ML20215K183

Rev 1 to Low Cycle Fatigue Assessment of ASME Section III Class 2/3 Piping for Unanticipated Loads

ML20215K183
Person / Time
Site:	LaSalle
Issue date:	09/12/1986
From:	Gibraiel S, Mahendranathan, Wipasuramonton SARGENT & LUNDY, INC.
To:
Shared Package
ML20215K172	List: ML20215K171 ML20215K177 ML20215K183
References
EMD-061058, EMD-061058-R01, EMD-61058, EMD-61058-R1, NUDOCS 8610280015
Download: ML20215K183 (16)

Text

.

SARGENT & LUNDY EMGINEERS CHIC AGO LOW CYCLE FATIGUE ASSESSMENT OF ASME SECTION III, CLASS 2/3 PIPING FOR UNANTICIPATED LDADS PROJECT NO. X-001 Calc. No.: EMD-060822 Acc. No. EMD-061058 September 12, 1986 Rev. 01 G

8610280015 861008 PDR P ADOCK 05000373 PDR PAGE 1 ENGINEERING MECHANICS DIVISION

4 i

SARGENTO LUNDY

l. EMjj$EERS Acc. No. EMD-061058 Page 2  !

?

f SIGNATURE PAGE i

3 l EMD

. Accession

! REV. NO.

{'

00 EMD-060822 PREPARED BY: R. MAHENDRANATHAN DATE: 07-02-86 REVIEWED BY: S. A. GIBRAIEL DATE: 07-02-86 APPROVED BY: G. T. KITZ DATE: 07-02-86 l EMD Accession f, REV. NO. ) //

01 EMD-061058 PREPARED BY: '

(4- V DATE: 7 IL R. MAHENDRANATHAN / /

REVIEWED BY: Nik N DATE: #!/1!hb 1

l i P. WI'PASURAMONTON APPROVED BY: '

• DATE: 9fl2./2G

/ S/ A. GIlRAIEL '

I

)

I i

i l

l

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

SARGENT O LUNDY EN j EERS

,co Acc. No. EMD-061058 Page 3 TABLE OF CONTENTS PAGE Title Page...................................................... 1 r

i Signature Page.................................................. 2 Table of Contents............................................... 3 l

1.0 Introduction.................................................... 4 2.0 Background...................................................... 5 3.0 ASME Code Section III NC/ND Fatigue Evaluation-Requirements......................................... 6 4.0 Basis of Section III NC/ND Fatigue Evaluation Requirements......................................... 7 5.0 The Proposed Approach to Fatigue Evaluation of Seismic Class 2/3 Piping.......................... 9 6.0 Use of Fatigue Evaluation Approach for Unanticipated Mechanical Loads.................................. 10 1 7.0 Conclusion...................................................... 11 8.0 References...................................................... 11 l

Appendix A: Example of Application............................. 12 l Figures......................................................... 14 QA Checklist.................................................... 16/ Final l

l

=

SARGENT & LUNDY E N $jj[cocRs Acc. No. EMD-061058 Page 4 ,

i

1.0 INTRODUCTION

The low cycle fatigue assessment criterion presented below is recommended for the assessment'of operating power plant piping for unanticipated loads. The.above criterion is not intended for use in design basis calculations.

During the scheduled snubber testing program conducted at the LaSalle County Station Unit 1 in early 1986, many snubbers failed to meet the acceptance criteria. The associated piping t

was evaluated for increased thermal loading due to snubber drag. i l

Some of the snubbers had been damaged due to unanticipated l

hydraulic transient loads. The source of the hydraulic transient was identified and the piping was evaluated. In all cases, the transient stresses were found to be within IE Bulletin No. 79-14 allowables, i

The majority of the increased piping stresses were caused by the higher drag loads on piping due to damaged snubbers resisting thermal expansion of the piping. Since no formal or informal rules have been proposed for static and dynamic overstress, a proposed acceptance criterion based on cumulative fatigue damage is presented below. It is intended for use in

, the assessment of piping for unanticipated static and dynamic loads in operating power plants. The above approach is a logical extension of the ASME Code Section III NC/ND fatigue l evaluation requirements into the low frequency range and could be used in place of Code equation 11 provided the total. number

- of stress cycles is less than 7000. Further, piping exceeding Code equation 9 allowables can be accepted using the above low cycle fatigue evaluation method provided all stresses including thermal are considered and piping meets equation 9 functional capability criteria GE Report NEDO-21985 (Reference 6). If the piping stresses exceed GE Report NED0-21985 requirements, piping is visually inspected to ensure no significant deformation or cracking of pipe has occurred.

[

SARGENT & LUNDY t EN RS jNE

, Acc. No..EMD-061058 Page 5 7

i

2.0 BACKGROUND

Fatigue-based criteria for the evaluation of piping were i introduced into the Piping Code, then ASA B31.1, in'the 1955 Edition. These criteria were based on moment fatigue tests on piping components by Markl (Reference 1), Markl and George (Reference 2) and Markl (Reference 3). The criteria involved the use of stress intensification factors (1-factors), and the stress limits related to cold (S c) and hot (S h) allowatle stresses; modified by a factor, f, which depends upon the number  !

of design cycles.

Stress indices were introduced in the 1963 Code for the l particular case of nozzles with cyclic pressure loading. ANSI B31.7, Nuclear Power Plant Piping, was published in 1969, and extended the concept of stress indices and fatigue assessment, for low cycle loads on piping. For Class 2 and 3 piping, ANSI B31 continued to use the fatigue evaluation method originally introduced in ASA B31.1 in 1955 which implicitly assumes at least 7000 cycles of loading. In 1971,Section III of the ASME Code was expanded to include vessels, pumps, valves and piping. With respect to fatigue evaluation methods for piping, l Section III adopted, with one difference, the rules contained in I ANSI B31.7-1969. The difference concerns the adjustment for stresses that exceed 3Sm (i.e., exceed.the shakedown limit).  ;

The Class 2/3 piping fatigue evaluation requirements are

, specified in subsection NC/ND 3652 Equation 11. The allowable stress in the above equation is modified by a factor f which depends on the number of design cycles. The~value of f.is defined in the domain of 7000 cycles to 100,000 cycles. The allowable stress at lower design cycles is defined to be equal to the allowable at 7000 cycles.

l l

\

I-

SARGENT & LUNDY ENC 3tNEERS Acc. No. EMD-061058 Page 6 The use of allowable stress at 7000 cycles for all cycles less than 7000 is not supported by any test data (Reference 4, l

page 60) and represents an arbitrary Code decision made to allow simplification of Code rules.

The recommended approach to fatigue evaluation for lower design cycles is based on the test data that form the basis of

the Section III N0/ND Code. Further, the factor of safety of 2 on stress and at least 20 on number'of cycles used-in the proposed criterian is identical to that of the Code.

30 ASME CODE SECTION III NC/ND FATIGUE EVALUATION REQUIREMENTS The code evaluation method involves the calculation of the j stress range S E by the equation l

c j

SE"I Z (I If the SE satisfies the equation SE s f(1.25 (Sn + Sc ) ~'S s ) (2) and Ss 5 S ,h the piping component is deemed to be acceptable from fatigue evaluation standpoint.

l Where f = Cycle dependent factor that varies from 0.5 at N = 100,000 cycles to 1.0 at N = 7,000 cycles.

= 0.5 for N > 100,000 cycles.

f = 1.0 for N < 7000 cycles, i - Code NC/ND St.?ess Intensification Factor Mc = Range of Resultant Moment Due to Thermal Expansion Sc = Code Allowable Stress at 100 F Sh = Code Allowable Stress at Hot Temperature l SE - Code Calculated Stress Range t

. _ , ~ .

l . .

SARGENT O LUNDY ENGl $EERS Acc..No. D3D-061058 Page 7 i

] S3 - Code Sustained Load Stress Z = Section Modulus of Pipe 4.0 BASIS OF SECTION III NC/ND FATIGUE EVALUATION REQUIREMENTS The Code method is based on the results of moment-loading fatigue tests given by Markl (Reference 1), Mark 1 and George (Reference 2) and Markl (Reference 3). Test assemblies were mounted in the fatigue test machine and subjected to a-preliminary load-deflection calibration. The assemblies were filled with water to provide a ready means for detecting failure, and then flexed cyclically (completely reversed displacements) through a predetermined displacement until a leak l

indicating a crack through the wall developed. In most cases, the component was then tested for 20% more load cycles to assess crack growth, etc.

The results were reported as points on log scale of Sp vs.

Np plots. Np is the number of cycles to failure (through wall crack). The corresponding nominal stress was computed by the ordinary beam formula: Sp - WL/Z. The load range W was.taken from the load-deflection calibration or. for loads causing plastic displacement, from straight line extrapolation of the elastic portion of the load deflection callbration. The lever arm L was measured from the point of load application to the point of initial failure (through wall crack).

The test method gives results that are consistent with the

- elastic analysis normally performed on piping system, even though stresses may be above the material yield strength and some plastic displacement may occur.

All tests were run on 4" nominal size piping components at room temperature. The material used to make the test specimens was ASTM A106 Grade B. Most specimens were tested "as-welded". A few girth butt weld specimens were stress relieved

i .

SARGENT & LUNDY EMGINEERS Acc. No. EMD-061058 i

cmcac Pa e 8 after welding and before testing; with no detectable difference in the test results. In order.to make the test information useful to the piping designer, Markl developed a correlation of the form:

-b 1 Sp = a Np (3) where a and b are constants developed fecm the test data.

Results of tests for girth butt welds are shown in Figure 1, in i which Sp and Np are plotted on log scales.~ Equation 3 is shown '

in Figure 1. The best fit of the test data was b = 0.2. The value of b = 0.2 was selected after evaluating test data for all types of piping products. For girth butt welds, the best fit value of 'a' is 490,000.

Sp = 490,000 Ng - .2 (psi, range) (4)

Mark 1's more general equation is:

1Sp = 490,000 Np - *2 (psi, range) (5)

where i is unity for a girth butt weld, and is the fatigue-based stress intensification factor for piping products other than girth butt welds.

Equation 2 and 5 are compared in Figure 2. Noting that Equation 5 is for cycles-to-failure (i.e., crack initiation) a l factor of safety is needed for design guidhnce. Figure 2

} includes a " design" line representing Equation 5 with a factor of safety of 2 on stress; analogous to the factor of safety of 2 on stress used in Section NB of the Code (The factor of 20 on cycles is also satisfied because the factor of safety on cycles is 25 - 32). The design equation is then:

iSd = 245,000 N ~ 2 (psi, range)

(6)

It is assumed, as in Code Section NB, that fatigue at temper-0 atures up to 650 F is not significantly different than at room 1

i temperature. Figure 2 indicates that even for S3 = 0, '

0 Equation 3 for A106 Grade B material at up to 650 F (S 3 , Sh" i

SARGENT & LUNDY EM j EERS

,o Acc. No. EMD-061058 Page 9 15,000 psi), has a factor of safety of 2 or greater up to about .

400,000 cycles. The f-factor, which varies between 1, for 7000 cycles to 0.5, for 100,000 or more cycles, is given in steps, between 7000 and 100,000 cycles. The factor of safety is high at low cycles; e.g., 5.2 at 100 cycles.

The high factor of safety at low cycles is due to the arbitrary decision of the code committee to use the "f" value j for 7000 cycles for all cycles less than 7000. The proposed approach given in the following section uses a factor of safety l of 2 on stress (32 on number of cycles) in low cycle fatigue evaluation. ,

I 5.0 THE PROPOSED APPROACH TO FATIGUE EVALUATION OF ASME SECTION III CLASS 2/3 PIPING The proposed approach to fatigue evaluation of Class 2 and 3 piping is a ~ logical extension of the Code rules in the low cycle range using the same test data and the factor of safety that form the basis'of ASME Code Section III subsection NC/ND j fatigue evaluation requirements for load cycles > 7000. The low I cycle fatigue evaluation requirements is satisfied using the following equation:

iS s 245000 N-0.2 (psi range) -

(7) where iS = Intensified Stress Range as defined in NC/ND 3652 Equation 11. Code Stress indices are used.

N'= Number of design cycles i

The above equation includes a factor of safety of two on stress l and 32 on the number of cycles.

1 If the piping is subjected to stress cycles with more than one stress level, then the cumulative fatigue damage is calculated using the Miner's rule as defined below:

l

.. . - - - , _- _ _ _ _ - , _ _ . _ _ _l

SARGENT & LUNDY EN j ,

E RS Acc. No. EMD-061058 Page 10 n)/Nj 2 2 +............+ n /Nk

+ n #N k 5I where nj =

actual cycles associated with ISj (j - 1, 2, ....., k) i N3 - Allowble cycles for Stress Level iS3 )

l iSj = Stress level associated with loading event j 6.0 USE OF FATIGUE EVALUATION APPROACH FOR UNANTICIPATED MECHANICAL LOADS l l

The main difference between mechanical loads included in l NC/ND Code Equation 9 and thermal or displacement load included Code Equation 10 or 11 is that mechanical loads are not displacement controlled. That is when the elastic analysis l

based stresses due to a mechanical load significantly exceed l yield strength, piping deforms and leads to gross failure.

However, the deformation due to thermal or anchor displacement load is controlled by the amount of restrained pipe expansion or anchor displacement. Therefore, failure due to gross deformation need not be addressed for thermal or displacement loads included in Code Equation.10 or 11.

l Stress due to mechanical loads can be addressed using cumulative fatigue damage concept provided, NED0-21985 functional capability criterian is met. If the pipe stress exceeds functional capability criterian visual inspection of the affected piping is recommended to ensure no significant deformation has occurred. If piping passes visual inspection as well as the low cycle fatigue evaluation method given in Section 5.0, then the piping is considered adequate. The stress range of the mechanical load should be combined with the stresses in Code equation 11 and checked using equation 7.

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

f SARGENT & Ll)NDY

$ E N j[7[,,E R S Acc. No. EMD-061058 i i Page 11 I

7.0 CONCLUSION

The low cycle fatigue evaluation criterion presented in l

Section 5.0 is recommended for the assessment of piping for unanticipated loads. Further, if the stresses due to mechanical 4

j loads exceed 79-14 allowables, visual inspe'etion of the af fected piping is recommended to ensure no significant deformation or cracking of piping has occurred.

The above criterion is recommended for the assessment of  :

1 piping in operating plants for unanticipated loads and is not

intended for use in design basis calculations. >

l

8.0 REFERENCES

8.1 Markl A.R.C., " Fatigue tests of Welding Elbows and i Comparable Double Miter Bends." Trans. ASME, Volume. 69,

! 1947, pages 864 to 879.

8.2 Mark 1 A.R.C., and George, H. H., " Fatigue Tests on Flanged Assemblies," Trans. ASME, Volume 72, 1950, pages 77 to 87.

8.3 Markl, A.R.C., " Fatigue Tests of Piping Components,"

Trans. ASME, Volume 74, 1952, pages 287 to 303 j 8.4 Rodabaugh, E.C., and Moore, S.E., " Comparison of Test Data and Code Methods for Fatigue Evaluation," ORNL-TM-3520 Phase Report No.115-10, November 1971.

~ 8.5 Rodabaugh, E.C.., and Yahr, G. T., " Comparison of Code Fatigue Evaluation Methods for Class 1 Piping with Class 2 or 3 Piping.

! 8.6 GE Report No. NED0-21985, " Functional Capability Criteria for Essential Mark II Piping," September 1978.

I

, -..,_,,.n--,. -

,,.rr. . . _ . , , - . , , . - , _ . _ _ . - _

- . .. - - . _ . . - _ . - . . . . -=.-. __ . . .. ~ _ . . -. - -. . .-.

SARGENT & LUNDY  !

{ ENGINEERS Acc. No. EMD-061058 5 l . CHICAGO j 4 [

APPENDIX A  !

I f

Example: Subsystem IMS-A2 LaSalle Unit 1 i 1MS-A2 is a small bore subsystem with 3/4 inch piping. It had t

i two snubbers (M-1302-23-140 and MS-C6-1005S) that exceeded the snubber i

. p testing acceptance criteria. The snubber M-1302-23-140 was locked  ;

rigid. The other snubber MS-C6-1005S had a drag force of 53 lbs.

which is -15% of its rated capacity. The piping was evaluated for the i 3

increased stresses resulting from the snubber failures. The stress at i one of the node points (50) exceeded the ASME Code Section III Subsection NC/ND Equation 11 piping allowable stress.

I The total number of thermal cycles to date is 85. There had been no earthquake. Other mechanical loads to date.are small and can be ignored. The defective snubbers are to be replaced prior to plant start-up. The following is the summary of the evaluation: i Prior to With ASME  :

Snubber Snubber Allowable ]

Failure Failure Stress i

l Code Equation 8 Stress psi 2802 2802 15000 r

l

{ Code Equation 10 Stress psi. 13307 47436 22500 Code Equation.11 Stress psi 16109 50238 37500 j Cumulative Fatigue Damage Calculation:

A. Fatigue Damage prior to Snubber Surveilence - Assessment

= 50238 psi Intensified Stress Level Allowable Number of Cycles = 2758 l

Actual Number of Cycles -

85 Usage Factor (85/2758) -

.031 4

i i.

I 1

l 1

s

,,,w. -. .,m .

,. . , , . , - . _ , . . _ _._yv.- _._. , , , , . m, ,.m_ ..._. _ . , ,.~,y , _- .. . _ . . _ . ,,

SARGENT & LUNDY EN EERS j1 co Acc. No. EMD-061058 Page 13 I

B. Fatigue Damage from now to 40 years - Design Basis Loads Intensified Stress Level = 16109 psi Allowable Number of Cycles - 813700 Actual Number of Cycles = 7000 Usage Factor = 0.009 C. Total Fatigue Damage:

The total cumulative Usage factor of 0.04 is less than the allowable value of 1. (However, if the usage factor exceeds 0.1, high energy line break must be considered.)

G

i i ' ' ' ' ' '

3ng , , ,

g i i l

l I

. ~

~ 200 cn a -

a 16n -

C C

,4

,,120 Sf

• 4 90,000 Nf-o.2 ,

e o ** *. .

c. 8 0 - - * *% * -

• e y *e e e s0 -

u

v. .*. ,, .

m * *

.e, e

.c 40 -

E e 30 = ' ' I' ' ' ' I ' ' ' l ' '

2 3 4 i 10 10 10 105 108

' y, Cycles to Failure, Nf --

jg as o

~ cn -

FIGURE 1 RESULTS OF HOMENT FATIGUE TESTS ON GIRTil BITIT WELDS, FROM REF. (3) b&

o w

CD

= =- -. - - = =

O 500 i i i i i i i I I Equo tion ( f), Cycles - to- Foilure Equation ( 5), with Factor-of-Sofety 10 0 of 2.0 on Str.ess -

m

. 50 -

Ss=0

~

r f I

E '

^ / Ss*Sh i I

l 3

E I r Equation (2 ), Code Stress Range Limits I a

un 10 for AIO6 Grade B up to 6SOF -

m Se=Sh = i l5 ksi per B31.1 5 -

l N

2?

8 I I .

l l l *5 10 lo t 10 3

10 4

lo s 10 8

Cycles FIGURE 2 COMPARISON OF EQ. ($) WIT 1100DE ALLOWADLE S'IllESSES FOR SA-106 CRADE D !!ATERI AI.

w c>

U L PROJE C T _ __

EMD FILE NO mp-Ob t O M h

b es et appl *C Abir tur th'% f alCulation?

1 4 is the Calculation techneCally adequate? 4 - - - - - - --

[,

if the design calCutat.on is not technically adequate recommend a Course of aClion N'% s it the ans*er to a is yes mdicate type of cal ulation (hand-prepared and or computer aided) and method of review y Hand-Prepared Design Calculation  ;

i

, 9 The review of this hand-prepared design calculation was accomplished 1 C1 y one or a combmation of the followmg (as checked)

C1 O A ,e.. . nt rep,eseniat..e sampie oi,epet.ie e '

a deiaaed re.se. oi ine origmai caicusaison 1 t aiC unabons O A ,e.ie= by an afte,nate simpie,eed o, app,o.imate A re..ew of the calCulat.on against a similar method of calculation Calculation pre.iously performed Computer Aided Design Calculation  !

I Program Acronym: _

Program No .

, Run i D Qg Yes No

_.. Date: ____

ves No a O O a ia. program apphCanie to, in.s caicuiai.on, e M ir the ans.er io d s no has ine program used m ihis

\, ,q [-@

o i;

ooes she input da,,

Conform ..tn the des.gn mout' calculation been ai dated m accordance with Appendia H CSD Standards and Procedures Manual

%s and es it on file m the EMD Library?

O O co.,eCriy devine ihe 0,ooiem eo,ine mai 's the 'e'e ao ' E MD _ _ _ .__ ._

-prngram algoothm1 "

x g O O i, tne ans er to d ,s no ha.e you. as ihe re.ie-er.

0 0 conta.nsoft.c.eneacco,a,,in oroduce. sulis with.n an. numcrical simitation of the program' .eritied ihrougn ene o,ogram eae iante oe conients ihai I the ABSOLUTE element time and date precede the t Aec the eesutts computer aided calculation time and date?

O Con % sient ..th tae mpui' C1 nOO l' a orogrammaeie Ca'cu'ator or mictocompu'er-0 0 correci and ..in.n the stated assumpi.ons and generated program was used m this analysis are

• you as ,e....., sai.st.ed .ein ine adeauacy o, lim.tations of the progr Am' the program file audst trail as used by the preparer 7

,1 O O i4.. tne 0,oosa.- em-n i-.. a ' rom Not Appi Cabic

• ~

• s

('SD e int #oHed teles?

. O O Are vou as ,c..e-er at ihes calCuiaisori satis' ed thai

()) e the methods used in the .alsdat.on adeQuategy .alidate 4 i 11 the anteer to (t es yes have the eacCute the program for this apphCatoon?

- "q seatrenrnt anit i nmpuer, output date been

[l Not ApphCable

.E - .ni inite.t in me n.itput ,

c' 1 MECHANICAL DEPARTMENT STANDARD STANDARD EMD CHECKLIST FOR REVIEW DOCUMENTATION OF CALCUL ATIONS FOR OFFICE USE ONLY - NOT TO BE SENT OUTSIDE OF SARGENT & LUNDY F ee.ois eesee