ML20090A222

Affidavit of Rc Iotti & Jc Finneran Re Consideration of Force Distribution in Axial Restraints.Prof Qualification Statements Encl.Related Correspondence

ML20090A222
Person / Time
Site:	Comanche Peak
Issue date:	07/09/1984
From:	Finneran J, Iotti R EBASCO SERVICES, INC., TEXAS UTILITIES ELECTRIC CO. (TU ELECTRIC)
To:
Shared Package
ML20090A206	List: ML20090A202 ML20090A212 ML20090A222 ML20090A235
References
NUDOCS 8407110306
Download: ML20090A222 (49)

Text

-.

1;_Uu . - _,, a ungact DUMETCD UNITED STATES OF AMERICA UP6s MUCLEAP. P.ECULATORY COMMISSION

.. moeno'-TM' **^"T" "*FE*Y *!T.4ICEMS7"C B ..Re#

. 10 P2:28 In the Matter of )

)

TEXAS UTILITIES ELECTRIC ) Docket Nos. 50-445

! COMPANY, ET--

AL. ) 50-446 i

)

(Comanche Peak Steam Electric ) (Application for Station, Units 1 and 2) ) Operating Licenses)

AFFIDAVIT OF ROBERT C. IOTTI AND JOHN C. FINNERAN, JR. REGARDING CONSIDERATION OF FORCE DISTRIBUTION IN AXIAL RESTRAINTS 1

We, Robert C. Iotti and John C. Finneran, Jr., being first duly sworn, hereby depose and state as follows:

(Iotti) I am employed by Ebasco Services, Inc. as Chief

Engineer of Applied Physics. In this position, I am responsible for directing various analytical and design projects in diverse technical areas, including analyses of the response of piping and support systems for dynamic events, including earthquakes. I have been engaged by TUEC to coordinate and oversee the technical activities performed to respond to the Board's Memorandum and Order of December 28, 1983. A statement of my educational a'nd

. professional qualifications is attached to Applicants' letter of May 16, 1984, to the Licensing Board.

i (Finneran) I am the Pipe Support Engineer for the Pipe Support Engineering Group at Comanche Peak Steam Electric

~

Station. In this position, I oversee the design work of all pipe

! 8407110306 840709 i PDR M K 05000445 O PDR

design organizations for Comanche Peak. A statement of my educational and professional qualifications was received into evidence as Applicants' Exhibit 142B.

O. What is the purpose of this affidavit?

A. The purpose of this affidavit is to address CASE's concerns with Applicants' method of determining the load distribution to axial restraints. CASE's concerns regarding Applicants' analyses of axial restraints are set forth in Sections XII and XVII of their Proposed Findings.

Q. What are axial restraint supports?

A. There are two types of axial restraints. The first type employs trunnions which distribute the axial load to the remainder of the restraint which is configured as a trapeze.

The second type distributes the' axial load to a frame support via lugs welded to the pipe. The purpose of both types is to provide an axial restraint for the pipe. Both types employ welded attachments to the pipe being restrained. (See Figure 1.)1 There are different configurations for both types. For the first type, which will hereinafter be referred to as welded attachments to trapeze supports, there are two basic configurations employed for both horizontal and vertical supports. One configuration employs a single trunnion welded to the pipe and also welded to a beam or tube steel cross piece which is then connected to the two legs of the.

1 Figures and Tables are appended at the end of the Affidavit.

trapeze (see Figure 1, type 1). These legs are either away struts or snubbers. The other configuration employs double trunnions (on either side of the pipe, which may run either vertically or horizontally) which are attached to the two legs of the trapeze (see Figure 1, types 2 and 3).

The second type of axial restraint will hereinafter be referred to as lug-type and exists in two configurations:

four lugs and two lugs.

1. Welded Attachments to Trapezes (Trunnions)

Q. What is CASE's concern with the welded attachment to trapeze supports?

A. CASE alleges that Applicants' design method for this type of restraint (modelling the support as a single support acting in the axial direction) is incorrect in that it ignores the rotational resistance of the restraint and, thus, does not account for certain effects on the piping and supports.

(See CASE Proposed Findings at Sections XII and XVII)

Q. What is your evaluation of CASE's concerns?

A. First, we do not agree that modelling of these supports as unidirectional supports, i.e., as a single support acting in the axial direction, is incorrect. As CYGNA has stated 2, and we agree, the modelling assumption employed by Gibbs &

Hill in their pipe stress analysis is generally appropriate.

Thic is so because the rotations are very small and 2 See Tr. 13081-83; 13105-10 and 13124-25. See also Board April 1984 Exhibit No. 1 (Testimony of Nancy II. Williams),

Response to Doyle Question 12, at 27.

accommodated by the play in the two legs of the support.

Moreover, when seismic analyses are performed using the response spectrum method, as is the case at CPSES, the resulting support loads are not dependent on the relative phase between the response motions, i.e., the axial and rotational motion. In fact, modelling of the rotational constraint of the support using a responae spectrum analysis would always add the peak of the response load resulting from the axial motion to the peak of the response load resulting from the rota~ tion. Therefore, this modelling technique would be very conservative and not necessarily a more realistic modelling technique. Consequently, Applicants' believe that modelling the restraints in question as purely axial restraints is adequate. As already noted, this view is shared by Cygna. Even though we do not believe the modelling technique propoasd by CASE is either more appropriate or necessary, we have evaluated the impact on piping stresses and support loads which could be calculated by modelling the supports as CASE would wish.

In order to assess the effect on piping stresses from mo'delling the rotational constraints, Gibbs & Hill performed a reanalyses of several stress problems for lines ranging in size from 4" to the 32". Table 1 (attached) shows a comparison of the results obtained for the pipe stresses under the two different modelling assumptions, i.e., with ,

and without modelling of the rotational constraint, for the

32" main steam line. As shown therein, the pipe stresses are negligibly affected by the modelling assumptions.

i Analyses of the other lines indicate identical results with respact to pipe stresses. Thus, these analyses demonstrate that excluding the rotational constraint of the trapene supports has virtually no effect on the pipe stresses.

Q. What is the impact on the loads computed for the supports themselves when the rotational constraint is modelled?

A. There is a change in loads on the supports themselves when the trapeze supports are modelled with the rotational constraint. However, that change occurs only for the trapeze supports themselves. The remaining supports are not significantly affected. Table 2 compares the loads computed for all supports other th.n the trapeze supports under the two modelling approaches for the main steam lines. As is evident from the table, the change in support loads is negligible. The same result was obtained for tha other lines reanalyzed.

For the trapeze supports themselves, however, the change in calculated loads can be much greater. This change would be expected when one models the rotational constraint of the trapeze support using a response spectrum analysis.

Under this circumstance there will be an additional load acting on the component in each side of the trapeze due to the rotational constraint since it is assumed that the peak load due to trunnion rotation is always coincident with the

peak load to due axial movement. This effect is illustrated in Table 3 for the trapeze supports included in the same stress problems from which Tables 1 _and 2 have been taken.

~

However, a completely bounding conclusion cannot be made as to the magnitude of the load increase resulting from the inclusion of the rotational constraint. This is so because in addition to the analytical technique, (i.e., response spectrum vs. time history linear analysis vs. nonlinear),

differences in loads are generally a function of piping flexibility, support rotational stiffness, and the free angle of rotation of the pipe as calculated from the non-rotational constraint analysis.

Q. Have you performed any additional analyses to assess the potential load increases and their consequences which may result from employing the modelling assumption suggested by CASE?

A'. Yes. Every double trunnion support employed in Comanche Peak Unit 1 and common has been evaluated against the loads which would be computed either from computer stress analysis or manual methods (discussed below), employing the rotational constraint.

For all of these supports the " free" rotation of the pipe (ccmputed in the abaence of rotellunal constraint) at the location of the support is'very small, i.e., less than 0.94 degrees. Accordingly, it is appropriate when ,

evaluating the loads resulting from this modelling l._

i technique, i.e., including the rotational constraint of the t support, to consider the rotation-which produces the increased load into either side of the trapeze to be self-  ;

limiting. In other words, that rotation cannot exceed the value which would occur if there were no rotational  !

constraint. Loads resulting from such rotation are, I

therefore, also self-limiting and may be characterized as loads resulting from the constraint of free end displacement. Section NF, Article NF-3231.1, of the ASME [

Code permits evaluation of such loads against an allowable equal to three times the normal allowable. Further, that Article requires no evaluation of such loads for emergency or faulted conditions. The total load experienced by the support can thus be characterized as being composed of the axial load, which gives rise to primary stresses in the pipe and supports, and the rotational load which is self-limiting and gives rise to secondary stresses in the pipe and supports.

t Q. . What are the results of your analyses of these supports?

A. The stresses resulting from the axial load have been previously evaluated in the normal design process and were found acceptable. The total stresses resulting from the L

combined axiat and rotational loads calculated in our reanalysis have been evaluated for each of the double ,

trunnions in Unit 1 and common against the allowable limit,s permitted by Section NF-3231.1. The total loads imposed on i

h

each side of the trapeze from modelling the rotational constraint have been found to be acceptable, i.e., in no

~

case have Code allowables been exceeded, when the increased loads have been factored in the support design.

Q.- What is the manual method used as an alternate to the computer analysis?

A. This manual method conservatively predicts the effect of the l

self-limiting pipe rotation on the distribution of loads to l

each side of the trapeze. It provides an appropriate method to readily calculate the change in load resulting from inclusion of the rotational constraint. To illustrate the l appropriateness of this manual method we present in Table 4 l a comparison of the additional loads 3 (due to rotation constraint) computed by response spectrum analysis and by the manual method for the 32" main steam problems of Tables l

1, 2 and 3. This table also shows the " free" rotation angle at the trapeze support points. As is evident from the I

results, the manual method always calculates additional ,

loads which are higher than predicted by computer analysis. I i

This conclusion was further confirmed by comparison of the results from computer analyses and the manual method for other piping systems.

l l l I

r 3 Additional loads here refers to the increment of load due to 1 i

modelling of the rotational constraint of the support, whi~ch is over and above the lood computed by the original analysis performed with no rotational constraint in the model.

f f

l Q. What is your conclusion regarding the validity of CASE's concerns with respect,to the modelling of rotational constraint for trapeze supports?

A. CASE's assertion that Applicants employed incorrect modelling assumptions for these supports is unfounded. 'As we previously indicated, modelling the trapeze restraint as a ,

single axial restraint is common practice in the industry, l

, and no basis exists to conclude that this practice is not .

1 >

appropriate or that another analytical model is more  !

realistic or better than the conventional analysis. As I demonstrated above, even if the trapeze restraints are ,

i modelled as CASE suggests, the resulting support loads and pipe stresses are within Code allowable values. Hence, CASE's concern ~that Applicants' modelling approach for these supports could have adverse consequences for the supports and piping is not valid. l

2. Lug-Type Restraints Q. With respect to the lug-type axial restrai'nts, do you agree with CASE's assertions that the method employed by ITT  ;

Grinnell to determine the loading distribution in axial e restraints is inadequate?

A. No, we do not. CASE presents two concerns which can be .

summarized, as follows:

L l

t

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

i (a) Construction cannot achieve perfect planes in the installation of the four lugs on the pipe. '2he re f o re ,

distribution of load according to stiffness of the support s,tructure is not valid (see CASE Proposed Findings at XII-6),

and ,

(b) Angularity of the pipe (due to thermal expansion at the point of support) will preclude four p,oint contact.

Consequently, the structure should be analyzed assuming single point contact at the extreme point of the structure (see CASE Findings at XII-6).

We will address the two concerns separately, below. {

Q. Have you performed any analyses to assess the validity of i

CASE's concerns?

A. Yes. With respect to CASE's first concern, we concur with i

CASE's premise that perfection in construction is not achievable. On the other hand, it is neither necessary nor reasonable to expect that the four lugs can be installed in a ,

t perfect circumferential plane with "zero" tolerance.

Nonetheless, we expect the lugs to be installed within [

F

" reasonable" limits and, indeed, have found that this is the case. ,

Construction practices in the installation of pipe lugs ensure that the maximum deviation in alignment of the lugs  ;

with their mating surfaces will be very small. We surveyed twenty-nine supports which have lugs welded to the pipe on ,

both sides of the support frame (see Figure 1). In only one i

i i

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

t instance was the measured maximum deviation (difference in distance between any of the lugs and the frame, on their respective sides of the frame) in excess of 1/16 inch. 'I n this instance, the deviation was 5/64 inch on one side of the support. In five other instances, the deviation on one side was 1/16 inch. Twelve supports had essentially no deviation.

. More importantly, we found that in most instances at least two lugs on either side of the frame are equidistant from the frame, and that the maximum deviation between two lugs on any one side of the frame nowhere exceeds 1/32 of an inch. In fact, 19 out of the 29 supports reviewed had at least the two closest lugs located equidistant from the frame on both sides of the frame.

Q. What do these findings regarding the location of the lugs demonstrate?

A. With maximum deviations at 1/16 inch, any overstress condition which may occur in the pipe, in the lug or in the frame will only be localized and self-limiting. If a local overstress condition does occur at a s, ingle lug, resulting local deformations will readily redistribute the load to other lugs. Because Applicants designed each lug to carry half the maximum load which could occur, even if some local deformation occurs the entire lead wil1 be fully resisted upon engaging one other lug.

Q. What have you found in your analysis of CASE's second concern _

regarding the distribution of loads between the lugs?

L

. 'l s

A. We first considered the situation which CASE claims should be i addressed. Specifically, CASE argues that the load should be assumed to be taken by the lug furthest from the support anchors (see Figure 1) on the support structure (see CASE Proposed Findings at XII-6,7). Under Applicants' original design assumption that two opposite lugs carry the load, the i l

load on the frame is assumed to act through the point where l imaginary lines connecting all four lugs intersect. This L loading condition will result in a given deflection of the ,

i i l frame. If, however, the load is applied further out via the l

extreme outboard lug (as CASE argues should be assumed), the  :

frame deflection can be larger, since the moment lever arm l between the frame embedments and the point of load i l-application is longer. Therefore, the frame may experience -

i l-larger stresses than would otherwise be computed on the basis of two lugs sharing the load. On the other hand, frame t

deflection w.ill tend to close the gaps to the other lugs.

Consequently, two cases are possible, if the load is initially not shared by at least two opposite or adjacent lugs.~ One case corresponds to the instance whereby the lugs l

are much stronger than the frame. In that instance the entire frame will either deflect sufficiently to bring

! e additional lugs in contact (if it is sufficiently flexible)  ;

or it will deflect or yield locally to accomplish the same thing. The second case corresponds to the instance in which l

l

the frame is stronger than the lugs, and the loaded lug deforms inelastically until an opposite or adjacent lug shares the load. Applicants have investigated both cases.

'To illustrate the case of the frame being weaker than the lugs, Applicants have performed a study of idealized frames loaded axially via a four lug arrangement. The typical frame used is shown in Figure 2. Two different cases were analyzed. One case utilized a M4 x 13 frame members with a 4" diameter pipe and the other case used W6x15.5 frame members with an 8" diameter pipe. For each case, four load combinations have been analyzed. The load combinations that have been chosen are, as follows:

1. Total load P, applied to the outboard lug (joint 6 of STRUDL Model)

2. Total load shared equally, P/4, amongst all lugs (joints 3, 5, 6, 9)

3. Total load shared equally by the horizontal lugs 4 only, P/2 (joints 3 and 9)

4. Total load, P, applied to the inboard lug (joint 5)

The results of these analyses are tabulated in Table 5 for each loading case and each configuration analyzed.

Q. Have you analyzed the effect of frame deflection on the capability of the support frame to engage additional lugs?

A. The two cases were chosen to simulate t ra nies that are I

relatively rigid, so that their deflection under these loads would not exceed the 1/16-inch guideline used at CPSES to .

design supports. One case was chosen to represent a frame i

i the deflection of which would be small, while the other represents the instance in which the frame deflection would approach the maximum 1/16-inch. Therefore, these frames represent the range of frame deflections that would be encountered at Comanche Peak and, thus, provide an indication of the ability of those frames to deflect so as to permit engagement of additional lugs.

If the frame is sufficiently stiff to deflect a minimal amount (as in CASE I) it will either carry the load having -

engaged a single lug or will deflect further until another lug is engaged. That additional deflection, however, is not i likely to significantly exceed 1/16". Alternatively, if the frame does deflect approximately 1/16" (as in CASE II),

depending on the relative distribution of the lugs, a second lug may be engaged before the final deflection is achieved or the final deflection may slightly exceed 1/16". Again, any excess loads would be self-limiting in that as soon as the required small deflection is achieved, the load will be '

l shared by at least two lugs, and hence, the deflection no longer increases for a given load.

1 We note that for the frames associated with the twenty-  :

(

nine supports which were ' reviewed for-lug spacing, the combination of pipe rotation, local yieldings'of lags, and frame motion will only have to result in a displacement of less than 1/32 of an inch for a second (or third) lug to -

! become engaged. In most instances, frame displacement alone l

i l

_ ~

will result in this displacement. When this is not the case, t

minimal yielding of the lug or the frame will bring a second lug in contact with the frame.

Q. Have you analyzed the effects of localized yielding of the pipe or lugs?

A. Yes. We have analyzed the effect of localized yielding in the lug and pipe surface which'would be necessary to bring additional luga in contact with the frame. This analysis, which has been performed using a non-linear finite element technique and the computer program, NASTRAN, is presented in Attachment 1.

Q. What are the results of your analysis?

A. The results show that minimal plastic strains, entirely localized at the surface of the pipe and lug welds permit 1/16" deflection of the lugs. These minimal strains are of ,

no consequence to the integrity of the pipe or the lug.

In addition, Gibbs and Hill has verified that the additional bending stresses on the pipe, which would occur if the two loaded lugs were adjacent rather than opposite, are ,.

acceptable. Attachment 2 summarizes the results of Gibbs and ,.

t Hill's calculations of these additional stresses for pipe sizes ranging from 3 inch to 24 inches in diameter .

Q. What are your conclusions regarding CASE's concerns with respect to force distribution in axial restraints.

- ~,y--- -

e-m ~~-m-.c <

A. We conclude that although the consideration of rotational constraint in modelling axial restraints could result in the calculation of higher loads in these supports, such a modelling assumption is no more appropriate than the conventional assumption of modelling the restraint as a single axial restraint. This is particularly so given Applicants' use of response spectra analyses. Moreover, even if the rotational constraint was included, the analyses discussed above demonstrate that the supports are capable of accommadating whatever load increases may be calculated by that technique. With regard to the CASE's second concern, (modelling of lugs), we believe it is premised on unrealistic assumptions. Nonetheless, even taking those assumptions as given, we have shown that a very small deflection or local yielding of either the frame or the initially contacted lug, will bring at least one other lug in contact with the frame.

This fact is consistent with Applicants' design approach of assuming at least two lugs will share the load. In sum, CASE's assertions present no concern for the adequacy of the design of these supports or accompanying piping.

17 -

/4"'

1W Robert C. Iotti Subscribed and sworn to before me this _9th day of July, 1984.

I

/4 Notary Public

, STELLK lltTE NOTARY PUK!C. STATE OF NEW YOM No. 31 1444786 Quel:f.ed in New York CowWy Commiss an Expires Mar.30,1905 ,

'A ,

4 i

l t

i I

I t

. YJohnkC.fwe r~ -

g Fimurran, Jr.y i

• WW W men to before me this fTday of July,1984.

jN? f. /%Q 3 '%

c e/ -1

^

~ , -- f; $ ""

4 '

. :- \ ./3) mtsry y

' M r- '8 '

'y~e

~-

f This is a telecopy facsinile. The' original will be cent under Senarate cover, o

6

?

I i

(

l

\

i h

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

O J

d FIGURE 1 TYPICAL TRUNNIONS & LUG SUPPORTS Trunnions

////> 11 1 11 giy Snubbers S I

Pipe e or Struts '

' i M nubbers or Struts f _ ) L 4

/ \ runnion T

Trunnion Type 1 '

< Type 2 ll l11 f) //1 .

Pipe p Snubber C or Struts e\ )

Trunnion!

Type 3 Lugs n

Two or C F Four Lugs O O '

]

u U

um W W

, y y,,,,- ,

CASE I

@ 8 @ q @ 10 S,7 g 5< > < >6

.o' A

~

9 .

,e sr  ? . .

.3 g

t, 2 @ '3 @

< 6.0 7 4.312 L, CASE 1: 9" PIPE ~

MEM. 1 + IO My X Il 8.625 m T =I00 F < n E= 27.85 X E6 PSI G=10.71 X E6 PSI Sy= 36,000 PSI ,

l CASE II

~ . ,

l

~

@ g @ q @ 10

.- 3 -h7

~~

-~

5, , , ,6

m t

6 n OmbO 7 ss M

~~

p m

d' it it h -

cy l g '2

'3 @

CASE II: 8 PIPE ,

15.0 .

,7.375 ,

T= 100 F 19.75 ,

E= 27.85 X E6 PSI ,

G= 10.71 X E6 PSI Sy=36,000 PSI MEM. I-->l0 W6 X 15.5

+

TABLE 1 MAXIMUM PIPE STRESS COMPARISON

• Prob. Node EQ 9. EQ.9 EQ. 9 EQ.9 EQ.10 EQ.10 No. No. Upset Upset Emerg. Emerg. (Trapeze)

(Trapeze) (Trapeze) 1-1 15 9152 9140 9844 9826 3190 3149 20 9969 9963 10816 10792 15218 15202 9 9690 9676 10499 10942 13457 13545 130 10259 10252 11307 11360 1992 2040 1-4 30 11475 11081 12167 11692 2179 2034 106 10965 10796 11554 11348 2830 3118 304 8421 8205 8616 8360 5240 5110 85 8819 8693 9933 9020 13981 14692

• Equations (9) and (10) are the equations of the ASME Code Section III which are used to compute stresses in the piping system for comparison tio allowable values. (Trapeze) refers to results which are obtained by the analysis which models the rotational constraint of the support.

1

I TABLE 2 CHANGE OF ADJACENT SUPPORT LOADS FOR ROTATIONAL CONSTRAINT AND NON-ROTATIONAL CONSTRAINT ANALYSIS Rotational No-Rotational Constraint Analysis (Kips) Constraint Analysis (Kips)

Hanger No. Fx Fy Fz Fx Fy Fz MS-1-04-004-C72K 33.63 32.37 MS-1-01-004-C72K 44.3 42.87 MS-1-01-005-C72K 28.52 2.75 28.71 2.76 MS-1-01-006-C72K 70.87 12.52 70.86 12.52 MS-1-01-007-C72K 33.27 33.83 -

MS-1-04-dO6-C72K 36.21 37.14 h

I s'

i L

TAELE 3 COWARlSm 0F TRAPEZE LOADS FOR R)TATIONAL ,

CmSTRAINT MD Nm-RATIONAL CONSTRAINT MALYS1S A. %tational B. Non-%tational  !

Constraint Analysis (Kips) Constralnt Analysis (K1ps) ,

Henger No. Fx Fy F2 Fx Fy Fz MS-1-01-003-C72K 53.59 27.77 '

MS-1-04-005-C72K 102 52.59 MS-1-04-007-C72K 86.4 63.43 1 MS-I-04-008-C72K 36.87 22.2 MS-1-04-009-C72K 70.66 41.23 1

1

y. ++-- n  % -

e--- -w-- - , --,----w - -, ,,y-- + , , - - y, ---r-y---,- e , -,

TAa.E 4 COWARI SO4 0F LNBALMCED LOADS IN TRAPEZE SUPPORTS FOR MMUAL MD COWUTER MALYSIS - SSE & FSAM 5

Free Fbtational Loads From Computer Hanger No. Ibtation Stlffness Manual Method Load (K1ps) 9 (deg.) k = ko6 (Klps) M/L j L

MS-1-01-003-C72K .111 4.6x10 16.02 11.80 MS-1-04-005-C72K .206 8.52x10 51.05 24.73 MS-1-04-007-C72K .054 9.84x10 16.56 10.76 MS-1-04-008-C72K .048 8.52x10 11.90 7.28 MS-1-04-009-C72K .1 39 8.62x10 40.22 15.26 I

1 j.

l

)

1 l

?

l

TAELE 5

. RESULTS OF FRAME MM.YSES (Figure 2 Frames 1 l l

s CASE I Loading Case / Joint Joint Joint Joint Ma g T Deflection inches 3 5 6 9 Members 1 & 4 *

l. All loads on outboard lug .0114424 .0046689 .0201173 .0114194 14.79  !

II. Loads shared .0086715 .0041157 .0140854 .0086628 12.07 4-ways lII. Loads shared .0087955 .0039680 .0139015 .0087955 12.06 2 weys I V. Loads on Lug .0056525 .0038580 .0084212 .0056409 9.37 inboard  !

i i

CASE 11 Loading Case / Joint Joint Joint Joint Max. T Deflection Inches 3 5 6 9 Members 1 & 4 #

i l 1. All loads on l outboard lug .0378060 .0194758 .0595369 .0378060 16.56 l I

l 11. Loads shared 4-ways .0294855 .0163781 .0440817 .0294855 13.85 L lil. Loads shared 2-ways -

.0296730 .0161213 .0436851 .02 % 730 13.85 f IY. Loads on Lug i Inboard .0207901 .0137941 .0294179 .0207901 11.15  ;

r I

t I

h t

i i

i kTTACHMENT1 i

PIPE LUG ELASTO-PLASTIC ANALYSIS i i

I. INTRODUCTION t

Typical pipe axial supports at CPSES consist of four lugs [

which are welded to the pipe on both sides of the support frame.  ;

i i

The design assumes that at least two of the four lugs function to transfer load to the frame. CASE has alleged that it is possible  !

for only one lug to be functional due to the fact that

{

installation tolerance may result in only one lug making contact l with the frame. This study investigates the local stress and l

, strain conditions in the lug and the pipe which might occur if a ,

[

i single lug carries all of the load, contrary to the design {

I assumption that at least two opposite lugs would share the load.

Inspection of several supports selected randomly have shown that a maximum deviation of 1/32 of an inch separates adjacent lugs. Thus, any loaded lug displacing more than 1/32 inches will cause the load to be shared by at least another lug. The same

! inspection identified 1/16 of an inch as the maximum gap between [

the frames and any lug. To account for a possible maximum f

! deviation between lugs of 1/16 of an inch, this study assesses the effects at a maximum deformation of a lug equal to 1/16 of an r

inch. r I I

.-.-_-_m . . . - , _ _ _ - - . _ _ . . _ , . . . . . , _ . , - _ _ - - - - . . .

i The analysis has been performed 1 sing elasto-plastic '

behavior of the lug and pipe material to closely follow the  :

i distribution in plastic strain in the pipe and the lug that might  ;

e

?

occur before another lug closes the gap to the framo and begins f t

to share load.

e i

i II. MODELLING l t

A finite element model of the pipe and lug has been i t

constructed utilizing the MSC/NASTRAN finite element computer i program. The pipe is modelled with sufficient length so that-the '

t local deformations are not affected by the model boundaries. l Since the load is applied to one lug, symmetry is employed so t

that only half of the pipe with the lug at center is included in j th'e model as shown in Fig. 1. As will be demonstrated, the strain effects are so localized that the use of a symmetric model ,

, is appropriate.

l The half pipe is modelled with two sides fixed and two ends i

with symmetric boundary conditions. Since the objective of this >

I analysis is to obtain the local strain distribution in the pipe f and lug when the lug displaces 1/16 of an inch at its load center, these boundary conditions are appropriate.  !

The lug and its welds to the pipe are modelled with hexahedron elements (CHEXA). The pipe wall is modelled with l t j shell elements (CQUAD4). Figs. 2-5 show dif ferent parts of the [

model. The pipe and the lugs are both assumed to be made of SA36 l

steel. The stress-strain curve of the material employed in the j l

l  !

i model is shown in Fig. 6. For modelling purposes, the curve has been approximated by a bi-linear curve. The slope of the elastic portion of the bi-linear curve is 29,000,000 psi and that of the  !

plastic portion, 140,000 psi. Yield stress is assumed to be 36,000 psi.

i III. ANALYSIS In an elasto-plastic analysis, the material behavior determines the stress-displacement pattern. Within the elastic limit (yield point), a linear relationship holds. Beyond yield, material strains in accordance to certain observed rules. -

Several criteria have been proposed that establish when and how a material yields. The most widely followed criteria are those established by Von Mises and Tresca.

As stresses exceed the yield strain, the stress-strain is no longer linear but changes with the increasing strain level. In a i

load-unload-reload loading pattern, it is observed that the new

' yield points' occur at different stress levels. This behavior is called strain hardening. Two of the most widely followed assumptions to account for strain hardening are the kinematic  !

hardening and isotropic hardening assumptions. The choices of yield criterion and strain hardening assumpt'.on depend on the characteristics of the material. For this model we chose the kinematic hardening assumption because steel has been shown to l

behave closer to this rule. For the yield surface, we have i

l chosen to adopt Von Mises. ,

The analysis is performed by applying incremental loads at i the lug surface. Before the elasto-plastic analysis a linear analysis was done to estimate the initial load and subsequent  !

incremental loads that should be applied to the lug.

The elasto-plastic analysis was begun with a load of 12,000 i

lbs applied to the lug. Subsequently, incremental loads of 2000 lbs were applied to the lug until the load reached 52,000 lbs, at which point a lug deflection totaling 1/16 of an inch was reached. The MSC/NASTRAN Solution #66 (nonlinear analysis) was utilized for this purpose. The solution provides element -

stresses and strains and grid point displacements at each 10ad increment.

IV. RESULTS Utilizing the computer output results, strain maps of the pipe and lug at selected load steps are plotted in Figs. 7-10. A j load displacement curve of the grid point 148 (outer periphery of the lug) is presented in Fig. 11.

The strain maps show that the plastic strain is highly localized. This confirms that the model chosen is valid, since boundary condition effects are considerably removed from the local plastic strain area. The strain maps also provide the i

patterns of progressive yield as the load increases. The load-displacement curve can serve as a guide to determine the datormation of lug under the applied loads.

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

i t

V. CONCLUSION The results of this study show that the plastic strains of the pipe and the lug are limited to the local area immediately adjacent to the lug. The strain levels are very low. At 1/lb inch lug deflection, the maximum strain in the lug is only .0009 in/in and in the pipe shell, .007 in/in. At such low levels of plastic strains, the pipe and the lug can carry the applied load without adverse effects until the load begins to be shared by the other lug (s).

The small amount of permanent deformation in the pipe shell would only occur in the first cycle of applied load, since subsequent cycles would be reacted by at least two lugs which have been aligned by the deformation of the first lugs.

t

" 4 y

i 1

b l

l

\ \ \ \ \ \ \ \ \ \ \ W

\ \ \ \\ \ \ \ \ \ \ $

\ \ \ \ \ \ \ \ \ \ \'

s5

\ \ \ \ \ \ \ \ \ \ fi si a:

". 5 N j

I j

I I j j I

l/I& x/ j I

j I I i

I j 5, :

't

/ / / ,K -

/ / / ;ai

/ /

/

/ 'sV=--

/ / / /

/ /

/

/ '

ig i / / / /  ;

V V V

\ h ,V.-,,.,,.-,._.,#

V V V V V

.\

A

)

/p 1

r .

! N #

/

1 l ,g N- ig7

• g 1 'l N les II' 111 332 113 314 21s p /

I "

23  % - l32 1< -

124 131 I I 4 12G 127 123 129 /

7 3B ' "

L47 11 33g g4g 1 \ I'*

141 144 I*S /

342 ~43

'l #

1 l 93 ' '

L5B 1 <

19 157 I N 8 52 97 153 155 156 /

H 154 p l #

1 I - lgg 168 If'

' \ 75 ggs 1GE I *' /

4j 1 y

5r ,g \

-$'164 - ,,

I f

T

,3]

M d

' 76 E5 GG E7 p 69 I

A g4 .7b - #

59 Ji q . .

1 8* 58 97 '

54 55 55 g / .

39 - - .8 i a j 4.

" o Ng g /

<{

k / 24

• 42 43 44 45 33 I N -

25 32 f

\ " 31 g /

~

27 28 23 3e

$ g N w .- # gg Z \ 18 N

17 j jg N g, I 12 13 34 IS /

/g I Fig. 2 N, P!PC WITM LDC STATICS 90M L u n . --.

s. n- .

Ca?>3 U.P. 2.86*1 1.13+1 1.86*1 La8EL5s ELEMENTS

h e

9 e.

e 4

b I E h ,

NH 5

~

! K R E

', 3 E :-

![l

~ ~

8 s

!  ! s- ,

m"'..

I 3_ 5 3 3 it 3 =

N.

~ -

A_ i. -

-  ; 5. .

E g e a c 5,:. -

!. E E 8  % *

• I z a -

E E a l

1

, = , , .

l J

t l

l A [

t l

9 6

\

4 G

3L /t 7)

/A / M/  !

, -vt d,\.M,\ 9 i; v i ; t n.).: 3

\\/\\ $.,

N- />

W 9

4 F

5 i

L e ,

I I:

I l

ll r

o 6

t I

l=

6 o l E '

N k J l.

.a w '

ins a Ch. ( ,

& L o,

U ==y u M.

e

=*

e G y

( bl e

• e l in us n - 1

/

I .

m .g *

.a p1 1

.a g N z. .

~

/

k ,

U, t' . .

k 'M A

e= .

)

k i.

t EK i O '

I f

i=

t X r

^

.,. - - - - - - - - - - , - - ~ <-m ---. - - - - - - - - - - - - , - -- . - , - - - - - - ~

1  ;

~

Stress-Strain Curves t For Specified Minimum Strength Properties The curves presented herein represent the minimum Testing of Steel Products," ASTM Designation A 370.

values that are guaranteed for the steels indicated. The curves plotted are for the following steels: USS l "T 1" (ASTM A 514, Grade F)', USS "T 1" type A The curves are indicative of the minimum stress strain patterns which may be expected from actual testing (ASTM A $14. Grade B)', USS "T 1"' type B USS con.

of specimens. pac", USS Ex TEN 60, USS COR TEN (ASTM A 242),

in general, tensile test results exceed the specified USS MAN TEN (ASTM A 440), USS TRI-TEN (ASTM minimum values for each steel. Many factors influence A 441),USS Ex TEN 50, USS Ex TEN 42.and ASTM A 36.

a test such as the composition of the heat of steel. The minimum yield points, or yield strengths, and location of the coupon, speed of testing, accuracy of minimum tensile strengths are indicated for the steels testing equipment, individual performing the test, use plotted.

of different testing machines, etc. Therefore, every l test coupon will not produce identical results, even * *" '

I though all tests follow the procedure outlined by *Thes,e

, , , , , sgels

. ,are,ava

,, ag in,pg,tes,.c tio[s a ofy$I*pYaNs$nty.

"Available in plates only.

l " Tentative Methods and Definitions for Mechanical r

I

'USS "T 1" (ASTM A 514, Grade F), :i, I15.000 psi. Min. re,s.ie strench; uss r.t-l USS "T 1,* type A (ASTM A 514, ..r.t type a. and -T.t tyo e 0.2% Offset Grade 8), and T.100.000 ps6. u.a. vierd strench; uss "T.t-

/ "T 1" type A. and "T.t" type 8 -

120 -USS "T 1" type B - @ 100.000 pse. Min. Tensile strength; uss com Pac I / - --- b 80.000 pse. Min. Yield strength; uss Ccw pac l

• y-----/:n-

• MN-@ g - - - -

.T 80.000 psi. Min. Tenssie strength, uss E n rtw 6o l ,

$ 60.000 pse. Min. Yield Point; uss En. Tem 6o

/ N l 'S rom psi. Min. Tensite strench; uss cea Tem s.

l / USS CON. pac \ Man. Tem. Tai.itM. and Es. Tim So 100 - e, ,,/ # p --- c. /T-4--- \-\ I 50.o00 psi. uin. vieia Point; uss cen.Tt= s.

'I Man-Tim. Tai Tim and Es. Tem So j

./.  %.N. \ USS Ex. TEN 60

{ 63.000 ps.. Min. Tens.ie e..nch; uss n rt= 42

i. 42.000 psi. Min. Yield Point; uss En Tim 42 llI /

N.N g USS COR. TEN C se 58.o00 psi uin. Tensile strench: ASTM a 36 ri 36.000 psi. uin. vierd Point; asTu a 36 ,

g 8 /

\ s;l\ s f.j 3 80

---___)+_1,._- , -N. _

u l '." \. N N' 1 USS COR. TEN (A and B) c '11

ry- - - -- g,- ,

:: --. .. - g, l ---_- '

(ASTM A 242),

2 }J. _ -- 4I '_ p _

g -'_~+ _ ;_

1. ' ' '* erUSS  % USS M4N. TEN

^ }' ~

}60 g_ _/d -,; ,,,

- t w i Ex TEN USS TRi- TEN u, g A A 6 ,,,,,

(ASTM A 441),

5 .

I f_ _ , __ y _ __L_

SS Ex. TEN 50

,, 'v', __ _____. . _ .c._ _ _ _ ___

l

_______ __9___ ___

I l i

l E=29 x 105 psi '

l 20 i I

. 1 I

O I 0.35 0.40

'g 0 0.05 0.10 0.15 0.20 0.25 Strain, Inches Per inch 0.30 Fig. 6 Stress Strain Curves For Specified Minimum Values wwmsm

0 4

4 m

1. 3 E 3l.o, h e i

I l

p +4,*D f' [ .d P f "v.it

/i /l*fA  !

,e

/ /g pg/ ,

/.  :

? J' ,C.' e

(

' )

r7 s,

Es'! '

=a A

/ e:t t( c /  : y N

• L ii

,/5/v /

l Ws.54./  % xi

%+ 1. ..-  : < .

\w n s n-x v  ;: E ,o g - ,

n. Q

$$ -Ql h r4

• SA O N$ .- 14-

. r -u r u g  ?~

y i t E G l o '

4 1

2 l

l N: I M

A O

I s

1 aJ I l

68 d .a p < a

+ p a

< u e ,

si f >-

I H .

1 l

l i

P

= 1

. r4 a

• a s.

s5)2 vsW4W a C, N]' W . VM'/ > .E. 19 o <

<h_l ... . - .

,. a 3

- /M  % a* ce

/ \

1 Msg 77shd N ' G, >gp t?

,a 1

g A ,

E G l 2

l d

I

• di l N-l l ^

O l \

0 '

• .a

- 4 VI i

a W i I

l

'h.

~

pr . + 'er a ,

.mefg 4 E- [ ?l t " L {i  :). ]

- "d$ T[ ' ili /k _

' f/k? O.'.

y

)

e,e- s., / /R .):./t/"

p.' ili j  %

.: 4. l-

, , . *- . - . e,e;n a

..! - L:.;

• f,*

g . gg

- . , xi ~ s,e w 3 b

a p  ;

, - _ r n s -v ^ \ AI g y (

e mI oe3 11.

.3 3 pi tn *

,4 z l.J ! S r u r av N D, kU 2

w 1 I i L

l N: 7 5 c

A

P L

E l

c. Y t

T L D S AI L A T E L R L E A Y P

k h

h *

"e f

l L /

E 1 D

O M

P =

.b S A 4

. i C

I M T 0

T 0A A

!h D 1 T S

D R M

'l V k

. G gL U L F

O I

iS F

E E H

l D

T I Y W G E U P

I L

P Za l

t

. i i

s l

!, r 1

l i I  !

t I f0p00 l ,

I i

3 m-i -.-.

QQQ c .

-i l '

j _ _ _ _ _. . _

j

c.  ; i l

i O 4000 i !i

. ;A i 2 i  !

3 I i  !

o I i

6.

v i p 30,00 0 l

'i

~M6 1

J l / l I  !

l i .

- 2o,ooo l ,

i

'/32. i l

1 -

i  !

Iopoo l t l

i I  :

0-

07. .04 06 .6,8 .10 DISPLACEMENT GMC,'HEO i r Fig. 11 LOAD-Pl5Pl.ACEMTrHT CORVE OF 6 Rib PolMT 148

ATTACHMENT 2

. Gibbe O Hill, fasc. Job No. 00-2.525-046 Clont TM.G Co .

R N ar* DENDt uG #FECTS OF u.NEVEMLY LOA DED GMEAR Lu.C,S ( 2 LUG 5 EM/,4(,E) rahdannn Number Sheet No. A. elZ w gr osse m o.= m osm m o= m o.m M I X X X X X pmp mr A% *Itris4

<>~*- w sk<dw

~ ~ ~

B E Mt> t A) d. E FF ECT.S bltE To M AEVE 4 LY LOA DF D SHF AR L465 (2tu.y o~<.emp pd.)

& ~ tetex.( g 774 we,tded adlacAeds ad etcltove, e d. b G $ H , w e+<. ca12 pri+ed 'n a.cco molau:e_

wik 4kt. tcge f adtachmi- ( n i n do y dc &>L M uon evtd skwv lug s, ) , o~J. 4ke. pigt sn e. . AppMg A i pvovW -ekt ovwviw of k olis G 6cd/o u p ctid.ava

~

kv EffW cdWAm+ 1p ad p/pt siw> ,. ,

l Tb peak 4Le enduod/ow o 2 budig e.ftcf r A to tvutvod lcadsd okens - Ut y , 2 (-tao) tuch edkthwads w e~<-

nebcied &s each o{ 5 (F& d?%d pip.e.

L6"OD M StEco ; Y OD t0" 0D 4

'2 4 " O Dj $ Vi]

p 6 " D. D , f l

~

Y iok0 Oh 1D 9m & hev E bi%, _ _ _

Tk lug enachads aladd, um 4Lt. ow> vi% L.

bp+ A tap, tw A. ucet 9ue pov ha au y 11_ wx dedp.ul W 4. bwet vak e d 4Lc_ be A g _

.Zt pcts me A.M oipFea4t,

~

e#

~ E dddfol clia.cAh to % ae4 Loc.d d*n uttbo A maa w t. L i ,aw.o=saa l Checking Methob# ; wh F-166, 7-82 l . _ _ _. _ _ . _ __,__ _

i .

Olbbs E Hl/1,Inc. Juo No. 00 - ESh - 0 4 6 Clent 4cCo.

&*wwt Pasu nt a c. EF FeCt1 op- u4EVEULY Lo40ro SusAR t.ar2 ( 2 tu32 cugop.) )

P'ain dahnn Number Sheet No. 4 o.f 2 j w .on 3lgr osse  % one w. om. a.v. o.as = om.

m L X X X X X Preparer AM fllf/84 cmemer %e

• 4/s I

~~ ~ ^ ca,$ ct & ( h sos f%5 es% sr> cW &

iA. O Cov a b ioi ol M% [.s h/2ow NL  !

edicwq.% .

% n ud w wa.hc asswap-t>s ukik kcreasc 4k safety facb of A; euchdi%

4. rke_ f acc. F ( 'Fg d) acii af 4b gud of -b /.4g , dcf  ;

4kt T b ct. dR adc. low fa' ch 4 -N- Q Ms u.d.uchy

& Wik gMed.

f. & i s ce.sW 4*h war. M'4tg s.% Mo F c.stNicLs.> wik Mk'. local s%) . d .s Stowh k rig.i fl.t 30% of mak". M'ag 5f% exi,1o c#dded ablM.

3. A pi p. A un a d 4Lc be4 3 h,cW.% v = aMa l

l oo ch44<_ ae . rkt_ w~A tu.4 c.oAi4*% op SR s3 i o-F- h u x G y & n en @r k. Wex s% vdu. .

1

- mes ee.g.e.*

~

. y

/

• , nu. srns>

7 b

]- ~ l l / \, j tug LocMiou we i

u. -

O

'l

~ ~

~w$ l

1. /v) s=

s Snv)(.as

= Snx k PA of exafged Lugs. itu. STitES) = Sg Checking Method # 1mmm- F-166, 7-82

! CMWm ---. ,-

PlPE Op.: 3" Gibbs B Hsil. Inc. Job No. C5ent Subject 15EMDIPJG EFf FC.T S oF uMEVFOLy LDAPED SHFA E LLl65 [2.l4GS EN64!L)

Calml*m Number SheetNo. Io4 F l mi.an T omie = om w. om m o.  % om m i X X X X X i er peror Aga r/ss/s4 ammer w dstis4  :

uipw, m :ua,, Le w v rgog. 00 : As-I-5sc B R th et-i-20s-oir-esse. u 4s sgy g ,: g , g i w i i pip e a 0.; s.s " , scarp.: 40 , MonroT ARM; 2. s " , secry,0put.s: 1,724 Broo.Honeur 83E AIDind Gu t6 s. TOTA L, AllDVA8LE CONDirl0M LOAD F ' (td> -ii4)- sTRen staE35 srREss. ' 51esss REMAP R S E q. 8 5 ~

17. 5 7. 3 1l94 1,20I rg 15,000 Eq.9(Xss( isl 377.5 269 2270 '2,487 tsy 22,soo Eq.9[gese) 151 3'I'1.5 219 2610 2,729 L85g 2Tooo Eq.ll 20 50 29 1721 1,750 ks4 3peo Eq9(sst) 229 57f 332 2767 3,099 216.%=32,oo

. ~c~ F ,k Xy - NEttigAL Arts oF BEND (MG R v' DRE To THE FoECE ~F.

Y

/.R -

t I

~

/ )p T IJoTH, itE tcH5 LF%Th PROB. 9 0 .: 45- I- 5f', S R H,; cc 28-004 -CSS R , tv.G sti! E. ; t/z , W., 1 ,

P IP E 0. D,: 3. 5 " 504ED.: 40 , Mcxept Aew,: 2.'25 SECTStt.T : l 72 4 _

BFNDW& 6EN tLOCAL TUTA L A lt.0WP.'.E BEMD.VcMEMT

' .werss - sitsc5 sTness Starss PEHARus CoNDit tou LO AD F (td s.aw)

Eq,.6 2. 2 S I 50 0 'l tooe h= 15,000 I

fq$Ans) 17 4 391.5 727 t,970 gl 97 i5y 22,s=

227 2,0S'5 i.sv 27,no Eq.9Ba) 17/# 331.5 2.310 302.2 179 1,9 so 2,129 V.4=57,5'*

Eq Il 137 2,272 2,5tc) 7.i6ss =32,40D Eq.$(a) t97 44S.2. 257 Checking Method # ;ma::::2,% ,

F-166, 7-82

? C ;*-. M y a y - ~~e- _

PIPE OD.: _ 6

~

- Gibbs S Hill,Inc. Job No. Cient l Mjad l3 EMDIMG. EFFEC.T S oF LAMEWOLy LDAPED SHFA R LR65 [2.LUS EN6ME.) f Cab _d*wiNumber SheetNo. 2 a0 F  !

w 93;r osa m os= m om m o.s. m om.

M r X X X X X l Paperw Mb f/15/6+

Checker W V2ghf4 man , m a.m, a w k \

MOg, se u. cT- - o17- 4o7- c124 Lt46styg,:  % , % ,3 f

00. As-i-Og ,

PiPG OD., f.62 5 ' SC W p.: 'to S MoHENT ARtt.: 4, 043 " secuoput.E: &6 I

/ )

/ l 6ENotNG Gen f LKAs. TOTA L.

I AttovAsLE ,

sfNo.nor.sur ' SWES$ EfITACRS CONDIT10M LOAD F ( b - m ). STRESS. . srREss. STRES.

i Eq,. 8 0 0 0 t,93I l,9 31 .g=16,600 Eq.9(Xss( 350t 14,253 l,676 8,757 10,It 3 I tsy24,'Joo j Eq.S[ga) Syor 14,253 I f6'16 11,691 13,'367 t.850 29,5 80 Eq.ll 13 3 54g 6/t 4,488 kg51 %)44,tro Eq.9(55E) 4669 18,310 2,232 11, t03 i 3, n s- ztsg y,as.6  :

l  :

'  % === ) oF ,x ._

,i I _

XX - NERTFAL Arts oF BENDING '

R , l', pas To we rogcE E

, 'i

._R , '

i

- / /pF .

x /

uoT4,nicHVtax74 ,

tuc. sisE ., I, I, 3 PRDB. NO .: AB-l-035A / B R N.; cr- t- 036 'co5-C728t

~

2 f#  ;

PirE O.D,k, 6,425" ,SCHEb.:405 , McHENT AElf.t 4.1'Y'SEcr , W L.~f: _

1! nom & 6sn + Local. Tttr4 L. A Lt.0WB:.E Cowo1 tons LO AD F BEND.e[oi4 (td ~unn- suscs. sTitess srREs 9EHARus

.Eq B. 73 215$ g g3 4:16,6oo 11 8.6 fq.%ns) 8, F83 l0 L 0 5,lt 39 6,44 9 15e 24,900 1990 16%' 29,880 Eq,9%) \*)n)g 8,583 lp g o Syb2 g,9 72 Eg. Il 2053 8,6 fi LO42 13,895' ll,9 i 31 $4 44,250 l Eq.$(se) 2519 IO,86 ft i7.78 6,634 7,912. 2%= 35,856 F-166, 7-82 Checking Method.#mj --, M::::::Ly

.r.r - -- ., -

_ y~ '

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

i PIPE Op.; 10" Uibbs a Hlif. inc. Job No. Client S@ 13 EA)DlM G- EFFTCT S oF Lt AJEVE/ULY LDAPED SHrA E LitCG [E L&S E Cah da*wiNumber SheetNo. 3 o.f F k Dam Rm Dan Rev. Dah h Dam ne,iemn T Dets X

f M 1 X X X X l r. ATA t/!5/84 checker he tulty unpu, ww, n w tr fEOB, D 0.: Aa-i-tro , y e 9 yA -1. oo6-oii- c s2 e u4ssgyt,: s, , , 4 PPG OD.'. 10. IS " SCRFp . 40 KohruT ARH.: 4 3.75, SECr.MODQ' . .2. 29. ')

Ge4 t'ec44. A LloVA8LE , i

. of40.Honeur SE Nol46 TOTA L.

CONDirtolJ LOAD F ~ (td> -m). STRESS. t sraE55 5795 55 ' STRESS f EffAP R S Eq,8 to 64 2 49l 433 sg g, goo Eq.90bs( 1773 11,303 S'78 4,534 49t4 tw 24,90o (I,303 3*l 8 5,2 58 5636 s.est ty,nD Ex.S(p) 1773 Eq.ll 3393 'Zi ,630 I23 5,3 61 4 024 %pq,zsd 12,7G 425 4,949 5 s 14 2164= >s;sr6 Eq.9(sst) 1995 qo oF ,x

~ '

XX - WEttTEAL Arts DF BhND(NG R V DwE To TME roecE E

'T R

, f I /F ocTH,itEtciGLF8xd4 If 3,4 Prob. NO .: AB-t-% B, B R W.; SF-X-00f-o30 F53P ttrc, 512 5 e, l

P LP iE o.D,: W.75" SCHEb.: 40 , MmENT AEH. 8 375 s SEcr A l.f: 29J -

)

BFNmW& 6EN tLou.L. TUTA L ALL.0WAT.E COMDtitoW LO AD F BEND.YCMEMT (tds..w) ~ ~ .w m.s . susc3 31xes3 STesss, Penus Eq.8 21 3l 17,847 597 2%7 3364 V I6PO*

3225 3,896 ist-27,ooo fq:%%ns) 23 95~ 20,058 67) 6Tf 7,436 t%= 37,0%

Eq 9gn)  ?,395 to,058 6]65 Eq.11 5280 (2,220 1,4 12. 15,4'2z. 16,834 sese}= 46000 Eq.8xus) iss3, 21,2l4 709 'b,4 58 4,167 2%r 3r,sro F-166, 7-82 Checking Method # ; ms::::2,m-

~ * ~ ~

- _ _ _ _ _ _ _ _ _ _ _ _ __WLWa~"'"' ----*v- _ _ _

PlPE OD.: 16 "

Gibbs S Hill. Inc. Job No. CGent Suhiect B EMDirJ6 - E Ff7C.T .5 oF (A A>EurAjLy LoAPED SHrA E 1465 [2.W.E E"6d45.) l Cah Jiabon Number Sheet No. 4 . 2 57

%.an 3=' one. m o.= w. o.m  % om. m om.

M i X X X X X Prop rer Am f/15/84 ch x*.r %_ rhtf# ,

unpu, ne w, tt aw l i

rgog, 00. : AB-l- 035A a e R cr- 1-oss. <tM. c 82 K Ltra se t,: iyz, l/2, 3

/

16" SECT M0ptit, E 70. ~5 f:PG OD.'. , SCM P.; 30 5T , KohrpT /,Rif.: 9. 5 " ,

SEUD.Honnur 6E nonNG Gen tl&AL. TOTA L- A LLD9AsLE ,

CDNDirialJ load F ' (td> -m). STRES5 SraEss, srREss.

• SwEss ff ffAP R S Eq. 6 0 0 0 M 97 3,4 97 re/6400 Eq.9(%sst 5,7tt7 35,'597 Sob 6 550 7,521 tsy24,yo Eq.9(ga) 3747 35;597 F04 g397 g,903 Lesg 2483o Ea, Il S' 47.5 .7 6 508 6,fo i Q.sv gtso Fo.9(sst) 37t-5 t 3 5;577 506 23.VO 8,856 2l6W 35;858

~o m F ,k XX - NEllTRAL Arts OF BEhp(NG R L' DwE To WE roRcE E

, f 'I' sR I

~

fr ooTH, yticHJ Lraxto PROB 90 .: A9-I-035A, a R H,; cr-i-ois - 415- c62tA , ti4c. sig E. ., I$, l'h , 4 dies 0.D,: 16" 95", SEcrP:WL.T: 70.3 _

scurb.

30 57, t4cwENT AeH.:

BEMD.fCKEift BFNOW& 6EN tLocal TOTA L A LL.ON Conotitou LO AD F ~ (tds. -w) ' ' mess . sTuscs. sTaess stars.s 9eHARus Eq.6 O O O 3,474 3,474 y [6,600 l

1 fq.%ae) )87-! 73,300 /,327 6,99f 8,322. i.51.e4,90o E.%*)

4. 9 8Zl 13, 300 1,327 Vl9 Io,846 i 8V 7',880 l Q.11 97l 8,27s' /I6 4,819 4,951 V-SM.2so M. 'Xsse) t3 g S/ 131,270 l, 6 74 S,360 l0,256 2%= M 24 Checking Method # j pd:::::2y,_ F-166, 7-82

- - - - - - - - - ~

PIPr op.; 24 " .

Uibbs S Hill. Inc. Job No. Client Subject 13Enotrac,.. eFFTC.T S oF uMEVFOLY LoAPED FHFA E L465 [2.LM.S EN6d4E D*ulation Number SheetNo. 5 af f .

%n w o.m an o= an om n= om m om m i X X X X X Preparer AFA 2/15/64 chemer er e +rw wipuij He u.Hj LtuTH rgog, 00 : AB- -60! ae R: cc -z-o50'-001- A dir t.us styr,: 4, i% , s Pipe CD.'. 24* SClf E p.; 20 STD MohrpT ARtt.: 13,7 5'" sect.Mor>ut.2: 161 9

/ / )

. DE40. Hor. tut 6E NoiNG GM f'KAL TOTAL- AliouulE ,

CONotrioM LOAD F (td > - N). STRF55 SrtE55 STRESS. ' 51RE55 I'f ffA P f(S oen ur.

Eq. B s,t 57 4S,40 9 268 3533 3.806 h= l 5.0" 1"UMJM.

Eq.9(X5s( NA UA DA- 4 '2 I t 9,21t L W 22,Too 8

Eq )(gc3r) h4 UA Al 4 4638 4,638 i%= 4000 E q ll DA MA n4 9190 5,l90 %)37, goo Eq.9(53r) DA NA 44 4645 6 *)/e5 2t6 W 32,400 F x

~c 8 n '

l - f XX - NE4TF'AL AriS oF BENp(NG R p*E To THE FORCE E j ,

/

1 W

/

'f F uoTH, it El?H{LFl%T4l tur, sta E ., f, 2 '/2, S PRoo. No .: As-i-6i s , E RN ; cc-i-070-002- A33 R. ,

PirE O.D,: 24" j scHED.: 20 ST , McHENT ARH.: 14 T ' SEcr N L.Y: 161. C) _

SEND.w.eur erwome ass +Locat. TpT4 c. A L t.0WB:.E

' mess . m ess steess Repeus ConDtitou LOAD F (Lds.-w) ~ sTusc5 Eq,.8 c)056 i31,312 8 11 Fil5 5,924 V IF oooi 6 Ilb i.sse 22,500 fq.be) 1184f 171, 738 1,06l 7, / 7 /

1.8%= 2 7,000

Eq % t) IlB44 171,738 1,06l 13607 /4,66 6 Q. ll 312.;4 4f2,603 f,7 76 34,'319 37, lls' V-M 37,5*

6,686 7,69/ 7.Ssc.32,40s

%. ##%E) (3458 l').5 14 I f,205 F-166, 7-82 Checking Method # jg '~ g _.,,

hppeetxx n o

c. ................. Gibbe B Hill,Irse nu.. Ces. . . . . . . . . . .

Cole 3y . . . . . . . . . . . . . . .

CN 5beetNo. .....Of....

New YtMest Chk'd/Apend. 3y . . . 4 . . . . G& HJobNo.........

Subjoet...............<...................................... Re f. Dwg./ Bees. N o. . . . . . . . . . . . . . . . . . .

s -

l N

s E

.==

N -

I

~

NNN\' EE h NN 1

l I i.

E 8

.t

• I si l w

k, k I

f

( 4 x i

h 4 , ,

k j

i

==

9 I E_ S ij i i

' ' l i

l

[5 i

, i Q n N i

{

i i l ,

i , , ,,

l l \l ;

mw-

. . E 1

1 -

I ,

I i l I I l t I i!

l

\

i , , i

i=4g p '

p . .Z I

l x h l kkkkkkM k E O k 4 8 O  %

• t amosa= _ _ _ _ . ___