Question 4 (transcript page 32)
Re: Use of 2500 F for 10" Pipe CYGNA has reviewed the thermal operating data for the RHR systems and has found that the inlet and outlet to the RHR heat exchanger can be at 3500F. This can occur under-normal (inlet) or upset (outlet) conditions, both of which must be included in any analysis., Please justify that the preload and stress levels due to a 350 F insulated pipe are similar to a 2500F uninsulated pipe. A. The answer to this question is best provided by first restating that the choice of 2500F for the 10-inch pipe temperature is a compromise choice wh.ich bounds the majority of the systems in the plant, and where used with an uninsulated U-bolt configuration is also representative of the case where the pire temperature may be 3500F but the U-bolt configuration is insulated. Second it is important to point out that there is a single cinched-up U-bolt which is used on the 10-inch portion of the RHR system. This is support RH-1-024-007-S22R which is on line 10-RH-1-24-601-4-2, which is connected to the outlet line of the RHR heat exchanger. The maximum normal temperature seen by the line is 2800F during initiation of RHR operation. Only under upset conditions, when component cooling sater may be lost, can the maximum temperature of this line reach 350 F. There are no cinched-up U-bolts on the inlet side of the RHR heat exchanger. laird it is germane to point out that the tests conducted on the 10-inch pipe specimens had a corresponding average temperature of the U-bolt equal to approximately 1500F. The 1500F is not necessarily the equilibrium temperature of the U-bolt, but the temperature reached by the U-bolt during the thermal cycle which required approximately 20 minutes to heat the pipe to 2500F. Finite difference thermal analyses-indicate that depending on the extent of contact of the pipe with the U-bolt and backing plate therewould be contact for a cinched-up U-bolt) the average temperature in the straight legs of the U-bolt may range from 175-1800F for little or no contact to 225-2300F for well established con-tact, when the U-bolt is uninsulated and the pipe wall temperature is at 2500F. The U-bolt portion in contact with the pipe would be es-sentially at 2500F. For a 3500F pipe with an insulated U-bolt con-figuration the corresponding U-bolt average temperatures in the straight legs would be about 3100F for the case of poor and good i contact. The curved portion would be at 3500F in either case. Re-l sults of the heat transfer analyses are shown as exhibits 1 through 4. l The effect of the temperature rise on the clamping forces acting on the pipe and the U-bolt for the two cases of 2500F pipe, uninsulated i U-bolt and 3500F pipe, insulated U-bolt, can be estimated by comparing l the relative growth of the pipe to U-bolt for the two cases, neglecting ! any deformation of the pipe. Since only relative growth is pertinent here, the one-directional growth of the U-bolt due to thermal expan-l. sion given as Yi where Y1 = t(4TL
where L is the projected length of the U-bolt which is given as 2R and 6 T is the temperature differential between the average U-bolt temperature and ambient (or a reference temperature), is compared to the diametral growth of the pipe, Y 2, which is given as Y2 = v( ATD
.The worst case relative expansion will occur for the stainless steel pipe and the carbon steel U-bolt. For the 10-inch pipe (10.750D),
coefficients of thermal expansions N,= 6.4 X 10-6 in/in/0F at 180-2300F or 6.65 X 10-6 at 310-3500F and % = 9.4 X 10-6 at 2500F or 9.53 X.10-6 at 3500F and a reference ambient temperature of 700F, the relative expansion for the two cases considered, i.e., 2500F pipe with bare U-bolt, and 3500F pipe with insulated U-bolt are as follows, where Case a refers to the instance of good contact between the pipe, the U-bolt and the backing plate (as would generally be the case for cinched-up U-bolts) and Case b refers to the instance of poor contact.
- 1. 250 F Case a: y = 0.00666 in. Case b: y = 0.00838.in.
- 2. 3500 F Case a: y = 0.00920 case b: y = 0.101 in.
- 3. 2500 F Test: y = 0.012 Finite El. An. y = 0.0141*
(* Finite Element Analysis used 2100F) As seen from the above, theoretical steadystate heat transfer analyses would predict that the case of 3500F pipe expanding against an insulated U-bolt could result in a differential pipe expansion which would be ap-proximately 30-40% larger than could be expected for a 2500 F pipe with uninsulated U-bolt. However, both the tests and the finite element analyses have been conducted in a manner that would encompass the case of 3500F, insulated U-bolt. As seen from the third row of relative expansions, both the test (by having a maximum U-bolt temperature of 1500F) and the finite element analysis, which used,a pipe temperature of 2100F but maintained the U-bolt temperature at 70 F, would yield relative expansions which are significantly larger. Another point to be discussed, is that the test has provided informa-tion on the transient thermal expansion differential between the pipe and the U-bolt. As seen from a sample of the raw data which is at-tached as Exhibit 5, the maximum tempetature differential between the pipe and the U-bol2 occurred when the U-bolt had reached a represen-tative temperature of about 100-1050 while the pipe had been heated to 250-255 , a difference in temperature cf approximately 150 F. This difference is well simulated in.the finite element analysis where a constant difference in temperature of 1400F. It should also be re-membered that for these temperature differentials, the amount of r Stress Chtsed by the thermal expansion is not very significant.
- 3. Q. - CYGNA Question 9 (transcript page 130)
Re: G&H Sample Size for Piping General Stresses TUGC0 has committed to provide data on the size of the Gibbs & Hill sample (transcript page 130).
=
- 3. A.. The sample size taken by Gibbs & Hill is given in Table 3 of our Affidavit.- frhat table includes all of the stress problems reviewed by Gibbs & Hill to provide random information to judge the adequacy and conservatism of Westinghouse's method of determining maximum piping moment stresses in straight sections.
It must be born in mind that this was the only purpose of the Gibbs
' & Hill review. .The purpose was not to determine what the maximum straight pipe stress is per pipe size.
- 4. Q. CYGNA Questions 6, 12, 18, and 19 Re: CYGNA Question 6 CYGNA has not received the A-36 steel stress relaxation graph and published report on stress relaxation (transcript page 77) nor a copy of TUUCO's answer to the NRC on this issue (transcript page
. 81). This information is necessary to complete our reviews.
A. The TUGCO's answer to the NRC on this issue has been submitted on .
' September 24, 1984 and a copy should have reached CYGNA. In any 4 case, we are summarizing below the pertinent portions of the an-swer, i.e., that relating to the A-36 stress relaxation.
There is scant data available on strain relaxation properties of SA-36 material. Some relevant data is reported in ASTM DS60 " Compilation of Stress-Relaxation Data for Engineering Alloys," for material hav-ing the same composition as SA-36 steel (note that this reference does not mention the material designation). Unfortunately not much data is available directly at the temperatures of. interest, i.e., less [ than 5000F although considerable information may be inferred from the data at the higher-temperatures as will be discussed later. In fact, only materials 2 and 25 have data at room temperature. Material 2 has the proper chemical composition but its physical properties are significantly different from those of A-36. Material 25 has physical properties similar to A-36 but does not quite meet all a f
-y es.--,-. ,y.., ,,v--- yw---,,w- -
, . - - - -v--- --, . - - - - ,v,, - *,, ,~,,--w, -- , i,,,~ r , , - - -----n.,,--,-,-y.-,v,--..-.-.+,,-,ef,---, - -r-m--
-[ s.
of the chemical specifications. Figure Al shows the stress strain . curve of material 25 at various temperatures within our range of interest, i.e. less than 500 F. This curve is used to illustrate the meaning of material relaxation (as opposed to overall mechanical relaxation which will be discussed later) for " monotonic loading, i.e. noncyclic. For the material to relax, plastic strain is required. Ferritic steels like A-36 exhibit a well defined proportional limit at which plastic strain begins. , The yield strengths of these materials are given at the 0.1% or O.2% elastic strain offset (in general it is the latter, although , for material 25 the former is used). In figure Al the details of the stress strain curve between the proportional limit and the yield point are not shown. From that figure, if the material is-strained below the proportional limit no material relaxation-will occur. Strains in excess of the proportional limit will result in relaxation, the amount of relaxation being proportional to the amount of plastic strain (or volume of material that has yielded). At room temperature the strain corresponding to the proportional limit is about 0.075 percent. At that level of initial strain, therefore, little or no relaxation should be oxpected. Figure A4, developed using the information on Material , 25 of ASTM DS60, shows that the relaxation is negligible. At 532 F, the strain corresponding to-the proportional limit point is 0.065 percent. Since the material 25 has been strained to ,.
.075% relaxation should be expected. Moreover, the heating of the material from room temperature to 532 F and the return to
~
L-s. room temperature contributes to relaxation. How this happens is explained by Figure A2, obtained via private communication with M.J. Manjoine, one of the authors of ASTM DS60 and a recognized authority in materials behavior. This figure is an expanded view of a portion of Figure A3, also provided by M.J. Manjoine. Figure A3 deduces the behavior of ferritic steels like A-36 at the lower temperatures from the fact that the behavior exhibited at the higher temperatures (above 700 F) for which the data is available is the same as that exhibited for mild austenitic steels which have data available at all temperatures. The behavior of austenitic steels is shown in figure A7 which is taken directly from reference 4 (see p. 27). As figure A2 shows a material which is strained to or above the proportional limit will lose load at constant strain simply as a result of the lower yield strength at temperature and the higher modulus of olasticity at room temperature than at temperature. Thus, if material 25 had been strained to yield at 532 F, upon its return to room temperature it could exhibit 35 percent of its initial otress. This would occur upon return to room temperature regardless of whether " material" relaxation occurs. If the I material is maintained at temperature, loaded for suf ficient time, material relaxation would also occur. This can lead to an* h cdditional 15-20 percent loss of load. However, for the latter time is needed to redistribute the load. Although we do not know for a fact, it is fairly obvious that the material relaxation characteristics of material 25 at 532 F must have been determined
~
g. at temperature, since as figure A4 indicates, there is some twenty percent relaxation. Similar significant strain relaxation should be expected at all temperatures for initial strains of 0.225 percent, and this is indeed the case. If the applied load results in a stress below 1/2 of the yield strength at temperature, the corresponding strains would be well below those corresponding to the proportional limits, and thus no relaxation should be expected. So far only monotonic loads have been discussed. To complete the discussion of material relaxation, it must he pointed out that the stress strain curve for steels are different between the cases of monotonic and cyclic loads. For the monotonic loads discussed so far,. the point at which mild-i i ferritic steel materials begins to yield is higher (by . l l l approximately 15 percent - private communication with M.J.
- Manjoine) than the point at which yielding will occur under cyclic loads.
The difference is shown in Figure AS. It is important that a distinction be made between " cyclic" ( loads such as are experienced by the U-bolts, whereas the load 1 can be cycled from a low to a high level without stress reversal, and " stress reversal" loads which are cyclic but for which the load causes the stresses to be alternatively tensile and compressive. The relaxation behavior for the two cases can be vastly different. Figure A8 (reference 5) shows that stress strain curve for ferritic steel under reversing constant 1
- s..
4 amplitude loads (reversing strain). Figure A9 (reference 6) shows an idealized curve for the kind of mild steel which is characteristic of both ferritic steels like A-36 and austenitic steels like A-304. Figure A10 (reference 6) shows the static (monotonic) stress strain curve and the cyclic (strain reversal) curve for a material like A-36. The cyclic curve is the envelope of the stress-strain curves exhibited during the cycling as shown by the dashed line of figure A9. It is important to compare the type of relaxation which one can experience under cyclic loadings with no strain reversal to those which can be experienced for the latter. To do so we will utilize Figure All, (provided by M.J. . Manjoine), which combines both types of loadings. In the case of cyclic loading with no strain reversal, the second cycle will have a proportional limit PL1 which is about 15 percent lower . than the monotonic proportional limit. However, if the cyclic is one of relatively large strain reversal (i.e., strains near yield here defined as .2% offset), then the proportional limit will be much lower as indicated by point PL2 in the figure. For strain reversal conditions, according to Mr. Manjoine there is little difference between the stress strain curve of ferritic steels like SA-36 and austenitic steels like SA-304. Thus, the material relaxation properties of SA-36 can be inferred for cyclic loads from those of SA-304 for which considerably more data is available.
Figure A6, reproduced from ASTM-DS60 (reference 4) shows the relaxation behavior of SA-304. It can be seen that for cyclic loading with strain reversal there can be always some material
- relaxation, but that for stresses below 1/2(y, the amount of -
relaxation is minor. Material relaxation, however, is only one of the parameters of interest in the overall relaxation of the U-bolt assembly. Relaxation of the assembly preload can be due to a combination of material relaxation and other mechanical relaxation phenomena that may manifest themselves during the various loading cycles, auch as wear, local yielding with load redistribution, etc. It is difficult to predict the amount of relaxation that might occur as a result of wear or. yielding of surface irregularities. It is for that reason that the long term, cccelerated vibration test was conducted, i.e., to simulate the number of cycles that the assembly would see during its entire lifetime of operation. It is possible, however, to estimate the cmount of mechanical relaxation that takes place due to local yielding, although it is impossible to tell how quickly it will occur since the time required for load redistribution depends on too many factors . Such overall estimates can proceed from a l knowledge of the stress state at each location of the assembly, I l which permits an estimate of the volume of material that might be et yield. This volume of material will relax over time, l redistributing load, and giving the appearance that the overall casembly relaxes. It is germane to estimate what amount of
relaxation could occur when the shank of the U-bolt is stressed . to a maximum stress of 1/2 yield strength. At such loads there - l l are portions, however small, of the assembly which experience l higher stresses and can in fact be at yield. These regions are - chown in Figure A12 as points A, B, C, D and E. Points A, B and C yield at the outer fibers when the U-bolt is cinched up and ) preloaded to relatively low value of loads as a result of otraightening the U-bolt legs. Yielding is, however, limited to the outer fibers near and opposite the pipe, and the material which yields occupies negligible volume, i l For consistency with future discussion of Westinghouse test data, we will use a yield strength of the material of the U-bolt I cqual to 36,000 psi, even though a,ctual material yield is about ! 45,000 psi. Test results obtained by strain gauges have all been roferred to the 36,000 nominal yield strength. When the stress in the s?.ank is equal to 1/2 the yield strength in the U-bolt chank area, for instance for the 10-inch assembly (refer to 4 Attachment 1 to the Affidavit) with the 3/4 inch U-bolt, the corresponding load is 7,956 lbs., which gives a threaded area ctress in excess of 1/2 of yield, i.e., 23,820 psi. However, as figure A13 indicates, the nut engagement results in stress concentration within the threaded area. Stress concentration can raise the average stress above yield. Since we have two nuts, a cimilar stress concentration profile will exist in the bolt within the other nut because of the nut engagement to the first For the 3/4-inch bolt, the nuts are 5/8 inch thick with six cne. s - . - - -
-~ '~ ~ ^~ ' .. --- -. . - .... . ~ . = -- :. . - . - , . - . - - . L. . . ~~ . ..
threads. Approximately half of the bolt volume within both nuts will have stress concentration in excess of 1.5. Thus, a total length of 5/8 inches will have stresses at or close to yield. The same is true in the other leg of the U-bolt. Thus, 4 about 1.25 inches of material out of a total of 31 inches will experience relaxation of the order 15 percent (relaxation from yield stress - see figure A2) if at room temperature. The remaining threaded area (approximately 5 inches) will experience l less relaxation since it is more lightly stressed. The amount of relaxation that it can experience can be estimated using figure 2, suggested by M.J. Manjoine. This additional threaded material would relax approximately 7.5 percent. Thus, one can approximate the overall mechanical relaxation,that would occur for loads [ resulting in stresses in the shank of one-half yield as 5 (.075) + 1.25 (.15) = 1.7%, or very low relaxation. 3.25 Perhaps more relevant than theoretical calculations to the question of when overall (material and mechanical) relaxation ceases for the U-bolts, is the actual data taken during the various tests conducted by Applicants (see reference 1). One such test is the thermal cycling test. Results of the thermal cycling test on the 4" Sch 160 otainless steel specimen indicated that the stress in the U-bolt was approximately 31,100 psi (or approximately 86.4% of the assumed yield strength of 36,000 psi and essentially equal to the cyclic yield strength). The total material would thus relax. i I
.., _ . 2_ _ _ __ . .; -
~
After nine cycles the residual stress was measured to be approximately 19,900 psi or 55 percent of the assumed yield strength. (Ambient temperature for pipe and U-bolt was essentially the same before cycling (105 F) and just before the 10th cycle (107.5 F). The U-bolt was heated to an average temperature of about 400 F (see page 16 of Attachment 3 to the Af fidavit) . From Figure A2 one can deduce that the temperature cycling would result in a relaxation of approximately 36 percent, of which the initial 25 percent would be due to the temperature cycling alone. The result of the thermal cycling test does in fact confirm that the room temperature stress before the thermal cycling, i.e., a nominal 31,100 psi, was reduced to 19,900 or a 36 percent reduction. . Another test which provides insight on the stress relaxation is the creep test which was performed immediately after com-pletion of tne thermal cycling test, without retorquing the bolts. . For the 4-inch specimen the microstrain measured in the two
- U-bolt legs at the ambient temperature before the creep test (77 F) were 856 and 775 microstrain for legs 1 and 2 respectively. (These microstrains correspond to a load of 4,870 and 4,409 lbs.) After the creep test with the ambient temperature being 91.4 F, the strains were measured to be 853 and 773 microstrain, respectively. When one accounts for the fact that at 91.4 there is a preload induced by the difference in thermal expansion between the stainless steel pipe and the carbon l
~~ ~ ^ ~ ~ ~ ~ .-.-....~.-l-- - T: ^
a l - steel U-bolt, and that had the ambient temperature returned to 77 F the preload would have been reduced by approximately 45 . Ibs., the final load at the completion of the creep test would be approximately 4,580 lbs. compared to 4,639 (or 1.2 percent 4 decrease). Since 4,580 lbs. corresponds to a stress of 23,367 psi (shank area), which is above 1/2 of the assumed yield strength of 36,000, this. decrease, if real and not due to instrument , uncertainty, would be due to the strain relaxation. The question y of whether it may be due to creep is addressed in the answer to the next question. For the 10" Sch 40 line, where the temperature is low (pipe 250 F and U-bolt 150 F) creep is clearly not a concern. The strains measure prior to the creep test (after the thermal I cycling test) were 283 and 280 microstrains respectively in legs
'I and 2 of the U-bolt (at an ambient temperature of 75.8 F). The initial microstrains correspond to a load of 3,625 and 3,578 lbs.
respectively. These loads correspond to a stress equal to 8,200 psi in the shank or 10,800 psi in the thread area of the U-bolt. In either case the stresses are well below the 1/2 yield strength, with the exception of highly local area in the thread , within the nut, and hence little, if any, relaxation should be 4 cxhibited. The strains after the creep test were measured to be 281 and 276 microstrains respectively corresponding to an average load of 3,567 lbs. ' 4 f
. :-. . _ _ . . . . . . . - . . . =_. . ._
The drop in load of approximately 39 lbs. is partly due to the lower environment temperatures after the test which was 66.9 F instead of 75.8 F. The drop in load corresponding to the 9 degrees difference is calculated to be approximately 11 lbs. Thus, relaxation (if any) was less than 0.8 percent. The seismic test provides further evidence of the relaxation phenomenon.- Initial information provided from the test, which is attached as Exhibit A3, indicated a reduction in load from 4,484 lbs. in both U-bolt legs to about 4,291 lbs. and 4,355 lbs. in legs 1 and 2 respectively, when the assembly was vibrated at 9 Hz with a constant amplitude of 7,000 lbs. This relaxation of approximately 12 percent could not be justified on the basis of the applied load which would result, coupled with the initial preload of 4,484 lbs. (50 ft. Ib. torque) in maximum load experienced by the U-bolt of approximately 6,100 lbs., and a i corresponding stress of 18,200 psi in the threaded area and 13,800 psi in the shank area. This led to questioning the validity of the 7,000 lb. load, and to the realization that the actual applied vibratory load had been higher, and to the results published in the Affidavit, which are included here as Exhibit A4. As seen in the Exhibit, the actual load applied to the U-bolt was in excess of 10,000 lbs during the peak portion of the cycle and initially in excess of 8,600 lbs. during the pull portion of the cycle. On the average the force seen by the U-bolt during the cycling was in excess of 6,600 lbs. (peak load of
, _ , , . - . . _ , . . _ _ . _ _ _ . _ , , , . _ _ . _ _ , , , - . , , _ , _ . ,_m.,,,,_._y ___..,______.______.m_ _ _ . _ _ . _ _ . _ . , _ , . ~ __ _ , _ . _ . _
v -- more than 8,600 lbs. plus preload of 4,484 lbs.) which would have resulted in a stress in the thread area of about 19,800 lbs. which is 11 percent higher than the nominal 1/2 yield strength, hence justifying the relaxation seen. Finally, the data obtained during the long term accelerated vibration test merits some attention. As stated in our Affidavit, the initial preload stress was equal to about 9,020 psi. After the initial reposition of the assembly which occurred approximately 5.15 minutes into the test (see attached raw data - Exhibit A5), and which resulted in an average loss of preload equal to 640 lbs, the preload was seen to decrease slightly, then increase again then decrease with a final preload being about 450 less than,the preload existing after the initial adjustment. During the period of time between the 4th sweep (21 minutes) and the 36th sweep (189 minutes) there was essentially no change in the preload. At the latter time is when the sudden cocking mentioned in the Affidavit on p. 30 took place, which resulted in some further preload decrease. Relaxation of the material discussed within the context of this reply does not change the total strain of the material. (See definition in 2 of Exhibit A2.) The preload at the end of the test is still sufficient to prevent loss of contact between the pipe and backing plate (see figures 17 and 18 of Attachment 1 to the Affidavit with an applied load of 1,500 lbs. and a preload of approximately 3,200 lbs.), thus the motion which resulted in further relaxation is most likely due to accumulated strain over
6 the more than 10 cycles experienced at an applied load of 1,500 lbs. These cycles represent the number that the support may experience during its lifetime, and hence the test results confirm that in spite of some relaxation, adequate preload would be retained throughout life. t cyclic plastic strain accumulation may occur at these loads, which are abnormally high for the period of time tested. An elasto plastic finite element analyses of a similar U-bolt, backing plate, pipe arrangement, conducted per an 8-inch pipe (same size U-bolt as the 10" pipe, indicates that for sufficiently high preload, the U-bolt can experience some i plasticity in the transition region between the straight shank and the curved portion and at the inner surface of the U-bolt apex. This occurs from the bending moment place on 'the U-bolt ' from the straightening action of the preload or full external This small amount of plasticity occurs even though the i load. average stresses through the U-bolt cross section is low, and in fact, for the particular case examined are only 2,000 psi. Under
- the large number of cycles seen by the specimen the accumulated plastic strain can result in sufficient permanent deformation to r
- permit relaxation. Also, wear and' yielding of surface imperfections can accomplish the same thing.
1 i m
J)
, ,, Figura Al
=. - ._ .
.._._.,~-: - - _. ~+-..m-.m. .
mm m.m
-- _ . .m..+_.+.
.m._-- _- 1. ._
4 .4j$: -..__ - -.-:
.-.w._..+o--
,.4++__ :o.++._
.e.e+4.t++ g e.. - -
..+.
; e, .-
----__:..o++++.. +.
., - -*m- __
-*+++,e*7f***+.P+ .=.41 + .k, . .- 999
***,s++.+ ;
+;- ,. _
4 .++o,. _o.
..- .:; ,. w++ .m.
- + m..--
/
. :.-+..
.-++.
- +*,.w.*++-
- ..+:'
9+e*+++.M*+++- _.+-+ .+ ; ; ; : .%.M
+++++,e - .
- + *,++++ :
, + + ,-._ _. --
w++
- n. .ww . , -- . + :
, * * . + + + + + + + * + . + . .
- --- , . - :__....._ . . . * ,+ .e.4
- .-.+4+u;,..,--.m. .
m *+w 4 &o++..*.++ + . * + .+o.+ . me :.+. ++4+ , _ .. , +
.. ...;+-,., .. ++ - +++* : - ; , . . w^: .+ + " + + * *
.w.*6t****+**** **,w.,*..:: * + + " -
+: :- - 4
.:+,++++-++: ; ,-- -
.w . aum.
,+.4.
. ..+ + + + + + . :_:: +.++,--..._::*
m __-
-+m4, ..4 + + -
4 .
+ -++ ;-F+*w - ,I +*+++ :- .
,w.---**.+mm,.++4,:!:-,-+..._
..+e+. .+++U +.,. +. m.+u .
44u + mo
.um ; -
+ + - .
+ ..,;;+ ,o .
- .-^^-- - ^m_:m ++++.:
;+ +..
++. m _. :++;
. .- : .+.
- .4+.+.+ , w.+o
- : . ++ +e,.++b.n.4,%+. +, . , * ..w+,+.o i ,+4.+.- . +++w.. ., .
.m m.eu,m.w.++.+4.+._..,
+
+
w+++
+
.*+ :
u+ _ *. - -:
.. ++. :.:1*: . . --
---+;
-_+._we.em.**,++++ w . ww
. .- ".t.++..'.. :-**
.+- . .._
._ +.:.+: -
7 :-.M y_,,,.
++.-.+-
. . - :++.+., .. ++. . _
--I.
.+'
- ^^'^^4
. _4_
++-.
- ++o - - - +-
+__
+:; , _
.+++4 --
- +.-. ---...+4+
_+e.-.-,.*.+***-*+.e++++++o .
+++.4+
. .++ m.. 4++ .++:
..,+.*t.* ..+. ++-
___ __ ._ :4 - -
++.-.;------- - -----+ * *^^^
++. . ::; :::;++ -
pg
;++;
w.+.+.++m.++..+:
- 4++
n.4 m +:-
+..
. :^
.... ..+
~-+.: 4. - -o .-o4o+I 1 m *+
---------t.4*...,+*wo*+.**.*++****.o::
^
gang J5 -
.. ..+ ..:++.--- +.
...*: - - .-- o,.,++4 we
-. .;:_:-- m
++M _. . . _:.+.;
.++;
+ -- + -+ ;+ ww
+ w+
....1.+.
+ aw . -
+4++:_ - -- ... .
++---
++e,
.:+o*++....m+4.+ ... . ++ ,m
+T +: +: .: . .: .; +_ +. . _ -
- w. - --
4 : . e.c o.. ,4++.++
- w-.-.+
+++.*9%.m++. m.u.4 +
-+ ,
;+<
. . ;4----
4
+4 ....+ + . m.+.w
- .4+.++
m+++.. - +: UT+. : : -. ., + . +-
.+++:.. ++^ ++
-.++ : .m._ + -
+.:e
----.++; ..- ,.---
- . - . +-
. * , ,.,.+. ,.
- ++ e.
+: - - - .._-_ .---.+::. +- :..+ ... .:
- ..+
..,: :_.+. +..
,-+.,
_y,._1- ,. . y^
-^
. . . : : 4.-
. - m. . . .o..w.++e,.*,t..m++**_
^9-,4 ,y_.4.
-, -. -, -.-. -* +:.. _-w : - - .----4^
- --.+ . 4.
..+.-***,,..s+m .+.
- +--
- : +.
w ..,+.++-- w+++.
.+.-
____m++ wo + - -.. . y .+* . -
-m--- .- <
Me...
.,,w+
_ . _ + *4. _M+ + . + . + + + _
+.,w.4
+-
- 4. . : ^:
- m. -- -
^^
- M *** t*+e r*-*-* **+=4* ***********+***+4 i.e,++,,w;. .. . . . .
- m. .._
,4--- -.-
... -..+__f. +. +..
.- . .--_ _ . ..._m g) - --+ >- . . .._ _ _ 4. m...4. ..+
**.m.+.5+&+=
r . .h, ::--- --
.pty +. + ... . . 97 _ ..
. , ,+. , ,. .:___ ,,,, .n..*. uem. .. .: 4+.- - -
- ++.: -,+.;--... ..
_:_-.-+.4,.:. m+, I
. -*-+++++ 4 - .mm 74 ~
.*4 ; ~*.4' ..
- : .* an.s +++o - ;::: . _ _ ,_ :
4m6+ , . : - , : : - _:, u. mu +: m .
- m,**+me. .. :. 9mwo*+. ----
... . .--+.: : +44 ,m- .. . . ..
+ + - + +
-+++ *
. .+ : :.
= -
+ ++*1m ,^. -_ y+.+.tt t_ -++ -- h-++h* ::.,w.,+ ** + + + "w*.e.+,*.-*++,+a.,.+m--- - -- - :* +**
-*: 4 T-- ****: .- _ _
^ ^^^^^
+++4 : --.
++ .,. . ;+ :------ --- :+:.;-- -- - -
.:. 4. ; - ..+ ys.G ,_- :: -.
. . g::. 4'. :
- - : 4. :
~
v 25 ._ ;I I .*,* - .++..!*".**.*.~. u +ou . .+:: .;. ;,.T.*f, - .: gm;
- +
m *+. <
.+ .:::__;--
+I . . +.o +e+++++ t.44,,;
; ^ .
- j
- .:*m:-*_+;;+++ ::. :: - ---.w*+m.--
+o w ++* : .
-m --
, i1 . ++*+++**++<
- ,.. ::;::.. .9
-:+ . * * -* -+ :**.* *+.*
.+ +.,* o o;.--
, . . - . +: -- --- - :,u,w++.;--: ++.* ;
- +- .--
; o .-
* - . .* ,++; -
o.,.
.++.o+ .d1++m.
- . . -o+++:... . . - --
---..:w+. + + + + ~ ~ .
M P.mimHI2 .
" ^ "
. p.
^^
M *" t * **"*- : ~+t+.+-++++**t**+*~+.+ ,
+.
.+
. +.+
- :-.^. :,+..,.*
.o.-
- mw.+ +
+.,
.___e. .+..m..-+-s .
.:.1. ... . . u t : - - - .++ . - - _
. 4.-p. LJ u I
*.m..4.mt .4y:e+.Qey.4- rg_.. .1.* ~ * ,-.- . . . - -+
, +. .. _ .
*T'-' ~. r ** .,.-:
wm.,. .
.. , ...- +4+4.f.by+4 b.*
-+
+.4.I.+C_
. .4_ _ . es ,
g-- g .n.... yy .. ... . u . _ . , .__-_ 4- *.+4.4-+4.4*++t**4+IA*E******NA**"".*+*+*+....****+f._73-.*+t*t*.*** y4.+.++.*. - - - --
.. - - - - + + + + I'.++++M#*** a,.- E**+**
-- - +:-
1
- +If***M.
20
- 3. m
.-e.*;
-+*.
7.+ :
,. _ _ ;+:
'+++.,a.- :_ :.
I.+.. .% - --- Q ,. , -
- - .+..;- t.- . m e wm.- ;
.-+--.
+.e, ; ..'+++++i E
gh. . . .. j _ _
- ++ -.?****++++g+----+* @,g;2g M. F , _____
.+ + :: - ;---. :
. + .o.*.+ ;,*' .+*.-..:o. .: . + - - : : .+.o
- o: .: .. + .:,+.. :;
g m -+
; 4+.-: .. - ..- :
m
- - .: , ,---., .4 m.e.,*mm,.- :..
- _:; . _ o:m .-
..o.+ m.h.o. .
U8 --- .+++:- *
-----:4.: + :--- --
^ - -
...w
-- ++ +w. -
- 3. g m - . . ,
- m .+ m1+._ *+.9.*+o.t+:.* - - -: --:+.
*.++,w
.-.--+. . -.
+.
. . : -,,.y +. + - +- ..- M..4.*+
- +-
-- ^
{ ;. : . _ - - -- - MI ---.- .:.. m , ..
._: w:: : w .-.- +- :--- : : * .*m,.
4 4 .; .: + m*. +*,e.,+* w P" C :
- , ....+; __ __
- : : ! *: : :!I.,.*+;: o t_ ". : "---
- **+
g{ . -- : *.+o +.*. .m . : . :.- o m.m , -: .- ..- . :. :. ._ _. .- 3 1
;+*-
i+t.+ ....9.n+***+ ::::::::.
- m.+. . w mm+++ m.u.m.-. 1.+.muu.+m. . - .- -.-- . :-- .
-.+..+...-.m+o..-
++ w :
. . , + . + . + + -
E, m .+:- 1
-----...:..:.;.:": 2::+o. .
.m.*,m . . + -
- .- -.+ -- : .
.o -
S ++ ,
..+:
+++.+..+eo.
m+m-m
.. . i . . . . ++ ww m .+: ....+.
- :+. .1 .+ -
..w:.. *1.m.+ .o 1 ;w: -- -+-
-+..--..+w:..-.----
..*.m,
- .:-e-
++.. _
------------n-*,.,+**.:.:..-.::--
+ *,*
.m
.~..m.m.+-- : ---- :
.m.
-m h -
+a-
- : : .+--
-m.+<
-a . - .
- _:.. . ++ ..+m++ .- - ::*.
. , - : - . - _ _ _ __+.;
+.
- :. : 4 ++. u .
* +,---
..- *+..,.+++..*o.,+,.:..L.4M..:.m-w-n..*.*7*.**.*4..+.
... -- ..-. .. . m...+ w ::::: ::: ;
o
.-+mu----- -
+
;+...~.+.::: . . . .
w: ---.-- . + fg
+
w .. m ; --------- nir._. .. :.,=:
.e
~.
; m . :.
+.--. : :::::.-
- -. ~.++m.~o...
m
.-....... a:m e
_ , . ~ . m..~.
.m
...m, .~ . .m. .~42:
me ::
. .+... .._.:: .,.._. o,-_.:::.. ::a.
..a. _...---- - . . m m
.m . .-mm.m m - .o
. m=:...t. -- ----- .=- . -_---
.~ .t _
i ._ m o*** ~ _f .
--- :. m
;: --.;4.; .. ~.t;..::-- :
.* a- ,.
**.....:::.1..*. .. . ****t:.m
.. .. .+
- m. ,
~ - : .:::: a..#.
*o..~..+.-.
.. +" +:::
. . . .m -
.-...u.h.
- .+.., ..
-:*..m~--+++*+++*** ***+
b.m: .~ o ~
- . : . : . m.e: m :. + 1*
+. .
....: -----* ..~-*.
++ - - - m - - .. -
m m
++
_. . _-.; ;: - - +"*t*** t +++ . .
~ . . ~
~. ..
-m . - .:_ : --
-~ ~. .. - -....a - 5
.m..- m~
t_ . u. .I*++*It+m *
++ . . *++2.. + + +.+ ..o. . . ++ . m.m.t.m uo ...
n..I.t
++ * + - + ~ + -
- o ; - ...
~ ~-- +
. ~: m: =-.-;.+.+.. .:-- .m
.: _ w..o t.mit.: .; .- : w: ..: ...:- :u -++ +.-. : u : : 3.~.**.**.+.. .. .. +.J++.+.:
f g+a.g..++. -wm.:.~.*m. . m m.* .~
. gg.m .: .m .: .: .: .:..:!. . .:.m.=
. .. . --+
+. +.. me. . +. a* *7! em. m.
'.-1::--a.+u
.:a.- s.k
...~.++.1.:
mo. .,*. ..------ ;;
...m . ~,~ . . m .. ..m,o.:----.:m:..~:
g +.y.
- ..- :-.-. .::- m. . ! : -- .
. to:::.. 2+.+.... .--.......?+ .. .
,. _; .g. - _: .---_.-.- :.
. _--=-:,.:-- ,.,.
m n.t:_ . - ----::'-._a.--................~.... i .: .. .
.m. .. ;:: m.~ .: .-.
- :.+ ttti. ..--- -m
.m-~~g.~
._ --~. - *'.t~*
1 u*t. . . _: -.
.. _. :. -... :: . . . ..+
- -. 1.mm ~. m. m. ..~.+.~...m..+++
- ++ ---
+ ++++ -_ +::,. + + +t.i
... +... .. ~ + .+ .. +.. + m+ . :+ +: +. - ::._: .m: : .: -_. ~~.. . ~-m~~- - - - : -
+ - - + - ~ - + ~ - - + - - +
_ . * .- +.
~.
- - . .m.1'I+.. .: m- 1*.+t++*++.++-+ +! - : +I .
.; .: .: : .. .1 -- ;:p _.a_.+ ..
m..m. Jin m
.m m.
m
..m.,.~-
. m m m.a.m: . .: .- . m .. - -
- ..m- : m m
. . + . ..: - .- ::: : .... - . _ .. ...
-i.m,-.. ..,*....+....+..4 -.~...~~.+..u.~:my..~~.~....m...mm._
-+ . - ~ .
.+ .. .
- , . .~. . 4 7 .. .__..--.
------..;.;.:--,+.i.--:_ ..*- t. . ~. .
.. m. . +
, . , _. ... _. - -:-: :_.. ~ ... .
. ~.*
a._ .. ,
.~._.~.*m..m_..,.I.~~...:.:; :
..- . . . - . .. -..m.m.~~
m .
- ...: .. ..:+
w...-...t..._.----_-.--------
.. .~ ...,m.,. ~.tn
.+m . . .. .
t .......m .-
.m..= _._
.. m. .~-.__~ .
2* m:.... m .. **
- m. - m. m m..~ .~+ +
u.. rt . ++?t++t?**?t***+1_*++t+++t*+++*+^-- ^ ::':;::!i7 . i*+++*t*++t+?*++h.*+++++++**-
.* * * * + + + + +
' 0/ . g O. ; O 2
- 24Z JJ' .n Z f . ?/
t.bthh ! & bEC .
'O Y)
... '. .,. N.
. . . . .. . Figuro A2
.., .. .. ..r.. ..A. .
_.._t........ ...... . .. ..._,. .. ... ...._..
.. ..~. ..2. .
.. .. . . . . .i f.. ... . .
t.
.* .....~...* . .
. . t. - . ..i. . . . . . , .
~% a.s g , y_,* * . ..
. a. .... ,.
. u...
. '.! .:t. *..
- t. . __
' A.ggg I .
- t
.- 1
,*. - - ,_g..- *
.....e __. . .y . ..
g #
, . . . - .-. . , _ . .g .
g' ... . .. _..
*?
l
- Dg .
l f- -g
,- i .- i.
- V . . . -. k. ..
.. Jo. _ . . . . . . . . . . _ . . _ . . . . _ _ ... . . . . . .. . . .
{2*T
. . . . . . . . . . . . . e .A _
@M, *
.. f. ..
! .A'A : e i 9 re --
.. eee-c-1 -
i
......n...._._....... . . . _ . . F+. wu., p,. m e,. . . . .i.
r - i . 8
.i X 5vr i t... ..
. ,towe **(
- __ ~ . _ _ . . . .... ...._.. . _.. .
- . ~.v. cm
- x. ,.
p
, s.
Sniss . x.tJ-. _ . _ _ . . . . . _ _ . _. _ .. . . _ . . . . - . . . . . _ . . . . . . . _ . . . . . . . I . . . . .
- s. . ;__....__..___ .... ._ . _ . . _ _ _ . . . .... .
2 n ss y. \ ssa.J-I ra ta u. , .s. .,.. .
# e-l-* W r-fNAs
- /e - - . .:_ _ - . .. . _... ._. . . _ . .
~
5 **. ?T.> L. .- . . A.
- h*I k f $.2 ]~
6~7 4
. . D U'*! A **fD b r* k $$ (4edse f /e'a'f, p - . -
O i.w 2oo yoo ,.o a,s No < ** 7*4*"As /t'Ed F6' A 4" At ./ A4 A es A a M f, 9 - J~. ? ar l
-- . - - - - , . w -. - . - - - - ---wcy ,,--r- -- - - - , , , - - y. ---,c- - - , +. - - . - , , , - - - - - - . - - ,. - - . -
c.. _ _ _ _ _ ._ _ - _ . - _ _ Figuro A3 R i % paa* os=&o N.- A 3 (e t* scw C M c'c u .17 tin
)* ,. \.,
(,a .
& g o..n arr's) d'* .v 9 .
SMas, 6~ .
. fyaeolho .
KS/ . I* i
..Jo . . . ._ _ _
, . . s .. . . . . . . . _ . . .._ .. .
Vy7 ) i i Lo m, k a,, 4 y .. . .: . . .
~
A c R ~ 4. WT+ 0 5 6~ l
~
G~ 4 . (C T d) r' ' -e. OO . . . ..
/Mfff a od otAdf ffdMEA!!dd SOh
*~k
_ seen Lod s<. sreesa es l . s,es).....l. \>
.y o_n.e.. -
. o-t
. n. e . u_ . . _ , . . _ _ _ _ . . _ _ _ .
~ -- . _ _ _ _. _ .,_) t y M a*** k .
e* . 2
- vee s..
-_.-.t._.-
c up
~I T d" # A ;. .[ T):
. !. .o.. ..
/o. .
*) :.
- . 4 i : ; :I
- : l . . ' , . : ..
!.. .! ... !.....: . !a .. . . ._ _ l= . _ . . . . . ._ ui ..._ . - L..i;:.. . , .
.. .pdr ;3*i . .s. . , .. ~ Ii. I l.
~
l ; e4 !. .d,.s~
.a{.,. #r! ,
5.GNATURE J M,44/Jd#ddf DaTE= 1-f7 CURVEMO.m_
Figure A4
,LO__. ,, . , ,
.u.: . .: : . : =.. =,., .:
,-.y:: , ....
- r. .
g: : . :.,J. . i. 9,,, p. . : , _ .u. . - z._ ,:- . :. . :..- q, . h n -
,=... :.. p. n. . . .
8 ~-~ 5 Y ' N ** " -
$ NY 5 *~ ' '- 5 W E
- = :n .n u _n
- ng un .- n:m : n .. :: ::n. .n: . .n u n..l'===:::- n . Ip.;=njn
== .:::. :
- -- :n: :d. ::n ::n in, 7_. ::::n nn.:. = :. u:n :: g ~:- - . I : .. -=.::dn u hT.N. r. ...
ta .. . . . :m = :m
.. . .. ... ........=...
"..J....._... .. ._
~ ~ - -_... ."_.
--~4
. ~...
... - ~ _,.-.<'~..4- :-
t~-
. 6__ - -
a .
-- +--
= + _--
......Z.- ... _ Z*. . . ~ .Am.- ._ .. . . . . . _ . ... _ . m.m. _ __-
. _ . _l - ._.* - .....
.u. t.. .. ...t_ __" -"-
_,. . _,1
,....+
. _~e_--
5 ---
. . -v.... ~.
...+-- ~ #..i .--- ~ -" ""."..,s.
N.,."..-..
,p.:. .
. . p. : . . . .: .x.-. n. .. _m_. . .n, .: -
- 2. :: v. . . -
.. t -: ::- u. . p. ... : - -. ;... .-
.: . =...
- w. ..
( lii M i;mWilli _,$ ' .* 4 __ . fi. IN5 li E i5 IH l ii Oi! !!! *
+
4 I- :19 EE f~
-": =. -
..: 1 ::. :n: n =- n- :... :n ..: u :n:
h:: =: :.:n n=..e..::.n n n:nu::n@:-kwn:. n nr ,. = := -
- ::d-:un-3:n:-4l :nu n.. : n- ;n un .- =@c _. n: .: e : :
I"_r
- w. . . .._. .
lu._ ... __ ...
=._... . A_ .4._.. .9. . ., .:.t = .
t ._.. .. _ = . ..
. . .. . . t. ... 4 t3 . .g _...
.g. _ . . . . _. . _ . . . _ . _ . .
g .. ...
... . ___.. 9. g _ .. . . + .._ -~e... .. . -- . .- + . ..
_. .. . , . _ . . . a ....w
. . . _. _.. .M.
~. ..,,_4 . _ . . . . . _
. . . .... _ i ... ... _ . . . . .. . _._. .M.. ..
.... ._.,.-.-....~_ ... . . .~ . _-.
. .. ., ~
._ -_ -1 . . . . _. .
- : ....% . .. b.
]. 7
. L nn. ;.
- s. +... ..
, .- , . . ... _ . . _ ,.y... . . _.
.:._. ~... ._ _,._ . -
e., . ., . . . p:y: .
- o. .
- [ -.., :
.. .+_.
2 __ 3
+ + . .y e .. +.. . . +. . t g _
g t . _ . .m
-j-
.m .
_ _ _ e is : ii , # i.
;ji3 .. - .
, : . g .- ' j[: I ;. , ..., I .- . 3 . . .
.. . I :. . .
ii- ii -; ;i L; ,ii, T*' . I+"I i :. "**' ' ' * * * "~ * * ' T**t , 1
,, , i;;, it+*~' *T* t T**t o ^--
- -- -C **~ --++--- :- -.-:
l, q III lIiI-I t lll st ,j t8 [T- {,., , , ,,., l..i. ..'- :1t-i ,
+ ,' I ,t
-- ..... . ....i.. .-e..1 8+. .. . _ . i - -- .
co ,
.i.
m I --- iei. l A r :
-2 g 9_ y. :. -- -_:
i:-- t.- .
.- .; .. , ...c. u; 2. . :
n.n... ....: :: .2.
.... 1
_:-r -- T g -- 5E$i'i-5I1'. ENd' -;T k: .$ I I.kN 5 I 55'. II :-E .~M @5
- IIS 1- O N ii'5 5.55 2-55N5 D 5
- n. ::1.n. . . :. n. . n.. n. :.n.u... ..h:.. ...n..-
._4.. ... .. . . . . . .. .. . . . . _ . . . . . . . . . . . __. ._..
- .. . ~:.
- r. . . : .::.=..:.... . .:: :n: : :. .n .:n.. n. .. a. .n. .:.=...n.. . . . .n.n. : :: =. =.._ =..u.-=.. n. _ :n..
._. t r5:.:.. :_.:.a. .. -.n. .
7
.._. __... . . ._.r..
_ ~ - ..= .. t. .
.n.__=.. .
. , _ .e . . ... ._
_ . . . . ___L.____.._
. . . . . ._ ,5 .._.1..
6 ._ ..... . up... . . _._ . .. . . . _ .. . . . ... .... . ... . . . _..... . . ..
._ . ... ..u_.. __
_..s _--- : + _.
...' ' ' '*'*l-*_..1 ..J__. .- - - *
.__ . . ..+
~_'-
----I.._.."
._4 5 '*-*- - * -' - - * -
P-" ' ' - - ' * + * ' ' - ---- '-' -~' * ' * - - * - - - - -- ' ' - - - ' - - ' - - --**+----' -*****--- - - - - * - ' - - ' - - -~ - ' ' * ' liii .. 4i .. iiiW:... . . . . liii Hi u
.a b'* ~-i- ..
fili Mi II" ' -
*: P. : . i
- =:
i :s Eii
... ' .a i...
Jii i'il m..
- = iE
- -- E )s
- n. . :n. .; 7. :n-- n:-h.4 :: =. . . . { n. : u. n. n. . :. :. n. . :.o n. . a. n. ;:. r=...:.n. . :. - n. .n. . n: n:: n. .n. .an. .:.:.
4 1:::_-:... .:=.. :.:.:. p. :. n... .
- .:. :. :.:. - i . . . . . . :_:n_. :. .=.:. ,-
iiE _:ii$"i_iLi5". tili
. - .. : ... .. n.. g1_ii. 'l 6... M,. . i.i.Ef t_i. E...". .f. . j.i_iii.i&. .i.E. - iiE::.i_ii. i._E.i'. :_i_i.'pii..: i..*ii .i.in_ i.E.i.j.ii..i.i " " . .' . ;_ b
'.l.iE.
3- -- _d_ .-- .
._. . :._t_.. ---
. ".-.... . ....4 e
- _. .p-- -e . . . . _ ._ . ,
_3 .:_. _._. _ . . .. - ~ ~ _ , ~ t,
- : ...,,p..- j .-:_. .. s; _.
._. .._.4 ._... . . . . ._y. .
- 4. .... _
1 1
+..
1 1 1
.... .~._.
.. s. _
.. ._ _.p_..
m n. 4
. _,.~ . ...~ .. . ...
1_
-4 ,
- . .. 4.-
_ ~
.._.~ -.
4.~ . _ - . .
+. q p= ;
H. e .:..-.. -;- 2 .- . . - j
; . t .
.%.9 _. , . ._ ._
G .,.+ .+
.=. ~v ._ . .. % f. .
3 n-
. _ - .. .-. , , .f ... .e. .n n .
&9 _
g
.1. . 4.4 g .. + ..
,.9 .-- . . - -
a ._.i- 4 '
, -. ,1 . . I. .. _.. -
e. i- _.[T -_ ' * * > a.w i:t:- . ,s. .. . o7
.i. : .
i
.~,._ .
,, r..
u, 3 i;- r,i i a
., +.. U+- y o 4
, 1 Ug j *' t+**
- hr p l 4 N ~Y mi ild f -ii dii 9, : i Md P-M h m it ih 3- :
' E - -- -
_ m in- -
-> l 8' N N N b 5 I# IN N F S % S: 5
~
E
~
i M! - - "#; 8 - d H' 5-l t I e 7 nn- un'
;:n: ::q r=n..n = =. n n:. .un::: -
h:;.j/:n .s n ntn:. p:t = - =; n:
.:n m -= nn - ; -- u n n-
.:- .- . :n .: :
n:. nn nn :n:. n:. ...
- nz:nt. - nnn n. = m-u=
n- . :n n:: =:-
. . . . . . . . ,... . . .... . . . __ .. . = . _-- :: n-4...t_. ..- ... -..-. . ..... . . ... --
.. ~.
l O ... ....
.....~
.. .. -+--. . .. ....- ~ - . ....
, A _,...T-..- - . - -...._
..-. ~~.. ... ..u--- -~.o. . ..
- --. ~ - -- g - .
t 6. - - r- m -- - [ _ _ . . 'h.. . ... . ,... I... . ~ . .. +. . .
...[_- - - - - . - ~ -
I_-.__...
.._ + - .
.. +.~
.- . ...... . + ._..- .__ .4-_
........._......b.. . _ . . _ + .
~~-3-_.
---'**4-*-*-** . . ..
**'-< - * ' '****4 * -. . .
3 5. *'*-*1---- * - - - * - - -- -- -- - - - * * - - - - - * - - - - - - - - - * - ' -- --- - -- ---- ----' N ii. fi- !!!! W iii!.1[ E* :4 S* * ' H. -i : it ..
- -H rM- k -
,-[n.i. ..-
... . ... .. . . . . .2 M. . *: ..*n' .- .
- n. ; n-... 2 : .. .... r.u.- .-: -. :::: ,-
4 .
.f.. ... .- .I. :*? :
a.:. - t
- n. =.. .. rn :n. . :.
.. g g'g-** a.. .. . . . . . . . , s ,. . . . . .. . . .. . -- .. . - ... .. . . . ...
.**=.
; ..*;*, .n! n.L. ... ." ..11: * .11= . * :* M.l..l..?. n.*. .n. *. ,. n. ! '**M... 1*
1** ;' l -*;.*.!~. *.1 .* !. =.0.,. : *
. . ...I..n71 n ."]^._- I'.
_ . ~ ..... ....."*:...
. . . ~ .. . .. . . . . .... .
t .
.. 1.. .... . . . . . . .
..4
..~..
. . .... . . .. . . .. .. ..~-. .. .. ..
. .. . . ...... .~ .-.... ..
~- . .
(
. . ...,-+
. . - ..n....n-
... . . . . .-. . .-- . .- . . _ . . g ..- ..
~1-.
r - - - ~ - + . .----
- 3. m- '
.a. .._;.,_~-.
, a
... . .1
_-. .,m. . .
. ..-. . . .../.. . . . . . .. . . . . " . -. -_
. . _ . . ..h ...I._".." _. - ..- . . ..q..
i
.. . ., +.
...-{..y.. ._..<
. . . .,.,..)-.6 ... .. i., ._m....
.4.... .. .. .. f... .-..+
. .. + .-
.+
.....p. . . . - .
.+ . . .. ..... e.
. . . . ...- +
u ... . _ . .w. _. . .... .
.. . -.+. .
.. . ..*.4.m.. . .
. .4,.. . .
L_ _ .. __ - t _.. .. ..L_ .. .__ .__._. .. _. -.__. ....
.H.._ _
. .. . ~ . . . _ . ..
. - . ~
..~_
.. ._._.. . .~. - . _
2 - 1 _.. . .u.. .. \ . . . , . ._ : + . . .
._.4
_ + .-. . . a _. {'.- _ +- - .._.p +._
.._.m
/.
. . . . . . . . . 4. .. ,_ . . - . . . . . . .-
_. .._ _. _.. L. .u .._ . . . . . . . _._ .s }_
/
. _. 9
_._ _. .. ... .. _. . . . . . . ._. _ L . l.._ ... . _ . _ . y
. .. 1. .
I . g-l
- o. .. ..
. , 7
. . . .~ . . .
.,7.... .-._y_. . . . . _ _
7 ,. s +- 3 R
- M l %
J4E-760U9 C}r.,ts :M c, .
.. Figuro A5 '
~-
v M* - -- pounsw e.- 1 13 . 1
. .) o . . . _. s
/t 7 _ . .--. 4g ... . .-
. $ MfS $ ,d' .
]
. . -. . . . f.r. s t . -
p
. . . _ _ . . . . _..._._._. .__. -a l--. . - - - - .
f cy,c ac n' A t.
' d, C : 9.t*/s l p , , . '
I .
- g. . _ . . . . . . . . . . . . .. . . _ . . ... . . . . . . ._._. -
. . t. .
g . . . _. .. . . . . , . . . _ m .. , ,
.,.. . ... .. . . .t._.. .
1..
....... . , .. . i. ..
, i -
. ; i . . . . ._ _. - ' .
, a . -
/-.. ..
i ._ i. . . .
' . . ./ . .
.. l l . .
t_ _ ! i . - t
.t
,.10.t . . p. 2 .
d3 . f. .. j t ; " l . , ' . . .j . . ., j..; .. .'7y,e/,s/.-f (,. ..
.i
.s. . ... . l .
. /. . *
. . _ ..t...._ . . .
.t. . .. . _ . ._........ . .
. ...e, t 4
- . .. .. .........l . - . . . ... ,.. .. .
.8 . . .
' ..7
-l .
's
.-l
- 1 -
- /
3 i
. . . . *-.._. . I.
i
. . , a. . . .
l . g . . g . , .g . l
- a t. <
....I. ...,
-. . . .s. . . .
l - 6 I , I.. . . r- -- - -- - - - - - - - * -* -- * *- ' l . t
~. . . . .. . . . . .. ....
. . . I , ....1..
-se
. s . I t
i .
. ,. t .
l.. ...
.._. .t
. t.
. ; .......g - . .
t* i. --'.-- P v .. . ,
. .t. .
. L. . ... . . .
_~
... .g. . . . . . . . .. . . . .. .
.. .... ._ . ..r.e. .- , -
sum ine . . 8800sATURE * * =/ #4aa m sf. DATE 9. fs* caservF aso
Figure A6 P- i Initial Stress. MPa 40 2 120 160 200 i i a i - i ' i 20 - g (1@ - 140
- 16 -
Umit of Relaration 120 2 O ' Afler '!irgin - E Monotonic Loading ,600F ( 310 1005 G 12 - g F _ g G E ' F 8 - Atter Stress Rewrsal. - c*. _ 3
~ 600F(315C1 gl a=
l 4- . 20 0 ' . ' ! '
. I ' 9 0
0 4 8 12 16 20 24 28 32 ca. Initial Stress ksi Fig. 2- Relaxation of solution annealed type 304 stainless steel-I i l _,___-.-e-,w*m-* - ^ * '"*' " _ _ _ _ . _ , . . _ _ _ . - - - - - - - - - -
- 22 'l Figure A7
~'
4
- \;
reTemperature Contigrada 200 3(C 400 - 500
- 600 700 s
O 100 3 4 j l 3 3 'l { ' e 8 3 l ' b
, - M R -
Eg /Ert 200 ,, 1 12 24 - . ,, e 4 Yleid Strungth l ,
. r, = 2 x 10-2/ min -
ig ,
)~ .
8 ' I 20 y
*3 c = 0.2 % Offset p a .
g' un l Remaining Stress t=0 , l J* 16 -
- 100 * ,
l Initial Stress = a . . 1 s . M -
'- , , , ~ ~r
- t = 100h - e ,
- N '
3 No Relantion For inillal Stress Below This Cury's
; -l t=1000h -40 Sol. An. T304$$
4 -
- 20 o .
' ' ' ' ' 0 0 20 1000 1200
- 0 200 400 600 Ternperature. *F .
09, t- Remaining stress for an anneeled type 304 stalnkss sisel bar at constant strain as a func6on of . ternperature and time. 1 l w . [ l *
. e j .
G P e
- - - . - ~ - . . . . , , , _ - , , - , . ,,- -- .- --,
.-.v--2, . - , - - - , - . . - - - - - _ _ _ . - . _ _ - _ - _ _ - _ _ _ _ - -
e, .. D I
. : fi! '
, . i :. fit .
l0l!!!!!P -
,ilMilliP i nsure ^9 7 . _ _
[ (Ref. 6) l l. I l r l '
""~
mosusa s ets am68 Smit
> Ka 'b#saserres. (bsessariasers of awesi,al.geef asuu sser!
- v .
$0- *
[ cucuc covc i. p
,t static nas 30 Figure A10
= .
30 - *
. . a vet is.uS . rou, (Ref. 6)
& as4ase steen emner, g IO -
- acot. maten i e+
en
* .e.
0 0 004 0008 OOL2
- STAAsN OR STRAIN AM8'UTUDC I K, v-- A,.e t. . hs asee... saima .or r. /or swea /. trit aus.a .eeet w
a
. - , , .--,_..,,,-.,,nw., -,n - , - - , , - . , . , . . , - , . _ , , - - - -.n,.-,,,,_....,,,., ,~ ----,,.,--,,ra,,- , , , _ , , , - , , --
.g.Ja - - , . - -
Figure A8 (Ref. 5) v . s
~
~~
s&/ O e?/
/
- f. ,, i ,-
e
'# * , . /
/
.#'l
/
/ x
,. / e
,s
- o
. 3,.</x, s'
s'
'.o,,,f. y.
//s'O w. .
/
f
- a. . OOg
/ . Q; O && Y & a l . . \
- t. ..
I
Figure All . I t-. so. ... _. . . . . . . . .. . B
/ ,. . . .
e
.** . ' .l .
s -
- t. .
4.o . ... / . . .. i
.I - ......
1 - s.
. 40 . .. .
o
. . . . a-
.y ./ft / .
/z A 2
,/
.o
- c .
C;, l ' g .
~e.=f ~Q 3. K2. s.Y Al-
- v. .
i
. . p.
f .. .-G o , a
* . : - - - ~
. -we .
A, - r
.i s & S.p r 0
, ga A(wn) 13 &a __.
i
/
/ 8(en ) f/3 82 /10 o*Y j . .
a.***o .
-Ro - - -
- g 4
O EW.. . - 4 .O- .M..N GG . . 8M 68. s .mo .. L I 4
;, .e e- ?? c ., s . . .
Figure A12 d e 9 8
, /
/
- j. f I
,C ee -
~ : .
i D,d L AU
- w . _ . _
e 9
.__ . _ _ _ _ _ , _ .-- _ _ . _ - - . - . ~f-n
/ m
~
e
, - - . k. .
w x n
]
j r mr mem t "L
!.h'l ' M',&
a . 7 8s I, j %. y fLk ,
.a ,
N - g
'i 4 'I k
/
$ .a8
- a' '. * -
,, y
.a .*
ou 1
.e .,
w'.,
- me t v ~
~ 'd
{' i
,as Anehrees of tunes longssis contrismating 8
-\ l
/ soownoisis,.aoe.p. '
~
* .--1 . ~ 1, __( , ; . . . , ._I .
y Ks K i Ki g,,_ 13 ' e
# ,{- . _. .. . .
m
,k s, a.
}g
't s , . .. . .
s}.. .< \ __ _:...izer r_s t.).E .4s.
'11
??%
w .
'"Y"1W. tut -
-n %ii ). e., ... ., ~... . .,. ..
f+nese slewel erwesod are showere for a 9/16 (. ..
.y.
- 1
- *
- 15 bo't peeso.eded to 100 K31. Positive . . , , , _ . . . . . , ,
nampb re are term.se stresses in K$l;n*0 stave '- ----- 2 n=+.r. e ew . . s .n x3,. ,
- ..__ a .
.{.- m. :
- e
- - -
- 7, - - -
prsics MC AGAINST TATIQUS . .. [ ., t
/. .__.____.a--
' tqJJ9'd_E.WI :.- - .
f \ '/ . . . . . . _
. s\ .
/ >
7 . . . _ _ _ . f ._ (4)
- _
{ ' - n i & h in - t-
- M ... .. c,;.. ,.in r. , .
! J \\ \ ' ,l~ l'2- .-
- l AN . ,
'5TPEss AT TilaEAD Roots tjJE 70 psUT EtoCMEWIIT
.-l .
C "h %
- l 3 i yac,; ( ) A low-etiNness bolt assembly; (4) Belleville
['
,,a w ,
,nde /m. n6s -iaL) Wj-f.
1 . 1 s
- e. A
~. _. e. m._ m . . - . . .. . . . . ..
'5. Q. Re: CYGNA Questions 12 and 19 Please provide U-bolt _ torque values that will be used in the field for all pipe sizes and the corresponding lower bound preload level expected as discussed on transcript pages 123 and 94, respectively.
Also, please provide preload versus torque data scatter and lower bound curves to be used (transcript page 100). A. The U-bolt torque values that will be used in the field has not been established yet for all pipe sizes and will be made available as soon as the information is finalized. However, the methodology that is employed in arriving at such values is provided herein, and is applied as an example to the values reported in our affidavit, so that CYGhA will be able to understand how all values are derived. The important thing to recognize is that we will determine the minimum preload level to be applied to the various pipe sizes and schedules. The torque to be applied is then derived from knowledge of the " mini-mum" preload necessary and the test data of preload versus torque. The data scatter obtained from all the tests where preload versus torque was measured is given in Exhibits 5.1 through 5.4 for the four specimens tested. Also shown in those exhibits are dashed lines which have been used to derive the linear relationship between pre-load and torque for each pipe size. It is to be noted that the cor-relations derived for each pipe size tested represent the condition whereby the lowest preload is achieved for the highest applied torque in the range of interest for each pipe size. The correlation is the usual l' = KID where t- is the torque in ft.-lb., T is the preload in lbs., D is the U-bolt diameter and K is a coefficient derived from test. For the four specimens tested, the coefficients, K, that would result in the lowest preload for a given torque are as follows SPECIMEN K 4" Sch 160 0.288 10" Sch 40 0.353 10" Sch 80 0.276 32" MS 0.403 Obviously average coefficient would be less and would vary between , , 0.25 and 0.35 as stated in our affidavit. To ensure that the minimum required preload is achieved in the field, the highest value of the coefficient is used regardless of pipe size, i.e., K=0.040. To arrive at the torque value, the following examples derived for the 4" Sch 160 and 10" Sch 40 pipes will serve as illustration. From the answer to the next question, the minimum value of preload necessary to
- maintain " stability" for the 4" Sch 160 assembly is 0.37 kips. With an average value of K (used in finite element analyses of the Affidavit) the torque value corresponding to this preload would be 5 ft.-lbs. With the maximum K value the torque is 6.16 ft.-lbs. Considering that the specimen is subject to thermal cycling and possible relaxation, a 40%
U0 c$4 U C/ .. . , 7.0 : - J 4 INCH STAINLESS STEEL i l' mLEG 1, BOLT 1 6.0 ' LEG 1, BOLT 1 LEG 2, BOLT 1 L
/
5.0 - I p (EG 2, BOLT 1
$ 7
$ /
E 4.0
/
) J 2 / 8 I * / g / _, 3.0 '
/
N 7 [ LEG 1, BOLT 2 _
/
/
! .. LEG 2, BRT 2 - 2.0 D , OTORQUE VERSUS PRELOAD
- eFRICTION .
i ' t
" sTHERMAL ,-
1.0 , t- .
~
- O 5 10 15 20 25 30 35 40 45 50 55 e 60 l
BOLT TORQUE - FOOT POUNDS l
r, w cz .- . P I . 10.0 ,, 10 INCH CARBON STEEL
, s 9.0 '
i LEG 1, BOLT 4
! 8.0 LEG 2, BOLT 4 1
i l g 7.0 2
) @
m 6.0 ' ,
/
! ' 5.0 - i u, o /
$ /
- d 4.0 -
/*
E / . i LEG 1 BOLT 4 ' 30- - l
/
LEG 2, BOLT 4 y- o TORQUE VERSUS PRELO W
/
2.0 , . FRICTION , j l 1.0 LEG 1, BOLT 1 i LEG 2, B0'.T 1 l 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160
! BOLT TORQUE - FOOT POUNDS i
l l
c w ne .. . f 9.0 < - 10 INCH STAINLESS STEEL i 8.0. I LEG 2, BOLT 2 1 l " o 7.0 LEG 1, BOLT 2 , 8 i t
) y 6.0 LEG 2, BOLT 2
} 8 - 5.0 - LEG 1, BOLT 2 d Len 1 AND LEG 2 g g gs g BOLT 3 ,
' LEG 1 AND LEG 2 d 4.0 -
/ BOLT 4
/
_ / LEG 2. BOLT 1 3.0 " /
- TORQUE VERSUS PRELOAD l
-
- FRICTION 2.0 " g
' A LOAD DISTRIBUTION
- 1.0
" M-8 THERMAL
- VIBRATION j e SEISMIC
) 10 20 30 40 50 60 70 80 90 100 110 120 130 l BOLT TORQUE - FOOT POUNDS' m , e
- l l :
l l .
4 e% s N e 9-
.u .
.S -
g W k=a = 5 _, g u
"N./ gss
~
. . .i m 42 E
m o E a 8 e 8 a.
~ - ~
e c # # m g a w w w w " W _. s a a g r W .m G W
- D 5
?
?
5 n f. a\
%?
\ ,
c x O E
- ~
e e w w
's -
a M .J J 1 1 i T a a
, a y W '
g g 4 g 4 e g m g 4 - O R - - - - - e
- N 000t/ SON 00d - GV013Hd 1108-0 7
margin is added to account for'it, leading to a torque of about 9 ft.-
-lb.. required for stability.
The total ~1oad experienced by the U-bolt with a torque level of 9 ft.- lb., coupled with a peak thermal expansion and peak pressure expansion load is approximately 4000 lbs. (see Table II-1 of Attachment 3 of Af-fidavit). For this load the stress in the U-bolt threaded area would be 28,200 psi, which is approximately 78% of the minimum yield strength at room temperature. With thermal cycling at this level of stress, a stress relaxation of between 30 and 40 percent may be possible (see figure A2 of preceding Question 4). In the' affidavit a value of'25 ft.-lb. was chosen as a compromise be-tween the 9 ft.-lb. and 35 ft.-lb. which tests showed resulted in good contouring of the U-bolt around the pipe (see Affidavit at 74). For the 10" Sch 40 pipe, as another example, the minimum required preload is computed to be 1.4 kips, which leads to a torque of 35 ft.- lb. Little or no relaxation would be expected in such specimen for such relatively low torque. Nevertheless, if one were to consider a 30% relaxation, then the minimum applied torque would be less than 50 ft.- lb. Note that a preload of 50 ft.-lb. (equal to a tension of 2000 lbs.), plus thermal and pressure expansion loads equal to 1600 lbs...and an external 10,000 pull load, the total tension with U-bolt would be 5800 lbs., which produce a stress of 17350 psi, which is less than of yield. Thus no significant relaxation should be expected.
- 6. Q. Re: CYGNA Question 18 What is the minimum level of preload required to maintain stability for the anticipated worst loading condition for stability (i.e., pre-load plus push at 50)? This question does not appear to have been answered by the finite element analysis (transcript page 122). Speci-fically, the first objective on page 1 of the finite element analysis has'not been safisfactorily addressed. The fact that " adequate fric-tional forces exist" requires a judgment based upon what are known to be the necessary frictional forces for stability under the anti-cipated worst loading condition for stability. Since the necessary frictional forces for stability under this loading condition have not been determined, it is not possible to know if an adequate margin exists between the minimum expected preload in the field nad the pre-load level necessary to maintain stability.
Without knowing the minimum preload required to maintain stability with a push load at 50, a judgment as to what constitutes adequate preload cannot be made. Maintaining a tensile load in the U-bolt legs does not guarantee stability. A. To respond to the question:
"What is the minimum level of preload required to maintain stability for the anticipated worst loading condition for stability (i.e., preload plus push at 50)?"
requires a two-part answer. The first, related to the finite element
=_
analysis performed, and the second, pertaining to piping systems that have lower normal operating temperatures and pressures than evaluated in the finite studies. Part one, the finite element analyses adequately addresses the first objective of the U-bolt testing and analytical program. This objectiie being to determine if;
" adequate frictional forces exist at the pipe / pipe support interface to balance a moment created when the U-bolt legs are not parallel to the strut so that the U-bolt strut assembly support is stable."
The analysis' program addressed the stability of four specific U-bolt pipe support systems assuming that the SSE seismic push force occurs at the system normal operating temperature and pressure conditions. Stable solutions are summarized in the finite element analysis report for minimum and maximum preload torque values. The minimum preload torque values given are not the actual lower-bound minimum values for stability but the lowest preload evaluated in the analytical study. The actual lower-bound minimum torque value required for stability could be much smaller. The second part, to assure stability for a push force, it is necessary to maintain a difference in U-bolt leg forces, this differe-ee pro-duces a couple which balances the induced moment due to the $ degree push force. This difference in U-bolt leg forces will result from a small amount of cross piece " rolling" on the pipe. If the U-bolt were a cable, no shear or moment capacity, the U-bolt leg tensile forces would have to be large enough to create fricitional forces between the pipe and U-bolt to maintain moment equilibrium. Since the U-bolt has shear and moment capabilities, the U-bolt leg tensions can be less than required for a cable. The effect of reducing the pipe system temperature and pressure is to reduce the total preload. The reduced preload will result in lower frictional force resistance capacity. Although the finite element analysis did not explicitly determine the absolute minimum U-bolt leg tension required for stability, they do show that the minimum - U-bolt leg tension for the stability approaches zero. It is necessary that the U-bolt leg forces be tensile (greater than or equal to zero) to ensure that the U-bolt legs will be active and resist the applied seismic strut loads. With one exception, the finite element analysis results are used to project minimum preload tension and torque values that ensure stability for the PRELOAD + PUSH (@S ). The exception is the 10" Sch 40 SS opeci-mens for which test data is direct 1/ available on preload reduction as function of externally applied push load (see figure 18 of Attachment I to the Affidavit). The projected values are given in Tables 1 to 4. As seen from these tables, the recommended torque values given in the tes-timony are equal to or higher than these minimum perload values except for the 32 inch U-bolt. Very large U-bolt leg tensions result from temperature and pressure in the 32 inch pipe /U-bolt system. Since the temperature and pressure preload effects are not present in the 32 inch pipe /U-bolt assembly, an additional preload torque is required for this pipe /U-bolt assembly. Therefore, for the 31 inch pipe, and similar pipe /U-bolt assemblies, care must be taken before assigning a torque
. . _ = - . _. _
T Table 1 4" Schedule 160 I s L h y% v' \
/ \
l \ l l
\
% /
%~ /
Tension T Tension T (kips)3 (kips) 2
- 1. Leg force needed to 0.0 .12
. provide resistance couple
- 2. Amount of unloading due 37 .'25
~~
- to push
- 3. Total preload necessary .37 .37 (Sum 1+2)
.37 Minimum Preload Moment : ----- X 60 Fr-LB = 5 FT-LB l 5.41 1
Note: This table is based on results given in Table II-1 of Attachment 3
- to " Applicants' Summary Disposition of CASE's Allegations Regarding Cinching Down of U-Bolts"
( I pg. 3
T Table 2 I g 10" Schedule 40 4 L y% N
' \
/ \
l
\
I l
\ /
\
s / . s / Tension T Tension 7 (kips)g (kips) 2
- 1. Les force needed to 0.0 .35 provide resistance couple
- 2. Amount of unloading due 2.79 1.05 to push
- 3. Total preload necessary 2.79 '.144 (Sum 1+2)
. n ' '"
Minimum Preload Moannt mM6.'47,(0[5) 140_0/17= 35 ft ibs.* _
,a . _ ..-
Note: '!his table is based on results given in Figure 18 of Attachment I to " Applicants' Sumary Disposition of CASE's Allegations Regarding Cinching Down of U-Bolts" It is important to note that this particular torque had been applied to the 10" Sch 40 SS assembly subjected to a 2 minute accelerated vibration test with a sinusoidal force input of 1000 lb. peak to peak, and the assembly was noted to experience no motion (see Affidavit at 29). The assembly had rotated and walked where the torque was only 20 ft.-lb. Hence, there is confirmation of the sta-bility of the assembly at the 35 ft.-lb. torque. For this test neither pressure nor temperature were present in the assembly. pg. 4
- ' ' - ~ ^ '
Table 3 d I A 10" Schsedulo 80" y-~~,
/
l
/ \
\
t t I i k% /
/
%,, e Tension T Tension T (kips)g (kips) 2 1., Les force needed to 0.0 .43 provide resistance couple
- 2. Amount of unloading due 1.51 '
1.51 to push
- 3. Total preload necessary 1.51 1.51 (Sum 1+2) l Minimum Preload Moment
- - *--- X 100 FT-La 20 FT-LB 7.51 l
Note: This table is based on results given in Table II-3 or Attachment 3 to " Applicants' Summary Disposition of CASE's Allegations Regarding Cinching Down of U-Bolts" I O [
Table 4 d I A 32" MAIN STEAM
,,,. - = . s -
/
- f
\
/ \
' \
I l
\ /
\
% /
% /
%= #
Tension T Tension T (kips)j (kips) 2
- 1. Les force needed to 0.0 8.1 provide resistance couple
- 2. Amount of unloading due -
29 0 20 9 to push
- 3. Total preload necessary 29.0 29 0
- (Sum 14) l
{ 29 0 Minimum Preload Moment ----- X 380 FT-2 s 1825 FT-2 6.04 l Note: This table is based on results given in Table II-4 of Attachment 3 to " Applicants' Summary Disposition of CASE's Allegations Regarding Cinching Down of U-Bolto" l pg. 6 L i
_ - . . _ _ . __- . . . _ _ _ _ . . _ . _ . _ . . ~ . . -_ value, to examine the normal operating temperature and pressure conditions of the pipe and their effects on preload. A further note needs to be added to the minimum preload required for the main steam line (32" MS). We do not consider it appropriate to determine the minimum required preload for this line in the ab-sence of pressure and temperature. This would only occur when the line is not functioning. Under such condition, the line fulfills no safety function related > to maintaining the plant in a cold shutdown condition. Further analyses conducted on the portion of the main steam line which has the cinched-up U-bolts, in the absence of these supports, indicate that no adverse consequences would result. For this reason Appli-cants have elected to retain 240 ft.-lb. torque as the minimum re-quired for stability.
- 7. Q. Re: CYGNA Questions 6, 12 and 18 Given that lower bound values of preload versus torque are to be provided in the field, how will these lower bound values be reduced to account for observed reductions in preload which occurred during the testing program (thermal cycling, vibration testing, etc.)? Also, what values of "necessary preload for stability" will these reduced values be compared to determine the margin against instability?
A. The manner in which the lower bound value of preload are " augmented" to account for relaxation phenomena that may occur, so that a cor-respondingly higher torque would be used in the field has been des-cribed in our prior answer.
1 014.// 919' TEXAS UTILITIES GENERATING COMinNY f g P. O. BOX 1002 GtEN ROSE, TEXA516043 November 8, 1984 U O ntd2 M o GL8/uknas<J Cygna Energy Services 8MR Pi= 101 California Street Suite 1000 San Francisec, CA 94111 Attn: Ms. Nancy Williams, Pro'ect Manager Ref: Telephone Conversation of Octoner 30, 1984 between J. Minichiel10 (CYGNA) and J. Finneran (TUGCO)
Dear Ms. Williams:
Attached, please find the TUGC0 clarification to the calculations for MS-1-002-005-5RR. If there are any further questions or coninents, please contact fis. Jeanne J. Van Amerongen. Very truly yours,
#fde L.M. Popplewell Project Engineering Manager LMP/JVA/1jh ','VG ,' A 7 p ... ...-. ....-
. .' a ard l OC J. Finneran , D.','iz EC'D/toccza: /j///,/ft/
- c3. s QQf r1:.2 s .;1. /. / $>y . c/(
cn m r.cr. titz .;) . / k. ti:. Ten I o ,,,......,,,,u ...,,,,,,,...... ......., ,
CYGNA QUESTION Cygna has reviewed TUGCO's calculations for MS-1-002-005-S72R (done in response to the 10/4/84 telecon between Cygna and TUGCO). Cygna has found the calculations correct. Cygna requested TUGC0 to provide documentation showing the AWS calculation is an appropriate method for evaluating this type of local stress. TUGCO'S Response We have attached a copy of the paper " Basics for Tubular Joint Design" by P.W. Marshall and A.A. Toprac. Their paper presents the background data for the local failure design criteria in AWS. Although that criteria is expressed in terms of punching shear, it also includes considerations of flange width to thickness ratio, branch member to main member ratio, and axial and bending stresses in the main member. Thus, as the paper makes clear, it is really a total joint design approach, not just punching shear. CYGNA should note on the third page of the paper that the authors state that for a V less than 7 the joint may be said to have a 100% punching shear efficiency. On page 5 of the paper, 1 for tube steel'is defined as b/t where b = half the width of the tube steel shape minus the thickness t. For MS-1-002-005-572R, ( eouals (4 .5)/(.5 + .72) = 2.8. i O
y.g.g- - .
.:q s' . * *
,e
** .p. ', , f, .
.-;2 kf y
- y. .. . 2.
E Basis for Tubu,lar Joint Design - -
'j 4
> =
t". ' . _j
- k. . .
i Design criterie of the codes that govern construction of , i offshore drillingplatforms are analyzedend evaluated , m .
- g. . -e BY P.W. MAflSHALL AND A. A.TOPRAC _
0 .
) . .' '
7
. t
'. Introduction Static Strength uterly where complete joint penetra. 5 Recently publiehod codes (Refs. simple and PuneNag sheerJeanse "8 fy*bufer I
( uetu (Rof 2) sed e't r tions for circular tuboe, which have d to Currently the moet popular style of welded connection for intersectfrq the ends of the branch ms,ebers. Althou:h the ca.?p e+e stress [ 3 L been in use for a num!.,4e of years in circular tubes se used in f;med c*f. pictwfe :s mei f.cre :er :Nu, tw e '# 'h0's structures is the **stmple'* loint concept of pun:rt!ng e.Te or, fi;. 2. has 4 offshots drilllog pistforms* The 3 ' ?-* of this r is to d nt litustrated in Fig.1. The tubular mom. been quite useful in correlating test a th ground det underlyl t ese bets are simply welded together, and data and formulating design criteria. The average (or nominal) punching ] l criterle. in terms of static and fatigits all load le treneferred from one , h strength, branch to the other via the chord. Sheer strees, v , acting on the poten. g p without any help from stiffening ringe .tlel failure su ,' ele calculated as: j-t or guseet plates. To prevent encoe. -
, ' , , gf*d eh) 3 L sively high locallred stresses in the (1) i L M '
chord, a short length of hoevier k ke ke j section (joint cen) is often used in the 3 i j connection area. In euch cases. the problem of joint design reduces to
?
,j 4/ that of slaing the joint can. partic. I J es... > st a es ee o vc(f t) : t - d b p r .r;. es ease N. i I i
- p. .
e3 8l.j
)
fg **m8 '7 I , I 8 ' e, res eman e a l ,
/
flp i - Shnple /eher i / ,e ela .
/ i l 'O ,' , ' , o*, 3
- f. W. AteMSHAll. le SteH Cirtl Engineer. l ' 'A
\ ' ~
' , Offshore Construttlers Shell Oil Com. l ,, s l
- r pony, New Orleans Le, A. A. TCPMAC le ProfesserelCivilEngineering TheUntrer.
l l l e
*e ren s e ne &y I .
' fi e#ref ferat etAusta j feeeris based en a survey seensoredby the WMC Subtemmittee en Welded Tabe.
@l 1 .
1y
.t,t t.I an.
tt t t
, ae. ,
m
'l
,$ krStruetwes . fig 2 - Punthing sheer flg 3-Intercettlan tine'eWetts
?
182.e l M AY 1974 . -
- r +9' u c * *
- .,. b ,*,
- 1 f *
~M. e .
- s 9- -
inst Loao g, ' tional effect of diameter rcio. ffp), se *t - i,,,g to ,
- 8 indidhted by Roark, was considered 'f F e n/d cLoggp agge tl O paradoxical in that test det] with [*
A j;, S
- c
-C
..j_
tubular connections did not show the same monotonic incres se in joint of'l. ~; gT - ciency as deo:cted ;n Fy. 4. In feet, T. - 4 g . I joint tgets cited by Toprac (Ref. 4) h, showed joint efficiency (in terms of . the ratio otshot spot stress to punch. : t# tit wtDTM simp etAna log shear) passing through a min - .A imum in the midrange of diameter I .' sophisticated analytical solution (Ref. 5) yields the more realistic plc. ture presented in Fig. 5. These results are consistent with those obtained
/
E8,$' Erg experimentally and with finite ele. ment analyses (Ref. 6). Insofar as ts stress levels in the chord and load i transfer across the weld (Q) are con. g I corned. For this joint. the stress con. . E ar# AT centration factor is 7.3, and the esl.
" QS - Yitt.DHp,9gg.[I. o.s r culated average punching shear stress, v,,nt which first yield at the 0
hot spot occurs (F,' a 36 kel) is only 0 "8 80 2.5 ksi. Comparable punching sheers
, faC/R for Roark and Kellogg would be 2.2 fic 4 -r Simplified punchirig ahear criterle kst and 3.4 ksi.tespectivel1.
Figure 6 summarizes the results of a parameter study made with com. puter programs based on Ref. 5. The Table 1 - Closed Ring sad Kellogg 8elutlena for Punching 8heer end Une Loed Capecities punching sheer stress, v ,at which yield stress is predicted for anlally Case
- Closed ring Kellogg loaded T. connections. Is presented as
- ' a function of chord thinness ratio, Y ,
y' and brace / chord diameter ratio, g , Punching sheer cepacity v,s a f(0) e y,2.34a 7 0 s As was previously noted su. ; O.57 Total joint capecity 18 alength a ftp) l'
- n perimeter perimentally, joint efficier cy (in proportkmal to terms of punching shear at yield)
- passes through a miminum for a diameter ratio in the range of 0.4 to gho t s g " 8 where f a t /t a ratio of branch Theoret/ca/ Approach. Solutions for f,h. gn or sn .
thickness to chord thicknese, . elastic stresses in cyl.rdrical shells pendent of diameter ratio, but varies inversely with the 0] power of chord 8 = angle between member sues av fabi for he ve s1rpelo i= r+ n cases shownin ng. 4. Tne closed ,ing iag,';ggg,,,,,,o,e,,,,,,,,,
' " ' "" " "* E'**
- f. and in a nominst axial and bending lr ca(puncI ear n line d 8*" ' ** ' ***
stresses in branch, respectively, capacities as shown in Table 1* perimeter (or intersection length) and
) dote that punching sheer capacity is gi.', where t is chord thickness - a l lt is to be noted that only the compo. defined in relation to the very impor* result which is surprisingly consis.
l nont of the branch member load tant nondimensional parameter y tent with the oversimphlied op. which le perpendicular to the main proaches considered earlier. rnember(chord) wellis considered be. (where 7 = R/t a chord thinnese ratio, H C* *'" cause this component is responsible Lafure cmm' for most of the locallied stresses. The redlus/thicknees n 'h' "" 'I II'll yloid_sa a
%Guastic theories seriously underprodlet_the -
terms K.andk, relate to the length nyallshis_sjetic strength _9Lpractical and section modulus, respectively, of This is analogous to the span to depth ratio of a strip beam, for which tubular _ connections._for example, a the tube to tube Intersection, which similar talationships may be derived - mild steel scale model of the connec. l la kind of a saddle. shaped oval (Ref, (See Fig. 4). tion in Fig. 5 actually carried the load I 3). Specifically the terms represent shown (appropriately scaled down). 7 the ratio of the true perimeter (or sec. These two relatively crude physical y models might be expected to bracket Naturally, a hot spot stress of 160 kev tion modulus) to that of the circular the behavior of air.iple tubular joints, for mild steel is unrealistic and the brace; they are plotted in Fig. 3. as a material as beyond yleid, and sub. function of f(defined above) andg, since the branch member loads the where chord along a combination of longl. lected to strains in excess of $3008 tudinal and circumferential lines. inlin. Under these circumstances, it
- Unfortunately they yleld divergent ,gppears that theoretleal elastle anaT-yses will be of hmited use in rotmulat.
#=R- a 'adlusI to chord diameter results bracereti* (or and tend to indicate disturb. ing practical _ design criteria 15 TITH 1T ingly high stresses in practical design To specify design ellowable values for situations. However, they both do ol quasi. static ioeding conditions.
the punching shear stress theoretical reflect the strong dependence of total 7mpsticar Approscrt shoular joints and experimental considerations are joint capacity on chord thickness and have a tremendous reserve espacity discussed below. . branch member perimeter. The addi. beyond the point of first yield (Ref. 7), t
. W El. DIN G R E S E A R C H S U P P t.E M E N T l 183.s
.c .
fth' Q f. NtMr;C N* W 2*bD W W ' M.: lgpun': ' gy. j g- -
~u ,,ir, a
p"g - . -ae ; E , PUNCteNG SMCAR as E AT FutsT YEt.D C.i .-
==
seraper{,,i vr,. e tr,m (= _ h ?g*;Q nuriYt.oacto Ts1 JoosT e.l.0 4 1
, ---. --yjpa ==( -
s - se w s ; e, r' a softse
-~
r3 8- Theoretiestelsetlestresses~ esielfi loededi.}eint as illustrated in Fig. 7. If a section
- ris s - Parameter seusy through the chord at its Intersection with the brace is considered for small loads in the elastic range, the distribu.
tion of circumferential stresses on diam, by extra strong pipe under 5 in. d2. Restraint to plastic flow caused by
- the outside surface are showrt as diam, and by double extra strong pipe triamiel stresses at the hot spot, a Stage 1 in the figure. Beyond yloid, through 12 in. diam. factor of 1,6 for the situation of the connection deforms (Stage 2) larger and/or thinner chords o Fig. 5.
While the applied load continues to should be treated on the Basis of a D. Strain hardening - for the mild incrosse. Finally, at loadga2.5 to_.8 reduced punching sheer capacity as steels represented in the test data, times that 31 fir 5GsRI'the 1010t f ails given by the curve in Fig. 8 and the ultimate tensile strength
-Tpullout failure as showF"f6'r (which is at least locally utilized I:nsion loads or by localltod collapse F, (2),, when a joint fails by separation of of the chord for compression loads Ultimateve
- o.s the material) is greater than the (Stage 3). specified minimum yield strength.
The average punching shear stress F. F,,(which is used for the empirical e at f llure', v,, has been plotted in Fig. *** * 'e
- o (2a) correlation and design formula) by
(' 8 relative to specified minimum yield 0.9 m 7 .: factors from 1.8 to 2.4. Corres. strength. F,, and as a function of pondingly, it is suggested that F, chord thinness ratio 7: 38 static tests Here, the design allowable punching used in calculating the allowable which failed in the punching shear sheer stress incorporates a safety v. should no,t exceed two thirds mode are represented, along with factor of 1.8 with respect to the (2/3)the tensile strength, two specimens which failed after empirical curve for ultimate punching 4. Further increases in capacity re. Cnly a few cycles of fatigue loading. shear,its intended range of applica, sult from the redistribution of load, l Th3 solid circles represent K. joints; tion is for the mid. range of diameter which occurs as the connection th3 rest are T and cross joints. Data ratios for which ve is more or less yields and approaches its limit cro from Toprac(Refs. 4,7) and other independent ofg,
- load. if the cylindrical shell is vis.
i sources (Refs. 8,9). Since the proposed empirical valized as a network of rings and For relatively stocky chord mem, design curve makes use of the poet, stringers, the sequence of events bets - thickness greater than 7% of yield reserve strength of simple mayoccur esIllustratedinFig 9. dFmeter or Y less than 7 - the joints tubular connections, it will be instruc. f.[tatle _ behavior, IrlAAIAkStt.Assas may be said to have a 100% punching tive to review the sources of this entra strain h=rrianina, load redistributtaa shear efficiency, in the sense that the capacity.These are: andfjatge deformation behavior plac_q. sh:ar strength of the materialis fully J.The difference between elastic 'straordinary_ demands on tne ductil. mobillred on the potential failure sur. and plastic bending strength (local, ify of the chord material. Some local.
. face.This criterion is met by ASTM A. 12ed) of the cylindrical shell, a N Y'eiding will occur 1t design load 53 standard weight pipe under 2 in, factor of 1.5. levels. These considerations should be kept in mind when selecting steels
, , for tubular structures (Ref. 8). .
- failure was defined es first eteek for tension leads This would functioneHy ' W"*****
impair the joint for subseeuent lerigue sera By and large, design codes repre. sent a consensus of engineering prac.
**The ukimate strength eroterre developedby MobertMe! Sireduces te- tices in a particular field. There was a Fr general feeling that, while the data of ummere e, e /fgj Fig. 8 (as replotted in terms of # in Fig.
Ogg ye 4 10) did not justify taking diameter (. AMsm% ele T Y ene K connections are tested en a commen besos. Although have lower elastot stresses then the correspondong i and y conneetoons, they stre have cat aKbeneficial connecroons ettect as the diam. v tss reserve strengtts so ther the ottimore especities come our sanoler. The chooldettsrence eter ratio apptoeches unity, as Indi. between Me6er's results and equatoonI2los in the degree of conservatism with respect to cated by the heavy dashed line in Fig. ' [ the seerter band shown by the test resunn Rober prosodes a good eversee fit wheroes the 10. i curve for eeuetion(21fests on the sete sode of most of the dets Rober's !($)shows relativeConssderable insight Square Tubes. er lott** mthrence of 6emeter retia ie, ttpj epn
- Into the etfett of $ on the ultimate 194.e l M A Y 19 74 .
~~ ' ' ~ " * *~
-. ~ 'r-~~~~
- _ _ _ _ ~. _ , _
PW81E O
<L 74 30 -
l J 6 N O N O it e
= =,so .
a (f . !. Fator - * *p* . s tc
- S . ..
'P , . . .
'o at os f.a na io I cancicR rat-> $
l
~ '> ag Fist f0-Steekstrength-p eMeets STAeti STAGE I STAGE 3 _
DEFLECTitBI '
!!) 7- Meeerve strength of a tubutor connection where # and Y ere defined in e -
t,0 s%n9er s9e'N3ws to the use;e for
% circw'se tubes. '.
InATE M LaniT The second term on the right of Y P
- FF /8 equation (3) is quite similar to the empirical punching shear, equation IA.TIMATE PUNCHING SHEAR (2); only the exponent of 7 is differ.
~'
eg e V F ent. The leading term corresponds to )_ C on P*a the g effect and has the following propeniet i I o [**C o o o o o o
- 1. Minimum value of 1.0, which e se . occurs at # = 0.5.
i 2. . Increasing punching shear effi.
. n , , , , , ,
, , clency at larger and smaller p .
'O 00 ratios; this is comparable to the 10 20 30 40 SO SO 70 i
theoretical results for circular T. I R/t ay CH0ft0 TH91 NESS RATIO jointsFig.6. p 3. Where4spproaches its limits (0 l (, fig. # - Emperece/ design curve - stetic strength and 1.0), punching sheer is limited by the sheer strength of the mate. . rlal (or by other considerations such as web crippling). Test data (Raf.10) for the specific [ ', , . . '. " .. '
- case of 5 = 5 = 0.187 chord are also
, plotted in Fig.11. Failur:s was defined as when joint deformation teached 3% of chord width. The strength in.
/ crease for #. ratios over 0.5 appears to be confirmed. with the test data y showing strengths ranging from 1.5 to 1.8 times the computed " upper f bound ** limit load. This reserve strength undoubtedly comes from
, g I I
some of the same sources discussed above for circular tube connections. 26f.625/ C. s Forg . ratios under 0.5, however, g the test data show equation (3) to be increasingly less conservative as # decreases. The darted line (Fig.11) represents a punching sheer criteria S0 / 1.25 which is independent of the # ratio, l fig 9 - Leed redostrnbutsett first yielding ecturs et het spot A. Cross hatched yieM line is giv*n by; onelegeus to plastor honge on a centinuous frame. Full strength elring AB is reached when p, Ve a vioWong else occurs et 8. efter tensaderable angle thenge et het spot. Ring A8 continues to f0f #< O 0 (3a) deform et constant feed whole rest ofjeont catches sa resultong in more uniform leed dis 0.5 1 tesborrers Limot leed et t eint is reached when ring CD and stringer CE else yield Deformed shape ss endscatedbr deshedtones Note that this stre!ght sloping line goes through the origin; total joint capacity goes to aero as the brace i ,' punching shear capacity of tubular of plastic design, the ultimate punch. perimeter and S ratio also approach i connections was geened from consid. Ing shear stress ve is obtained as: raro. The combination of equations << ct: tion of a limit analysis of square a25 F, (3) and (3a) results in criteria with ' tubes. Using the yield line pattern of ' vj s more or less consistent safety f actors l Fig.11 and the upper bound theorem #(1.#) 0.5a7 (3) throughout the range of$. 1 WELDING RESEARCH SUPPLEMENT l 195-s
- L 1
1 , - , , . . . ..~,-,e.. .....,.,,. . . . . .*.n -~. = . * * . = = **=
GC Jepensee Ressereh j-; L
- WECFIC RESULTS A * *8 *d *" *"* 'Y* *" a,'.
FOR Sn58.187 CHolt I joints with circular tubes has been re- MATEllAL t.lMIT ported (Ref.11), which cmploys the 50 "- y p e a4 Fyw/ _ _ $'),"/g physical model of Fig.12 to derive an empreselon for theoretical ultimate 4 .- . p j ...-
,, strength which can be reduced to the ,
usu following: g ,, 7, LIMIT ANALYS13 J2b 08 !, s. g)
- 1. gag py e
6memas# ,, , M) , 0.5 y , 2rR 30 - . Vp e$(1-$) .0.5y When the effective length B. la taken
, se equal to the chord circumference.
- a. -p, the last term becomes unity, are to - , , i equation (4) becomes identical with ,
. I equation (3). With a term for the basic verletion of v with F, and 7 . mod.
, i p.YlELD ified by a terra empresalne the # .
10 - F i LifES offset. Yp'l Q57 Test data were used to justify an
/ pop p 4 0,3 l l empirical modification of the empree.
o#,,,... j,f
- 4I sion for ultimate punching shear. .
0 ' ' ' leading to the results plotted in Fig. O at Q4 Q4 QB t.0 12.and S- RATIO font !I - Ukimere strength enetysis - square te60s v, e
- 0.304 y (44)
$ (10.833 $) ,
in this empression the term for # . effect has the following properties andimplications: sehPLIFED 1. A value of 1.0 for$ e 0.6 Fy P L! hili v e QS 8a 2. Increasing joint efficiency for p/RI-#1 0.5 y trR
- t ( ANALYSIS larger $. ratios. up to a llmiting in.
.- : HINGE g creese of 1.8. fold for$ a 1.0.
!- " J LNES . Note that for the mld. range of diem. p eter ratios (#from 0.26 to 0.75) the assurnption of constant punching . !' sheer afse prevides a reasonable fit to t .= cen 3 of F s 12, in imi. u ..r ha.
} ,'as ett, re , wry sman /J.,.po., w.r, g p . ne o .. ... m e r..t vii.e.wi.o ta, th. t,.:;. .ncr e... n p.- et. cioxy .
l; 40 precicted by sne # . mod!!!st in equa. ' c tion (4a). Accordingly, it has been rec. ommended th'at a modifier of unity be c used for values ofpless than 0.6. This
< le consistent with the results for square tubes and appears to be con.
30 . o.3 FF servative with respect to theoretical WP
- i .304 y results (Fig. 6).
Og Propeeed4.Effset N
** Applying the modifier. O s, for the o
l y 20 " effects of diameter retlo, to the punch. l f% o . Ing sheer crit:rla of ee, nations pro. gg posed earlier (equatione (2) and (2a)
** one obtains:
F, lo . Ultimate v, e Q , . , (5)
' "k CONSTANT vp FOR
Allowable V, a Q 8 I 0.255S10.75 0,g xyor where
,, o . .
4.;/ 0 02 0.4 0.6 0.8 LO e DIAh8ETER RATIO S a, 0.3 forg>0.6 font 12 ~ Jeoo'nese resons - tress soones 0(1*0.833f) 194.e l M A Y 19 7 4 . e .
C and 0001.0 for#$ 0.6
'{..
.- Li - 8 o o These criterb. locluding Os. . ira plot.
n* '% '33.* '
- 1. _ .e
- a. ted se the heavy dashed line in Fig.
[* g 1 o 10. O ! 08 - ,
.,.' t Og e LO latereeden Effsete p
- AI M Jopenese date (Ref.11), showing ,
g gy h the extent to which omtal load in the i g chord member reduces its cepecity to '
. E i gg . O, e m.asM carry punching sheer, are plotted in g l PORM> 0.44 . Fig.13. The proposed modifier 0, for
[ g, 1,48 - p'- f: p. Interaction effects would be used in design se follows: NOAO Allowable v* e (6) Q -
/ Paea Q, O,.
2 al - 0 9 = 7:' (
- V s at -
where Q, a t.22 0.5 lU! for;U;> 0.44
]
, Q, = 1.0 forlUl $p.44 0
e 14 08 -04 -Q4 -02 0 at Q4 QS G5 to N SSON TENSON andlUlm chord utillaation ratio at the CHORD UTIUZ.ATION RATO connection.
. U* @p F4 13 ~ Interaction erkets of stress in chord 1 NEGATIVE ECCENTRIOTY ZERO ECCENTRICITY POSITIVE ECCENTRICITY
/ 2 c rC ;
, b/,e#
i l I b/ 9HEAR ON 2.5" /0" EAR ON 9" $ l { 40M C 7 ' HEAR ON 6= l40 N 4 70K
' VFRT. WELD *j l OVERLAP WELD OVERLAP WELD BEAH NG ON LEG U $ji70"' 4
\
-h U W \
- . +.
\ O 3
' 5 V e b k
-C {j -.
l\ ) 3 h' COMPARISON OF JOINT EFFICIENCIES 1 a CALCULATED g {!
. TYPE OF 8ASED ON 8ASED ON $ .i JolNT NOM. YlELD ULTIMATE 137K IN 6984 255K IN 6% $ -^'r POSITIVE l
ECCENTRICITY 41 % S4% , 62 % 82 Y. CENTRICfrY
; NEGATIVE $
,' 86 % 10 8 %
ECCENTRICITY __
.s bg 14 ~ .leonts of various eccentricorses ?d ,
h,. 4 1- [ WELDINO R E S E ARCH SUPPLE M ENT l 197.e
some e ti, us e c;g + s.: . ferred directly from one brace tD cn. the braces *
. ggWgg g*,g other through their common weld. t ,a throat thickness for the One advantage of such joints le that, common weld betwe:n j[,,,_, -
since the chord no longer must trans. braces *
.: , f, _- for the entire load, its thicknees can e the projected chord length la
.: , [
be reduced and ** joint cans" elim. (one side) of the overlap. h.,Ol .;p insted.The amount of overlap can be ping weld, measured in th a g : controlled by adjusting the eccentric. plane of the braces and po . {L J
. j lty of brace conter!!nes, as indicated pendicular to the main g in Fig.14. Negative eccentricity (Ref. member"
- 12) can be used to increase the A comperleon of computed capa.
\ ** I amount of overlap and the static load transfer capacitv of the connection, cities, in terms of brece axial load, P, A crude ultimate strength analysl: using ultimate ve and yield v.,a t .
versus test results is given in Fig.14.
% is proposed (see Fig.19), in which the punching sheer capacity for that pot. Equation (6) appears to be consente.
//pt 18 - Centponents st realelance hr tion of the brace reaching the main live in predicting static joint capa.
" #8 M "Il'd"'* member and the membtano sheer cities, provided there is sufficient duc.
capacity of the common weld be. tility that the stiffer element (the over. tween braces are assumed to act lap)does not fall before the rest of the in designlulwould be taken as the Joint catches up. At elastic loed levels simultaneously. Thus, the total capa. A!SC ratio for the chord at the tubular city of the connection for trans' erring the overloo is so much stP'er that It connection two re p+ct to critons losen pernndeever to tne chord ce. * * '*** bes*d on yield). Equat.on (6) Incluces gor.,,, ee o e...!ce.st ;
- cm * '. 4e k' A ir^'v.*-
safety factors and corresponds to a tional:y us.c, sor e ces gaers i .e to symmetrical failure envelope, as proportion the overlap to carry at shown by the solid line (Fig.13), P sin d e v, t i + 2v,,t ,l (7) least 50% of the acting transverse , Where heavy well joint cans are used . load. at tubular connections, the ut!!! stion where Where extreme amounts of overlep ratio will often be less than 0.44 for are used. It may become necessary to the joint v e" a!!owable punching shear check the capacity of the connection
"" "redITCtidhe,cafi, due corresponding to Interaction. to Forno stress equation (6) for the for transferring loads parallel to the highly stressed K anc Xjoints' main member main member as well as transverse without joint cons, but with equal dl, t a main member wall thick. loads. Both may be accomplished ameters, the increase in joint offic. ness with vector combination of the lency over equation (2a) will be verlous strength elements, as limited to about 30%, when both Os li e circumferential length for suggestedin Figs.14 and 15.
, , . and Q, are considered. that portion of the brace which contacts the main Fatigue Cveriepping Jointe member and Few members or connections in in overlapping joints, the braces conventional build;ngs need to be de.
Intersect each other as well as the y,, e allowable sheer stress for signed for fatigue, since most load chord, and part of the load is trent* the common weld between changes occur Infrequently or pro. duce only minor cyclic stresses. The full design wind or earthquake loads are sufficiently rare that fatigue need c,000 not be considered.
- However, crane runways and sup.
I " portlog* structures for machinery are l 0# often subject to fatigue loading condi. I 10 0 = s tions. Offshore structures are subject [ lo a continuous spectrum of cyclic D' - 2,000 a. wave loadings, which require consid. l 60 - . l e erstion of cumulative fatigue damage g 3 *
, (Ref.13).
u - g -- 1000 Welded tubular connections, in par. e -A a ticular, require special attention to fa.
. 30 .
F* % " ligue, since statically acasptable de.
' ,, ' N ,
s - Soo signe may be subject to locallied plastic strains, even at nominally
% CP%-
O." it' % ., ' X allowable stress levels.
- l k Fatigue may be defined as damage 3
. 7 ****.,,,%..,,**%.., -
200 S that results in fracture after a suffi.
- ,g, .
g
% ,,, %=,,'. - 100 gr l mo % **
*! stopt that the Ime lead rapstett v,,e t,,
; fr, s ' ' '
% , '* % y shouldnet enreed the sheering reperoty of
! %W I'*
E0 SCI l0* l0* C* 801 00
- the rhonner edi omong bese metal "trioterred therd length us proportoonal to C'tCt.ts Of L.0A0 the resuttons et membrene streer, acrong at poen value along the tutt length et she
!Ist I6-demot 01 t fatigue design rurvesIsee febte f) avorteppong weld.
198 s l M AY 1974 ' 8 i
......,p............: ..-n.. . . ~ *" --*= ~ ~ **-~e*~- < * " = " * * * ~ * *
- Tabist 2 - Peelgue Cassgestes
,- e - , .
,. 33ges * . .
' . essegory . Situation 8 Kinde of strese *8
. .A' Plein onweided tube. TCSR 9.;. A Sutt splices, no change in section, full penetration groove TCSR wolds. groutul flueh, and inspected by a.tay or UT.
3 Tube wet's lomei udMel seem. TCin st 8 5.r so ces. Mi p,.e ,,g.o.: g,,c., ies, g ,,,io tig,n. ;33 , .f
!'* 9 tierrbarts ets cer.tenee,vs.t wriu.es tong iAinsi eteen.rs. 1:2it h C ha ep'ees, ts!t nretist.s.i geon
- meics, es wait,o
- 7:1!): F j.
O Nte->ars weia treoversa friest er efrie'a, or mesce ierw.o.,s
,f ~ g TC:M
~ ..t.s.oies ., ~ .e,..em D Tee sad eweiroern t. arts wt3 NI prei,eteen welds TCC;t O',less.c et rubular connectionat. ,
Seate T. v. ce K cennectises em L.I cenerese:oa TC3'4 en erace*i e r** team enmu, m,si ta c%d a h* Mat s% ve ww4, ssp ,aiese per 04.egy, si :,s 1) 8 ( S$'ance<t T a %J crs,c *crv=+ sorets w.
- esmet p.re,,si.g,i TCB4 in ,pe-twr taaet %st a su be to oaJ r, r Cais .v GL ! '
;,we .* % pe IWei w Je f eece;,i se tutW :
connectos,:isk , 9.". Mornbers where doublet wrap. cover ptetes. longitud 6nel E TCSR in member. , a:Wfeners, guseet pletes, etc., terminate teacept et A E', tubular Simpleconnectionek T, Y, and K type tubuter connections with TCSR in beench member (main member in almoie T. Y. or partial penetration groove weeds or fillet weide; eleo K connections must be checked seperstely per Category comples tubutet connectione in which load itene:et le K of T; weld must also be cheched per Cetegory G'L accomplished t y overlap (negeths eccentricity.)guseet plates, ring ettffeners etc. F End wold of cover ple'te et doublet wress welds on Sheet in wold. guteet plates, stiffeners, etc. O f end etuciform iointe. toeded in teno6on or . Sheet en weld (regardless ol dir ection ol loeding:
. bending, having fillet or partial penetration groove weide.
O' Simple T. Y, or K connections having fillet or partial Nominal sheer in weid(P/A e M/S) penetration groove wekse, X Mein member at simple 7, Y, and K connection. Hot spot, stress or strain on the outside f(v' outface of the rnein .nember, et the toe of weld joining branch rnember - measured in model of a, prototype connection, of calculated with best evellable theory. X Unreinforced cone. cylinder intersection. Hot spot stress et angle change. X Connections whose edequacy is determined by testing Wores meneuted hot spot strain, after shehe down. l en occurately scaled steel model. (, KM Sitnple K type tubufet connections in which gamma Punching sheer on sheet stee888 of ano6n member, t ret 60 R/T of molti member does not onceed 24. , i T
- Simple T and Y tubufet connections in which gamma retto Punching sheet on sheet etesiaof main member.
! R/T of main member does not encoed 24. ie 1 set takest Empe=ae nsasses swwe C. seseex n nenei.ee .eo.aeas.a sena es e ewereas ev s d seems seen seas.nem.a seaw,, er men wee i enmas . mas =n ne se own a . se se e,se es
.I see tesues t lt
[ cient number of fluctuations of stress. to a safety factor of 3 on computed ual stresses develop. What is usually Where the fatigue environment in. fatigue life. An alternative approach, measured on the actual structure (or a . l- volves stress cycles of varying magni. which will be presented here, is to scale model)is the strain tange, with tude and varying numbers of applica. use fatigue curves which fall on the the aero point undefined. The con. tions, failure is usually assumed to safe side of most of the data,it might stant strain range sp;;tesimation is in l .' occur (or reach a given probability be noted that a linear cumulative falt agreement with the results of fa. ., level) when the cumulative damage demoge rule is consistent with the figue tests on practical as. welded io ta, D, reeches unity, where fracture mechanics approach to joints, particularly in the low cycle i fatigue ctack propogetion (Ref.14). tange. . D e 1 n/ N (8) Stress fluctuations will be defined Fatigue criteria are presented as e and n e number of cycles applied et a in terms of stress range, the peak.to. set of S.N design curves pig.161 for given stress range trough mognitude of these fluctus. the various situations categotlied in 1: N e number of cycles et that tions. Mean stress is Ignored. in weld. Table 2. stress range corresponding ed structures we usually do not know Curves A. 8. C. D. E. F, and G are to failure (or a given probabil. the aero point, as there are residual consistent with AISC fatigue criteria s pM lty of f ailut stresses as high as yield which result (Ref.15), which appear in turn to re. Some designers,e) limit the damage from the heat of welding. Where flect the data published earlier by ratio to D 33 when using median or there is localized plastic deformation WRC (Ref.16). Curves rather than best fit fatigue curves, corresponding during shakedown, a new set of resid. tabufsted (step function) ellowables e W E LD IN G R E S E A R C H S U P P t.E M E N T l 199.s O . .
- . - - - ~ -. -
l ., Rea.
.' e J' .
'd l*<e"* *
- lgy . -
~g o ,
a M
'es os me ce go or es ce es sec es a ss m* er es cmss.m
, o i' FU l 7- ferique curve C - nominalstress agatent to weM y.e to se 4
e 1
* ,f7'e
. fig 20 - Punthing sheer fatigue strength e/K tennections j d' '
E , , j'h' c ,.*. *. Q ,
% l
$. g .
*o. 's' %
.,, a sus on es os o' or se s% %,
cTct.cs O
'en ga g4 es os o r gs O '
cra.ss h fig: 18 - fatigue curves C and X - het spot strain satrocent to wood ,We,, I.
- Fog 21 - fatigue curves O and 0*- nominst member stress et
, ,* R ec . tuttponettelsen T welds endsimplejoints .
.c .
,, eg . '7 1c
, ,. A g c' . . m itw N p
+
*.**\ , ,
@ ,f,
- f
*e e, e. y ,l ,, ,e
~~,.,_< ^? .
i cm :, . y ga
's. peea s e ..
F.se to se gygggg e
*
- ome,w FC 19 - Punthing sheer totigue strength el T.tennections i
sweet : cre used because they are more op. proprtt3 to tubular structures eu. O posed to a continuous spectrum of cyclbe loede, in these simple eltus= fog 22 - ferrgus turves t and E*- nemonalmember stress et til. tions the nominal member stress (f.
- Ier welds endremplesje nts f.) fairly well represents the actual ett:ss as would be measured adja.
c:nt 13 the weld. See Fig.17. . perimental(Refs. 4. 7). analysis of the actual as. welded hardware - tubular Curve X le based on current design connection. Category X is consistent connections. pressure vessels, lab. practices for offshora structures (Ref. with category C since the local oratory models and prototype failures 8). The televant stress for fatigue fall. transverse stress adjacent to the - from a variety of sources (Refs.13,
, uro cf tubular connections is the hot weld is considered in both cases. In 14. 18. 18, 19. 20, 21). In the low spot stress measured adjacent to the the range of inelastic stresses and cycle tange. the design curve corre.
i g. weld, as shown in Fig.18. This is low cycle fatigue (Ref.17) it is more sponds to roughly 96% survival (5% h realistic to deal in terms of hot spot failure probability) based on test data
! g nom.cIly considerably inal member stress, higher than the and would which are spread out over a scatter strain tathet than stress. , normally be determined from a de. The data plotted in Fig.18 repre. band more than one log cycle wide.
1:iled theoret6 cal (Refs. 6. 6). or en. sent hot spot stress (or strain) from Wsthin this range, all structJral qual. 200.s l M AY 19 74 . - l e
.. . . . . .. . ., . . . . . . . . . - .. . . , . . . . . . - . . , . - . . . - . - -.. -- ~ - - --
S ity steels show similar fatigue sino that for some connections of thi) ere/#eevirements for the MeterAre went. J'. . . behavior, independ:nt gf yi ld type curve E b too conservative but int e/Structur:tSte*t ru6es tsc.s. 777s.
..'
- strength in the range of 36 to 100 hei: unfortunat:ly at this steg3 no distinc. 4. Toprec. A. A., et al.. 'W;lded Tubuter
,. . Differences which show up* for tion can be made. Connection,e- An invest'9: tion of 5tr::see emooth polished laboratory spec. Curvee D. E. F. and G are limited to 'n T.Jomes we/deg Journet Vol. 45. No.
a 1:none in the high cycle range simpfy situations in which nominal member f[enuary 1966. Res. Suppl.. pp.1.s to
,do not apply to practical as.we stresses represent actual load (notched) hardware subjected to loc transfer across the weld. Curve G is 5. Dundrove. V stresses et Intersec.
laed pleetic stralne in the presence of rien of ru6es - Cross ed T.Jomts The shifted down to a factor of 2.0 to University of Temes. S.F.R L Technical I a corroelve environment (e.g sea. account for the uneven distribution of Rooort P 850 5(1966L water). load transfer across the weld at the d. Greate. Ojers. A Computer Program 1 Little data are available fof.the high tube.to. tube intersection (Ref. 5). Ier the Ane/rsis et rl,aulu r, Joints. . l cycle range.*over 2 m 10* cycles. In The dets supporting the empirical University of California Structural Engl. the presence of initial flows and/or design curves. T. K. D*, and E' general, nowing 1.ab. Report No. 691911969L ;. corrosive en*on,nents. e,e is no ly show ,nore scener een the ,nore 7. = o e g'e.-' e^- e a< roarea. ^. ^ endu,eme n, nit. . and ine fsogue basic dois of rig. i ,,,,e,ny be. w - strength continues to drop off, cause they neglect some of the rele* Unfortunately, use of curve X re. Council Bulletin 125. New yott. N.Y Oc. - ! vant factors, and only represent iober 1967. stuires knowledge of stress concentra. "typicar* connection geomstries. s. Marshett. P. W., et al.. "Materiais i tion factors and hot spot stresses Where actual stress concentration Proolems m Offshore Pteiforms." Offshore 7 within the tubular connections - factors are known, the use of curve X Tec5nology Conference Proprmt No OTC i information which would not be avall* 18 to be preferred, t04311969k ~. able to many designers. However. Because of the uncertainty and 9. Rober. J. B.. " Ult,1 mete Strength De. ,
, anyone should be able to calculate scatter involved. calculated fatigue '
punching sheer (equation 1) and 8," D",'*hd*Q're,nfN
, TC 4 lives should be taken with a healthy make use of the empirical design amount of skepticism, and should be 09721 curves T.and K (Figs.19 and 20) for to, crett, w. j., wided Tubular Con.
- viewed more as a design guide!!ne noctions of nectangurer and Circuiar Not.
cyclic punching shear in, respec* than as an absolute requirement of low sections." paper for presentation to 4 tively.T and K connections. These are i the code. ' the Tens section. ASCt. El ruo, October ? based on data assembled by Toprac s.to,1970. l (Ref. 21) from tests in which the Concluding Remarks 11. Tootec. A. A., et al . tudies on i chord thinness ratlo.y . was limited 7,3,f, j,,,, , j,,, _ py, s, _ 3,,, - to the range of 18 to 24. Thus the The criterla presented have been *I 8e8e8'8h 8eP88 'epo'8 p'eD*'ed for I curves may err on the safe elde for developed primarily on the basis of re. Weidmg Rueerch Council. Tubular Struc. very heavy chord members ( y under search and experience with flued off. '
g*2. O u ,, 18), and they could be unconservative shore platforms. These structures are em r'd m b for chords with 7 over 24. Since the ruouter Connections in structure / Wort. '
highly redundant, and locallaed tubu. weieing neseerch Council Bulletin No. 71, theoretical elastic punching sheer lar joint failures can occur without i
- efficiency (Fig. 6) varies inversely 1961. .
I leading to collapse of the atructure. 13. gell. A. O., and Walker. R. C.. with 7 0 7, it is suggested that, for One purpose in presenting this "Stresne Esperienced by en Offshore
+
- ' chords having y greater than 24. the !
paper is to let potential designers of Mobile Drilling Unit." Offshore Tech. l allowable cyclic punching sheer be re* nology Conference Proprmt No OTC 1440 other classes of tubular structures ,i duced in proportion to (24/Y). U87D l see just how the data fall relative to #
.y Once failure of the chord in the the proposed criteria, and what the punching shear mode has been pre
- of efd T b ter Jo nts O shore T h wanted, by the use of heavy well scatter is, so that they may be in a noloov Conference Pruormt No OTC 1228 '
** joint cans,' or by means of other position to evaluate the suitability of 119701 l joint reinforcement, the problem of the criterla for thelt particular applica. 15 American Institute of Stul Con. k
- tion, struction, Specs /,cer,ons for Design /e6 possible fatigue failure in the braces 11 Also, it is hoped that, as additional rication end frection o/ Structurel Stw/ 5 remains. In simple joints, localized data become available, they will be /or duWmps. New York. N.Y.. Festuary 12.
stresses in the brace may reach 2.5 1969. $ times nominal f, e f,, due to non uni. compared against the criterla and y dets given herein. Such comparison
- 16. Wen. W. H, and Gme. Le /s. 5 form load trsneler (a factor of about 2. discussion, and re.examlnstion rigu' el WeWed I'est Structuret Weldmg 1.
Fig. 6) restraint to Poisson's. ratio R ueerch Council. New York. N.Y.1964, 2 breathing (a factor of 1.6 for perfect should eventually lead to a better 17. Peterson. A t.. "Fetigue of Metals 3 asisymmetric restraint). and continu. design. In Enginntmo and Oman." ASTM Mer. 1 ity with the severely deformed chord, The authors are Indebted to their burg Locture.1962. , Accordingfy, curve D' (Fig. 21) when colleagues in the various API. AWS, 18. Koodstra. L. F., Longe. E. A . and WRC. and ASCE task groups con. P$ chert. A. G., " Furl sise Prusure vesset
]
applied to nominal brace stress takes a these factors into account. Data corned with welded tubular struc. Tming and its Apohcotion to Dwsn," _t tures, whose prodding and comments ASME Peper 63.Wo.293.1963. points are for thick walled simple 19. Bouwtsmp. J. G . Tubuler Jovits joints tested by Bouwtamp et al(Refs. helped shape the guidelines present. f ed here, Under Static and Alternetmg (oedt Uni. _- 14,19), for which f ailure occurred in versity of Cohfornia. Structures and Mete. 9 the brace (branch member) rather Ae/etences nels Research Report No 6616.' Berk. ' than in the chord (main member). 1. AP/#ecommendedpractice /er Plen. eley, June 1968. Where some other form of joint re, ning. Designmg. and Constructmp /ned 20 Toprec. A. A., and Netereien. M . An i inforcement is used (such as brace C# shore Flettorma AM RP 2A. Founn Innstigerie et WM fuouter Jomrs ; EditionllS73L Progress Report. Internet,onel Institute of overlap. gussets, or rings) localised 4 stresses in the brace may become 2. Amence Weldmg Socierr Structural Weidmg Comm XV Doc XV.265 69. June j Wedding code. AW5 01.172l1972t 1969. i larger and more difficult to ascertain 3. British Stenderd 4491959 Appendia 21. Toprec. A. A .. Design Conssd. and thus have to be desgned accord * 's : C. "Determmation of the Length of the eretions /or Welded rubular Connections, l. 6ng to curve E'(Fig. 22) which emplies Curve of intersetton of a Tube with An. Report propered for Woldeng Research 4 stress concentration factors as high other Tube or with a Flat Piete". and Council. Tubulet Structures Committu. 5 as 6. However. :: should be stated British Standard 938 1962. Spec. /or Gen. December 1970.
- YV E l.O I N G R E S E A R C H S U P P L E M E N T : 201.s }
g
, . l
- -- -- - - - ' ~ " " ' "
TEXAS UTILITIES GENERATING COMi%NY 'GET .Fl E
, P.o. BOX 1002 Gl.EN ROSE. TEXAS 70043 November 16, 1984 liOTED NOV 9 1984 " 5
- 3 Ms. N. H. Williams (AN.)
Project Manager , CYGNA Energy Services 101 California Street, Suite 1000 hi *- g o San Francisco, California 94111-5894 COMANCHE PEAK STEAM ELECTRIC STATION fJ.b Independent Assessment Program Phase 3 g9 t Cinched U-Bolt Testing & Analyses Program M8 = f {I L Additional Information gqgqg fp REF: 1) J. B. George (TUCCO) letter to N. H. Williams (CYGNA), dated November 1, 1984 - same subject
- 2) N. H. Williams (CYGNA) letter to J. B. George (TUCCO), " Status of Cinched U-Bolt Testing and Analy 's Program", 84042.018 dated October 1, 1984
Dear Ms. Williams:
Reference 1 provided in its attachment the information requested by Reference
- 2. Included in the attachment as part of the answer provided to Item 2 of Reference 2 were results of a finite difference heat transfer analysis con-ducted for an uninsulated and an insulated U-bolt configuration on a 10-inch pipe.
A rechecking of the modelling of the contact areas between the U-bolt and the pipe and the pipe and the crosspiece has indicated that the contact be-tween the pipe and crosspiece was overestimated and that the contact between the pipe and the U-bolt had been incorrectly assumed to extend for an arc of 1800 Accordingly, we are providing in the attachment to this letter the results obtained for the uninsulated case of the pipe at 250 F and the in-sulated case with the pipe at 350 F, where the boundary conditions of the model are changed to reflect the more realistic contact areas. We will be glad to discuss the details of the model, if CYGNA so desires. Please call if you have any questions. CYGNA Very truly yours, - [' p TEXAS TII,ITI S GENERATING COMPANY JOB NO _ Jcfg gjg, DATE REC *D/LOCGEDs _ }}//q) pct i J hg f LOG l'O. s fj Q 7
. Obor / FILC Vice President / Project General Manager p, /,j g , g C30SS EIF. FIIE gg _ g g JBG/RCI/gh f' ,
cc: S. Burwell R. Iotti J. Van Amerongen D. Wade a snvensenx eur snam essa. storm nursuor venus Asr
h- f
.,e Ei ATTACHMENT 1 3-
'? Revision to Item 2 of Reference 1.
Please replace Item 2 response with the following: i- A. . The answer'to this question is best worded'by first restating that the. choice of 2500F for the:10-inch pipe temperature is a compromise choice !' which bounds the majority- of the systems in the plant, and where used with _an uninsulated U-bolt configuration is also representative of the case where the pipe temperature may" be 3500F but the U-b'olt configuration is insulated.' Second, it is important to point out that there is a single cinched-up U-l bolt which is used* on the 10-inch portion of the'RHR system. This is support RN-1-024-007-522R which is on line 10-RH-1-24-601-R-2, which is connected to the' outlet line of the RHR heat exchanger. The nr2ximum normal temperature i seen by the line is 2800F during initiation of RHR operation. Only under i upset conditions, where component water cooling may be lost, can the maximum temperature of this line reach 3500F. Therp are no cinched-up U-bolts on the inlet side of the RHR heat exchangers. . I Third, it is germane to point'out that the tests conducted on the 10-inch , l pipe specimens had a corresponding average temperature of the U-bolt equal l to approximately 1500F. For the particular configuration examined here, i.e., ! stainless steel pipe and carbon steel U-bolt, the approximate 1500F represents l the equilibrium temperature of the U-bolt. The following describes the tem-perature history during the thermal cycling test and the creep test for both' the U-bolt and the crosspiece. l Thermal Cycle 1: l The pipe reached the test temperature of 2500F at 30 minutes, but then con-tinued to climb to over 2800F before settling back down ot 2580F. The U-l L bolt radius and leg stabilized around 1950F and 1500F, respectively, near the end of the cycle. See Figure 3. l Thermocouples 2, 9 and 10 on the crosapiece reached temperatures of 1290F, 1360F and 1440F, respectively, at the end of Cycle 1. These are less than the equilibrium temperatures reached du;ing the creep test. Figure 4 shows that temperatures had not leveled off. Refer to Figure 9 for location of thermocouples. Thermal Cycle 6: ,_ The pipe reached an equilibrium temperature of 2500F within 20 minutes. The ! U-bolt radius and leg reached 1830F and 1440F, respectively, around I hour. ' l See Figure 5. Therecouples 2, 9 and 10 on the crosspiece reached temperatures of 1250F, 1320F and 1390F, respectively, at the end of Cycle 6. These are less than the equilibrium temperatures reached during the creep test. Figure 6 shows that temperatures had not leveled off. l ts-
e , af Pass 2
'. ~,
te Creep Test: = The' pipe reached an equilibrium temperature of 2500F in less than I hour. The U-bolt radius stabilized at 1850F within I hour. The U-bolt leg sta-bilised at 1480F within 2 hours. See Figure 7.
>iThermocouples 2, 9 and 10 on the crosspiece reached equilibrium tempera- '
tures of 1380F, 1460F and 1540F, respectively, around 3 hours. See Figure 8. 10" Specimen Summary: With a pipe test temperature of 2500F, the U-bolt reached thermal equili-
.brium during each cycle of the thermal cycling test, but the crosspiece-didn't. The entire assembly reached thermal equilibrium shortly into the creep test.- A summary is provided in Table 1.
Results of finite difference thermal analyses are very sensitive to the assumed , E area of contact between the pipe and the U-bolt and-the pipe and the crosspiece. When the U-bolt is cinched, the.line contact between the pipe and the U-bolt extends for an arc which is less than 1800 , and the precise extent of which
~~ depends on the cinching force and the spacing of the bolt holes in the cross-
' piece.- Similarly the cinching process tends to produce a loss of contact at some points between the crosspiece and the pipe due to either bending of the crosspiece or. local deformation of the pipe. This loss of contact, however s' -small, profoundly affects the heat transferred from the pipe to the crosspiece.
A hest transfer model has been executed for the uninsulated U-bolt configura-tion with the following assumptions. Heat transfer from the pipe to the U-bolt is along an are near the apex of the U-bolt. At the diametral location there is a small gap (less than 1/16") between the pipe and U-bolt. No gaps are assumed between the U-bolt and the crosspiece (the assumption is believed to.be inconsequential since both elements are roughly at the same temperature at that location). Heat transfer between the pipe and the crosspiece takes place through a line contact extending 2 inches along the pipe, and via gap conductance, along the circumference of the pipe and through a gap increasing from zero to 1/128" linearly from the end of the contact area to the end of the plate. Likewise, the heat transfer between the pipe and the U-bolt also considers the gap conductance with areas immediately adjacent to_the line of contact and extending out to the U-bolt radius. This model produced results
-which more closely match the results of the test.
Results of the analyses are shown in Figure 1 for the uninsulated case. In Figure 2 similar results are shown for the insulated case. The only difference between the-latter' analyses and that of the uninsulated configuration are the pipe temperature, which in the latter instance is 3500F, and the presence of insulation. For the uninsulated case the average temperature of the U-bolt in the curved portion is 175-1800F, while the straight portion is at about 1500F. For the insulated case the corresponding temperatures are approximately 3000F and
/2600r res~pectively.
- , . ., . ,+ .-_ - . _ . .
... - -. ~ _ - . .. . _ ..
o p -
< s Page 3 , 1 h
- .(The effect of the temperature rise on the clamping forces acting on the pipe 0 'and the U-bolt for the two cases of 2500F pipe, uninsulated U-bolt and 3500F i ; . . pip'e. insulated U-bolt, can be estimated by comparing the relative growth of
[f the pipe to U-bolt for the two cases, neglecting any deformation of the piph. t Since only relative growth is pertinent here, the one-directional growth cf ]. the U-bolt due to thermal expansion given as Yi where Yi= Q TL + where L is the projected length of the U,-bolt which is given as 2% and T . is theitemperature. differential betweentthe average U-bolt temperature and ambicht (or a reference temperature), is' compared to the diametral growth of
~
the pipe, Y7 ,,which is given as ,
,' Y2= 4(ATD 1
g i u, The worst case relative expansion will occur for the stainless steel pipe-and the carbon steel U-bolt. For the 10-inch pipe (10.75 OD), coefficients of thermal expansions (= 6.3 X 10-6 in/in/0F at 150-1800F or 6.6 X 1076 at 0 260-300 F t.ud A = 9.4 X 10-6 at 2500F or'9.53 X 10-6 st 350 F and a reference ambient temperature of 700F, the relative expansion for the two cases con-sidered, i.e.1, 2500F pipe with bare U-bolt, and 350
- F pipe with insulated U- '
bolt.are as follows: w
- 1. '250 0F. AY = 0.011755 inches . .
- 2. 3500h AY. 0.0137 inches
- 3. Finite Element Analysis B/ = 0.0141*
I: (* Finite Element Analysis used 2100F.) As seen from the above', theoretical, steadystate heat transfer analyses would predict that the case of 35_00F pipe expanding against an insulated U-bolt i could resalt in a differential pipe expansion which would be approximately 17% larger than could be expected for a 2500 F pipe with uninsulated U-bolt. How-ever, the finite" element analysis has been conducted in a manner that would ; encompass the case of 3500F insulated U-bolt. As seen from the third' row of relative expansion, the finite eierent analysis, which used a pipe temperature of 2100F but maintained the U-bolt *emperature at 700F, would yield a relative
- expansion which is comparable to tt case of.3500 insulated.
Another point to be discussedg is that the test has provided information on 4 the transient thermal expansion differential between the. yipe and the U-bolt. As'seen from the data which is attached as Figures 3 and 5', the maximum s temperature differential between the pipe and the U-bolt occurred when the l .U-bolt has reached a representative temperature of about.100-1050 while the
. pipe had been heated to 250-2550, a difference in temperature of ipproximately ,
1500F, This difference is well simulated in the finite element analysis
.w here there is a constant difference in temperature of 140 0F. It shon1d also 1
.be' remembered.that for these temperature differentials, the amount of' stress caused by the thermal expansion is not very significant.
s 3 v g
.,,..,_,,,,..,,,_.-,~.--d- ,-,,_..,,_,.a,-_,.-
1 . .. i TABLE 1 : l f l
~
l 4 U-BOLT THERMAL AND CREEP TEST . , I DATA EVALUATION l TIME REQUIRED TO REACH EQUILIBRIUM TEMPERATURE, "F EQUILIBRIUM TEMPERATURE, HOURS l' U-BOLT U-BOLT U-BOLT U-BOLT PIPE RADIUS LECS T/C 2 T/C 9 T/C 10 PIPE RADIUS LECS T/C 2 T/C 9 'T/C 10 , , INSULATED SPECIMEN 2.5 2.5 * * *
* *
- 2.5 THERMAL CYCLE 1 559 498 451 T THERMAL CYCLE 6 560 530 440 * *
- 2.0 2.25 2.75 * *
- g 564 495 451 322 340 365 2.0 2.0 2.0 3.0 3.0 3.0 TT] -
! j.- CREEP r. r r
} 10" UNINSULATED SPECIMEN 1.5 * *
- 250 195 150 * * * .50 1.5 1
THERMAL CYCLE 1
* * * .25 1.0 1.0 * *
- 3> i-THERMAL CYCLE 6 250 183 144 .T, l' 3.0 3.0 ',
250 185 148 138 146 154 1.0 2.0 2.0 3.0 . s% i CREEP 4
$e .
! 32" 1NSA ATED SPECIMEN
- THERMAL CYCLE 1 560 * * *
- 4.0 *
* * * *
- 5.0 * * * *' *
- THERMAL CYCLE 6 560 l 251 4.5 11.5 12.5 14.5 14.5 14.5 154 175
]l ' CREEP 563 440 353 I
- THERMAL EQUILIBRIUM WAS NOT ACHIEVED.
.___...,m_....-.
_.7.. O a
~
EBASCO SERVICES INCORPORATED
. .,M ?ZunA ..,, io[2 2/sp -
. . . , o, CMED. DY o . . . r.
DATE OPS NO. me. CutNT TN bO p=0;ter CNw b 'h
. . , , , , m wAu i so" wh set pg u- setcr og eek .& as6% _Ab4l sudaTis FIGURE 1 c
l4C IU,
. . N I,
~154 ;
7 .. ' 50 . . .
-2 \ gno.
/ . s 14 9
/ \ \'
i
..s -
(
/
" c = 161 9
J
\
\$
. . . . - - . -a . ..
~
EBASCO SERVICES INCORPORATED yy 74 DaTg I A b ~ suggt OP h 0887. CMKO. DY DATE OF5 No. No. CLIENT TO bE P40 JECT b EAM w...et h ub 'a so" tub sht a~n V - sd LT A mth C P'f G O 25 c* G ^~~ M SA~\ug Q yg FIGURE 2 N d i 350 \ l
/'
i N. . f i l .
,! 's d>M
/ \
N _ l l f w3
.)
t 99
- . , _ , _ . . m.
4 e.h* es h44_w w. A te w ,t.44 4.>+s..s . . ..A- . .
- 9. - _$e . e4 . eee ** . .*
e G ppsuMiR@ O 9
.e 10"55 - uuth46 C%EL p- .
15PE M 9 e L 2~~ g-hT m g e .. _ ._. . . j l i u. .w. . =. . . . . -_ . I . _. . . . _m... . . _ . . e l t
, y. . . . _ ..-_ . . - .- ._._ - . .
. e . .e egnae.e=.e. . O_e
. emee ._. 6 .m.
- e.g &es. @ WM-D .
pM e..sep 4eae e .e > l 0 .
. ._. .. . . _ _ ._. . L ___ --_ _.1 _ _ . _ _ __.. . _ . .S t
. _._. _.. . _TME,. h f5.-___ . _ _
. M W. . M M m e
.= 9. ee.4 6s aueeux e amm. == .. DM w w .e e _e. e ee .e 6 . MS N Gh 6..
FIGURE 3' i i l l
gsM h O 9
- PREbgrNARY e
9 9
. .e e . . m e.
.__. . . . . _. _. .. 10 -
U na 4"11ED .. _ .. u ._ . ._._ . .__._. _ _ c vc u!. 1. __ . . . ..
. _ _ __ _ . _ . . . CECSSRECE _. . ..
gw .p. _& .. g .pq ee
. 8.D . . . -D h.. _ e e &6& 6.ee- h m.. O 9
ggo . . ... . _ .__ _ _. - _ . _ . . _ _ 10 9 u.. __ . . _ . _ .. . . -. . . _ . _ t em - -. - . _._ . . , . ,_ M po .. . . _ . . _ . . ... .... . __ .. e- .e wb4 e. W s..M_H. em s
- m. 66. 6 M
>h*w.6 m ..W__
Se _ l -
.. _ _So M b'MM*. .M
.' .. Oe. Me
- WM...HWO.4m. M
. -e. 4 m .w . ... ..66_
l l l i
& ^
i . D.S ._. . t.O 1.5 i l l * . -.. - - _
- M_... . . . .- _. ._ ...
FIGURE 4 i l l I I l d I k
..s_... .e . ._ ..
E
- N g a "p ? , F CI L'{G g iiM T 4I I C[\L f
d D'n - uuM$p4Tiilb cic 4 4 i w .. . i g
- ^
Y g. . . - . . . - ( .. .. . .. . l .. . . . . . _ . . . . . . . . ! c
. . . L. .. . . _ 2..._.. . 3 s
l FIGURE 5 l i t t l I
. _--.....- -_ _ _ _ . . . . . . , _ _ . _ . _ _ . , _ . , , , . . , . . _ . , . . ~ . . , _ . , . . , _ . . . _ . . . . . . . ~ . . . . , _ _ . , _ . . _ , . . , - _ _ _ _ . . . - _ - . - _ _
3.__._-...=_.
?RELLMiNAksi
_ . _ _ . \0Y - uu 415uAn ._
.._.g.,-_...._.___ c.M Ct.E I. . . . _ . _ _ . . _ .
._ .. ._ _ _ _ _ _ _ _ _ __.._ cccE6PiELE __. .. . _ _ _
... H 50 - - - - - - - --
9. 4 N I_.m__. 4
.*.*e
. _ .g . _.. __. ..
[ _ _ _ ... ( t t I l '
- - . . ._ _4 i
as. LO 1.5 ( ! _ .- ___.__ _. "T w , modes __ i l FIGURE 6 l i L I l l i
PREUM:i%Ry _ G y - - . .
~ -
f
. . .g
. . ..qp
.t 7 ,
e-
~' -
3* 'N li .. . b S $ . . . . . . . . . .. . el sv m a - M 3 o N _ . _ , .__. -- .. _ . . . _ _ . . - _ . . . . . g _-. ... . ... . . . , _ . . _ . . _ . . . . . . - .y F
.. . . . -,, ,a.
.m.
.k - - . . - ._ M ._. -. . . .
- s. %mm w w w e- eg wa-----.--.---.-w-.-m- , - w. - + , - . - . - , - - - --e-+-. - - -+ - -
~ . me ,
ER E_lMI NARY e 9
=*= m.m**.-=** * '-- * * * * * * * * * *
- mm. e se e. aoemeemeeeeeoe . - -
e0 -. ._-..
.. . . . . _ _ . . _ . .+ A Y -
Mom m * - ep.m one e om . .
..yao< e om.
m m* _e= Ge = e6 e.6. . 9.N.Oh-
.m m e e es e4 m....oh. .e.e=. .a p g . .pe e.-
ew.-6 -ee .M
.,p,ggg g en ag gppm a e me e>,g ,p 4.m g gg g,m W
_. ._ . . __. . . __ . _ . . _ _ .- . -._g D- . .W46M. 6..
.m e&
g .O pp m g
== .*.s.=.- .m=*=*
.- .m=
6 . . _ - - . - - . . . + . 6 g . ._ _ _ _ _ . m co
- -._ g M
2 W b R N - ___ ____ __ . . ss: 3 . ..
& -. [
_ ..U y-y .- _. - - . s - . - -
.Q ... _ . _ _ _ ..
.q.----
-M e em.. M W> .gp G e
e.6 e m .ame. *W.. 6EF6 e_
--a -
ee ,e. m Oe .- g O _.e De . . M e em-e e A-- e a'4h en e e.pe e I i
, a '
V a R o
-. o _..
n
- . . . . . . . - . -. . . . A y gyg.--
g._..._..__ __
*. e s a f e
)
6 4 T/c 9 1 TN 10 Tt 2.
~ j, -
/
s s u.eoo eT/c
+ ' -
l u , U*NY WNb i
$ g gC.EE1976 ~
FIGURE 9 l . l l t}}