ML20116K657

Forwards Cygna Ltr Summarizing Status of Cygna Overall Conclusions for All Four Phases of Independent Assessment Program

ML20116K657
Person / Time
Site:	Comanche Peak
Issue date:	01/29/1985
From:	Williams N CYGNA ENERGY SERVICES
To:	Bloch P, Jordon W, Mccollom K Atomic Safety and Licensing Board Panel
Shared Package
ML19283E951	List: ML20116K657 ML20116K678
References
NUDOCS 8505030348
Download: ML20116K657 (41)

Text

__ . _ _ _ _ _____

.e A

[ T6TO .

even

• 415 397 5500 101 CaMornia stat. sete 1000, san Francisco CA 941115894 January 29, 1985 I 84056.052 i 4

i P. B. Bloch, Esq.

Chairman, Atemic Safety and Ocensing Board i U. S. Nucler" Regulatory Comission i 4350. East Wt.4 Highway, 4th Floor i Washington, P. C. 20814 ,

'Dr. W. H. Jordan 1

.881 W. Outer Drive -

l

.0ak Ridge, Tennessee 37830 ,

I

~ ~Dr.~ K. A. ' McCollom .

Dean, Division of Engineering, Architecture and Technology Daklahoma. State University S-111 water,. Oklahoma 74074 Gentiemen:

Enclosed for your information is a copy of Cygna letter 84D56.050 suimarizing #

the status of Cygna's overall conclusions for all four phases of the Independent i Assessment Rogram. This letter wts transmitted to the NRC Staff last week.

Very truly yours, i

N. H. Wi!Iiams i

-Pro. ject Manager -

I

, Enclosure '

_cc::See Attachment  :

1

'.1 -

1

.i h5030348850422 ADOCK 05000445 -  ;

A PDR ,

i i

i

. san Francasco . Ansaan .3anZhsgo . Cheago 3henland

84055.052 Massrs. Bloch,~ Jordan and McCollom- Page 1-p, ._:.aary.29, ISS5

l. .

i tes.."Juanita:Ellis President, CASE I l 1426 S. Polk.  !

l i r.

. Dallas,', Texas.'75224, w/o attachment I ~ Nicholas S. Reynolds Esq. i

-Bishop, Liberman, Cook, Purcell & Reynolds 1200 Seventeenth Street, N.W. i

Washington, .D.C. 20036, w/o attachment-I

~

Robert Wooldridge, Esq. i Worsham, Forsythe & Sampels 4 2001 Bryan Tower .

. Dallas, Texas 75201,.w/ attachment ,

Mr. H.'R. Rock s Gibbs & Hill,.Inc. .

i

.393 Seventh Avenue New York, New York -.10001,w/attactment i Mr. A. T. Parker ~

Westinghouse Electric Corporation

.P.O. Box J55

- Pittsburgh,- Pennsylvania ~15230, w/ attachment Renea Hicks, Esq.

Assistant Attorney General .

i

.. Environmental Protection Division 1 P.O. Box 12548, Capitol Station

! -Austin, Texas 78711, w/ attachment l Mr. Junes E. Cunnins Resident Inspector / Comanche Peak Steen Electric Station  !-

c/o U.S. Nuclear Regulatory Consnission l i l 'P.O. Box 38 i

. Glen Rose. Texas 76043,w/ attachment l i I 4

S Mr. S.' Burwell Licensing Project Manager U.S. Nuclear Regulatory Commission l'

7920 Norfolk' Avenue -

[

, Bethesda,: MarylandV20014 -w/o :st+=* ment

.Mr. John 7. Collins-

~.U.S. NRC, Region-IV

.611 Ryan Plaza Drive

. Suite 1000

. Arlington, Texas- ~75011, w/ attachment .

< Mr. Lanny Al'an Sinkin 1114 W. -7th,. Suite ~ 220 L

!' ' Austin, HTexas' -737D1 w/att whmint

84056.052-Messrs. Bloch,' Jordan and McCollom Page 2 January 29, 1985 a

'Mr. B. T. Clements

• Vice President Nuclear. ,

T;xas. Utilities Generating (Company Skyway: Tower , '

400 North Olive Street L.B. 81-i' - Dallas Texas 175201, w/ attachment Stuart'A. Treby, Esq.. '

Office of the Executive Legal Director U.S. Nuclear Regulatory Commission 1

7735 Old Georgetown Road - -

Bethesda, Maryland 20814, w/o attachment Mr..J.'B. George ]

Texas Utilities Generating Company 4 I

i Comanche Peak Steam Electric Station f Highway FM 201 .

l Slen Rose,. Texas- 76043, w/o attachment

- Mr. David H. Wade Texas Utilities-Generating Company j-400 North Olive Street, L.B. 81

' Dallas; Texas t 75201, w/o attachment Mr. David R. Pigott

.Drrick, Herrington .& Sutcliffe -

I 600 Montgomery Street .

San Francisco,--Californin' 94111, w/c attachment-Ms. Ellen C. Ginsburg Atomic Safety and Licensing Board 4350 East West. Highway -

Bethesda, Maryland 20814,<w/ attachment i Mr. V. Noonan i U. S. Nuclear Regulatory Commission .!

7920 Norfolk.- Avenue . l Bethesda Maryland :20814(w/o attachment .

i  ;

' l' l l )

-! I l

t

~

r

- er w

1

.(

ici cawoma str

t. sun.1000. sen Francisco, cA e4iii-ses4 .

415'397-5600 January 31, 1985 84056.053 C

Mr. J. B. George

. Project General'Nanager Texas Utilities Generating Company Conanche Highway Peak FM 201 Steam Electric Station i Glen Rose Texas 76043

Subject:

Phase 4 Open Items - Punching Shear .

Texas utilities Generating Company Ccmanche Pak Steam Electric Station Independent Assessment Program - Phase 4 Job No. 84056 i

Reference:

Open items - Punching Shear,*.B4056.D51, knu Der Mr. George:

following three areas:The attachment to the above referenced e

letter ha 1.

The attachment refers to pipe support drawing A

wascopy of this drawing was inadvertently omitted when the letterMS-1-0 distributed.

statement: " see attached support drawing)."Cygna has revised Att The drawing was pro--

vided as an e(xagle only and is not integral to the made.

2.

In the last graphic error).sentence of connent one, change'"all" to "an" (typo-3.

line analysis of the finite element results" and inse -

l words shown, which batter describe Cygna's approach to the evaluation.

e

. San Francaco Boston SanDege Ceucago Actdand

-w

-. =

l l

S ,

sumer M IS

'Mr. J. B. George 84056.053 j January 31, 1985 Page 2 i

This letter and attachment therefore supersede letter 84056.051 and its attach-ment. Please call if you have any questions. .

Very truly yours.

C N. H. Willians .

Project Manager i

NHW/ajb ,

Enclosure cc: Mr. V. Noonan Mr. S. Burwell Mr. S. Treby Mr. D. Wade Mrs. J. Ellis i

M l

I i  !

t 6

i i

I i

l l

4 sl '

Mr. J. B. George 84056.053 January 31, 1995 Page 1 of 2 3

ATTADGENT (Revision 1)

Calculation of Allowable Punching Shear / Joint Capacity in Tubesteel with Holes For support MS-1-002-005-S72R, Cygna performed a finite element analysis to determine the effects of tube warping, and check the stresses in the tubesteel and coverplate in the region of the U-bolt hole. On October 4,1984, Cygna requested that TUGC0 review this support and provide calculations justifying the design. TUGC0 submitted calculations on October 18, 1984, which showed that stresses in the tubesteel were acceptable using the " punching shear" method of AWS D1.1-79 Section 10.5.1. Cygna further requested justification for use of " punching shear" as an appropriate. check. TUGC0 provided their justification in their letter dated November 8,1984. (L. Popplewell. TUGC0 to N. Williams, Cygna).

Cygna has reviewed the TUGC0 justification and has the following connents:

. A._ The AWS equation for calculating the punching shear al-lowable for tubesteel connections is based upon the re-sults of a limit analysis assuming a specific yield-line pattern within the chord of the tubesteel. When a hole is placed in the tubesteel and the edge of the hole is .

loaded, limit analysis would predict a different ~ yield-line pattern. This new yield-line pattern will result in

- a lower allowable punching shear. The presence of the coverplate further complicates the problem of determining '

punching shear allowables since one cannot autostically _.

expect an increase in the AWS punching shear allowable proportional to the increase in thickness provided by the ,

addition of a coverplate. In addition, the close proxi- 1 mity of the load to the edge of the tubesteel also influ-- i

. ences< the calculation of an alloieble punching shear.

-B. In the actual problem nodeled and reviewed by Cygna, our  !

, q finite element analysis predicted very little nargin to I allowable in the coverplate using an average of the fi-

. -nite element results along a line between the hole and the edge of the coverplate. The TUGC0 calculation re-

- ceived on October _18,1984,- clearly shows a mergin of approximately 6:1 (12.76/2.21). Thus, the TUGC0 calcu-lations.would predict that this joint is acceptable for approximately six times more load, a fact not borne out by the finite element analysis. While Cygna did not con-sider plate plasticity effects-in the finite element analysis Cygna is, nevertheless, concerned with the i

o

' '4

=-

• • • • • - 1 Mr. J. B. George 84056.053 January 31, 1985 Page 2 of 2 i ATTACWENT (Revision 1) large difference in predicted capability, and attributes much of it to the use of AWS D1.1-79 without assessing

• the impact of the deviations from D1.1-79. That is, one must cons.ider.that:

1. D1.1-79 assumes the brace and chord are welded '

together. Thus, the brace locally stiffens -

the chord. This is not the case for the nut loading the tubesteel. .

2. D1.1-79 assumes the chord is solid. This is not the case for tubesteel with a hole in it.

3. . D1.1-79 states that yield-line_ analysis can be used if 8 < 0.8, which is true for this joint i

(8 e .6). Thus, AWS does recognize that yield line theory can also be used to predict joint.

. strength-in configurations. pictured in AWS.

Bhsed on the above, Cygna does not accept the use of AWS D1.1-79 as an appropri-ate method for establishing an allowable punching shear / joint capacity in the '

case of tubesteel with loaded holes (with or without coverplates). Cygna re-quests that TUGCO. provide further justification on the design of such unique '

j oints.-

W i

h I

i i

4 8

s t

I l

l

s 101 Cahfortua Street, suite 1000, San Francisco, CA 941115894 4151397 5600 January 31, 1985 B4042.025 Mr. J. B. George .

Project General Manager Texas utilities Generating Company Conanche Peak Steam Electric Station i Highway FM 201 Glen Rose, Texas 76043

Subject:

Phase 3 - Walsh Doyle Allegations Richmond Insert Allowables and Bending Stresses Texas Utilities Generating Company Conanche Peak Steam Electric Station  ;

Independent Assessment Program ,

Job No. 84042

Reference:

M. H. Williams (Cygna) letter to V. Noonan (USNRC), "Open Items Associated with Walsh/Doyle Allegations," 84042.022, dated January 18.1985.

Dear Mr. George:

Cygna- has additional concerns with TUGCO's method of developing Richmond insect allowables and evaluating bending stresses when used in combination with tube-steel. These concerns are listed in the attachment to this letter.

This-letter completes Cygna's consnitment listed under item 12 of the Open Iteins List attached to the above referenced letter. If you have any questions or wish to discuss the subject, please call. -

Very truly yours. -

t a

• N. H. Williams '

Project Manager i.

NHW/ajb

. Attachment i xc: Mrs. J. Ellis Mr. S. Treby

.i , Mr. S. Burwell

• i , Mr. V. Noonan g Mr. D. Wede t-SanFrancuso esseo's Ducago phetunne

l

, s - ,

Mr. J. B. George 84042.025 i January 31, 1985 Page 1 of 4 ATTAC00Elli Richmond Insert

Reference:

(1) Affidavit of John C. Finneran, Jr., Robert C. Iotti. and R.

Peter Deubler Regarding Design of Richmond Inserts and Their Application.to Support Design.

(2) Calculations attached to the TUGC0 letter to Cygna dated May 8, 1984.

Cygna has performed a review of the above referenied TUGC0 Affidavit and May 8, 1984 calculations. The Affidavit included test results and calculations provided as justification for the Richmond insert allowables and a detailed analysis of the tubesteel connection designs. As a result of the testinD and calculations performed by TUGCO, Cygna has. drawn the following conclusions:

1. . TUGC0 performed testing to justify the single insert.allowables published in specification 2323-55-30. .While Cygna is in tesic agreement. with the testing procedures.and results. Cygna has.two questions related to the tests:

• Cygna has seen a minimum concrete compressive strength of 4000 psi throughout our reviews of calculations for pipe supports and cable tray supports. Cygna requests documentation which shows that the actual concrete compressive strength in all buildings with Richmond

. inserts exceeds the 4900 psi concrete used in the tests (Affidavit, page 14).

e Cygna has not found sufficient justification for the safety factor -ef 1.8 for emergency / faulted conditions. Cygna is, however, pursuing this question as part of our Phase 4 cable tray reviews and notes it

. here only for inforsultion.

2. 'TUGC0 performed a series of tests to justify their method of calculating the axial force in a bolt / insert due to torsion of tubesteel. TUGCO's evaluation of the test results concludes that a safety factor on.the order i of 3.2 to 4.0.is justifiable. . Cygna, however., has the following questions

.and comments:

-*- 'The tests of tubesteel/ bolt combinations do not includrthe larger sizes of tubesteel, for example . support MS-1-002-006-C72K (8x8x 1/2") j or CT-1-039-413-C42A (10x6x 1/2"). TUGC0 should justify that similar '

safety factors are expected for the larger sizes. p

• Only one test was run on each tubesteel/ bolt combination. TUGC0 should-justify that the results are repeatable, i

I

l

=-

sa w ns _ _

- Mr. J. B. George 84042.025  ;

January 31, 1985 Page 2 of 4

• ATTADOENT  :

(continued) e TUGC0 should clarify the location above the concrete surface of the applied load in the shear and torsion tests. The Affidavit (page 29) -

indicates it is two inches above the top of the tubesteel for the tor-sion test. This appears to be physically impossible since the dis-tance from the top of the tubesteel to item four (Attachment F-4) is 2.75 inches. p

• The moment listed in Table ~ 1 (Part 8) is referred to in the Affidavit i as torsion. Please clarify the axis about which this is taken.

e What concrete compressive strength tes used in the Richmond insert /

tubesteel tests 7 ,

I l

• Referring to the torsion test on the 4x4x 3/8 tubestsel with 3/4" off-  !

set, the Affidavit states (page 29): "When the shear equals 3 kips, the corresponding torsion is 21 in-kips." This implies a moment arm of 7" about the tube longitudinal axis. If the applied load is only 2" above the top of the tubesteel, the moment arm to the tube longitu-dinal axis is only 4", and the torsion is 12 in-kips. While Cygna realizes.that there is a logical alternative, i.e., that the torsion refers to a moment at the concrete surface, there must be correlation '

betweer. the value of torsion in the test and that reported in Table 1 (Part B) of the Affidavit (assuming the moment referred to is tor- t sion). Cygna has checked the calculations in their scope and finds that, in each calculation, the moment used as torsion for the Richmond insert is the moment at the concrete, which tends to support the use of 7". However, Cygna has not been able to match all the data in Table 1 (Part 8) against the calculations. For example:

~

A. MS-1-002-013-C72K (Drawing Revision 9. Calculation Revision 7 d dated 2/20/84), sheet 33: -

Tension = 15.4 kips (Table 1 = 23.975 kips)

Shear = 8.94 kips (Table 1 = 4.301 kips) ,  !

Torsion = 44.7 in-kips (Table 1.= 3.295 in-kips)  ; )

5. MS-1-002-006-C72K (Drawing Revision 7, Calculation Revision 6 i dated 12/20/83), Sheet.17: l Tension = 4.069 Kips (Table 1 = 4.069 kips)  !

Shear = .455 Kips (Table 1 = .455 kips) 3 Torsion = 5.676 in-kips (Table 1.= 2.048 in-kips) i Please clarify where the data in Table 1 is taken from and how the i 2

moment reported in Table .1 correlates to the torsional noment obtained  ;

from the tests. 1~  ;

I i

a=s**

Mr. J. B. George 84042.025 January 31, 1985 Page 3 of 4

' ATTADOUT (continued) e TUGC0 performed the tests in the absence of an applied tension load.

TUGC0.should justify that the tests do envelope the possible combina-tions of tension, sheer and torsion exhibited by the supports.

• In performing the torsion. tests TUGC0 welded a plate to each open end of the short piece of tubesteel (item 3, Attachment F-4). These  ;

plates stiffen the tubesteel against the applied torsion and may not give results representative of field conditions. Please justify that the configuration tested does represent the actual effect of the tube-steel on the bolt, especially in regards to prying.

3. For tubes subjected to both force perpendicular to the axis of the tube (i.e., pullout on the bolts) and torsion, Cygna reviewed the TUGC0 finite element analyses results concerning release of the bending moment restraint at the bolt as an appropriate modeling technique. The analyses results show that the tube deflection due to bolt elongation and tube deforation exceeds the calculated tube end displacement due to rotationlit the STRUDL pin supports, which.would justify TUGCO's assumption. Cygna has, hewever, two items requiring clarification:

• There appears to be text missing in the Affidavit between pages 33 and 4

34. Please provide the missing text. '

• On the second page of Table 2 of the Affidavit, the bolt interaction for two supports (FW-1-095-700-C62K and FW-1-098-700-C62K) increased when moment MZ was released. Not only does this seem contrary to what one might expect, it is also contrary to the data shown for the Rich-mond insert interaction. Please explain this discrepancy between -

insert and bolt and, if the bolt interaction is correct, why the in-crease is reasonable.

4. In sizing the bolts for use with-the inserts. TUGC0 has neglected the'pos-l sible bending stress in the bolt (i.e., bending due to treating the bolt as l a bass). In standard bolt / baseplate connections, this bending is not con-

.sidered since the bolt effe:tively acts as a sheer member (in a bearing connection) or the shear lead is taken by frictional forces-between the

. . concrete and plate (in a friction connection). In.the TUGC0 design, how-ever, the sheat load on the bolt my be offset from the concrete by the height of the tubesteel and.the thickness of the washer plates. if any.

~ Thus, a bending moment my be built up in the bolt which standard bolt e design formulae do not consider. In addressing the concern of axial tor-

. _ sion (2 above) TUGCO's tubesteel/ bolt- test investigated thjs phenomena as

, pointed out in their affidavit. These tests showed a significant factor of safety when compared to TUGCO's method of considering bolt bending stress

- in the bolt design as described in their Affidavit. Cygna, however, has

.two. concerns on TUSCO's evaluation of bolt bending stresses:

i

- ---- __.-_. , , - - - . . . - w w =- <w --

=

4 1

o =, ,

Mr..J. B. George 84042.025  ;

January 31, 1985 Page 4 of 4 .

ATTADGENT (continued) e In arriving at the 1.75 factor noted on page 25 of the Affidavit. .

TUGC0 relies on a factor of 1.33 which is derived from a comparison of .

.the finite element to hand calculated results in Appendix E-3. How- i ever, the finite element results are centroidal, not surface results. ,

Thus, the longitudinal location of the bending moment in the bolt model is 1/8" from the end of the model. The radial location of the

" average of element centroids" is not at the outer fiber, but at an l' interior location, since STARDYlE only reports stresses at the cen-

'troids of its solid elements. This underestinates the' outer fiber stress by the ratio of outer fiber radius to centroidal radius. These errors my not have a significant effect on the TUGC0 conclusions ,

since the 1.75 is used as a cutoff ratio only and TUGC0 does examine connections with higher interactions. TUGC0 should comment on the effect of these errors.

e TUGC0 does examine the bolt " bending stress" effect for. one support (CT-1-053-408-C62R) with a high (4.12) interaction and shows a factor ,

.of safety greater than four when compared with the tubesteel/ bolt test results. However, TUGC0 does not address the supports with higher interactions, especially CT-1-054-430-C42R which has the highest in-teraction ratio shown in Table 1 or support CT-1-053-418-C62R which j has 46 in-kips of moment. TUGC0 should justify the factor of safety '

for connections with interaction ratios or loads higher than those addressed in the Affidavit.

6 1

4 i

I

\;

L f

F l

4 e

8"888 , ,_ 415 397-5600 101 Caldo nia Street. Suite 1000. San Francisco. CA 941115894 t

- February 8, 1985 84042.021 Mr. J. B. George ,~

Project General Manager Texas utilities Generating Company Comanche Peak Steam Electric Station

-Highway FM 201 Glen Rose, Texas 76043  :

l

Subject:

Phase 3 Open items ,

Mass Participation and Nass Point Spacing Comanche Peak Steam Electric Station i Independent Assessment Program l Job No. 84042 *

Reference:

(1) L.M. Popplewell (TUGCD) letter to N. Williams (Cygna) '

"Cygna Potential Finding Report Mass Participation and the Mass Point Spacing Error in Problem AB-1-61A."~ Dec. 7,1984 (2) R. E. Ballard (G8H) letter-to J. B. George '(TUGCD),. "Itass Participation," GTN-69454, September 14, 1984 (3) N. H. Williams (Cygna) letter to J. B. George (TUGO), " Phase 3 Open Items - Mass Particpation," 84042.017, ,

. . September 21,.1984.

(4) N. H. Williams (Cygna) letter to J. B. George (TUGCO), " Phase 3 Open Items - Mass Participation," 84042.019, October 2,1984.

l

Dear Mr. George:

)

(

Section 5.4 of the Independent Assessment Prograin Final Report states-that )

l Cygna's conclusions are contingent upon the satisfactory resolution of specified l

. open items. A sumary of TUGCO's completed actions with respect to two items, l mass participation and mass point spacing, were provided to Cygna in reference (1) above. Cygna has reviewed the completed program and 1s-submitting our iconnents for your review in this letter. , l

. Before detailing any of the Cygna review results,- a' historical sumary of nejor )

milestones and correspondence is useful in orderOriginally,to fully understand Cygna's after a preliminary perspective of the asss participation program. f .

review of 5 stress problems, Gibbs.A Hill _ submitted a plan to TUGC0 in re erence '

I (2). ,Cygna reviewed the plan and. approved it in reference (3), but noted that I

SanFrancisco Boston Ctucago Achland

F saeu'ss Mr. J. B. George ,

February 4,1985 Page 2 i

more detail was needed before Cygna could comment on the adequacy of the screening criteria. Some of this detail was obtained during Cygna's visit to Gibbs & Hill on September 28, 1984. During that time, Cygna reviewed a portion i of the work that had been done to date and the criteria that was being developed to evaluate the scope of the re-evaluation. Noting some apparent deviations to .

the Reference (2) plan, Cygna submitted a letter to TUGC0 (Reference (4)) which  !

contained a. set of suggestions and concerns surrounding the implementation of r the plan. These concerns would require consideration before Cygna would be able I to close.the mass participation issue.

The results of the study which TUGC0 conducted to determine whether or not the inclusion of the missing mass had any significant effects on the pipe support However, i loads were not formally documented as comitted to in reference (2). ,

TUGC0 apparently concluded that further piping analysis was required since 205

- of the 271 pipe stress problems associated with the CPSES Unit I design were eventually reanalyzed.

Cygna conducted their review of the completed program during the week of There November 26,1984 'for both the pipe stress and pipe support- 1) disciplines.

the adequacy of

-were two points Cygna wanted to assess during that review:

the criteria used by TUGC0 to determine when to stop reanalyzing stress problems, and 2) the thoroughness by which the pipe stress / pipe support reviews

- were conducted and documented. This letter contains the results of that last ',

review and our recomendations for future action. It appears that TUGC0 my not have understood Cygna's reference (4) coments and, as a result, a less rigorous criteria was developed than the one Cygna anticipated. .

All of the reanalysis was done using Version-D of ADI. PIPE which permits inclusion of the missing ness or 2PA effect. In addition, Gibbs & Hill reviewed .

all 271 problems for the possible existence of mass point spacing / selectionTable '

errors similar to those described in Cygna- Phase 3 Observation PI-09-01.  ;

1 provides a listing of all the piping problems and notes.whichTheproblems revised pipe contained nass point errors and which problems were rerun. {

support loads resulting from this reanalysis were then . transmitted to the pipe support design groups for review and disposition.

f In order to evaluate the pipe stress portion of the nass participation program,

.Cygna first reviewed the Gibbs* & Hill work instruction for the piping reanalysis and found that it sufficiently addressed all aspects of the original Cygna

--finding. --Then a sample of 32 stress problems were chosen ih accordance with statistical sampling techniques described in MII. STD 105D for detailed reviews. This spot check was performed to ensure that Gibbs & Hill had properly  !

implemented their_ procedure. JTable 2 contains a list of.the stress problems Cygna selected.

L

=--

Sp,VICIS _

Mr. J. B. George '

February 4,1985 (

Page 3 f

Cygna found' that Gibbs & Hillparticipation.

ith lowest had considered the missing ass effect by re-The rean l

. running the prob ems w  ;

problems which had exhibited 30% or less ass participation Gibbsplus allthen

& Hill problems in '

which the systems were subjected to steam / water hammer loads. In reviewed the results to ensure that piping stresses were within allowables.

addition, Gibbs & Hill reviewed all welded attachments on the liain Steam /Feedwater lines, based on the rationale that these are the most highly ,

loaded systems. They also reviewed select penetration loads and valve  ;

accelerations for compliance with project criteria.

Following the reanalyses, Gibbs & Hill transmitted computer output sheets to the .

site which listed the original support loads, the new (i.e., corrected) loads l and the percent increase. Tne two site lead engineers reviewed the output sheets and automatically approved any supports showing either load increases less than 5% (PSE /ITT Grinnell. field group practice) or insignificant increases This applied (NPSI field group practice) without reviewing the. support package.

to all supports except the main steam /feedwater supports within NPSI scope, which were reviewed for any load increases. TUGC0 engineers reviewedWhile the TUGC0 remining supports against the " mass participation" load increases.

directed their engineers to use their standard design procedures in conducting their review, a work instruction or review procedure was not issued.

In order to review the pipe support portion of the ass participation program, Cygna selected 270 pipe supports from the approximately 1,100 pipe supports These

' associated with the 32 pipe stress problems in the Cygna review sample.

supports represented 19 of the pipe stress problems. Cygna chose the supports to cover both PSE and NPSI site design scopes.

In reference 1 TUGC0 has presented their position that the 205 reanalyzed problems represent a statistically significant sample which was biased towards Since l the problems most likely to be impacted by the effects of missing ass. j the reanalyses have shown that no piping problems were overstressed and no I supports required modification, TUGC0 has concluded that the remaining 66 stress problems need not be analyzed to close out the ass participation / mss point '

spacing issues.

The results of Cygna's review of reference 1 and the actual pipe stress and pipe In some instances the review was l'

support evaluations are presented below. inconclusive since insufficient evi

• conclusions. i

7

=- .  ;

Mr. J. B. George February 4,1985 i Page 4 -

Pipe Stress Results 1._

Of the 66 stress problems not terun to assess the effects of ass participation,13 contained mass point errors of the type identified in Cygna Observation PI-09-01, i.e., failure to include a mss point between supports. Since a reanalysis of problem AB 61A during the Phase 3 reviews demonstrated that the addition of a mass point between supports may significantly affect the local support loads (300% increase in one support load), Cygna does not find sufficient justification for not reanalyzing these problems.

Although these 13 stress problems have higher participation factors. it is not obvious how this compensates for the omission of a ss points. While reference 1 presents arguements to support the mass participation sagle,- it does not attempt to justify the inass I point spacing sample.

2. In the sample'of 32 problems Cygna found three with ass point errors which were not included in the= reanalyses models. These were:

a. Stress problems 1-19A and 1-66B were missing a mss point between an anchor and an adjacent support.

~

b. Although a Gibbs & Hill reviewer noted that a ass point should be included between two supports on problem 1-29V, this was not done in the reanalysis.

Using a normal inspection acceptable quality level of 1.5, two errors constitutes a rejection. Although Cygna concurs with the methods Gibbs & Hill egloyed to address the mss point spacing issue, the methodology was not thoroughly executed.

3. Of the 66 problems not terun to assess the effects of mass participation,14 contained a nass point error with the input of concentrated weights in ADLPIPE Version C. Gibbs & Hill categorized concentrated weights as a mass point problem because they had independently discovered that ADLPIPE Versien C did not lump.these properly in the dynamic analyses. This program error was corrected in version D and thus was automtically accounted for I inall Teenalyses. Since neither Gibbs & Hill nor'TUGC0 have j demonstrated in reference-1 or elsewhere whether this effect is -

significant, Cygna does not find sufficient justification for not

- remnalyzing-these problems. 4

0 I

=. -

Mr. J. B. George February 4, 1985 '

Page 5 l

4. Cygna found no evidence that the review considered the effects these reanalyses would have on the following areas of concern: ,

i i

e Valve acceleration generic study (inentioned in reference 1 as

" sampled")

• Flange loads generic study e Welded attachments in break exclusion zones e Areas with stress intensification factor errors as noted in Phase 3 Observation PI-00-01

• Functional capability for stainless steel elbows

5. The study of welded attachments only considered attachments to ,

lines which are subjected to steam / water hansner loads. In these configurations, the seismic load is usually a small percentage of the total design load. Thus any increase in seismic loads is unlikely to impact design. It is more likely that other welded attachments my be affected significantly since they do not include steam / water hammer loads, and, thus, seismic loads represent a .

large percentage of the design ined for.those. supports.

Pfpe Ssp y t Review

1. In reviewing support RH-1-005-016-C42K for the new loads, Cygna '

noted that certain welds in the calculations dated 3/2/84 had a sell argin. This support is a large gang hanger with three large bore supports (RH-1-005-016-C42K, -018, and -019), six small bore supports, and three conduit supports. The large bore support loads ,

increased as a result of the mass participation study, as did two small bore support loads. The review for the new loads provided no calculations to show that the welds are acceptable. Cygna was unable to establish the acceptability of the welds through .

inspection. .

?

2. 1teference 1 indicates the support review is complete. Yet, Cygna's  !

understanding from site personnel (as of November 30,1984) is that j auch of the study work had neither been checked nor Jipproved and, '

therefore, was not complete. In fact, for support AF-1-043-001-Y43R (Problem 1-156), the PSE reviewer noted that another design document, SA-4210, was still required to complete the review of .

this support.

3. Cygna-did not find sufficient justification that relatively sell load increases could be written off without inspecting the

' calculation. While any of the supports had large argins due. to .

i

sT Mr. J. 8. George i February 4,1985 Page 6 originally conservative design loads, there were still scoe supports within Cygna's scope which exhibited very little margin such as described in item 1 above.

4. None of the ' calculations were updated to show the latest loads ,

since TUGC0 apparently considers these loads for study purposes only, not as loads of record. CygnaIf does not understand the preliminary assessment the had {

justification for this position. demonstrated that the mass participat observations were of no concern, then this is a logical position.

Such a conclusion cannot be drawn since the mass participation / mass .

point spacing discrepancies do significantly affect pipe stress and  ;

pipe support margins. The analyses should be formally updated to reflect the corrected analyses.

5. Cygna did not find evidence that the effects of the errors noted by

' .Cygna in Phase 3. (e.g., three-sided welds and incorrect composite section calculations) were accounted for during the pipe support load increase reviews.

6. Cygna did find that some supports had large margins, as described in reference 1, and that the original weld designs conservatively ,

.used 18 ksi versus an actual strength of 21 ksi.

Sumary Based upon Cygna's review of the mass participation study by Gibbs & Hill and lVGCO, there are a number of actions which Cygna recommends in order to properl close this issue. Some of these items arise from the fact that 1UGC0 is treating the results of the reanalyses as a study. Since the new loads are based upon more accurate analyses, Cygna feels that any increases in design loads due to the reanalyses should be incorporated as the design loads of record with complete documentation as required by the 1UGC0 quality assurance progrm. Only by doing this, will 1UGC0 ensure that any future evaluations This would of '

'the piping or supports will be based on the latest information.

f

avoid any possiblity of future judgements being formed using unconservative

. data.

In addition,.since the use of 305 mass participation as a cutoff criteria does i

not meet the CPSES' FSAR requirements, it is still necessary that evidence be .

furnished that would allow Cygna to conclude that the reanalysis of the remaining 245 of the piping problems to. include the effects of missin5 mass would not result in any design deficiencies or significant reductions in margins,. The fact that all problems with 305 or less mass participation in a i

~ - * "

8**

~

Mr. J. B. George I February 4,1985 I Page 7 i

given direction were not reanalyzed does not eliminate any possible systentic error in the problem selection process. The selection process used by Gibbs &

Hill does not provide any assurance that those problems which were not reanalyzed have a high percentage of supports with low argins. TUGC0 had originally proposed to present a correlation between design ergin and ass participation (reference 2). Reference 1 only presents a correlation of percent load increase or decrease to ass participation. It does not discuss a ss point ,

spacing at all, nor does it present any comparison for pipe stress results. ,

Thus, it is difficult for Cygna to be reasonably certain that 761r, of the problems reanalyzed envelop the potential support or stress wargin reduction that would occur if the remaining 24% of the problems were reanalyzed.

In order to close this issue, Cygna reconnends that TUGC0 initiate the following actions:

1. Reanalyze the 13 stress problems which had missing ass points but were not originally rerun as part of the study.

2. . Perform a review of all. stress reanalyses to determine if. they are missing ness points between an anchor and an adjacent support.

This deficiency should be evaluated for its effect on the stress results. ,

3. Review all reanalyses to assure that all corrections specified by the Gibbs & Hill analysts have been incorporated.

-4. Reanalyze the 14 stress problems which had concentrated weights I input but were not originally rerun as part of this study.

5. Evaluate the effects of reanalyses on the following:

e Valve acceleration generic study -

e Flange loads generic study

~e Welded attachments in break exclusion zones i l <* Welded attachments in general i

~e Areas with stress intensification factor errors as noted in Phase 3 Observation pI-00-01 i e Functional capability for stainless steel elbows i I

6. Provide evidence .that 7911sf the problems analyzed envelop the 8 potential stress / support u rgin reduction that would occur if the .

- other 245 of the problems not analyzed were analyzed.

I 1

4

ss Z Mr. J. B. George February.4. 1985 Page 8 t

7. Revise the Gibbs & Hill as-built QA binders to reflect the latest analyses.

8. Revise the pipe support design calculations to reflect the loads fram the reanalyses. <

9. Incorporate the generic coments made by Cygna in the Phase 3 ,

a report (e.g., three-sided weld, composite section, welds in '.

combination with bolts) into the support adequacy review

. calculations.

I Overall, Cygna believes that the stress analysis portion of this study was reasonably well documented and executed. The assiissions noted appear to be more attributable to scheduling constraints rather than to lack of procedures or understanding of the engineering principles. On the other hand, the pipe support reviews were conducted without any procedures. The only evidence of ,

completed support reviews for the load increases was a signature. - In some cases the. supports were tco complex to use engineeringJudgment as a Aasis for acceptability in lieu of calculations, and therefore, r.alculations. should have been performed. Also, we would like to ' emphasize that this analysis was

necessary to bring the piping analysis into compliance with the FSAR. As such, it is Cgyna's opinion that the reanalysis should be the analysis of record. ,

Please call if you want to discuss these conclusions in greater detail.

Very tru1y yours, ,

t . '^ $Q N. H. W1111 mas Project Manager i

.cc::Mr.:D. Wade Ms. J. van Amerongen (telecopied) f Mr. R. Ballard ,

Mr. D. Pigott .

Mrs. J. Ellis i Mr. S. Treby -

Mr.S.Burwell(telecopied) 3 I I I'

i

I swoor seevas TABLE 1 1 m 29N

• 44A M 62Y *M 77 N 96C *M 16K 2-67I M

• M NO

• 165F

• 2 68T 2 *M 290

• 465

• 622 8 84 78 3 ett 29P

• 475 591 63A N 79A *M 97A

• 165G 2 68I M 4 em 295 51A

• 63C/s *M 72 4 975

• 165H

• 2-970 M j 5 *M ffT 51C 8 84 63D *M 79C

• 97C

• 164A

• 2 994 M SID *M 64A *M 790

• 970

• 166B

• 2-99B *M 6 4 290 * '

7 em 29V 4 52U M 648 *M 79E 8 00 135A

• 166C

• 2 150F

• S 9 29W

• 52V 8 94 64C

• 79F m 1358 1660

• 2-150G

• 9A 291

• 52W 8 94 640 *M 80A *M 135C

• 167A *M 2-150H

• W *- 29Y

• 52X

• 64E

• SOB *M 1350 *M 1678 894 2-1501 9 9C 292

• 52Y M 64F

• 80C

• 135E 167C

• 2-150J *

• M 2-151

• 90 30

• 52Z

• 66 *M 800 m 135F 1670 10A 884 31 M 55A *M 64A 4 81

• 150F

• 167E *M j 10B *st 32 *M 568

• 668 8 98 86A m 1506

• 167F *M J 10C *M 33 4 55C 8 94 65C M SEB

• 150H

• 168 *st 10D

• 34A *M 560 *M 67T *et - 86C

• 1501

• 169 *M 11 A *M 348 *et 57 M 67U M 47A N 1500

• 170 *M ]

lis 4 14C 4 58 47V *M 875

• 151A 171 *M 11C *M 35A 9 59A M 67I 87C

• 1518 172 m  ;

151C M

• 12A m 35 55 *M 67Y M SBC

• 174 151D *

• 125 M 35C

• 59C *st 67Z *It 880

• 175 120 *tt 350 590

• Esf 9 WE

• 152 804 178A 12E *It 3K

• 60 480 *M 88W

• 153

• 17 5 *M "M * *M 154

• 179

• 19A m 35F

• 61A ESV 88X 15

• 36 *eq sit 9 48X *et 88Y m 155

• 180 M 19C

• 378

• 61C 4 68Y 8 84 IBZ 156 *M 186

• 188

• l 21 M 37W 9 610 m 68Z 89 est 157A 23A M 37I

• 61E M 69

• 90

• 1575 m 189

• 23 M 37Y

• 61F 4 70 est 91 8 84 157C M 2-51A *It 23C M 372 824

• 71A MA 8 84 158A

• 2-510

• TE ' N 40 62R M fit est E2B 91 15 3 9 2- 528 24

• 42A *st $2C est 72

• 93A N 15E 2-52U

• 17 *el 42R *st 42D

• 73

• SM est 163 2-613 *st fB een 460 est aff M 74

• 94 M 1ALA

• 2 61E 19r

• 4M *M 82F 75

• N 4 16 5

• 2-620 29L

• 455

• B6 M 76A M 96A

• 165C

• 243 M 294

• 45T 821

• 7N N 95

• 1450 *~ 2-677 N 271 problems 9etal

• Problems teenalysed in 88H R.T.'(203)

' N

• Problems with sess point specig er seneenttweed ut. Wiscrepencies (141)

• M

• Assaalysed with andiftsattaas to fr and er sencestrated ut. (105) 4 l -

i l

?_______ - - __. - . _ . .-__ .- - - . _-.. .__

o mummmmmme umyu - - -

I

< TABLE 2 Problems Reviamed ty Cygna i

AB-1-D1 AB-1-672 AB-1-06 AB-1-68T l AB-1-10C AB-1-6BY AB-1-12A AB-1-71B l

. AB-1-19A AB-1-72 AB-1-2B AB-1-79A AB-1-29V AB-1-79F I

AB-1-34A AB-1-86A AB-1-36 AB-1-88X AB-1-428 AB-1-95

-AB-1-61A AB-1-135D AB-1-61B AB-1-156 AB-1-63C/B AB-1-167B AB-1-64D AB-1-171 AB-1-66B :AB-1-1788 AB-1-67V AB-2-520 i

}

i i

4 4

l l

i l

l i

l

esse ,__

101 Cantorrua Strest Sune 1000. San Francisco CA 941115894 415 397 5600 8 February '12,1985 -

r84056.041 I

Mr. J. B. George Project General Manger Texas Utilities Generating Company .

Comanche Peak Steam Electric Station Highway FM 201 Glen Rose. Texas .76043

Subject:

Cable Tray Support Review Questions Comanche Peak Steam Electric Station Independent Assessment Program - Phase 4 Job No. 84056

Dear Mr. George:

We have reviewed our correspondence and telecon files to ascertain the status of the cable tray support review questions asked to 1date. Attachment A contains a sunmary of currently unanswered cable tray support design review questions.

Additional questions may be asked as we complete our documentation of the TUGC0/Gibbs & Hill responses to previous questions and evaluate the responses to ,

the attached questions. Perhaps some of these questions and their relative '

effect, on design adequacy any be answered more efficiently as part of the dynamic analyses being conductd on selected systems. Pisase advise us of any questions being addressed as part of that effort.

If there are any questions while preparing responses, please call. _

1 Very truly yours, ,

l N. H.' Williams 9ro3ect- Mannger NHW/rmk tc:-*. 5. Burwell (USNRC) .

.*. S. Treby (USNRC)

Mr. D. Wade (TUGCO)

Ms. J. van Amerongen (EBASCD/TUGCO)

Mrs. J. Ellis (CASE)

Mr. R. Ballard (G&H)

Mr. R. Kissinger (TUGCD) t SanFrancisco Speeon Ctucogo McNand

4 Mr. J. B.' George Febraary 12. 1985 84056.041 . Pa ge 1 of 13 ATTACitENT A- i .1 s

CABLE TRAY SUPPORT REVIEW 1lLESTIONS  !

l l

l

1. EMBEDDED PLATES  !

2 ,  !

References:

1 (1) Gibbs & Hill Specification 2323-55-30 Revision 1, "Struc- -

-tural Embedments," Appendix 4 (2) TRW Nelson Division, Design Data 10, *Embedment Properties of. Headed Studs," 1977 h. ,

(3) Gibbs & Hill calculations SCS-113C, Set 1. sheets 18-19,  !

42-46 i!

-(4) Gibbs & Hill drawing 2323-5-0919 Revision 3 <

Several cable tray supports within Cygna's review scope are at11 ached to embedded strip plates.. These supports are Types SP-7 with brace, SP-7 ,

~and Detail F (drawing 2323-El-0601-01-5), which-is similar to a multiple SP-7 support. Each of these supports consists of channel . sections canti-

- -levered from the embedded plate. These supports resist vertical, trans-verse, and, in the case of Detail SP-7-with brace, longitudinal tray loads. . The connection between the cantilevered channels and the embedded -

i plate is an all-round fillet. weld'. Such a connection provides full '

moment transfer.

.l

-The design of the embedded plafes for'the support types -listed above was.

performed in 1979. In 1981, Gibbs & Hill specification 2323-SS-30 revi-sion 0, " Structural Embedments," was issued. ' Appendix 4 of this specif1:

, cation lists the criteria and allowables for attachments to embedded strip. plates. , Revision 1 of this specification (Reference D]) did not '

alter the criteria or allowables for embedded plates. Cygna has noted '

that the criteria listed in the referenced appendix are more stringent than those used in the original design.of the embedded plates for the i support types listed above. - .

1 Cygna ; requests a' confirmation of the. assumptions used in the generic .

embedded plate. analysis. Gibbs & Hill has stated that 'a factor of safety g .

. of two ees.used in the analysis. . Aigid plate assunptions.as'well as prying action effects were also considered. Cygna has calculated the i

allowable stud tensile Ioads based on a lues and reduction procedures reported in Reference .(2) above and on a factor of safety of two. .The values calculated by Cygna were identical to the allowable . loads for point loadings ' appl.ied along the plate centerline at stud points. This-indicates to Cygna that' prying action has- not been' considered in the

. development of embedded plate allowables.

n&

I l

'- Mr. J. B. George February 12, 1985 84056.041 Page 2 of 13 ATTADOENT A CABLE TRAT SUPPORT.EVIDf SESTIONS I

With regard to the design of-the embedded plate connections for the sup-port types listed above, Cygna has noted several inconsistencies between the designs and the criteria reported in Specification 2323-55-30, Revi-sions 0 and 1. The embedded plate connections that are used for the listed support details are moment resisting connections. Section 3.4 of i Reference (1) above states that pin. connections are to be used for load l 4

transfer to embedded plates. ~ This section also requires that when a mo- i ment is to be transferred, the embedded plate connection must be stiff- , ,

ened. Further restrictions on moment transfer are listed in Section 4. l Gibbs & Hill has stated that all moment connections must be evaluated per  !

these sections. Cygna has not seen any evidence that such evaluations l J

were performed for the cable tray supports listed above.-

Sheets A4-1 through A4-9 of Appendix 4. reference (1), list the allowable

. values for various locations of applied point los'ds. Section.3.1.11sts various reductions for locations other than those shown in the Appendix

' sheets 1-9. These reductions include interpolation.of allouable. values ,

for attachment locations between the midspan and stud pairs along the longitudinal plate centerline. Interpolation of plate allowables is also required for attachment points between the longitudinal centerline and the stud lines at the p. late edge. In addition, a 40 percent reduction is required for end-span loadings. It appears that none'of the above-reductions was considered in the desigm of the embedded plates for the

. . listed details. -

< Section 3.3 notes that- the full allowables shown on sheets A4.2 and A4.4-

.are applicable only when the attachment is within i 3/4 inches of the

~ ~

longitudinal plate centerline. The centerline tolerances allowed in Ref-erences 3 and 4 for beam connections of Details SP-7 and SP-7'with brace, are 1-1/2 inches and 2-1/2 inches-for eight'and ten inch plates, respectively. A'1-1/2 inch tolerance is allowed for the brace. connec-tions of Deta11'SP-7 with brace. These tolerances exceed those specified in..Section 3.3..even though the full allowables~ were used.

l

.Plasse provide ~ Cygna with the following:

1.. Documentation which evaluates the. effect of prying action on embed-'

ded plates and studs; l '4 I

-wr. +v

, , - - . . , .-w - ,.-- -., . . ,,  % .-- - .--,-* .,,, - __

)

Mr. J. B. George February 12. 1985 84056.041 Page 3 of 13 ATTAD9ENT A .

. CABLE TRAY SUPPORT REVIEW iltESTIONS i

k '!

2. Documentation and justification for the acceptability of the monent connections for Details SP-7 SP-7 with brace and Detail F; and

3. Documentation which verifies that the details listed in question.2 above meet the criteria set forth in Appendix 4 of Reference 1.

'2. AFFECT OF TWO-INCH THICK ARCHITECTURAL TOPPING ON HILTI' BOLT D1BEDt1ENT LENGTH

References:

l '

3

-(1) 'TUGC0 SDAR-CP-80-05 dated 8/8/80 '

(2) Brown & Root Instruction CEI-20 -Revision 9 (3) CMC 6114.. Revision 4 dated 10/12/83, Support-No. 2998 (4) CE 85720. Revision 0, dated 2/2/83, Support No.13080 Two cable tray supports within Cygna's review scope, support numbers 2998 and '13080, use floor mounted base. plates. These supports are located in areas where 2" thick architectural _ topping is present. A talkdown was

~

performed to verify that sufficient Hilti anchor bolt embedment was pro- '

vided. Based on the length code stamped on the exposed ends of the in-stalled Hilti Super Kwik bolts and the measured projection, (i.e., the

-distance from the flonr slab to the top of the bolts), the bolts in-stalled are not of sufficient length to provide the minimun embedment -

specified in References (3) and (4) for these supports.

i  !

Cygna calculated minimum embedment length before torquing for these' two ,

I

. supports as follows:

\

l I

1 \

1' l q

.e d

i

9 Mr. J. B. George February 12, 1985 Page 4 of 13

- 84056.041  !

ATTAcitENT. A

- . CABLE TRAY SiffdRT BEVIEif IREST1055 l

Embedment -- (bolt length) - (topping) - (bolt projection) +-(nut thickness)

Support gBolt. Bolt projection Minimum Embedment

' No.- Code Length Above Slab Calculated Required 2998 U 13" 4-3/8". ~7- 7/8" 8-1/2"

.' 13080 1 16" 4-1/4" 11" 13-1/8"

. Cygna es provided with a copy of Reference (1) above, when this concern 1

was first discussed with TUSCO. The corrective action indicated in Ref-erence (1) requires a case-by-case evaluation of all Hilti bolts in-stalled in areas with two-inch topping. Cygna interprets this as apply-ing .only to bolts-installed before the date of Reference (1). Since both

' floor mounted supports within Cygna's scope were installed after that date, they would not have been included in the evaluation.

Reference (1)'also indicates that Reference (2) above uns revised to reflect the effect of topping-.on embedment length. Section 3.1.7.1.2 of reference (2) states:

- " Expansion bolts which have less than the specified designed embedment length into structural concrete but greater than the values indicated above in 3.1.7.1.1 shall be evaluated by-the responsible desip engineer. If found to be acceptable "as-is," appropriate design change' documents shall be issued. If '

found to be unacceptable, the expansion bolt shall be reworked

[ in accordance with 3.1.7.1.1 a. or b." -

Assuming thet-this criteria was followed for the installation of support numbers 2998 and 13080.:the Quality Control inspection travelers should

, show that the "specified designed embedment length" was not . met, and.that -

l . an evaluation es performed by thei" responsible design engineer."

'please provide copies of the existing desip change documentation for~

, these supports indicating the acceptability of the reduced embedment I length. . If no documentation exists, plasse assess the impact-of this 4

. issue on ather floor mounted supports in the areas where.two-inch topping  ;

is used.

I i

I- + -== s--> g p g ---e

.- j l

~*

Mr. J..B. George February 12. 1985 ,

i 84056.041 Page.5 of.13 1

ATTADBENT A TABLE 1tAY 3UPp0RT ltEVIEW QUESTIONS

3. CONTROLLING LOAD CASE FOR ONE-BOLT BASE ANGLE ANALYSIS

References:

(1) Gibbs & Mill Drawings: 2323-5-0903 Detail 1 2323-5-0908

-2323-5-0909 (2)- Conference Report dated 11/14/84, 2:15 p.m.; Bhujang, -

Chang, Berry, Horstaan and Russ (3) TNE Calculation.by J.C.C. dated 11/15/84, " Investigation  :

of Single Clip Angle" .

(4) Gibbs & Hill Calculation' Binder SCS-215C, Set 4

~

.During the discussion regarding the analysis of Detail 1, one-bolt base angle connections, as noted in Reference (2) above, Gibbs & Hill ens to determine the controlling load case for this. connection, i.e., the max- ~

imum loads. The calculations provided. Reference (3) above' use the loads from a B-2. type support, but do not provide any comparison to show that this is the controlling case. Based on Cygna's review of existing calculations, it appears .that other support types may develop larger loads for this connection detail. . An example of a support type with

-larger loads is case E-4, per Reference (4) above. . sheets 16-20. -

~

i Please provide documentation indicating the controlling . load case for the  ;

one-bolt base angle and an analysis considering these loads. -

1 i

4. . WORKING POINT DEVIATION CALCULATIONS .

i

Cygna

has reviewed the Gibbs A Hill .uorking point analysis performed in l March and April 1984. This reviou identified problems.with boundary l

conditions (unjustified restraint.of frames in the . longitudinal-direction) and the effects of closely spaced modes. Gibbs 8.H111-revised

• u the working point analysis to 1 correct these discrepancies and:resutznitted *

' theur to Eygns .for review. Sased on discussions with site personnel, Cygne -understands.that any work associated with the closure of the ,

- working point stu(y has- been suspended until the NRC mandated as-built

- program is coupleted. Cygna. however, has concerns about.the analysis i

4 i

Mr. J.'8. George February 12, 1985 page 6 of 13 84056.041 ATTADGENT A

. CABLE DAY SWPORT EVIEW tlLESTIONS I

and its application which would not be addressed by an as-built program alone.

The working point study establishes cut-off elevations below which the supports may be considered acceptable for given working point

. deviations. This evaluation is _ based on assumed accelerations, 8'-6" ,

tray spans, enveloping aspect ratios, and maximum permissible working  ;

point deviations. Above these elevations, Gibbs & Hill evaluated the supports on a case-by-case basis using design ' documents only. No l

consideration uns given to trays where the unit weight exceeded 3610/sf due to-added fire protection or to the as-built support configuration as reflected by applicable CMC's and DCA's. Cygna has identified spans up i to 12 feet (reference Cygna letter 84056.01g. dated 8/10/84) in length 4 which indicates that there would be a problem in justifying the qualification of trays below a given elevation using an assumed 8'-6" span. . Further, for supports located above the cut-off elevation where the map drawing shows spans granter than 8 feet, an additional 6" t

. installation tolerance must also be considered.

The effect of possible variations in aspect ratio, fire protection weignt and actual working point deviations must be considered in the same. manner i as- the above concerns regarding span violations. Further, Quality.

Control's use of a working point deviation criteria as the sole means of

. ensuring compatibility with the Gibbs & Hill. analysis will not unto

itself ensure design adequacy.

please indicate the plan of action TUGC0 will use to resolve ~ this issue. '

. 5. 1)ETAILS F-H, DRAWIN6 2323-El-0601-01-5 A!O SP-7 EFFECTS OF SMALLER WELD

- SIZE AND UNDERRUN 4

Reference:

(1) Conference Report dated 11/17/84, 8:00 a.la.;' Chang,  !

Huang, Horstaan. Russ and Williams l

-From the referenced conference report:

1

l-.

. Mr. J. B. George February 12. 1985 l .84056.041 Page 7 of 13

ATTADOENT A

_ CARLE TRAV.5MPPORT SEVIEW SESTIONS "Cygna has reviewed the Gibbs & Hill calculations which l address the change in fillet weld size (3/16" vs.1/4") for l Details F-H. Cygna has noted that when the effects of l eccentric loads are considered, the welds-appear to be

overstressed.- The Gibbs & Hill calculations did not consider these effects. Gibbs a Hill inquired if the member was also overstressed. .Cygna responded that for a 4'-9" cantilever length details SP-7 and F-H there is a resulting 4% over- ^

stress. Cygna provided the following list of details which appear to exhibit overstress conditions in the weld and/or member.

For. 3/16" fillet weld without underrun:

i (a) Details F-H without brace with 30" tray and maximum moment arm.

(b) SP-7 attached to embedded plate- with 24" and 30" trays.-

l (c) SP-7 with brace attached-to embedded plate with 24" and 30" trays. (Gibbs & Hill noted that allowable tray spans for embedded plates is 7'-6".)

(d) Details F-H attached to embedded plate with 24" and 30 " trays.

For 3/16" fillet weld with 1/32" underrun for all tray sizes:

(a) Details F-H.

(b)- Details F-H attached to embedded plate. [

-(c) SP-7 attached to embedded plate. ,

i .(d) SP-7. with brace attached to 'aubedded plate.

Patential member.overstress examples include:- i

- (a ) Details F-H with 30" tray. i (b) Details F-H attached to embedded plate. I (c) . SP-7 attached to embedded plate. - .

(d) - SP-7 with brace attached to embedded' plate." 8 Il J

% , - ,,w, .- - ~n --a, ,. , ,w.- --,w,-e--+- -,---w, , - - , - , , - ,e - a-,a

..Mr. J. 8. George February 12. 1985  ;

84056.041 Page 8 of 13 ATTADGENT A CABLE 1 RAY SUPPORT E VIEW QUESTIONS ,

1 1 .Please indicate the plan of action TUGC0 will use to address the

- potential weld and member overstresses indicated above.

6. WELD DETAIL FOR CONNECTION OF CHANNELS TO CLIP ANGLES 4

References:

i -j (1) Gibbs & M111 Drawings: 2323-5-0901 8 4

2323-5-0902 '

2323-5-0903 (2) Brown & Root Drawing FSE-00159

! (3) Conference Report, dated 11/17/84, 8:00 a.m.; Chang, Huang, Horstaan. Russ and Williams (4) Conference Report, dated 12/19/84, 9:30 a.m.; Keiss, ~

Warner, Mercer, van Amerongen, Horsten and Russ 3 (5)- Gibbs & Mill Calculation Binder SCS-215C, Set 2 2

(6) Gibbs & Hill Calculation Binder SCS-122C, Set 3 (7) Gibbs & Hill Calculation Binder SCS-146C, Set 1 (8) CMC 82988, Revision:0, dated 11/15/82 Revision 1, dated i 3/7/83 arid Revision 2, dated 7/11/84 (9) -DCA 20228, Revision 0, dated 4/10/84 and Revision.1, dated

~

j 4/30/84 (10) RFIC EH-1842.-dated 11/5/82

. A reviou of the cable tray support desip drawings, Reference (1), showed i that no minimum length is specified for the weld attaching a support 4 channel to the angle section used as a base plate. Support types for-j . which the weld length is not specified include Details 1, 4..B. C, D, , ,

Case SP-7 plan, etc. The hanger assembly drawings Reference (2), typi-cally indicate the distance between the face of concrete and the end of .'ll the channel to be 1-1/4", but specify no tolerance for this dimension _.

The 1-1/4" dimension results in a 3-3/4" lap between the channel and an l 15 x 5'a 3/4.

~

i '

!l

, Referring to the discussion in Reference (4), TUGC0 indicated that the 1- '

,-1

- c l/4" dimension was used as a annimum distance for installations where the l

' channel es attached on the outside of the angle (Type II).- It uns also l f

j .

4

,,..-,y - ,-o. ..._,.p- . - ._y._ ,y. ,-_,,-,-. , -, , , , . - - - , - - . - - , , , .--m, ,, , , . .,,,,.y,,_- , , , , . -

l Mr. J. B.. George February 12. 1985 *

, .84056.041 Page 9 of 13 ATTADDENT A '

CABLE 1 RAY JUPPORT REVIEW SESTIONS  !

9 used as a minimum distance' for installations where the channel-was at-

.tached on the inside of the angle (Type 1). (Due to the internal radius

=on the angle, a member can not be attached any closer than "k" from the heel of the angle, e.g., k = 1-1/4" for.L6 x 6 x 3/4 and L5 x 5 x 3/4.)

i If the channel laps into the fillet at the corner, it does not lie flat ,

! against the leg of the angle and a gap will result. Mr. Warner assured Cygna that the weld fit-up inspection prevents this from occurring and .

that the weld inspection assures that the proper weld, including ~ end "

returns, is installed.

If 1-1/4" is used as the minimum distance for Type I connections, the resulting weld length could be less than 3-3/4". In Gibbs & Hill's eva-j lustion of weld size underrun, a weld length of 3-3/4" ans used per Ref-erence (6) for.the standard connection details'and per Reference (7) for SP-7 and SP-7;with brace. The use of a shorter- weld length could. result '

in an overstress in the welds used in this detail. per Reference (10),

! clarification en .this atter es voquested by site personne1'and as a result CMC 82988 [ Reference (8)] was issued to give specified tolerances on the connecting weld length. CMC 82988 ms issued in November 1982, by which time the majority of the Unit 1 cable tray supports had already i l been installed.

f Further review by Cyya noted that due to the radius at the toe of an l

angle section, at gap will exist between the web of a channel and the an-

gle section at the toe. (See Figure 1.) In order to achieve a 3-3/4" _

. weld length and the necessary return, the. fillet weld sust bridge this ,

gap. Cygna believes that without the use of a special welding procedure.

' l a fillet weld will not achieve its full effective throat at this

! location, and thus the weld section properties.will be less than considered in the. design calculations. '

i i

Plese: provide the following: '; i p -e '

~ The documentation .from engineering used by Brown 8 Root to establish ' l the:1-1/4" distance between the face of. concrete and the end of ;I channel shown. on FSE-15g drawing sheets. This tes previously L .. requested in the Reference (4) monference report.

l t  ;

__ _ _ _ _ _ . ~ _ . . _ _ _ _ _ . . _ _ _ . . . _ _ _ _ . , . _ . _ , . _ _ -.._ _ ._- _..-.__- _ _ .__ -

v.

l Mr. J. 8. George February 12, 1985 l l

.84056.041 Page 10 of 13  :

ATTADOENT A M M T 3 WPORT E VIEW ipK5TIONS i

I

-e 'The installation tolerance for the-connecting weld length that was used prior to the issuance of CMC 82988 and provide assurance that a

' minimum weld length of 3-3/4" tes always provided.

• The welding procedure which applies to performing a fillet weld ~over the gap between the channel and the toe of the angle section or i documentation to show the acceptability of the connection. 1

7. WELD DETAILS.AFFECTED BY GENERIC CHANGE DOCUMENTATION

References:

1 (1) CMC 58338. Revision 0 (2) Gibbs & Hill Calculation Binder SCS-122C. Set 3

('3) Conference Report dated 11/17/84, 8:00 a.m.; Chang, Huang. Russ..Horstaan.and Williams

- As 1 discussed in -the referenced conference report. Cygna indicated that in order-to evaluate the effect of the use of 1/4" fillet welds for cable tray support- fabrication, ' generic CMC's and DCA's which address changes in welding requirements must.be reviewed. Cygna provided a list of CMC's .

and DCA's affected by this concern, but noted that the list should be checked for completeness.

During additional review, Cygna located CMC 58338, which any be effected

, by this concern. This CIC allows an alternate weld pattern to be used for the connection between the beam and hanger members. An evaluation of

' the weld. pattern using a- horizontal run along with top and bottom flanges of the (C x 7.25 results.in an appromiente 205 reduction in-tne torsional modulus compared to the value used in .the weld evaluation per Reference

~

l . (2). The'CVC for- this CMC -indicates .that no salculations were required. '

l Please; provide the following:

e Justification for.the adequacy it this alternate weld pattern, f

i. taking inta_ account the potential of weld size underrue and the use i

of.1/4" weld ~ size for this connection detail.

t i

I' L_.________ __ _ _ _ _ _ _ _ , _ _ _ _ , _ _ _ _ _ _ _ . _ , _ _ _ _ . , . _ , . _ _ _ _ _ . _ . . _ _ _ _ _ . . _ . , _ . . _ _ _ _ _ . _ _ _ . .

I l Mr. J. B. George February 12, 1985

. 84056.041 Page 11 of 13 ATTAC K NT A CABLI TRAY SIPPORT REVIEW- QLESTIONS

• The status of Gibbs & Hill's investigation into the:effect of the

reduced weld size on the design review of the generic CNC's and DCA's for weld . details.

I J

8. INSTALLATION TOLERANCES FOR CABLE TRAY SUPPORTS

References:

(1) DCA 20228, Revision 1  :

I (2) DCA 9738. Revision 3 '

(3) Specification 2323-55-168

! DCA 9738 provides the following revision to specification 2323-SS-16B:

9.6' TOLERANCE

9.6.1 In general, all structural steel work shall be plumb and level within the tolerance 1

500.

9.6.2 Hangers for electrical raceway systems shall be installed within following tolerances. The cumulative effect of rolling (mill).

fabrication and erection tolerances shall not exceed those given below: _

,

• A hanger. shell be considered plue if the l

~

angle between the longitudinal axis of the main member (s) of the hanger is at right

. (90') angles to the supporting surface 1 (ceiling or floor). A tolerance of

• 2*, ,

ur.less otherwise noted, shall be acceptable -

provided the integrity of the supported raceway system and its. attachments to the  !

. hanger remains intact. -  !

.* Hangers supported on vertical surface ,

(wall column, side of a beam, etc.) shall -

be considered level with seximura tolerance sf 1/2" in-10 feet.

t

O e

i Mr. J. B. George February 12, 1985 l

84056.041 Page 12.of 13

• l ATTAOWENT A s  !

. CASLE 1 RAY SUPPORT EVIEW SESTIONS a

i I

l

• wUse of extreme boundaries of such '

tolerances may be at times aesthetically .

unacceptable and the owner or its representative may provide stricter tolerances on case by case basis..

Cygna is concerned by the potential effects of the

• 2* tolerance for the ,

pluseness of the hanger on the axial load carrying capacity of the mem- -

ber. For a. typical 12'-0" high support. 2* slope results in a 5" offset .

of.the lower end of.the hanger. This offset could result in additional '

bending. stresses in the vertical members when axial loading is consid- -

ered.

! Please provide the following:

e A copy of the design review. calculations for DCA 9738 considering

the effect of the 2' tolerance m support design.  :

e The procedure used by the respo'nsible engineer to determine if a I

finer tolerance is needed to assure that the " integrity of the supported raceway system and its attachment to the hanger remains inta ct."

9. DETAIL "5". DRAWING 2323-5-0905 SUPPORT NO. 3136

- \

References:

(1) CMC 8229. Revisions 0 - 13 (2) .Gibbs a Hill Calculation'51nder SAS-1341. Set 3

Cygna review of the " Design Review" Calculations for DC.8229.

Reference

. (2), indicated that there are several potential errors'in these calcula-tions. . Cygna provided a list of questions and comments, pertaining to t these calculations, to Mr. S. K. Shujang on October' 20, .1984. No re-sponse has been received.from Sibbs & Hill.

Please'. provide Cygan with the status of this review.

i I l

3 1

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

l

. l Mr. J. B. George February 12, 1985 i 84056.041 . Page 13 of 13 j

ATTADOENT A .

j CABLE TRAT.SFPORT EVIEW SESTIONS ,  ;

j

. ,10. DETAIL "H", DRAWING 2323-El-0601-01-5 SUPPORT NO. 734 i

References:

(1) Brown & Root Drawing FSE-00159, Sheet 734 (2) CMC 164. Revision 4 (3) Conference Report, dated 10/27/84, 9:15 a.m.; Keiss, van Amerongen, Chang, Huang, Russ, and Williams Responding to the discussion in the referenced conference report, Cygna noted that this. support has deviated from the generic design for detail H without brace by rotating one of the channels 90* from the standard orientation. Heavy duty clamps were installed per Reference (2). No consideration uns given to the additional of longitudinal loads to this support. The conclusion from Reference (3) was for TUGC0 to-provide Cygna: with calculations evaluating this support's as-built condition.

Please Provide the calculations generated by TUGC0 in response to the discussion.

I

~

1 4

l c .

1 l

4 4

l l

t

,- ~ - - . , , , -. , , ,- , . - - - - . - , . . - , ,

l e

Calculation .

4L i 4 . Sheet

"'""""' l

'IINAS Mc, CPSE6 3AP P'w*c 4 '*""'3 b W " I ?. - M C3C/

Checked By Date Subect Cema. tam seux tzaun l 1

System Joe No p g Fise No Analyses No Rev.No Sheet No F6 vat i h actweca C.nwma A, c 4 we gy %3cc_

t

_. x.q- - ,

.P=

Ns X_.-l

./ (

s gs

.s, s. l

__ 3 _. _ .

_'s -

.' ~

\  : t- cm -

4 taw .i i -

\ M M M M W W m 43 m A7

' 1 Sreme A- A u oc , _ :

I t

. / . . _

b L6454 ov i AY L4vh k Vg I

I

.g i I '  :

t tt.st

• .5

• 7 4

g

._ - . _ _ _ - - _ . = .- - -- .-

?

9

=-

assnas 101 Caldornia Street. Suite 1000. San Francisco. CA 94111-5894 415'397 5600 January 31e 1985 84056.053 Mr. J.-B. George Project General Manager Texas Utilities Generating Company  !

Conanche Highway Peak FM 201 Steam Electric Station Glen Rose, Texas 76043 ,

Subject:

r

' Phase 4 Open Items - Punching Shear  :

Texas Utilities Generating Company Conanche Peak Steam Electric Station Independent Assessment Program - Phase 4 Job No. 84056 i

Reference:

Open Jtems - Punching Shear.".84056.051e 29, 1985. An Daar Mr. George:

following three areas:The attachment to the above referenced rect the letter has b 1.

The attachment refers to pipe support drawing A

wascopy of this drawing was inadvertently omitted when the letterMS distributed.

statement: " see attached support drawing)."Cygna has revised At vided as an e(xample only The drawing was pro--

made. and is not integral to t i

2. In the last sent graphic error). ence of causent one, change'"all" to "an" (typo-3.

line analysis of the finite element results" and ins .

~

words ava1 shown, cation. which better describe Cygna 's approach to the 1

San Francasco Boston SanDoge C8 mengo Achland 4

ER E__s.s Mr. J. B. George 84056.053 j

. January 31, 1985. Page 2 This letter and attachment therefore supersede letter 84056.051 and its attach-ment. -Please call if you have any questions.

Very truly yours,

^%1Wh . L N. H. Williams Project Manager i

NHW/ajb , i Enclosure cc: Mr. V. Noonan Mr. S. Burwell Mr. S. Treby Mr. D. Wade Mrs. J. Ellis i

M f

I i

I i

i l

I l i

I i l

l sanwess_; ^

Mr. J. B. George 84056.053 January 31, 1995 Page 1 of 2 ATTADGENT (Revision 1)

Calculation of Allowable Punching Shear / Joint Capacity in Tubesteel with Holes for support MS-1-002-005-S72R, Cygna performed a finite element analysis to determine the effects of tube warping, and check the stresses in the tubesteel and coverplate in the region of the U-bolt hole. On October 4,1984, Cygna requested that TUGC0 review this support and provide calculations justifying the design. TUGC0 submitted calculations on October 18, 1984, which showed that stresses in the tubesteel were acceptable using the " punching shear" method of AWS D1.1-79 Section 10.5.1. Cygna further requested justification for use of " punching shear" as an appropriate. check. TUGC0 provided their justification in their letter dated November 8,1984. (L. Popplewell, TUGC0 to N. Williams, Cygra).

Cygna has reviewed the TUGC0 justification and tus the following connents:

. A. The AWS equation for calculating the punching sheer al-lowable for tubesteel connections is based upon the re-suits of a limit analysis assuming a specific yield-line pattern within the chord of the tubesteel. When a hole .

+

is placed in the tubesteel and the edge of the hole is loaded, limit analysis would predict a different yield--

line pattern. This new yield-line pattern will result in

- -a lower allowable punching shear. The presence of the '

coverplate further complicates the problem of determining punching shear allowables since one cannot autonatically _

expect an increase in the AMS punching shear allowable proportional to the increase in thickness provided by the addition of a coverplate. In addition, the close proxi-mity of the load to the edge of the tubesteel also influ-

. ences the calculation of an allotable punching shear.

B. In the actual problem inodeled and reviewed by Cygna, our

. j finite element analysis predicted very little surgin to allowable in the coverplate using an average of the fi-nite element results along a line between the hole and

. the edge of the coverplate. The TUGC0 calculation re-

. ceived on October 18, 1984,- clearly shows a margin of approxistely 6:1 (12.76/2.21). Thus, the TUGC0 calcu-lations.would predict that this joint is acceptable for approximately six times more load, a fact not borne out by the finite element analysis. While Cygna did not con-r sider plate plasticity effects in the finite element l . analysis, Cygna is, nevertheless, concerned with the i

I

4 s: I Mr. J. B. George 84056.053 -

January 31, 1985 Page 2 of 2

• ATTAQWENT (Revision 1) j large difference in predicted capability, and attributes much of it to the use of AWS D1.1-79 without assessing 1 the impact of the deviations from D1.1-79. That is, one aust consider.that: .

1. 01.1-79 assumes the brace and enord are welded +

together. Thus, the brace locally stiffens the chord. This is not the case for the nut loading the tubesteel.

2. D1.1-79 assumes the chord is solid. This is not the case for tubesteel with a hole in it.

3. . D1.1-79 states that yie'.u-line analysis can be used if 8 < 0.8, which is true for this joint .

Thus, AWS does recognize that yield i (S = .6).

line theory can also be used to predict joint strength in configurations. pictured in AWS.

Bhsed on the above, Cygna ~does not accept the use of AWS D1.1-79 as an appropri-ate method for establishing an allowable punching shear / joint capacity in the case of tubesteel with loaded holes (with or without coverplates). Cygna re-  !

quests that TUGCO. provide further justification on the design of such unique .

joints.

O m

i i

I e

s l 1

i t

a

r[ ~

4 Es .

415 397 5600 101 Cahtorma Sweet. Swse 1000. San Fratesco. CA 94111-5894 s

~ January 29, 1985 ,

84056.051 i

q 4

Mr. J. B. George Project General Manager Texas Utilities Generating Company ,

j Comanche Peak Stean Electric. Station ,

Highway FM 201 Glen Rose Texas 76043

Subject:

Phase 4 Open Itans - Punc'hing Shear

' Texas Utilities Generating Company Comanche Peak Steam ElectricJitation

- Independent Assessment Program - Phase 4 Job No. 84056

Reference:

M. H. Willians (Cygna) letter to L. M. Popplewell (TUGCO), "Open Items Associated with Walsh/Doyle Allegations " 84042.22, January 18, 1985 Dear Mr. Georgei Cygna and 1UGC0 have corresponded on several occasions regarding the punching shear / joint capacity of tubesteel with holes. The attachment to this letter sunnarizes Cygna's differences with 1UGC0 on the acceptability of using an AWS D1.1-79 methodology for checking the adequacy of these designs. Cygna does not consider this to be a standard design - particularly when punched tubesteel is used as the backing. plate for a cinched U-bolt. We believe that careful consideration should be given to applying AWS without considering the basis for the standard.

a 1.7.-

. 4

• Sen Frarcoco esseen San Dogo . Cfhcager McNend

'seeEi . _

84056.051 Mr. J. B. George Page 2 January 29. 1985 i

This letter coinpletes Cygna's comitment listed under item 4 of the Open Items List attached to the above referenced letter. Please call at your convenience if further discussion of this position is necessary. i Very truly yours.

l, N. H. Williams Project Manager Attachment '

cc: Mr. V. Noonan M . S. Burwell Mr. S. Treby Mr. D. Wade Mrs. J. Ellis Mr. D. Pigott t

4

• s I

6 h

i l

t i

e

f i

t

/ umummmmme ,

Nb C ATTACHMENT

• 4 84056.051 Mr. J. B. George Page 1 of 2  :

January 29, 1985 l

Calculation of Allowable Punching Shear / Joint Capacity in Tubesteel with Holes For support MS-1-002-005-572h, Cygna performed a finite element analysis to determine the effects of tube warping, and check the stresses in the tubesteel  ;

l and coverplate in the region of the U-bolt hole (see attached support '

drawing). On October 4,1984, Cygna requested that TUGC0 review this support and provide calculations justifying the design. TUGC0 submitted calculations on  ;

October 18. 1984, which showed that stresses in the tubesteel were acceptable Cygna further -

using the " punching shear" method of AWS DI.1-79, Section 10.5.1.

requested justification for use of " punching shear" as an appropriate check.

TUGC0 provided their justification in their letter dated November 8, 1984 (L.

Popplewell, TUGC0 to N. Williams, Cygna).

Cygna has reviewed the TUGCD justification and has the following connents:

f

1. The AWS equation for calculating the punching shear allomble for tubesteel connections is based upon the results of a limit analysis assuming a specific yield-line pattern within the chord of the tubesteel. When a hole is placed in the tubesteel and the edge of the hole is loaded, limit analysis would predict a different yield-line pattern. This new yield-line -

pattern will result in a lower allowable punching shear. The presence of the coverplate further complicates the problem of determining punching shear allowables since one cAnnot automatically expect an increase in the AWS punching shear allowable proportional to the increase in thickness provided by the addition of a coverplate. In addition, the close proximity

' of the load to the edge of the tubesteel also influences- the calculation of all allowable punching shear.

2. In the actual probleermodeled and reviewed by Cygna, our finite element analysis predicted very little pergin to allowable in the coverplate using a yield-line analysis of the finite element results. The TUGC0 calculation received on October 18, 1984, clearly shows a argin of approximately 6:1 (12.76/2.21). .Thus, the TUGC0 calculations would predict that this joint is acceptable for approximately six times more 1 cad, a fact not borne out by the . finite element analysis. While Cygna did not consider plate

, plasticity effects in the finite element analysis, Cygna is, nevertheless, concerned with the' large difference in predicted capability and attributes much of -it to the use of AWS D1.1-79 without assessing the Impact of the deviations 'from DI.1-79. Ttat is, one mst consider that:

i

, -- m _ , --___ _,,-,

g. .__ _

' ATTACHMENT Mr. J. B. George 4 84056.051 ,

. January 29, 1985 Page 2 of 2 ,

. a'. D1.1'-79' assumes the brace and chord are welded together. Thus, the brace locally stiffens the chord. This is not the case for the nut loading the tubesteel.

b. D1.1-79 assumes the chord is solid. This is not the case for -

tubesteel with a hole in it. i

c. D1.1-79 states that yield-line analysis can be used if S < 0.8, which '

is true for this joint (6 m .6). Thus, AWS does recognize that yield '

line theory can also be used to predict joint strength in configurations pictured in AWS. ,

i Based on the above, Cygna does not accept the use of AWS D1.1-79 as an appropri-

• ate method for establishing an allowable punching shear / joint capacity in the ,

case of tubesteel with loaded holes (with or without coverplates). Cygna re- ,

quests that TUGC0 provide further justification on the design of such unique ,

joints. ,

e

- I l

t i

I a

! i I, l 1

l l

)

e I

.