Pipe Stress Analysis SEMINAR COADE

'Of
..
I
iJ
•
•
•
''J
fJl
•
•
•
.0
Q
.c'
•
•
•
•
•
•
•
•
•
.C:,
o
0
•
•
•
Q
U
,
•
•
•
0 SECTION 1

\,

"

,

\

'J
•
@
•
,
•
•
,0 1.0
0 1.1
•
•
•
0 1.2
0
.CJ
•
•
• 1.3
1.4
•
1.5
•
•
•
•
•
0
•
0
e
•
•
0
0
0
•
•
• COADE Pipe Stress Analysis Seminar' Notes Section 1

Table of Contents

Introduction to Pipe Stress Analysis 1

Theory and Development of Pipe Stress Requirements 8

1.1.1 Basic Stress Concepts 8·14

1.1.2 3-D State of Stress in the Pipe Wall 14-15

1.1.3 Failure Theories 16

1.1.4 Maximum Stress Intensity Criterion 18·19

Fatigue Failure 20

1.2.1 Fatigue Basics 20

1.2.2 - Fatigue Curves 22

1.2.3 Effect of Fatigue on Piping 24-25

1.2.4 Cyclic Reduction Factor 25

1.2.5 Effect of Sustained Loads on Fatigue Strength 26

Stress Intensification Factors 28-33

Welding Research Council Bulletin 330 , ~ 34

Code Compliance ' 43

1.5.1 Primary vs. Secondary Loads 4345

1.5.2 Code Stress Equations - 4546

1.5.3 BS1.1 Power Piping 46

1.5.4 B31.3 Chemical Plant and Petroleum Refinery Piping 47

1. 5.5 ASME Section III, Subsections NC & ND (Nuclear Class 2 & 3) 49-50

1.5.6 B31.4 Fuel Gas Piping 51

1.5.7 B31.8 Gas Transmission and Distribution Piping Code 52

1.5.8-Canadian Z183/Z184 Oil/Gas Pipeline Systems 54

1.5.9 RCC-M C 55

1.5.10 Stoomwezen 56

1.5.11 Special Considerations of Code Compliance 56-59

\

1.5.12 Evaluation of Multiple Expansion Range Cases 59

\

i0

.. "

-. ··:i

• o

:)

19

• • • o

e

.(J

• • • • • • • • • .e)

o Q

• • o

Q

COADE Pipe Stress Analysis Seminar Notes

1.0 Introduction to Pipe Stress Analysis

In order to properly design a piping system, the engineer must understand both a system's behavior under potential loadings, as well as the regulatory requirements imposed upon it by the governing codes.

A system's behavior can be quantified through the aggregate values of numerous physical parameters, such as accelerations, velocities, displacements, internal forces and moments, stresses, and external reactions developed under applied loads. Allowable values for each of these parameters are set after review of the appropriate failure criteria for the system. System response and failure criteria are dependent on the type of loadings, which can be classified by various distinctions, such as primary vs. secondary, sustained vs. occasional, or static vs. dynamic.

The ASME/ANSI B31 piping codes are the result of approximately 8 decades of work by the American Society of Mechanical Engineers and the American National Standards Institute (formerly American Standards Association) aimed at the codification of design and engineering s tan d ai-ds for pi ping sys terns. The B31 press ure pi ping codes (and their successors, such as the AS:ME Boiler and Pressure Vessel Section III nuc1earpiping codes) prescribe minimum design, materials, fabrication, assembly, erection, test, and inspection requirements for piping systems intended for use in power, petrochemical/refinery, fuel gas, gas transmission, and nuclear applications.

Due to the extensive calculations required during the analysis of a: piping system, this field of engineering provides a natural application for computerized calculations, especially during the last two to three decades. The proliferation of easy-to-use pipe stress software has had a two-fold efi'ect:first, it has taken pipe stress analysis out of the hands of the highlypaid specialists and made it accessible to the engineering generalist, but likewise ithasmade everyone, even those with inadequate piping backgrounds, capable of turning out officiallooking results.

The in ten tion of this course is to provide the a ppropria te background for engineers entering the world of pipe stress analysis. The course concentrates on the design requirements (particularly from a stress analysis point of view) of the codes, as well as the techniques to be applied in order to satisfy those requirements. Although the course is taught using the CAESAR II Pipe Stress Analysis Software, the skills learned here are directly applicable to any means of pipe' stress analysis, whether the engineer uses a competing software program or even manual calculational methods.

Why do we Perform Pipe Stress Analysis?

There are a number of reasons for performing stress analysis on a piping system. A few of

these follow: '

1 . In order to keep stresses in the pipe and fittings within code allowable levels. 2 - In order to keep nozzle loadings on attached equipment within allowables of manufacturers or recognized standards (NEMA SM23, API 610, API 617, etc.).

1-1

• •

• a

~W~'t4~'~ric~"'~~¥if,~¥;tf':?W~~~~~~tt~~W~~~f~?f.?~~t;~~~~!

COADE Pipe Stress-Analysis Seminar Notes

3 ~ In order to keep vessel stresses at pi ping connections wi thin AS:ME Section VIII allowable levels.

4 . In order to calculate design loads for sizing supports and restraints.

5 In order to determine piping displacements for interference checks.

6 . In order to solve dynamic problems in piping. such as those due to mechanical vibration, acoustic vibration, fluid hammer. pulsation. transient flow, and relief valve discharge.

7 -) In order to help optimize piping design.

Typical Pipe Stress Documentation

Documentation typically associated with stress analysis problems consists of the stress isometric. the stress analysis input echo, and the stress analysis results, output. Examples of these documents are shown in Figures 1·1 through 1·5 on subsequent pages.

The stress isometric (Figure 1-1) is a ske tch, drawn in an isome tric coordinate system , which gives the viewer a rough 3·D idea of the piping system. The stress isometric often summarizes the piping design data, as gathered from other documents, such as the line list, piping specification, piping drawing, Appendix A (Figure 1·2) of the applicable piping code, etc. Design data typically required in order to do- pipe stress analysis consists of pipe materials and sizes; operating parameters, such as temperature, pressure, and fluid contents; code stress allowables; and loading parameters, such as insulation weight, external equipment movements, and wind and earthquake criteria.

Points of interest on the stress isometric are identified by node points. Node points are . required at any location where it is necessary to provide information to, or obtain information from, the pipe Stress software. Typically, node points are located as required in order to:

1 ~ define geometry (system start, end, direction changes, intersection, etc.)

2 - note changes in operating conditions (system start, isolation or pressure reduc-

tion valves, etc.) ,

3 define element stiffness parameters (changes in pipe cross section or material, rigid elements, or expansion joints)

4 • designate boundary conditions (restraints and imposed displacements)

5 specify mass points (for refInement of dynamic model)

6 ~ note loading conditions (insulation weight, imposed forces, response spectra, earthquake g-factors, wind exposure,snow, etc.)

7 . retrieve information from the stress analysis (stresses at piping mid spans, displacements at wall penetrations, etc.)

('7)-

• o

rJ

o • • ~

()

• •

~.

• •

• •

• •

(),

@ a ,.

• •


o

at • • •

o

o .()

• • • • • • •

• •

• C)

® Q

• • •

\~

• •

• 'I

The input echo (Figure 1-3) provides more detailed information on the system, and is meant to be used by the engineer in conjunction with the stress isometric.

The analysis output provides results. such as displacements. internal forces and moments, stresses, and restraint loadings at each node point of the pipe, acting under the specified loading conditions. CAESAR II provides results in either graphic or text format; Figures 1~4 and 1.:5 present stress and displacement results graphically. The output also provides a code check calculation for the appropriate piping code, 'from which the analyst can determine which locations are over stressed.

SSEMl

HateriaJ A18& Gr.B

SH @ 188 deg. = 16,588 psi SC @ 79 deg. = 2B,9ae psi T = 78B dBg. F. Flue Gas

P = 125 psi

Dia = 28" Std.Uall

Insul = 2" CalciUM Silicete

110

:I':~US

Tower- 128

'~:."'" Support Froo Tower

~3S

~ ..

COMputed therMal expansion or the vessel is ~t45 17.2&8E-6 in/in/deg.F. at a telnp of 828 deg.F. 1;

, Exchanger

Hode 188 is 28.88 rt. above vessel skirt

Disp. e lBB = (82B-7B)deg.F(17.Z68E-&)in/in/deg* (2B.BB)(12)rt.in/ft. = 3.121 in.

Disp. @ 12B =

(8ZB-7B)(17.26BE-6)(2B.88+6.S-15)(12) = 1.8 in .

Figure 1·1

"._:: .,

" '

" .: ~ .: - ..

r~ _.

G
•
e
~.
~
"'~':/
•
0
0
0
0
•
e
'''Ji
0
(_).
•
•
.A_a. •
.~ •
•
•
•
•
•
()o
I,:)
0
•
•
0
0
.1)
•
•
• COADE Pipe Stress Analysis Seminar Notes'

-< e~
I .. · .
sl :I .,; ... · . :r :3
an 5 e .. .. : : .. :
~ ~
\0-
J • .. ..
0 Iii ... ... · . :l :J
z
0: I ..
0: l:3 ... J .. ; .. : =-
>- ~a.c
; ~~l .. I ;- I
;! ~ :!
§ ::
i.l ... a ; ; ; E
::e !
1£ ~' 9!
5 \ ~=-= t ! .. ~ ; ~
• s \ u~ =
s· !~ ! § ~ s .. ~ ; ;
~~ .1
~~ ~ ~ .. ; ~ ~
! § !
..,'" .. .. 2 ..
85 j !l :! .: ;:; ~ ~
~~ ~ .. D ~ .. .. Ia .. .. ..
2 2 11: II: ~ 1i lil 00 R .. .. .. .. i 2 ~ : 2 ; .. .. ..
i!i3 ;i .. ::. ;i = :: .. .. ..
:3 .. ....
......
it;;: ~ ..
=>- J! .. 2 " .. .. " 0 a a .. " ..
) ~ ::. 2 22 . = = ;i :::: = ::
::l:l
!i J .II t I .. . ~ .. .. 4 .. -~ ~ . . ~
""III t ~ 2 fi 2 It :; :: ......
a ... .. ......
22 i
RI !!:J ...-J-,
=::0 :: ::: :: ....... ,... ..... : t t ~ ::: .. :: :; =
-a .!t .,. .". ..... - ..
;-
!fi ~:Q
,.l =-:!
.( !J I i i , ,~s, :I : : 1: !; , :S :S :S
~
~ _= .1;, .. ;; ··11:::=:: ~ ::: ~.::: ? ? ::: ? .. .. : . ..
e~l , 'T'l' , , , • , .. 'I' ...:.. "t 'II
..I
~ ,. ...
"''''1
~ ~;S
......
.c ......
r=; 1IOi1l~ ; ; ; ~ a .. ~ ;
~~j ~ HS;SEE i i :: a .. 3 E :! § ~
;:::t;; - . ~ .. -
III
..... I
:a J .. ~
~i : ;:: .... t t
. .. , ! i1a i a :I ':
.. .. .. ....... ,. . UI ,. • ,... .. .. ,. ..... ~
Z -cz
0 Ii .c.:IIi .... :5 i .: e e ....
i! . ,. :I ~
a .,... ...... - '" II1II .... ..
..
'" J
: j .. .. :, ; '" &: .: : ...... ~ ... :t- i ;t .. i 5 i 5 5 t ;t ~
... :;=:~~s ;:$! s ~ : i ~
::i; I :; .. c .. .. .. 'III 41 .... .....
iii .. !i!
'" .a Jl
!~ j 1 ;; i :a :a 11 : ::I
li ! !
~~ J; ~ ,:j.:i.i" J;
51111 =-= • 11 tI I $:;1 :~
c:~ ! : : A C <II <ill: "III;
<II ~~ <II
:3 • ii s as 1:1 ,13 3
Figure 1.2 @ e

• • • . :)

o

• • • o

o

tI..,'_)

• • • • • • • • • ,(:)


CAESAR II VERS 3.18 JOBNAME:SSEM1 PIPE OATA

DEC 10. 1992

3:05 am

Page 1

From 100 To 105 DY- 3.500 ft. PIPE

Dia- 20.000 fn. Wall- .375 in. Insul- 2.000 in .

GENERAL

T1- 700 F P1- 125.0000 lb./sq.in. Mat- (l)LOW CARBON STEEL E- 27.900.000 lb./sq.in. v - .292 Density- .2899 lb./cu.in.

RIGID Weight- 3.290.00 lb. DISPLACEMENTS

Node 100 DX- .000 in. DY- 3.121 in. DZ- .000 in. RX- .000 RY- .000 RZ- .000

ALLOWABLE STRESSES

B31.3 (1990) Sc- 20.000 lb./sq.in. Sh1- 16,500 lb./sq.ln.

- - - - ~ - - -.;. - - - - - - - - - - - - - + .. - _ - - - - - - - _ - - - - _ - - - - - - - - _ - --

From 105 To 110 DY- 3.000 ft. BEND at ~TOH end

Rad i us- 30.000 in. (LONG) Bend Angl em 90.000 Angl e/ Node @1- 45.00 109 Angle/Node @2- .00 108

From 110 To 115 DX- 12.000 ft. BEND at MTO" end

Radius= 30.000 in. (LONG) Bend Angle- 90.000 Angle/Node @1- 45.00 114 Angle/Node @2- .00 113

From 115 - To 120 DY- -15.000 ft. DISPLACEMENTS

Node 120 DX= FREE DY- 1.800 in. RZ- FREE

DZ- FREE RX- FREE RY- FREE

From 120 To 125 DY= -3.000 ft. BEND at "TOW end

Radius- 30.000 in. (LONG) Bend Angle= 90.000 Angie/Node @1- 45.00 124 Angle/Node @2- .00 123

From 125. To 130 OX- 35.000 ft. RESTRAINTS

Node 130 +Y

From 130 To 135 RESTRAI NTS

Node 135 +Y

OX .. 35.000 ft.

o From 135 To 140 OX- 35.000 ft.

RESTRAINTS

() Node 140 +Y

• • • Q

o

• • • t)

From 140 To 145 OX- 20.000 ft. BEND at ~TO" end

Radius- 30.000 in. (LONG) Bend Angle- 90.000 Angle/Node @!- 45.00 144 Angle/Node @2- .00 143

From 145 To 150 DY- -12.000 ft. RESTRAINTS

Node 150 ANC

Figure 1-8

1-5


•

CASE 3 (EXP)D3=DI-DZ

FIU::SSDl1

DEC i.l~ 1Z:17a~

'11=+,1%_ QUIT

HODES

o·

" ... ' .

(

2~

NODE" 123

OVRSm MAXSTR

BNDING TORS AXIAL STRESS

SYMBOL BHDIHG TORS AXIAL STRESS

o

o

•

•

COLOR BNDING TORS AXIAL STRESS

o

OUEJ!STRESSED NODES

•

•

Figure 1-4

•

•

FILE:SS£nl

DEC 1.199Z 1Z:19am

RESE! QUI!

HODES DEnT SPECFY MGHtF GROU ClILORS

•

•

•

OR IGt'tL BLANK

•

Cl

'iGk'I~'1 HRDCPY

o

•

•

o

HODE" 1Z5

HA)L DISPS. X

o

•

•

•

CD

()

• • •

o

e

• • •

o

o .C'.)

• •


What are these Stresses?

The stresses calculated are not necessarily real stresses (such as could be measured by a strain gauge, for example), but are rather "code" stresses. Code stress calculations are based upon specific equations, which are the result of 8 decades of compromise and simplification. The calculations reflect:

1 Inclusion or exclusion of piping loads, based upon convenience of calculation or selected fa'rure. In fact the result may not even represent an absolute stress

value, but rather a RANGE of values. '

2 • Loading type - these are segregated, and analyzed separately, as though they occur in isolation, even though they actually are present simultaneously.

3 Magnification, due to local fitting configuration, which may in reality reflect a

decrease in fatigue strength, rather than an increase in actual stress. -

4 . Code committee tradition - every code is a result of a different set of concerns and compromises, and therefore may appear to be on a different branch of the evolutionary ladder. Because. of this, every code gives different results when calculating stresses.

• A summary of significant dates in the history of the development of the piping codes is presented below:

• • • • • •

.. ~)

o Q

• • • o

o o

• •

"

1915 1926

1955

1957 1968

1969 1971 1974

Power Piping Society provides the first national code for pressure piping. The American Standards Association initiates project B31 to govern pressure piping.

Markl publishes his paper "Piping Flexibility Analysis", introducing piping analysis methods based on the "stress range".

First computerized analysis of piping systems.

Congress enacts the Natural Pipeline Safety Act, establishing CFR 192, which will in time replace B31.8 for gas pipeline transportation.

Introduction of ANSI B31. 7 code for Nuclear power plant piping. Introduction of ASME Section III for Nuclear power plant pi ping. Winter AddendaB31.1 moves away from the separation of bending and torsional moment terms in the stress calculations and alters the in~ensi.

fication factor for moments on the branch leg of intersections. '

1978 ANSI B31.7 is withdrawn,

Welding Research Council Bulletin 330 recommends changes to the B3l.1. B3L3, and ASMEIII Class 2 and 3 piping codes.

1987


1.1 Theory and Development of Pipe Stress Requirements

1.1.1 Basic Stress Concepts

Normal stresses: Normal stresses are those acting in a direction normal to the face of the crystal structure of the material, and may be either tensile or compressive in nature. (In fact, normal stresses in piping tend more to tension due the predominant nature of internal pressure as a load case.) Normal stresses may be applied in more than one direction, and may develop from a number of different types of loads .. For a piping system, these are discusse~ below:

Longitu~nal stress: Longitudinal, or axial, stress is the normal stress acting parallel to the longi tudinal axis ofthe pipe. This may be causedby an internal force acting axially within the pipe:

----- FAX

---------- __ ..

--

SL = Fax/ Am
Where:
SL = longitudinal stress, psi
Fax = internal axial force acting on cross-section, lb
Am = metal cross-sectional area of pipe, in2
= mdo2 ~ dj2) /4
= 1t dm t
do = outer diameter, in
di = inner diameter, in
dm = mean diameter, = (do + dj) /2 1-8

G
·G
0
O·
,
·
•
"
()
0
•
•
~
,"'l-
.....
~
(J.
•
•
": .. - ... •
"
~ ;.
•
••
•
•
•
C-l
0
(,)
•
•
0
0
0
•
•
• o 8

-. • .,

-8 IS>

• • • o

e

.... )

• • • • • • • • •

tf.J

Q o

• • • If)

CJ

• • •


A specific instance oflongitudinal stress is that due to internal pressure:

Figure 1-7
SL = PAil Am
Where:
p = design pressure, psig
A = internal area of pipe, in2
I
= 1t di2 I 4 -:

Replacing the terms for the internal and metal areas of the pipe, the previous equation may be written as:

8L = P dj2 I Cdo2 - dj2), or:

For convenience, the longitudinal pressure stress is often conservatively approximated as: 8L = P do I 4 t

Another component of axial normal stress is bending stress. Bending stress is zero at the neutral axis of the pipe and varies linearly across the cross-section from the maximum compressive outer fiberto the maxim urn tensile outer fiber. Calculating the stress as linearly proportional to the distance from the neutral axis:

Variation in Bending Stress Thru Cross Section

( r--------c=7 Max compressive stress 1/2 max compressive stress

I Neutral Axis

M \ Zero bending stress

\ 1/2 max tension stress

stress

Figure 1-8

1-9

COADE Pipe Stress/Analysis Seminar Nares

Where:

Mb = bending moment acting on cross-section, in-lb

c = distance of point of interest from neutral axis of cross-section, in

I = moment ofinertia of cross-section, in4

Maximum bending stress occurs where c is greatest - where it is equal to the outer radius:

Smax =

Where:

Ro = outer radius of pipe, in

Z = section modulus of pipe, in3

= l/Ro

Summing all components of longitudinal normal stress:

SL = Fax / Am + P do I 4 t + Mb / Z

Hoop stress: There are other normal stresses present in the pipe, applied in directions orthogonal to the axial direction. One of these stresses, caused by internal pressure, is called hoop stress. This stress acts in a direction parallel to the pipe circumference.

Figure 1- 9

The magnitude of the hoop stress varies through the pipe wall and can be calculated. by Lame's equation as:

SH = P (rj2 + rj2 ro2 / r2) / Cro2 - rj2)

1-10

Q
0
e
().
~" ......
•
0
0
0
•
•
e
0
0
CJ.
•
•
.- •
", •
•
•
•
•
•
(~
@ 'I
0 I
• I
I
• I
0 I
!
o
()
•
•
• COADE Pipe Stress Analysis Seminar Notes

• Where:

• • o

Q

• • • o

C)

.()

• • • •

• •

• • • ~)

e o

• • • Q

o

• • •

SH = hoop stress due to pressure, psi

q = inner ramus of pipe, in

ro = outer radius of pipe, in

r == radial position where stress is being considered, in

The hoop stress can be conservatively approximated for thin-wall cylinders, by assuming that the pressure force, applied over an arbitrary length of pipe, I (F = P di 1), is resisted uniformly by the pipe wall over that same arbitrary length (Am = 2 t 1), or:

SH = P di 1/ 2 t I, or:

SH == P d, / 2 t, or conservatively:

SH == P do / 2 t

Radial stress: Radial stress is the third normal stress present in the pipe wall. It acts in the third orthogonal direction, parallel to the pipe radius. Radial stress, which is caused by internal pressure, varies between a stress equal to the internal pressure at the pipe's inner surface and a stress equal to atmospheric pressure at the pipe's external surface. Assuming that there is no external pressure, radial stress may be calculated as:

= -p

Figure 1·10

Where:

SR == radial stress due to pressure, psi

Note that radial stress is zero at the outer radius of the pipe, where the bending stresses are maximized. For this reason, this stress component has traditionally been ignored during the stress calculations.

Shear stresses: Shear stresses are applied in a direction parallel to the face of the plane of the crystal structure ,of the rna terial, and tend to cause adjacent planes of the crystal to

1·11

,

,iCOADE Pipe Stress Analysis Seminar Notes

slip against each other. Shear stresses may be caused by more than one type of applied load.

For example, shear stress may be caused by shear forces acting on the cross-section:

~l \\

S h eo r Distri buti a n Profile

------......)

----------~ !

__________ ! _tv

IMAX

Figure 1-11

'tmax

VQ/Am

==

\Vhere:

'Cmax ==

maximum shear stress, psi

v == shear force, lb

Q shear form factor, dimensionless (1.333 for solid circular section)

These shear stresses are distributed such that they are maximum. at the neutral axis of the pipe and zero at the maximum distance from the neutral axis. Since this is the opposite of the case with bending stresses, and since these stresses are usually small, shear stresses due to forces are traditionally neglected during pipe stress analysis.

Shear stresses may also be caused by torsional loads:

Figure 1~12

'tmax ==

0' • C

o·

• e:>

(0
C
•
•
0
I.)
0
C) •
. ~
•
•
•
:.~ • • • • • •

(--)

0-----0

()

o • • o

o c

• • •

COADE Pipe Stress Analysis Seminar Notes i

• Where:

• • o

e

• • • o

e

.()

.-

• • • • • •

•

= internal torsional moment acting on cross-section, in-lb

c

= distance of point ofint.erest from torsional center (intersection of neutral axes) of cross-section, in

,/

R

= torsional resistance of cross-section, in4

= 21

Maximum torsional stress occurs where c is maximized - at the outer radius:

'tmax =

Summing the individual components of the shear stress, the maximum shear stress acting on the pipe cross-section is:

'tmax =

Example Stress Calculations:

As noted above, a number of the stress components described above have been neglected for convenience during calculation of pipe stresses. Most U.S. piping codes require stresses to be calculated using some form of the following equations:

Longitudinal stress: SL = Mb I Z + Fax I Am + P do I 4 t

Shear stress:

=

Hoop stress:

=

Cross sectional
properties: Piping loads:
do = 6.625 in Bending moment (Mb) = 4247 ft-Ib
di = 6.065 in Axial force (Fax) = 334881b\
\
t = 0.280 in Pressure (P) = 600 psi
Z == 8.496 in3 Torsional Moment (MT) = 8495 ft-lb
Am = 5.5813 in2 • Calculations are illus tra te d for a 6- inch nominal diameter, standard wall pipe (assuming the

~ piping loads are known):

e o

• • • e

o

• • •

1-13


Longitudinal stress:

SL = '4247 x 12/8.496 + 33488/5.5813 + 600 x 6.625/4 (0.280)

== 15547 psi

Shear stress:

l' == 8495 x 12/2 (8.496) = 5999 psi

Hoop stress:

SE " 600 x 6.625/2 (0.280) == 7098 psi

1.1.2 3MO State of Stress in the Pipe Wall

During operation, pipes are subj ect to all of these types of stresses. Examining a sm all cube of metal from the most highly stressed point of the pipe wall, the stresses are distributed as so:

SR

M ~ /SH

T "( tii------------p ~ T s~:. t T

-- ------ -- : Sl

\ ITIITDl__________ ) ~ _ I· .. , -

I

SH S

R

Figure 1-13

There are aninfini te number of orientations in which this cube could have been selected, each with a different combination of normal and shear stresses on the faces. For example, there is one orientation of the orthogonal stress axes for which one normal stress is maximized, and another for which one normal stress is minimized - in both cases all shear stress components are zero. In orientations in which the shear stress is zero, the resulting normal components of the stress are termed the principal stresses. For 3·dimensional analyses, there are three of them, and they are designated as 51 <the maximum), 82, and 83 (the minimum). Note that regardless of the orientation of the stress axes, the sum of the orthogonal stress components is always equal, i.e:

5L + SH + SR = 81 + 82 + 83

The converse of these orientations is that in which the shear stress component is maximized (there is also an orientation in which the shear stress is minimized. but this is ignored since the magnitudes of the minimum and maximum shear stresses are the same); this is appropriately called the orientation of maximum shear stress. The ma:rimum shear stress

G .. C

o

•
•
.,
;)
>3
C=-:.
•
•
- •
,
.--.
•
•
•
•
•
(l
',-
0
0
•
•
0
o
U
•
•
• 4J tiJ

• • • o

o

• • • C)

o

.(_)

• • • • • • • • •

.0

COADE PiP!; Stress Analysis Seminar Notes

in a three dimensional state of stress is equal toone-halfofthe difference between the largest and smallest of the principle stresses (81 and 83).

The values of the principal and maximum shear stress can be determined through the use of a Mohr's circle. The Mohr's circle analysis can be simplified by neglecting the radial stress component,therefore considering a less complex (i.e., 2~dimensional) state of stress. A Mohr's circle can be developed by plotting the normal vs. shear stresses for the two known orientations (i.e., the longitudinal stress vs. the shear and the hoop stress vs. the shear), and constructing a circle through the two points. The infinite combinations of normal and shear stresses around the circle represent the stress combinations present in the infinite number of possible orientations of the local stress axes.

A differential element at the outer radius of the pipe (where the bending and torsional stresses are maximized and the radial normal and force-induced shear stresses are usually zero) is subject to 2·dirnensional plane stress. and thus the principal stress terms can he computed from the following Mohr's circle:

T

The center of the circle is at (8L + 8H) / 2 and the radius is equal to [[(8L - 8H) /2)2 + 12 ]112. Therefore. the principal stresses, 81 and S2. are equal to the center of the circle. plus or minus the radius, respectively. The principal stresses are calculated as:

Sl = (SL + SH) /2 + [ [(SL . SH) / 2J2 + t2 ]1/2 and

As noted above, the maximum shear stress present in any orientation is equal to (Sl . 82) / 2, or:

o --82 = (SL + SH) / 2· [ [(SL - SH) / 2J2 + t2 ]112

e

• • • o

o

• • • (It

tmax =

[(SL • 8H)2 + 4 t2 ]112 2

COADE Pipe Stress Analysis Seminar Notes .....

1.1.3 Failure Theories

To be useful, calculated stresses must be compared to material allowables, Material allowable stresses are related to strengths as determined by material uniaxial tensile tests

. ,

therefore calculated stresses must also be related to the uniaxial tensile test. This relationship can be developed by looking at available failure theories.

\'

II> II> <II ... ..... Vl

Strain

Tensile Test Results

Unixiai TensileTest Machine

Tensile Test Specimen

There are three generally accepted failure theories which may be used to predict the onset of yielding in a material:

1 - OCTAHEDRAL SHEAR, or VON MISES THEORY 2 - MAXIMUM SEEM or TRESCA THEORY

3 - :MAXIIvIUM STRESS or RANKINE THEORY

These theories relate failure in an arbitrary three dimensional stress state in a material to failure in a the stress state found in a uniaxial tensile test specimen, since it is that test that is most commonly used to determine the allowable strength of commonly used materials. Failure of a uniaxial tensile test specimen is deemed to occur when plastic deformation occurs; i.e., when the specimen yields.

0
.,
C
O·
.~
•
,.,
0
0
•
•
n
i)
e
l1.
•
•
._ •
.. •
•
•
•
•
•
c~
Q
0
•
•
0
r)
U
•
•
• Q g _.

• •


/

The three failure theories state:

Octahedral Shear- Von Mises Theory:

"Fail ure occurs when the octahedral shear stress in a body is equal to the octahedral

C) shear stress at yield in a uniaxial tension test."

<.) The octahedral shear stress is calculated as:

• • • o

e

In a uniaxial tensile test specimen at the point of yield: 81 = SYieJd; 52 = 83 = 0

Therefore the octahedral shear stress in a uniaxial tensile test specimen at failure is calculated as:

.() 'toct = 1/3 [ (5Yield ·0)2 + (0 - 0)2 + (0 . 5Yield)2 ]112

• • • • • • • • •

r/:)-

-~ Q

• • o

o 'J o

• • e

= 2112 x SYield / 3

Therefore, under the Von Mises theory:

Plastic deformation occurs in a 3-dimensional stress state whenever the octahedral shear exceeds 2112 x Sr-ield / 3.

Maximum Shear Stress· Tresca Theory:

. "Failure occurs when the maximum shear stress in a body is equal to the maximum shear stress at yield in a uniaxial tension test."

The maximum shear stress is calculated as:

'tmax =

__ Ina uniaxial tensile fest specimen - at thepoin t of yield: 81 = 8Yield; 82 = 83 = 0

80:

'tmax =

(8Yield· 0) / 2 = SYield / 2

\

,

\

Therefore, under the Tresca theory:

Plastic deformation occurs in a 3-dimensional stress state whenever the maximum shear stress exceeds 8YieJd / 2.

1-17


Maximum Stress ~ Rankine Theory

"Fail ure occurs when the maximum tensile stress in a body is equal to the maximum tensile stress at yield in a uniaxial tension test."

The maximum tensile stress is the largest, positive principal stress, 81. (By definition, 81 is always the largest of the principal stresses.)

In a uniaxial tensile test specimen at the point of yield: 81 [' = 8Yield; 82 = 83 = 0

Therefore, under the Rankine theory:

Plastic deformation occurs in a 3-dimensional stress state whenever the maximum principal stress exceeds SYield.

1.1.4 Maximum Stress Intensity Criterion

Most of the current piping codes use a slight modification ofthe maxim urn shear stress theory for flexibility related failures. Repeating, the maximum shear stress theory predicts that failure occurs when the maximum shear stress in a body equals SYield/2; the maximum- shear stress existing at failure during the uniaxial tensile test. Recapping, the maximum shear stress in a body is given by:

1;max = (81 ~ S3) 12

For the differential element at the outer surface of the pipe, the 'principal stresses were computed earlier as:

81 = (SL + SH) 12 + [[(SL ~ SH)/2J2 + 1;2 J1I2

82 (or S3)

=

'As seen previously, the maximum shear stress theory s ta tes that during the uniaxial tensile test the maximum shear stress at failure is equal to one-half of the yield stress, so the following requirement is necessary:

tmax = [(St ~ SH)2 + 4 1"2 ]112 2

<

2

Multiplying both sides arbitrarily by two saves the time required to do two mathematical operations, without changing this relationship. Multiplying by two creates the stress intensity, which is an artificial parameter defined simply as twice the maximum shear stress. Therefore the Maximum Stress Intensity criterion, as adopted by most piping codes, dictates the following requirement:

[(SL ~ SH)2 + 4 -c2]112 < SYield

1-18

0
(I
G
u·
-
•
()
:.J
0
e
•
e
()
0
(\
'-~
•
•
•
,j
-, '.
•
•
•
•
•
(~
0
0
•
•
C
I,)
_u
•
•
• e ~

• • • o

e

• • • C'>

o

• • • • • •

• • •

.{.J

r.':'II. V

o • • • c

G

• • •


Note that when calculating only the varying stresses for fatigue evaluation purposes (as discussed in the following section), the pressure components drop out of the equation. If an allowable stress based upon a suitable factor of safety is used, the Maximum Stress Intensity criterion yields an expression very similar to that specified by the B31.3 code:

[ Sb2 + 4 St2 ] 112 < SA

Where:

Sb = longitudinal normal stress due to bending, psi

St = shear stress due to torsion, psi

SA = allowable stress for loading case, psi

Example Stress Intensity Calculations:

Calculation of stress intensity may be illustrated by returning to our 6-inch nominal diameter, standard wall pipe for which longitudinal, shear, and hoop stresses were calculated. Reviewing the results of those calculations:

Longitudinal stress: SL = 15547 psi

Shear stress:

=

5999 psi 7098 psi

Hoop stress:

=

Assuming that the yield stress of the pipe material is 30,000 psi at temperature, and a factor of safety of 213 is to be used, the following calculations m ust be rna de:

[(SL - SH)2 + 412 ]V2 < 2/3 x SYield, or:

[(15547 - 7098)2 + 4 x 59992]112 < 213 x 30000, or:

14674 < 20000

The 14674 psi is the calculated stress intensity in the pipe wall, while the 20000 is the allowable stress in tensity for the material at the specified tern perature. In this case, the pipe would appear to be safely loaded under these conditions.

1-19

COADE Pipe Stress/Analysis Seminar Notes

1.2 Fatigue Failure

The fail ure modes discuss ed above were sufficient to describe catastrophic failure based upon one time loadings. However, piping and vessels were also found to suffer from sudden failure following years of successful service. The proposed explanation for this phenomenon was fatigue failure of the material, resulting from propagation of cracks on the material crystal structure level due to repeated cyclic loading.

·i

1.2.1 Fatigue Basics

Steels and other metals are made up of organized patterns of molecules, known as crystal structures. However, these pa ttems are not main tained throughou t the s teel producing an ideal homogenous material, but are found in microscopic isolated island-like areas called a grains.

Inside each grain the pattern of molecules is preserved. From one grain boundary to the next the molecular pattern is the same, but the orientation differs. As a result, grain boundaries are high energy borders. Plastic deformation begins within a grain that is both subject to a high stress and oriented such that the stress causes a slippage between adjacent layers in the same pattern. The incremental slippages (called dislocations) cause local cold-working. On the first application of the stress, dislocations will move through many of the grains that are in the local area of high stress. As the stress is repeated, more dislocations will move through their respective grains. Dislocation movement is impeded by the grain boundaries, so after multiple stress applications. the dislocations tend to accumulate at grain boundaries, and eventually becoming so dense that the grains "lock up", causing a loss of ductility and thus preventing-further dislocation movement, Subsequent applications of the stress cause the grain to tear, forming cracks. Repeated stress applications cause the cracks to grow. Unless ab ated, the cracks propagate with additional stress applications un til sufficient cros s sectional strength is lost to cause catastrophic failure of the material. Figure 1-16 ill ustra tes this process.

.'

1-20

• • ",

@
c.,
•
•
--- •
-\ .....
•
•
•
•
•
(~
@
o
•
•
S
'.)
0
•
•

,

I

• •

•

o

•

•

•

•

•

• •

•

•

•

•

•

•

•

•

•

•

•

Molecular pattern in unstressed grain

• ._

Slipping of one molecular surface over another after first application of stress

-+:;-.tion

-+ §§§§§§§§ Slip"

\... Slipping of a second' molecular surface after a

second application of

. '\ stress

Dislocations beginning

to interact and tangle

~ ~

After many repeated applications of stress the dislocations are completely tangled and the grain is 'locked",

With another application of the stress, the grain 'tears" and a fatigue crack is initiated.

Figure 1·16

One eyel e

\

TEST LOADING CURVE

Tens il e Test Specimen

Figure 1·17

1-21


One important consideration is the fact that fatigue cracks usually are initiated at a free surface. Corrosive attack on a material often produces pitting ofmetaI surfaces. The pits act as notches and produce a reduction in fatigue strength. In those specific cases when corrosive attack occurs simultaneously with fatigue loading, a pronounced reduction in fatigue properties results which is greater than that produced by prior corrosion of the surface. When corrosion and fatigue occur simultaneously, the chemical attack greatly accelerates the rate at which fatigue cracks propagate.

Unfortunately, fatigue failures can occur even when the stress in a material is below the yield stress. This is because localized stress concentrations can cause plastic deformation in a relatively few grains despite the fact that the stress over a gross area of the section may be far below the material yield stress. If the section is subjected to a sufficient number of stress cycles, cracks can initiate in highly stressed grains and then propagate throughout the material, ultima tely resulting in a fatigue failure of the section as a whole.

The fatigue capacity of a material can be estimated through the application of cyclic extensive/compressive displacement loads with a uniaxial testmachine.ias shown in Figure 1·17.

S am ple results for typical ferrous rna terial ( with a yield stress of5 7, 000 psi) are shown below:

Applied Cyclic Cycles to
Stress (psi) Fa il ure
300.000 23
200.000 90
100.000 550
50.000 6.700
30.000 38.000
20.000 100.000 1.2.2 Fatigue Curves

A plot of the cyclic stress capacity of a material is called a fatigue (or endurance) curve. These curves are generated through multiple cyclic tests at different stress levels. The number of cycles to failure usually increases as the applied cyclic stress decreases, often until a threshold stress (known as the-endurance limit) is reached below which no fatigue failure occurs, regardless of the number of applied cycles. The endurance limit (for those metals that possess one) is usually quantified as the value of the cyclic stress level which may be applied for at least 108 cycles without failure. Typical ratios of the endurance limit to the ultimate tensile strength of various materials are 0.5 for cast and wrought steels; about 0.35 for several nonferrous metals such as nickel, copper and magnesium; and 0.2 to 0.3 for rough or corroded steel surfaces (depending on the degree of stress intensification).

An endurance curve for carbon and low alloy steels, taken from the AS1-IE Section VIII Division 2 Pressure Vessel Code is shown in Figure 1 ~ 18.

1·22

o·

"'":'" ....... J.

• 1$

o

o

• • e

o

'CJ.

• • -- .

'. •

• • •

(J. e

e e

• • C

o

o • •

•

(I)

f) -_.

• •

o

e

• • •

o

o .("--J

-'


• • • • • • • • • .c)

o

e

• • •

o

o

e

•

• a

-

1'.,1· ......

UI __ uq .. ",.

OIT "'1'"fO." ......., _,..~ .. ._ .......

,~~~~-LWU~ __ ~-L~~~ __ ~~~~~~~~~~ __ ~~~~~ CYCLES ,,.,

'.

"'. !>-llO.l DESI," F ... nGIIE CIIII\IES FGlI c.utIlON, LOW A1.I.GY, Stll:lES (x:o:. III1lM AU.OY STEW ""D ~,,~ TENSILE STULl fII. TDI P£1Uo TURts .OT OU;a:DI~ ~

Figure 1-18

Note that according to the fatigue curve, the IDa terial doesn't fail upon ini tialloading, despite enormously high stresses that appear to be well above the ultimate tensile stress of typiCal carbon and low alloy steels. The reasons for this are:

1

The highly stressed areas under fatigue loading 'are normally very localized. Catastrophic failure under one-time loading will normally occur only when the gross cross-section is overloaded.

2

Fatigue curves are usually generated through cyclic application of displacement, rather than force, loading. Displacement loads are "self-limiting". If a pipe is overloaded-with an imposed displacement, plastic stresses will develop, deforming the pipe to its displaced position. At that point there will be no further tendency for displacements to occur, and therefore no continuation of the load, or further deformation leading to catastrophic failure. In the case of an applied force (which is not a self-limiting load), deformation of the pipe does not cause the force to subside, so deformation continues until failure.

The stress shown in a fatigue curve is a calculated stress, based upon the assumption that Hooke's law is applicable throughout the range of applied loading; i.e., S == E e , where:

f.

3

E

modulus of elasticity of material, psi

=

e

strain in material, inJin

==

1-23

j;,i~&~,;;~f~iili£i~iii~~~~'f~~~~;jl~~:f:~WJ1B~~~%~~;E§;ff;i\:1i;~~'i~~f~!~~~~~i.jw1~;f!f!K;,':i"{~~;t~;i,~i~~@;:k~~~r'~ii;R!~!~~~

. ~ ,:" .". ~ _ '

:;.'h.>~ ..• ~.>'.:.~C"'::.~ ·~··j~!~\'L~~:t..:l~~~~;~)'.....~~.~~·'~~~~~~-i~·~:~~~/.,"J~~·::·_.~ ~,~;;"_~;;~~·~·~~~·i/~2~1~'~.i~.';:_~~~.::.~~.:~~"':-1:\- .. ~;~~·kt:··>~'i:;~~\·~i:";"Wtai,).{H,b.~~ ~~~Lo:~,_,p.~':L.: .. \~:::~:~;..t~~.i~~J.:.L. ~~';'~h.~i;K::"V.~~ ;'~ici.l6,

-' .


In reality, once the material begins to yield, stress is no longer proportional to the induced strain, and actually is much lower than that calculated.

1.2.3 Effect of Fatigue on Piping

A R. C. Markl investigated the phenomenon offatigue failure ofpiping during the 1940's and 1950'5, and published his results in papers such as "Piping Flexibility Analysis", published in 1955. He tested a number of configurations (straight pipe, and various fittings, such as pipe elbow, miter bend, unreinforced fabricated tee, welding tee, etc.) by using cyclic displacements to apply alternating bending stresses. Plotting the cycles to failure for each applied displacement, he found that the results of his experiments followed the form of fatigue curves.

16'

~ 41' f '

I (fYPJ rr P -a- -

~J Y ) , Range 01 imposed displacements to

~I--_":::-' -'1 impose complete stress reversal.

, Girth butt weld

61 I S~angeOlinplane/~ /Z

,_-.,J-. -------'. displacements' _ '

J!I---.,., Range of ouiplane

displacements

,-r--r'--r------, -a"""A,ange of inPlane/ ~ dlsplacements?- / 'Flange of outplane

I displacements

--r-------,-B,_Aange of inPlane/ 7

~ I disp[acement~ 'Range of outplane'

nlJ_ <' disp'ac,m,"s

Figure 1-19

If an initially applied displacement load causes the pipe to yield, it results in plastic deformation, producing a pre-stress in the system, which must be overcome by subsequent stress applications, resulting in lower absolute stresses during later load cycles. Because of the system "relaxation", the initial values of the thermal stress are allowed to exceed the material yield stress, with the aim being that the system "self-spring" during the first few cycles and then settle into purely elastic cycling. This "self-springing" is also called Elastic Shakedown. As shown in Figure 1-20, the maximum stress range may be set to 2SYieid (or more accurately, the sum of the hot and the cold yield stresses) in order to ensure eventual elastic cycling.

1-24

• •

o G

<OJ.

• • • • • • • •

~"')"

i

-, 0

() o

• • o

o o

• • •

Q


• •

o

e

• • •

o

e .(~)

• • •

•

•

•

•

•

• .c)

o

Q • • • e

• • • o

Stress

Time---+-

Figure 1-20

Based upon this consideration, the initial limitation for expansion stress design was set to the sum of the hot and the cold yield stresses - the maximum stress range which ensured that the piping system eventually cycled fully within the elastic stress range. Incorporating a factor of safety, this resulted in the following criterion:

SE <= F (SYe + 8Yh)

Where:

SE = expansion stress range, psi

F = factor of safety, dimensionless

8Ye = material yield stress at cold (installed) temperature, psi

8Yh = material yield stress at hot (operating) temperature, psi

1.2.4 Cyclic Reduction Factor

At some point, in the vicinity of7 ,000 cycles, the (SYe + 8Th) limitation intersects the fatigue curve for carbon and low alloy steel. The allowable stress range must therefore be reduced

to fit the fatigue curve for cyclic applications with 7,000 cycles or more: \

SE <= F «s-, + 8Th)

1-25


e

Where:

f = cyclic reduction factor, as shown in the accompanying table

CYCLIC REDUCTION FACTOR TABLE
Cycles N Factor f
;\ 1 7,000 1.0
"."\
'\ .
1\ 7,001 14,000 0.9
14,001 22,000 0.8
22,001 45,000 0.7
45,001 100,000 0.5
I 100,001 200,000 0.5
200,001 700,000 0.4
I 700,001 2,000,000 0.3 1.2.5 Effect of Sustained Loads on Fatigue Strength

o·

• e

o

o • • o

o e

';-~~

• •

• • • • • •

In almost all cases the material fatigue curves are generated using a completely alternating stress; i.e., the average stress component is zero. Research has shown that the magnitude of the mean stress can have an effect on the endurance strength of a material, the trend of which is shown below:

~ ,~ "3-<

~ ~ Sa from anduranca

::! as /' curve for completely ¥ alternating stress

e Yield

Mean Stress Axis

" Yield

~ a-""
..
..
~ ~~:~
..
Go
.s
;
'e
!
;
~IIII < a"IIIZ ( "'1111 < orlll4
! I
ro~ !o' la'
C ,tie s 10 foi 1111"
(61 For Desig

Figure 1-21

1-26

Tensile

• ()

I, e

o o

• • o

Q

-0

• • •

Q) e

• • •

o

e

• • •

o

()

e(J

• • • •

• • • • •

"c)

o (ji

• • • Q

Q • • •


/

Note that as the mean stress increases the maximum permissible absolute stress (Sa + 8m) increases, while the permissible alternating stress decreases. The relationship between the allowable a1 tern a ting stress and the average stress is described by the Soderberg line, which correlates fairly well with test da ta for ductile materials. The equa tion for the Soderberg line is:

8a(Allowed) = Sa(for R=-l) x ( 1 - Sm/SYield )

Where:
R = Sminl Smax
Sa = (Smax - Smin) 12
8m = (Smax + Smin) 12 Note that during the development of the ASME Boiler and Pressure Vessel Code Section III rules and procedures for analysis of nuclear piping, the Special Committee to Review Code Stress Basis concluded that the required adjustments to a strain-controlled fatigue data curve based on zero mean stress, occur only for a large number of cycles (i.e, N> 50,000 - 100,000) cycles for carbon and low-alloy steels, and are insignificant for 18-8 stainless steels and nickel-chrome-iron alloys. Since these materials ~~~the majority of the piping materials in use, and since most" cyclic loading events comprise much fewer than 50,000 cycles, the effects of mean stress on fatigue life are negligible for piping materials with ultimate strengths below 100,000 psi. For materials with an ultimate strength equal to or greater than 100,000 psi, such as high strength bolting, mean stress can have a considerable effect on fatigue strength and should be considered when performing a fatigue analysis.

For a piping application, the implication of the Soderberg line on the . fatigue allowable is implemented in a conservative manner. The sustained stress (i.e., weight, pressure, etc.) can be -considered to be the mean component of the stress range after system relaxation, and as such is used to reduce the allowable stress range:

SE <= F f (SYc + STh - Ssus)

1-27


1.3 Stress Intensification Factors

As noted previously, Markl's fatigue tests generated endurance curves for various fitting configurations, such as straight pipe, butt welded pipe, elbows, miters, welding. tees, unreinforced and reinforced fabricated tees, mostly using 4" nominal diameter, size-on-sise fittings. Mark! noticed that the fatigue failures occurred not in the middle of'his test spans, but primarily in the vicinity of the fittings, and in those cases, they also occurred at lower stress/cycle combinations than for the straight pipe alone.

Earlier theoretical work pointed to a possible explanation. It had been shown that elbows tend to ovalize during bending, bringing the outer fibers closer to the neu tral axis of the pipe, thus reducing the moment of inertia (increasing flexibility) and the section modulus· (increasing developed stress).

Ovolization of Bend

x

Section

Figure 1·22

The stress in tens ific a tion factors (the ratio of actual bending stress to the calculated bending stress for a moment applied to the nominal section) for elbows was known to be:

10 = 0.75/ h213

ij = 0.9 / h2l3

Where:

10 = out-of-plane intensification factor

ii = in-plane intensification factor

h = flexibility characteristic

= t R/r2

t = pipe wall thickness, in

R = bend radius of elbow, in

r = mean radius of pipe, in

.-

1-28

o o e

n-

,'- ~;,.'

- 0'

o

•
•
~
,oj
g
':~).
•
-
. ...._ •
_.
-".
•
.'
•
•
().
'.
•
0
0
•
•
C
o
0
-
•
•
• e (fJ

• • • o

o

• • • o

{)

e(~~1

• • • • • • • • • tC_)

o o

•


Markl found this to correlate fairly well with his test data and so adopted it. Tests on mitered bends correlated well with those for smooth bends, providing an equivalent bend radius R was used in the above equation for h. Marld's estimates ofequivalent bend radius are shown below:

lie = rCI + 0.5 sIr cot D) (for closely spaced miters)
Re = r(l + cot D) I 2 (for widely spaced miters)
Where:
Re = equivalent bend radius, in
s = miter spacing at the centerline, in
D = one-half of angle between cuts Marld found that the unreinforced fabricated tees could be modeled using the same formula as that for single (widely spaced) miter bends could be used, if a half angle of 45 degrees was used. This produces a flexibility characteristic of:

h = tlr

For butt welded tees (such as ANSI BI6.9 welding tees) Marld again adapted the bend equations, this time computing an equivalent radius (Re) and an equivalent thickness (te). Markl's equation for welding tees was:

h = c ( te R, I r2 )

'Where:

c = ratio of tee-to-pipe section modulii, dimensionless

= (tJt)3/2 (Markl's recommendation)
te = equivalent pipe wall thickness, in
= 1.60t (Markl's recommendation)
Re = equivalent bend radius, in
= I.35r (Markl's recommendation) •

• Inserting these values into the expression for h yields:

h = 4.4t1r

Q This is precisely the expression used today for ANSI BI6.9 welding tees.

o • • • Q

1-29

C'OADE Pipe Stress Analysis Seminar Notes

For reinforced fabricated tees, Mark! used the expression he had previously used for welding tees, with different equivalent wall thickness and bend radius:

h = c ( te RJ r2 )

Where:
c = (te/t)3/2 (Markl's recommendation)
te = t + tp
tp = thickness of reinforcing pad or saddle, in
Re '= r The following tables compare the stress intensification factors suggested by Markl's test results versus the values calculated with his equations (results are for 4" nominal diameter, standard schedule pipe):

Bend in-plane (ij)
tR/r2 Test Calculated
0.062 4.49 5.7428
0.210 2.17 2.5476
0.129 4.38 3.5238
0.320 2.02 1.9238
0.319 2.10 1.9286
0.316 1.90 1.9381
0.328 1. 70 1.8904
0.331 1. 53 1.8809
0.324 1.36 1. 9095
0.332 1. 28 1. 8762
0.328 1. 46 1.8904 -'

1-30

G
«t
0
O·
·
•
0
Q
0
•
•
e
0
0 I
I
'C), ,
•
•
•
_".- •
-.
•
•
•
•
•
()(J
0
0
•
•
e
0
0
•
•
• e o

- .•

• ..

o e

• • • o

e .r-~I W\-~

• • • • • • • • •

c, .... /

o o

• • • c

o

• • •


Un rei n forced tee (i 0) :
t/r Test Calculated
0.0390 11.04 10.84
0.0455 6.12 7.06
0.0947 2.95 4.33
0.1111 2.34 2.89 Reinforced tee:
in-plane (ii) out-plane (io)
tpad Test· Calculated Test Calculated
0.12 2.21 2.63 2.43 3.17
0.237 1. 78 1. 74 1. 83 1. 98
0.5 1.10 1.14 1. 08 1.18 These form ulas for intensification factors were adopted (and expanded) by the piping codes. Specific formulas and/or fittings recognized by the individual ASME/ANSI B31 codes are usually shown in Appendix D of those codes (see Figure 1~23).


I

/

APPENDIX D

FLEXIBILITY AND Sl'RESS INTENSIFTCATION FAcrORS

UBI.! 0-1'

FlEXIaIUTY FACTOR t.I.NO STRESS INTVlSIl'IC.l.TlOk FACTOR I

-' __ llI,llI)

OII\ ...... ~

i" i;

'W ..... __ ~J ...... r-. '~I, , .... >11

~.

"

tlIl ~,_.-..._

~ -<.r} n + ...... tt "'_ m, ('1, I'll, /111

._.

;m

.... , If

-l-;;;

..l." .~I:J.,~

~

I Jtt"~

$1 ............... "'"..; ... --J:='t('+t.M1J

(N_ Il>. "/, on

0.'

;;:;

I .. , .. r f --2--;;

0.'

~

!'>I ,..., ............ "-$101~

tn~ • ..."

'. l: ""',-

T, ~ 1,.1 r

,~ .... 12', (~), ,.." un, fill]

Q.' "'''

f

._

~

III ~_.,_, ......

I~ 'MtJII.,.... .,!I.IiIIIfMt

(NOIn '11. "'. lS), (1.21. I U~

1',' =s= =-~~"

~-;$' 11:lLl.J1-- .,.

.'" ._ T,

101 _,_, ...

Ito! I" .... '2). 1'1. 1111,' un

r 't

1111 l-.oI ............ _ ,~ 2: 0·""1

T, ~ 1.:1"1'

1_ Ill. "1. WI!

~.'

;;;;

( t.) f

1+;; i;'

-r=-, wtjCkG·"'~"'''1''''C """

'.l""O ..

r~ l U J

1_ .. 12., '.lJ. lUI. lUll

(01 _

1 __ 1

I ...... W. I<I.!U 11.211

f

,.,-

~

$IoN

"..., __ h,-;"

._t ~_I(_(m

__ Jo;oo,_," __ """"

. '

Gut" I ............

--_ .. _- """"

Woo _ ,.". , ...... A$MI .1 .. , ;.. ..... 1

-_-. .. _ .....

_.-_ .. __ ", __ Ul_IIO»

La L2

_'HI U f.J U

1-32

~~*'.)
00
g
0
o·
,
•
f.!l
0
0
•
•
0
,"\
· .,
S
!~~).
•
•
•
:~ .,.
•
.'
•
•
I:)
Q
0
0
•
•
0
0
c
•
•
• e
SJ
-.
;
•
•
0
0
•
•
•
0
0
.c)
•
•
•
•
•
•
•
•
•
.J_)
@
()
•
•
•
0
"0
•
•

Subsequent research has demonstrated that Markl's formulas, having been based on a limited number configurations (significantly having omitted reduced outlet tees) and disregarding any need to intensify torsional stress, are inaccurate in some respects.

The major problem with reduced intersections tees lies in the out-of-plane bending moment on the header. Stresses due.to these moments can never be predicted from the extrapolation of size-an-size tests. Figure 1-24 below illustrates the origin of this problem.

Mob

Mob

Size-an-size

Reduced Intersection

, Figure 1·24

Errors due to these moments can be non-conservative by as much as a factor of two or three. Furthermore, when the rlR ratio is very small. the branch connection has little impact on the header, so use oflarge stress in tens ifi ca tion factors for the hea der can produce unreasonably large calculated stresses.

R.W. Schneider of Bonney Forge pointed out this inconsistency for reduced branch connections. His paper on the subject states that the highest stress intensification factors occur when the ratio of the branch to header radii is about 0.7, at which point the nonconservativism (versus Markl's formulas) is on the order of two.

i from Markl

-1.0

0.7 1.0 rlR

Ratio of Actual i to Morkl's i vs Ratio of Branch to Header Radius

\.

Figure 1·25

1-33

;....;;....,.....::..i::.~. ~~ .;;.-~~~J/~~~ ~!.~~ ~~~ ~;..d al~.i~h;L~.~~~io_:~..:.,.....~~ ..... it:~~ :;.> ~':~ ~ ~~;; ~. : ~~:~ ,'~":':":"::-..;"':~. !S~ . .:..:';; =;;":_"~,.._,:g~.~~,-*~: ... ';"'~;':;~w.~~i,:j.:U~""~-"~·J i~~ ~"!"j ......... .,,_~~;-;.;.~ "~::" !..s·'_;"r·t;.;~_~ .;,,:,.;.:.~O:"" ... ~; .;_ijJ ~·;~·~j:!t·;~~l~:~·~.~,J';~~i..i:.~~~ '~ '!.i-J1. \,j:'

- .. _-. -


1.4 Welding Research Council Bulletin 330

The Welding Research Council's Bulletin 330, "Accuracy of Code Stress Intensification Factors for Branch Connections" documented a major attempt to re-assess the existing code requirements for the intensification of stresses at tees and other branch connections. The difficulty of this task was summed up in the bulletin by author E. C. Rodabaugh, who stated:

'We would rate the relative complexity of i-factors for pipe, elbows and branch connections by the ratios 1:5 :500. These cornmen ts on relative com plexi ty, we think, are relevant a t this point because at least some readers will be looking for simple answers to what they perceive to be a simple subject. They will not find any simple answers in this report."

,

Summarizing the findings ofWRC 330 in order of increasing importance:

1) The following note should be added with regard to branch connection flexibili ties:

"In piping system analyses, it may be assumed that the flexibility is represented by arigidjointat the branch-to-runcenterlinesjuncture. However, the Code user should be aware that this assumption can be inaccurate and should consider the use of a more appropriate flexibility representation."

2) ASME 2/3 and B3l.1 users can use the "Branch Connection" expressions for unreinforced fab ric a ted tees whenever rlR < 0.5. (Markl's formulas specified that the same stress intensification factor be used on both the branch and header legs of a tee, regardless of relative sizes. The codes noted above permit the reduction ofthe stress intensification factor at the branch forrela tive diameters. CAESAR II automatically considers the effects of reduced in_tersections on the stress intensification factors for these codes unless directed otherwise by the user through the setup file.) .

3) B3l.1 erred when including the calculations for branch connection stress intensification factors; instead they should have included the calculations as they appeared in AS~ III. (Further clarification of this note is given in note 10 herein.)

4) B31.3 should include the stress intensification factors for branch connections as per AS:ME III. (B31.3 uses Markl's original formulas, thus specifying the same stress intensification factor for both the branch and header of a tee, regardless

of relative sizes.) .

5) B31.3 should intensify the torsional moment at branch connections, with the torsional intensification factor estimated as: it = (rlR)io'

6) B31. 3 should eliminate the use of ii = O. 75io + 0.25 for branch connections and tees. It can give the wrong relative magnitude for header moments, and may underestimate the difference between Mo and Mi for rlR ratios between 0.3 and 0.95, and perhaps over-estimates the difference for rlR ratios below 0.2 and for rlR = 1.0.

0
•
0
n·
,., .... J'
•
Q
o
C'
•
•
~
()
e
C~
•
•
•
",
•
•
•
•
•
/J
",
0
G
(;)
•
•
0
0
0
.
•
•
• ~


9 '. •

• o

e

• • •

7)

B3l.3 and B3l.l should add restrictions to the stress intensification factor tables indicating that they are valid for RIT < 50.

8)

The codes should add notes that indicate that the stress intensification factors are developed from tests and/or theories based on headers being straight pipe with about two or more diameters length of ' pipe on either side of the branch.

9)

The codes should also add notes to indicate that for branch connections/tees the stress intensification factors are only applicable where the axis of the branch pipe is within 5 degrees ofnonnal to the surface of the header pipe.

10)

The stress intensification factors for unreinforced fabricated tees, weldolets, and sweepolets should be changed to:

o For CrlR) < 0.9:

• • • • • • • • • .r:_)

Q e

• • •

c

" •

• •

Ib = 1.5(RlT)213 (rlR)1/2 (rlr p), with ib(tIT) > 1.5

For (rlR) = 1.0;

0.9 (RIT)2/3 Crlrp), with ib(tIT) > 1.0

And:

lr =

0.8 CR/T)2/3 (rlR), with ir > 2.1

'Where:

intensification factor for branch (to be linearly interpolated for rlR ratios between 0.9 andl.0)

R = mean radius of header pipe, in
T = thickness of header pipe, in
r = mean radius of branch pipe, in
rp = outer radius of branch pipe, in
t = thickness of branch pipe, in
lr = intensification factor for run (header) pipe \ Additionally, if a radius of curvature r2 is provided at the connection, which is not less than the larger of tI2, (Tb'+ Y)/2. or T/2, then the calculated values of ij, and ir may be divided by 2.0, but with the restriction that ib>1.5 and ir>1.5.

Also, where reduced outlets are discussed, branch ends should be checked using Z = p (r2)t and i(tIl') in place ofi, with i(tIT) > 1.0.

1-35


11) There was not sufficient data available on reinforced fabricated tees for Rodabaugh to make any definitive recommendations regarding them. Rodabaugh does however suggest that the normal usage whereby the width of the pad is taken to be at least equal to the radius of the nozzle should be observed even though not explicitly directed by the code ...

12)

For tIT ratios of about one or more, stresses tend to be higher in the header. and are fairly independent of the wall thickness of the nozzle. As the tIT ratio gets much smaller than one, the largest stresses shift to the branch. (This finding originally came out of the research for WRC 297.)

,

'·1 \'

Comparisons ofWRC 330's proposals for stress intensification factors for various types of tees, versus B31.3 calculated values are shown on the following pages.

·6
•
..
~.
v., j
•
(j)
e
0
•
•
"
',j
Q
C~
•
•
.- •
e:
•
•
••
•
•
(J
"0
G
0
•
•
0
,."\
. ..J
i.;)
•
•
• G o

• • •


!iQ INTERSECTION RADIUS

~831.3~ VS. "WAC 330~ UNREIN FORCED, FABRICATED TEE STRESS INTENSIFICATION FACTORCOMPARISON

o

@ • • • o

o

.(~)

• • • •

• • • • • .c:)

13 Q

• • •

" o

• • • Q

HEADER NOM

1 48. ... .1.

3 4e.

;. 41l.

-~ 4;4

, 41l. I 41!.

'il. 411,

5 4~1 5 49t 5 "I~

6 411.

~ 40.

6 411.

~ 48. 6 40.

8 -4a. a 4B. B. 411. g ~8, 8 41.

10 n. 16 U. IB 4B. U ~Il. !I! 4B.

12 ~e, l2 40.

12 ~el 12 40. li 40.

H 411. 14 48. l' n. H 40. 14 4B.

BflANCH WRC

SCH

40.

1 41l. t 40.

4B. 2 _48. !- 40t

I 411.

2 ~Il. 3 48.

I 411.

2 41l. l 4~~ ., 48.

s 41l.

2 4~. 3 40. 4 U.

5 41t 6 4~.

s 40. 4 48. 5 43.

., 4e.

5 4e.

b 46.

a 48. U! 42. ,

5 U, b n.

B 48. I 16 n. -

12 48.

tJ u. e 4e. ll'> u. 12 U. 14 40.

330 b

40. I 2.U3

I:.: (.aS4 3.3~'1

-831.3-·

i ib iob

2.874 2.~33

1.769 3. lSI!

i.769 ~.3sq

3.479 2.960 l.~8B

4.769 2.B60 3.488

3.48a 2.86!! 3.488

3.416 4.M12 5.69. J.992

3.349 ~.~89 5.5BI! 6.359 .11.255

4.m S.444 fl. 2112 6.9JQ 4.540

!'.127 S. He 6.592 7.2111 A,94'1

5.642 6c.2H 6.81j4 7.875 S.2B"

6. I,m 6.69a 7.S4!! 8.443 5.523

6,383 7.382 8.16b B.56'? -5.5Q'1

_).1611 3.169 :;'.lb9 3.169

3.441 3.441 3.44J 3.441 3.441

1.6~,5 3.b55 -3.655' 3.655 s.sss

3.9bl 3.901 3.961 3.961 3..'W

-4.213

4.213 4.213 4.zn

- 4.213

4.392 4.3n ~.392 ~.3n ~.3n

~.458 4,458 ~. 4511 4.458· 4,453

4.255 '4.255 4.255

(.5~8 4.541l 4.5411 ~.54~ 4.5411

4.9'~ 4.949 4.949 4. 9~9 4. '?49

3.892 3.8'i2 3.g~2 3.8n

4.255

4.255

5.284 5.284 5.284 5.284 5.284

5.523 5.523 5.523

~.523

5.523

5.5~9 5.599 5.5'19 5.599 5. ~j!19

i ib

• 330 b .853

1.82B .7S~ .b17 .541 .ae9

.b79 .824

.822 1. 0118

.IIU .13m

.821 uu

.928 .b77 .557 .8H

.S16 .671 .5S!! .528 .895

.747 .669 .6J2 .535 .n:

.728 .666 .582 .52~ .795

.b97 .699 .545 .51Q .795

330b '330h

1.0811 2.162

.822 2.m

I.BU 2.986

1.139 .B31 .684 I.BSIl

J .271 .1127 .76.) ,669

1.0B9

1.1!1~ .8:H .732 .656

I.BII0

.954 .831 .75J .b86

1.81111

.1136 .839 .768 .671

I.B60

.'HS .837 .732 .654

1.ge@

,877 .767 . bab .653

1.11118

WRC

-931.3-

_ib

i ih iOh 1 330 h

2.074 2.433 .95'

2.769 3.359

2.769 3.359

2.J08 2.868- 3.488

2.1116 _ 2.86P 3.488

3.893 2.868 3.488

·2.m 2.I&e 2.665 :;.4611

2.1118 2. U@ 2.342 3.948

3.16'i 3. J69 3.1£:9 S.l~!!

'3.441 3.4~1 3.HI 3.44J 3.441

3.b55

3.655 3.655 3.655

~.~bl- 3.9(,1 3.96l '3.961 3.%1

4.213 4.213 4.213 4.213 ~.2!3

4.3'1:: 4.392 4.392 4.392 ~. 392

4.45ft ~.458 ~.4511 4.458 4.4515

3.8~2 3.81?2 3.8'12 3.892

'.255 4.2SS 4.255 4.255 4.255

4.548 4.549 4.548 454{1: 4.S4B

4.949 4.949 4.9411 ~.9~9 4.9~9

5.284 5.284 5.284 5.284 5.284

5.523 5.523 5.~23 5.523

5.5911 5.5911 5.599 5.599 5.599

1.319 .927

1.362 J .302 .9~5

1.569 1.:58Q 1.189

.916

1.639 1.639 1. 478 1. n:2

.91B

].74l 1.141 1.349 1.983

.986

1.9B6 1.754 J.410 1.178

.911

2.@8b J.7:56 1. .6B 1.122

.997

iOh 330 h 1. 125

.764 .67e • bill ,549 .Bat

3.78:5

2.m 2.1111 2.111 3.374 4.&36

2. J98

2.259 -

2.SIe

3.161 4.399

2.U8 2.399 2.818 3.755 4.M7

2.m 2.523 3.382 4.138 4.9B9

2.328 3.M7 3.811 4.538 4.971

2.fl82 1. 741 1.338 I.B63

.B95

1.911 1.461! 1.168

.9112 .89'

1.688 1.125

1.657 1.657 1.125

1.853 1.853 1.468 1.125

2.826 2.826 I. B17 1.4BII 1.125

2.162 2.162 I.b74 1.346 1.125

2.356 2.191 1.161 J .472 1.125

2.516 2.282 1.B41 1.497 1.125

2.618 2.IB9

1. 673. 1.337 . L 125 \

\

2.405 1.938 1.469 1.236 1.125

COADE Pipe Stress Analysis Seminar Notes ----~~~-~~~~~~~====~~~----------------------------------;

NQ INTERSECTION RADIUS

"B31.3~ VS. "WAC 330~ UNREIN FORCED. FABRICATED TEE STRESS INTENSIFICATION FACTOR COMPARISON

o·

-

F· .. •

•

HEADER NOM

16 48, tb 48. I~ ~a. 16 .f@t: jb 4B.

IS U. III ~~. 18 411. !~ ill. IS 411.

29 411. 29 48. 20 4~. 2~ 41!. 2~ 48.

2~ 48. 2; 4@t 24 4', 24 4a,

BRANCH SCH

S 49. Ii! U. 11 48. 14 40. ~D 46,

c,

13 4B.

11 40. !4 48. Iii 48. !3 AZ.

12 4a. 14 4B. 16 41!, Ie 411. 2~ 411.

16 411. IS 48. 2@ 49. 24 49.

311 n. 2-44i!. ~ 48. ~0 40.

J2 4a. 32 49. 32 411.

3' 411. 34 411. H ~0.

Sa 48. 10 48. 3-1. 41. 3'6 48.

42 U. 42 41. 42 411. 42 4@. 4, '8.

WRC 330 b 0.825

8.121 8.123 5.595

i.2BI 7.850 8.22Q

e.ni 5.598

7.711 U.~B2 8.048 9.165 S.R\l!

E.lm 8.560 1i.1I37 ~.~43'

-1331.3-

i ib iab

4.446 5.595

4.446 5.595

4.446 5. sss

4.446 5.515

4. 4~b ~. S9S

4.4ff 5.578

4.4H 5.599

4.Hg 5.598

UH S.S9B

(.H7 5,598

4.681 5. Bal

4.6U 5.861

40691 5.801

U~I 5.881

U.RI 5.881

ua1 4. H1 4.197 4.n7

5.943 5.'m 5.1j14S 5.9U

330 b .651 .593 .m .5111

! .795

I

! I

I I ,

.618

.5J2

.502 .!?J

.583 • S4':' .521 .791

iob 330 b .829 .733 .sn .MI

l.aB8

.717

.71~ .6119 ,60360

I ,91l~

.752 .718 .071

.633

WRC

3.3"32

3.961 ~ 4.352 I 4.973

I

2.96f i.sz: 3.371 4.m 4,976

3.21!B 3. blI4 4.119 4.633 S,IS6

• n6 J.S!2

.694 3. 9S1

.658 4.S~7

1. e8i . S.281

-831.3-·

i ih iah

4.446 S.S9S

4.446 5;595

4. H6 5.595

4.H6 5.595

~.H6 5.595

4.449 S.S9a

4.4411' 5.593

4. H1 s.sss

4. H~ 5.5g3

(.H'1 5.598

~

330 h

1.667 l.:m 1.123 1.922

.1174

J.59J 1.263 l.Hg

1, ~B6

.9g~

4,6~1 S.9~1" 1.4g3

4,681 5.BBI 1.277

4,6al S.BBI 1.117

4.6al 5.8BL .993

4.681 5.881 .a~2

4.7a7 4.797 4.]97 4.7117

'1.7i2 S.14e 6.521 _" . 539 .;72 4.619 5.141

0.. S2Il 5.148 0..5211 .7118 1.988 S.796 5.148

24 411. 18.394

38 48. 18.134

32 41. 7.227

5.67B 5.6711 5.67;

s.eqq 5.899 5.899

5. iSB 5.7ea 5.7B8 5.7a9

6.'81 &.411 6.-'&8 6.418 1I.4IB

,

7.227 ! 7.227 I 7.227 !

7.5321 7.532 •

7.532 \

~ 7.384 i 7.384 ~ 7.334 J 7.384

9.211 ; 8.288 ' 9.2ag ,

B.211 , 8.21a

.546 .m .79S

.6115 .713 1.1I8\!

~.783 6.861 11.424

5,879 6.293 6.695

5.446 5.831 6.213 6.563'

.713 .bS? .668 • 649

1.8811

5.[08 5.5~ s.asil 6.228 7.289

S.b7B 5.679 5.6711

5.il99 5.999 5.:199

5.788 5.788 5".798 ~.7aa

b.481 6.41B 11.411 &.41 • 6.481

!:i.'HS: [':m 5.1)-\3; 1. I'll 5.'143; I.m 5,943· .891

6.S'S i 1.111 6.~2' i .887

7.227: I. I all

1.227 .945

1.2,7; .883

!

7.532 t 7.532 I 7.532 :

7.384 j 7.384 ! 7.394 7.S8"

8.lla 9.211 I a.211 I

8.281 I 9.288 t

1.~83 .937 .S91

l.a~3 .993- .'33 • as::!

ioh

330 h 2. US 1.6" J.H3 1.286 1.1~5

11. 76:;

11i.:m 34 4'a. ~ 7.532

38 ~8. 11.210

32 49. 11.599

j~ 49. 9.9112

1.38-\

311 48. 32 48. 34 'I. 36 43. 42 48.

U.Sla 11. 981 12.231 12.o:i~

8.2111

.501 .572 -

.793

.6-48 .733 1.&118

o

• • 9

.5th .m .585 .794

.659 .&37 .746

I.BIII

1.989 1.599 I.H6 1.21>b l. 125

1.769 1.6111 1.4!9

.55& .-531 .521 .Si7 .7BI

1.252

L 125

• • •

--.

• • • •

-1.&92 . [, SZ4 1.~51 1.125

1.411 1.12~

1.511 1.294 1.125

1.281 1.197 J. 125

(J. -, t)

o c

• • o

o IJ

• • •

i..-!S& 1.1ab 1.198 1.125

1.238 1.157 1.187 1.118

.979

1.587 1.462 1.393 1 • .116 1.125

e e ~.

• •


.t:lQ INTERSECTION RADIUS

"B31.3~ VS 'WAC 330W WELDOLET STRESS INTENSIFICATION FACTOR COMPARISON

e • • • o

e

.C)

•

•

•

•

•

•

•

•

• .c)

o

o • • • o

o

• • • G

HEAOE>i NOM

2 48. .. 40.

;> 4ft. 3 40. :\ 4~.

~ n. 4 41l. 4 ~I!. 4 48.

5 U. 5 4~, 5 U. S 48, 5 4tt

b 48. 6 "8. 6 46. o ~O, 6 n.

E 46. a 40. I! ·n. e 48. a u.

III 48. le -40. re 4B. 10 U, Iii 40.

12 'lI, I? 40. 12 40. i2 'B. 12 411.

!4 4il. I' 40. tA 48. 14 411. l4 40,

SCH

BRANCH WRC

, 330 0 I

4i!. i 2. 43-~

, :: I;::

A~.

411.

Z 411.

~+ 48.

1 4B, 2 411. 3 ~Il. 4 411.

J U. 2 41, ·e, 411. j 41l. S 411.

3.~ES 4.769 , 3.488 I

! 3.3b6

,

) 4.682 I

I 5.6Q4

i 3.1192 !

!

s.si: 4,51.19 5.588 6.359 r.zss

-83\.3-

UW] 1.897

i io iob

!.516 1.516

1.516 1.516

1.5711 1.57B 1.578

1.756 1.750 1.756 !.756

1.92B L'l28 I.ne I.ne t.928

2 40. 4.47; 2.0~B 2.048

3 40. 5."4 2.B4S 2.148

~ 41l. : b.m 2.114S 2.B48

5 40. 6. m 2.114S 2.048

6 -49. 4.S4e. 2.048 2.~4S

3 49. . S.IS7

.. 411. ; 5.910

~ 4i1. : 6.m

b '411. : j. 210

8 48. 1 4.949

,

!

.. 48. i 5.642 S 41. : 6.294

C 48. • 6.IIS4

!! 40. t 7.875 Itl n. i 5.234

s n, : 6.834 b 48. I 6. bill 8 U·. i 7.S4'? Ie 48. i 8.m 12 411. '5.52:;

b 40. 8 48, III 41. 1.2 46. H 46.

! 6.38:; 7.m 8.166 8.569 5.599

2.133 2.2"33 2.233 2 • .233

"':I "..,., 1.loL.')'rJ

2.384 2.384 2.:;84 2.SS4 2.384

2.492 2.492 2.m 2.492 2.491

2.526 2.526 2.52'; 2.526 2.m

-2.m 2.233

1.570 1.570 1.570

1.756 1.756 1.756 1.iS6

1.916 1.926 1.92e 1.920 1.920

2.233

2.23.3

., '1,L..Lw~

2.384 2.384 2.38~ 2.384 2.3:34

2.m 2.49"2 2.m 2.491 2.491

2.520 2.526 2.m 2.526 2.526

.375 .315

.3BB .308

.451 ___d_51_

WRC

-B31.3-

i il1 ioh

I.rm 1.1197

t. 516 1.516

j, 516 1. SIb

~

330h

.58B

.722 .588

1.57m 1.578 .748

1.570 1.:571 .74B

1.578 1.576 .56B

2.1110 1.756 l.7Sb • B3!.

\ 2. lSI I. 75~ 1.756 • B36

2.665 1.756 1.756 ,659

1 3.46S ._~ _1~ 7S6 ..:~I!B

2, IDS 2.119

:;.0411 3.783

.uo .378 .331 .310 .451

i 2.1811

!. 2.258.

1.810 3.361

A.399

2.148 2.1148 2.@48 2.048 2.148

2.233 2.233 2.231 2.233 2.233

2.384 2.384 2.384 2.5114 2.3~

2.492 2.492 2;492 2.n2 2.492

2.526 2.526 2.526 2.526 2.526

1.128 1.92~ .914

t.m 1.92B .914

1.~7? 1.928 .82H

1. m L 928 . 632

1.928 1.920 .SH8

.371 .451

.467 .467

.329 .329

.451 .451

.57'1 .579

.418 .413

.344 .3H

.3112 .302

.451 .451

.458 .376 .330 .296 .m

.431 .378 .339 .3U .451

.422 .379 .346 .383

, .451

,

I .4J3 I .378 \ .338 ; .295 l .4~1

I .396 1'1 .346

.:519

1 .295 I .451

330h 2.162

2.1n 2.986

.458 2.IBI 2.848 .376 2.1119 2.&48 .331 2.711 2.848

.296 ~ 3.374 2.848

."1 . ~.~36 2.14S

.371 .451

'[' 2, IBI!. 2.JBS

I 3.89;,

.422 .m .346 .303 .451

2.1118 2.399 i 2.871 i 3.755 ! 4.697

i

j 2.118

12.52"3 ~ '3t382

2.233 . 2. 233

2.233 2.233 2.233

2.384 2.394 2.384 2.394 2.394

2.4Q2 2.491 :;:.m

2.492 2.492

2.m 2.S211 2.526 2.526 2.5211

.?7S .97~ .755 .bS7 .588

1.863 .989 .7?~ .664 .5iB

1.135 .9114 ,831 .0.35 .588

!.lSI .987 .755 .603 .5eS

I.B85 .829 .bb3 .558 .SIB

,..,.,

......

.51B

.748 .748 .5ea

.sss .836 .659 .sea

.914 .914 .S2B .632 .51B

.413 .378 .331 .295 .451

4.131 4.919

.975 .975 .755 .607 .s~a

1.063 .989 .m .6M .588

2.328 :;.IU 3.811 4.538 4.977

1,135

.994 .831 .635 .518

1.181 .987 .755 .b0} .SIS

.396 .346 .389 .295 .451

1-39

LISS .829 .1163 .558 .50S


Q • C

.t:!Q INTERSECTION RADIUS

"B31.3~ VS 'WAC 330~WELDOLET STRESS INTENSIFICATION FACTOR COMPARISON

HEAOER NOM

16 48. 16 4B. rs 48. 16 4e. 1~ 48,

- IS 4l1. 19 40. (9 4iJ. 1~ 49. Ie ~~.

2e 4~. 2a ~8. 28 48. 28 43. 28 U.

24 43. 14 411. N 48. 24 4@.

33 41. 31 48.

.J2 48. 32 41. 32 43.

34 -4iJ. H 4~. 34 U.

:.~ 48. 36 n. ~I! 4f1. H 48.

42 48, U 48.

42 "', 4Z 48.

42 48,

8RANCH SCH

s U.

JS U. 12 49. 14 49. 16 4~.

~~

!R 411.

i±~,: 4B.

14 411. i

16 40.' 18 40. J

i

12 ~IJ, I

14 411.

16 48. f [S 43. i 28 48.

WRC 330 b 6.S2S 7.633 8.322 B.723 5.595

7.29l 7.aS~ 8.219 B.rn 5.59<1

l,71! a.~a2 8.646 'i.lb5 5.881

!~ ~0.· s.lIn

18 43. 8.566

18 40. Q.837

24 43. S.Q43

-831.3-

i ib iob

2.52' 2.524

2.524 2.524

2.524 2. S24

2.524 t.S24

2.524 2.52~

2.526 2.526 2.526 2.-526

2.52~

2.617 2.617 2.617 2.6\7 2.617

2.691 2.681 Z.bBt 2.6a!

3.2bl 3.261 3.261

3.:399 3.:178 3.3~a

3.NI 3.m J.~3'1 3.·m

3.~77 3.6~9 3.Q~9 3.699 3.m

2.m 2.:in !

! [

! I

I

I

UI7 . 2.617

2.611

2.il17

2.6\1

USt

2.681 2,68!

1.52b 2.526 l.S2/)

i ib iob

330 b 330 b

.3711 • n-8

.33"1 .331

.3&3 .393

.289 .289

.m .451

.35! .312 .387 .m .~51

.339 .324 .3113 .211b .451

~~'7

.~,) ...

.313 .2r!

2.b81 i .451 i

!

!

2. '42j .3&3 2.942 .451

1-40

.351

t322 .311i • Z81 .m

.324 .3113 .286 .451

~~'7 .~.} ..

.313 .197 .451

.H4 .322 .451

.289 .329 .451

.297 .2B7 .336 .451

.m .311 .391

.293

.451

WAC 330 h 2.664 3.332 3.961 4,~S2 4.973

-831,3 .. ·

i ih ioh

2.52Q 2.524

2,524 2.524

2.524 2.524

2.:m 2,524

2.52~ 2.524

2.526

,. lEO""" I. • .J~~

2.526

2j52~

2.526

2.61i 1.61i 2.617 2.6!i 2.6!7

Z,b!!! 2,b81 2.681 2.b81

2.942 2.9U

J.2111 3.261 3.2111

3.398 J.39B 3.398

3.331 r.m 3.331 3.331

3.~99 3.6Q9 3.6~"1' 3'.b99 3.6.9Q

2.526 2.526 2.S2t

2i5i6

2,6P 2,6l1 2,b17 2.1.11 2.611

2.68: 1.681 2.6al 2.681

3.21:11 J.2111 3.2111

3.398 3.399 3.398

3.331 3.3'31 3.331 3.331

3.b~9 3.1t'?9 3.699 3.699 3.6-99

~

330 h

.9H .iS8 .637 ,586 .5~!I

ioh

330 h ,H1 .15S .631 .'j81 .5as

CY.

" ,,, .• j-

• c?

o c

• •

-,,0

24 4a.; 3S 4o,_-1

I

24 43'1' 3S 49.

32 40.

3S 40. 32 40.

34 41}. i

,

9.792 fr.52a

U.394 Ill. 134 7.221

11.763 lB.:m

7.532

38 48.! 11.21B 32 4B. I I 1. 599 34 48.! ~.902 ;)b ~8 ~ 7.384

. l

,

3~ 4!!1. j 11. SSB 32 48. i 11.9;7 ·H ~!,! 12.231 3& 4!.! IZ.633 42 41!. i S.2SI1

1

3,261 3.261 3.Z61

3.398 3.398 3,398

3.331 3.3.31 3.331 s.m

~.im 3.6911 J.II~9 3.4'11 3.6~9

.322 .1 .451

j

I

I

,

.289

.329

.451

.2H .2B1

.330 .451

.3Z1 .JIl .3al .293 .4~1

2.964 3.523 3.1171

4.423 <4.1?6

,8S2 .711 .6S3 .571 .sas

.799 .m .635 .S6S .s~a

.163 .679 .bH .sse

.b37 .:5iB

.692 .543 .SSB

.SiB .sn • Sse.

.612 .571 .m • SiB

.aS2 .717 .653 .571 .S8i1

,ns .i26

3.28B 3.61H ~.IIB

5.156

J.S!2 3.951 4.397 5.282

4.&19 S.796

4.783 b.BSI 6,424

5.S79 6.293 6.695

.716 .069 .629 , ~94 .51118

.753 .679 .

.61i1 .5~8

• • •

••

• • • •

15.4-4b S.83a 0.213 I 0.56J I

I S.I&B ! 5.533 I S.BSb

,. b.na

7.28'

.637 .5118

.S7a .5411 .51lB

(J

e

o e

• • e

.1112 .571 .537 • SilO

• 7J,~ .1I1t9 .028 .594 .519

-0 o

• • •

@ COADE Pipe Stress Analysis Seminar Notes 9

.. ~

N.Q INTERSECTION RADIUS

• "831.3" VS 'WRC 330~ SWEPOLET STRESS INTENSIFICATION FACTOR COMPARISON

•

HEADER BRANCH

NOM SCH

o

G • • • o

o

.C)

• • • •

• • • • •

C·'

",

· )

(3 • • • c

o

• •

• e

~ 49.

J 46. I 48.

3: 48, 3 48~ 3 48.

4 U, " 48. -4 48,

.; 4e.

S 4~. 5 ~e~ S 48. 5 48~ ~ 4~1

e 'Ie. II 48. b 4e. to n. 6 U.

e 4B, a 41. a' 4!. B 48. 3 4~.

III 4e. a ~~. l~ AB. U ~e. IB411.

12 A~. 12 Ag. 11 4e. 12 411. 12 41l.

14 48.

1 ~ 41l,

14 48, I ~ 48. 14 UI.

1 48. 2 "B.

40, 2 49 ..

3 40, I !

48·1'

.2 41!~

3411. I 4 n. !

" I

411-. ,

2 49~ ! 3 411 •.

-4 "iii. : 5 48r

2 40. :3 40. 4 n.

5 40. : 6 4B ••

3 4et - • 48. I 5 40. \ 6 - 411. i 8 41l.

WRC 330 b 2,433

4.1184

Z;,3S9

2.01S 4.155 3.486

2,563 4.0118 ~.M4 3.892

2,4a~ 3,Q40 5.58B 6.358

-831.3-·

i ib iob

.n'i .'1116

1.188 1.H18

1.222

11222

1,222

it33?

1.337 J.337 1.·i37

1. 439 [.439 1.439 1.439 1.439

3.736 1.518

5.W 1.518

6.2112. 1.518

e.91, 1.518

4.540 uta

5.187 ~,.~m 1:.5112 /.210 ~. 949

"48. :.>.642

5 48. 6.294 II 4B. i 0.834 B 4B.! 7. S75 18 48, I 5.294

I !

1.1132 1.632 1.632 Lb32 1.632

1. 726 !. 726 1.726 .1. 726 1.726

1.793 \.193 1.793 1.1Q3 1.7113

1.814 ! ,814 J.8H l.914· ;,914

).251 i.25J

1.290 1.296 1.2911

1.458 1.45~ 1.45i' 1.~5~

.1.585 1.585 1.585 J.SB5 I. 5a~

1.6~! 1.691 1. 691 1.6'71 Lb'11

\.843 i .315 !.843 ! .276

1.114;) I .1411

1.843 ,226

1.1143 .330

1, QbB 1.963 1.968 1,968 1.968

2.BS7 2.~5i 2.857 2.857 2.BS7

2.885 2,B8~, 2.e8S 2.885

2.885

jib 'ob

330 b 330 b

.382 .372

.2"1 .:;54

.467 .m .351

.m .312 .312

.566 .356 .255 .372

.638 .482 .284 .149

,4~2 .311 .273 .2H .372

.312

.280 .256 .372

WRC

330 n 2.1112

2.liI@ 2.986

2.189 2.lilB 3.893

2. I iii 2.11'~ 2.6065 3.46i

2.liI~ 2.10B

2.342

3 •• n 3.7SJ

2;11@ 2.10& 1.711 3,:m 4.9311

2.IH· 2.2~8 z.au :'.361 4.39~

.349 2.110

.313 2. 39~

.286 2.8!8 .25i! 3.m

.372 4.697

-831.3-

i in ion

.919 .91l6

1.188 !.lSS

1.222 1.222 1.222

1.337 1.33i 1.337 1.337

1.439 1.-43'1 1.439 \ .43'1 1.439

I.SIB 1.51B 1.518 1.518 J.S1B

1.632 1.632 1.632 J .632 1.1132

1.726 !. 126 1.726 1.726 I.nil

1.793 1.7f/3 1.793 1.7113: 1.793

loll 14 1.814 ! .81~ 1.814 1.814

1.2~1 1.251

1.296 1'.296 1.296

1.458 1.4Si! 1.458 1.45B

1.585 1.595 1.595 1.595 1.5B5

1.&91 1.691 l.im I.blll lob91

1.843 'J.84:5 1.943 1.843 1.8~3

1.%B I. 1I~8 1.968 1.%8 1.1I1i8

2.157 2.957 2.857 2.857 Z.8S7

2.685 2.185 2.8B5 2.8SS 2..885

~

330 n .4:56

ion 330 h .419

) 413. (, 4B.

1.1 ~13,

113 U. 'I 12 4!.

6 48. i

8 48. 18 46. 12 48. U 4!.

b.2:54 b.il81l 7.549 8.443

,522 .529 .235 .344

.57Q .365 .258 .216

.m 2.110 .:m 2.523

.272 3.382

.244 •• m

.372 4.9&9

.2B4 .248 .222 .212 .m

,

.J27 ,; 2.~29 .280 '\ 3.m

.255 3.811

.24~ 4.538

.372 i 4.1177

1-41

.5611 .:JIII!

.592 .592 • 39~

.637 .637 .582 .3a6

,695 .b85 .614 .4i3 .38B

.723 .723 .5hB .458 .st«

.777 .723 .561 .486 .371

.822 .n9 .' .m ! .468

I

I

I

I I

.858 .7HI .5~3

.434 .3115

i_

.77'1

I

.595 .476 .468 .364

.617 .617 .4111

.698 .698 .544 .U9

.755 .• 677 .521 .419

.905 .885 .624 .5e! .419

.879 .816 .656 .548 .419

,937 .82e .68b .524 .419

.975

. 91~' .1.23 .498 .419 \

.996 .684 .547 .4b8 .419

6.383 7.382 B.I66 8.569 5.599

.486 .27'i .245 .219 .m

. i

.386 .274 .251 .219 .:m

.297 ,272 .237 .212 .325

...... ~'."h'L;'4

_ . ",,!5)Jt _ _ .. _ . 4. (~JAl


. _.auz.

..

!$: AU

ioh

330 h .78: .~2S .526 .47'" .419

o Q C

~.

' .. ,,/

• o

NO 1NTERSECT10N RAD!US

"831.3" VS 'WRC 330" SWEEPOLET STRESS INTENSIF1CAT10N FACTOR COMPARISON

HEADER NOM

16 4@.

[6 U, 16 ~e, 16 4i1, 16 4ll.

III U. II! 4e. 11! 41!. 1 ~ 4~. 18 40,

211 48. 29 U. 21!i U. 20 ~l!. 29 U.

2~ 49.

24 4@.

14 48. ~ 40.

31 ~3. 38. 48.

32 ~9. 52 48. ·:;Z 48.

:H 46. 34 48. H 40.

Sb 48. Sb ~e. 10 411. 3b 48.

41 48. 4: 4e. 42· 48. 42 48. ': 49.

.'

BRANCH SCH

I! 46.

Ii 49. 12 ~8. 14 46. 16 .8.

II

U!,', .3. ; 12 "4\t, J l~ 4Q.! 16 49r:

18 ~~.!

12 49. 14 49. 16 4g. 18 48.

I

29 41L ~

WAC 330 b 6,825 7,635 8.~22 8,723 5.595

7.l1H 7,853 S.229 8.197 5,518

7.711 a. Bat IU4iJ 9,165 5,981

-831.3-·

i ib iob

1.813 2.@84

L81J 2.984

1.813 2.984

LS13 2.994

1. 81J 2. ~84

L8I4 LaJ4 I.an 1.::[4 I, 81 ~

1.87i1 1.878 1.87B 1.87B LS7ff

1.'?IB J.910 1.119 1.91B

,2.m 2,m

2.209 2.269 2.2b9

2.354 2.354 2.354

2.312 2.m

2.111

2.312

2.SU 2.548 2.~48 2.548 2.H8

2.692 2.b92 ,.692

2.B85 2,385 2. B8S 2.~as

2.160 2.160 2.1011 2,1&~ 2.160

2.213 2.213 2.213 2.213

2.42B 2.428

2.895 2.811S 2.965

2.75'9 2.758 2.7~1 2. 7~1

3.IS4 3.154 3.IS4 3.154 S.U4

i ib iOb

330 b 330 b

.~66 .30S

.n7 .27J

.218 .259

.298 .239

.·124 .S7Z

.23J .229 .2~6 .:324

.243 .231 .216 .m .322

.2911 .266 .2Si

~-.

I .:.'~~ i

.2ell .267

.216 .372

.274 . Z58 .24S .372

.258 .372

.259 ,266 .372

,238 .272

,2.~ .n7 .2711 . lit

.265 .2Si: .2'9 .242 .n!

WRC

330 h 2.664 3.332 3.961 ~.3S2 4.97·3

-831.3-

i ih ioh

I. 813 .2.084

1.9iJ 2.1194

1.813 2.S84

I. 813 2.1184

1.813 2.884

1.814 i.s: , 1.814 L9!4 1.814

l.BS5 ~.a8S 2.085 2.11115 2.985

~

330 h

.6911 .544 .458 .~17 .3H

(:) o

.734

• • ~

o Q

(-~

16 ~II. j 8.1176

18 _48., g,566

29 ~e, 9.~)7

2~ 411.' 5.943

24 411, I· 39 48.

24 49. 311 48.

32 411. t

i

I 30 49.!

32 491 1 _~ ." I

,:, ., ... t

; 33 42.; 32 49. i 34 49. i

36 43.!

:~B 49'1 32 48.

34 "8.! 36 at. i

42 'I. i

9,n2 6.52!

UJ.:)94 18.m 7.227

! 1. 7&3 ie, ;:;[1

11.21B 11.599 9.9a2 7.384

u.ssa 11.187 12.2at 12.633

8.218

.23b .m .z: 1 .321

.21:1 .318

.213 ,224 .su

,m .m .313

2.'164 3.523 3.971 4,413 ~. 971:

3.21!~ 3.604 4.116 4.6.JJ 5.15b

J.an (.67~ 1.373 Lan l.on

2.160 2. il,a 2.16'a 2,1 bi1 2.16Z

.612 ,SlS .41:" ,'18 .364

I :~;:

.454 , 4~4 .363

.544 .483 .4J4 .362

.,448 .3:17

tS,;:! .539 ,4i! ,Hq

.659 .599

.466 .H9

• • • • •

.' •

• •

(~

,236 .199 .m ,313

.221

,213 .211 .291 .319

.3.512 3.951 : 4.:m

I

j 5.282

!

4.01~ 5,796

4.793 6.181 6.424

5.871 6.293 6.695

1. tjll~ L9Ja t. 9J~ L91l!

2.213 2.113

.H4 .37B .353

.4il0 .374 .352

.425 .397

.352

•• 92 .459 .432 ,439 .3~Q

.b31l .S6il .58J .4111

.563 .449 .419

.417 • ,HI)'" .H~

.5f1::! .HZ .UJ .419

o o

• • c

o 0'

• • •

I 5.446 5.831

16.2113' 1i.563

,

I S.16e i 5.533 ! 5.886 I b.l2a : 7.2a9

z.an 2.428 2. 971 2.428

2.269 2.092 2.269 2.692

2.269 2.692

2.354 " Ulft5

2.35~ 2.805

Z.3S4 2.885

2.312 2.159 2.312 2.758

2,312 2.158

1.312 2,751

2.548 l,n4

2.541 3.354 2.548 3.1:4 2,541 3. il54 2.HI 3.9S4

.59l .552 .519 .499

• till

G COADE :Pipe Stress Analysis Seminar Notes

o

~.

• • e

e

• • • o

o

,()

• • • •

• • • • •

()

o @

• • • Q

(.)

• • •

1.5 Code Compliance

1.5.1 Primary vs. Secondary Loads

Markl's investigation of the fatigue problem, following the earlier recognition of the maxim um stress theory of failure, led to identification of the basic problem in the design of piping systems. Not one, but two different criteria must be satisfied, one for primary loads, which may lead to single application catastrophic failure, and one for cyclic, displacement~ driven loads that may lead to fatigue failure (especially in the vicinity of fittings and other discontinuities) after repeated applications. The main characteristics of these two different types of loads are described below:

Primary Load Characteristics:

1 • Primary loads are usually force driven (gravity, pressure, spring forces, relief valve, fluid hammer, etc.).

2 ~ Primary loads are not self-limi ting. Once plas tic deform a tion be gins it continues unabated until force equilibrium is achieved (through change of the external boundary conditions or through, material strain hardening), or until failure of the cross section results.

3 . Primary loads are typically not cyclic in nature (and those that are, such as pulsation or pressure variation, show characteristics of both primary and secondary loads).

4 • Allowable limits for primary stresses are related. through failure modes such as those advanced by the Von Mises, Tresca, or Rankine theories, to the material yield stress (i.e, the point where plastic deformation begins), the ultimate strength, or, for sustained loads only, to time-dependent stress rupture properties (such as creep characteristics).

5 ~ Excessive primary load causes gross plastic deformation and rupture. Failure may occur with a single application of the load. Note that failures that occur due to single load a pplica tions us uaily involve pressure (hoop stress) design failures and -are not directly addressed by CAESAR n or by the flexibility stress requirements of the codes. Such pressure design requirements are encompassed in the minimum wall thickness requirements discussed in detail in separate sections of the codes.

Secondary Load Characteristics:

1 ~ Secondary loads are usually displacement driven (thermal expansion. imposed

anchor movements, settlement, vibration, etc.). \ '

2 . Secondary loads are almost always self-limiting, i.e. the loads tend to dissipate as the system deforms through yielding or deflection.

3 . Secondary loads are typically cyclic in nature (except settlement).

1-43

, ~_ ~ • I , .. ~~. .,.,;


4 ~ Allowable limi ts for secondary stresses are based upon cyclic and fatigue failure modes, and are therefore limited based upon requirements. forelastic cycling' after shakedown and the material fatigue curve.

5 A single application of the load never produces failure. Rather catastrophic failure can occur after some (usually high) number of applications of the load. Therefore, even if a system has been running successfully for many years,it is no evidence that the system has been properly designed for secondary loads.}

Several examples should help illustrate:

Ii.

~ Stress Failure: Springs were improperly sized to support the weight of the valve operator on a system. When the line was filled for hydrotest, everything (stresses and displacements) appeared fine, since the pipe could support the moment imbalance at ambient temperature. However, heating up the fluid (and pipe) during startup, the valve sagged and the guardrail was crushedin less than 30 minutes due to the decrease in strength at the operating temperature.

Steps to failure:

1 Weight loads were improperly accounted for. (The primary stresses were too high.)

2 ~ At operating temperature there was a resulting drop in material strength.

3 Gross deformation began almost immediately and continued until force equilib-

rium was achieved (the spring bottoming out). .

•

Secondary Stress Failure:

After 12 years of successful operation, inspection of the inside surface of a vessel revealed fatigue cracks in the vicinity of a piping nozzle connection. A subsequent analysis showed that a temperature increase in. the adjacent vessel and piping system (alongwith a relocation of pipe restraints for the new operating conditions) made several years ago caused the stresses to exceed the expansion allowables. Even though the calculated stress range at the

144

0
•
0
O·
,."
•
0
.
0
c
•
•
n
0
C!
•
•
.. _ •
"i
•
•
•
•
•
('
0
()
0
•
•
0
o
V
•
•
• junction was well over 470,000 psi, thejunction survived several years because of the selfrelieving nature of the thermal load. and the fact that the system cycled fewer than a dozen times over the two year period.

o COADE Pipe Stress Analysis Seminar Notes

• • • • • • • • •

JJ 1.5.2 Code Stress Equations

o • • • o

e

• • • o

o

eb

G o

• • • o

o

• • •

Steps to failure:

I - Thermal allowables were exceeded by mistake.

2 - After about a dozen applications of the excessive load. cracks formed on the

highly stressed inside surface of the vessel at the junction with the nozzle.

Therefore. code compliance requires that the piping system be checked for both types of - loading - primary and secondary. The basic steps involved in doing code compliance are outlined below:

1 - Compute the primary stresses, i.e. the stresses due to the sustained primary loads, usually weight and pressure, and those due to the occasional primary loads. such as earthquake, wind, fluid hammer, etc.

2 - Compute the range of the varying stress, i.e. the expansion stress range.

3 - Compare the primary stresses to their allowables, which is based on a factor of safety times the hot yield stress.

4 - Compare the expansion stress range to its allowable, which is a factor of safety times a value varying with the number of cycles such that it fits the material fatigue 'curve (adjusted for mean stress), but never exceeds the sum of the hot and cold yield stresses.

Note that due to the shakedown effect, and the fact that the primary and secondary stresses have different failure criteria, these two load types are reviewed in isolation. Therefore, it should be stressed that, as far as most codes are concerned, there is. no such thing as "operating stress".

The piping code stress equations are a direct outgrowth of the theoretical and investigative work discussed above, with. specific limitations established by Mark! in his 1955 paper. The stress equations were quite similar throughout the piping codes (i.e., between B3l.1 and B31.3) until the winter of 1974, when the power codes, having observed that Mark! was incorrect in neglecting intensification of the torsional moment in a manner analogous \0 the bending component, combined the bending and torsional stress terms, thus intensffymg torsion.

It should be noted that the piping codes exactly calculate the stress intensity (twice the maximum shear stress) only for the expansion stress, since this load case contains no hoop or radial components, and thus becomes an easy. calculation. Including hoop and radial stresses (present in sustained loadings only) in the stress intensity calculation makes the

145

t~~t!\!.;5i;?!~{li~~~~;f;;~f{f~;;~J~~~t7~01t,{':¥;~~~I~J;~;;,;,(g~~:~';K~g~'lfA!~~~~.~.;;_~~~;;~fff-,:~~'~~:~~*~;:'::::~):j;-~~tr~:·~w;!l~f1¥~2 .

. '"' ' ... ":.'-


calculation much more difficult. When considering the hoop and radial stresses, it is no longer clear -which of the principal stresses is the largest and which is the smallest. Additionally, the subtraction ofSl-83 does not produce a simple expression for the stress in tensi ty. As it turns out, the inclusion of the pressure term can be simplified by adding only the longitudinal component of the pressure stress directly to the stress intensity produced by moment loadings only. This provides an equally easy-to-use equation and sacrifices little as far as accuracy is concerned.

The explicit stress requirements for the piping codes addressed by CAESAR 1] follow below. Note that most codes allow Pdi2 / (do2 - di2) to be used in place ofPdo / 4t.

,I

1.5.3 831.1 Power Piping

The B31.1 code requires that the engineer calculate sustained, expansion, and occasional stresses, exactly as defined below:

Sustained:

0.75i MA P do

+

z

4t

Where:

Ssus,81 = sustained stress, psi

1 = intensification factor (single factor for all types of moments), as per

Appendix D ofB31.1 Code (note that 0.75i may not be less than 1.0)

MA = resultant moment due to sustained (primary) loads, in-lb

Sh = basic allowable material stress at the hot (operating) temperature, as per

Appendix A ofB31.1 Code: Sh is roughly defined as the m.in.i.mum of:

1) 114 of the ultimate tensile strength of the material at operating temperature;

2) 114 of the ultimate tensile strength of the material at room temperature;

3) 5/8 of the yield strength of the material at operating temperature (90% of the yield stress for austenitic staiDless steels);

4) 5/8 of the yield strength of the material at room temperature (90% of the yield stress for austenitic stainless steels); and

5) 100'10 of the average stress for a 0.01'10 creep rate per 1000 hours.

1-46

0
J to
0
n·
., .. )
•
"
Q
0
•
•
(')
()
G
(~'~
•
•
-~ •
•
•
•
•
•
/>-.
(~'
0
0
•
•
0
G
~
•
•
• it COADE Pipe Stress Analysis Seminar Notes o

.• Expansion:

iMe

• •

e • • • o

o

• •

z

Where:

SE = expansion stress range, psi

Me = resultant range of moments due to expansion (secondary) loads, in-lb

SA = Allowable expansion stress, psi

Sc = basic allowable material stress at the cold (installation) temperature, as per Appendix A ofB31.1 Code.

Occasional:

0.75i MA

0.75i MB

Pdo

Soec =

+

+

z

z

4t

• Socc= occasional stresses, psi

Where:

• • • • • •

MB = resultant moment due to occasional loads, in-lb

k = occasionalload factor

= 1.2 for loads occurring less than 1 % of the time ._

= 1.15 for loads occurring less than 10% of the time

,[_) 1 .5.4 931.3 Chemical Plant and Petroleum Refinery Piping

o o

• • o

Q

o

•

•

• Q

t1t~j¥t~~~~tS~~:r~ __ ~~&~Z~~_41~~:~~~~~

- , ~ '.. :-; . 1 ~ ~ ....

Sustained:

B31.3 does not provide an explicit equation for sustained stress calculations, but only requires that the engineer compute the longitudinal stresses in the pipe due to weight and pressure, and then ensure that these do not exceed Sh. This is most commonly interpreted

to mean: '

Pdo

+

+

z

4t

1-47


. ,

Where:

Fax = axial force due to sustained (primary) loads, lb

"

Mj = in-plane bending moment due to sustained (primary) loads, in-lb

Mo = out-plane bending moment due to sustained (primary) loads, in-Ib

Ii, 10 = in-plane, out-plane intensification factors, as per Appendix D of B31.3 Code

\:

Sh \':,= basic allowable material stress at the hot (operating) temperature, as per Appendix A ofB31.3 Code. Sh is defined as the minimum of:

1)

1J3 of the ultimate tensile strength of the material at operating temperature;

2)

1J3 of the ultimate tensile strength of the material at room temperature;

3)

213 of the yield strength of the material at operating temperature (90o/c of the yield stress for austenitic stainless steels);

213 of the yield strength of the material at roam temperature (90% of the yield stress for austenitic stainless steels);

4)

5)

100% of the average stress for a 0.01% creep rate per 1000 hours:

6) 7)

67% of the average stress for rupture after 100,000 hours; and

80% of the minimum stress for rupture after 100,000 hours.

Expansion:

Wi Mj)2 + (io Mo)2 + 4MT2}1J2

.,..
Where:
Mj =
Mo =
MT =
Sc = z

range of in-plane bending moments due to expansion (secondary) loads, in-lb range of out-of-plane bending moment due to expansion (secondary) loads, illIb

range of torsional moment due to expansion (secondary) loads, in-lb

basic allowable material stress at the cold (installation) temperature, as per Appendix A ofB31.3 Code.

1-48

fl-

o

• • e

("';

. ./

• •

"- ..

·t

_,.

• • • •

c,

'0

e o

• • c

o

.... 0

• • •

4) COADE Pipe Stress Analysis Seminar Notes

o ~ ..

• o

e

• • • o

o

.)

• • •

• • • •

.P

o Q

• • • Q

• •

•

Occasional!

The equation for calculating occasional stresses is undefined by B31.3, which simply states that the sum of the longitudinal stresses due to sustained and occasional loads shall not exceed 1.33Sh. The default interpretation of this requirement is to calculate the sustained and occasional stresses independently (as per the equation given for sustained stresses above) and then to add them absolutely.

Note the differences between these two codes:

I - B3l.1 intensifies torsion, while B3l.3 doesn't.

2 - B3l.3 calculation methods are undefined for sustained and occasional load cases, while they are explicit for B3l.l.

3 - In its most common interpretation, B31.3 neglects torsion in the sustained case, while B3l.1 includes it.

4 - B3l.l neglects all forces, while in the default interpretation, B3l. 3 inel udes Fax in the sustained case.

5 - Allowable stresses are different for each code.

6 - Stress increase for occasional loads are different for each code.

Note that both codes additionally cite a conservative value of SAf(1.25Sc + O.25Sh), which may be used instead of the more liberal allowable off(1.25Sc + I.25Sh - 8n. This is a carry over from pre-computer days, when sustained stress calculations were rarely done, so 81 was not known explicitly, and conservatively estimated to be at its maximum allowable level of

~ - ,

Specific requirements of other common codes are shown below as well.

1.5.5 ASME Section III, Subsections NC & NO (Nuclear Class 2 & 3)

Sustained:

Ssus = B18lp + B2 Ma / Z < 1.5 Sh

Where:
BLB2 =
SIp =
Ma =
=
Sh = primary stress indices for the particular product under investigation

longitudinal pressure stress = P di2 / (d02 - di2), psi

\

resultant moment on the cross-section due to sustained (primary) loads [Mx2 + My2 + Mz2JlI2, in-Ib

basic allowable material stress at the hot (operating) temperature, as per ASME ITl Code. Sh is roughly defined as the minimum of:

1-49


1) 1/3 of the ultimate tensile strength of the material at operating tempera. ture;

2) 1/3 of the ultimate tensile strength ofthe material at room temperature;

3) 213 of the yield strength of the material at operating temperature (90% of the yield stress for austenitic stainless steels);

4) 213 of the yield strength of the material at room temperature (90% of the yield stress for austenitic stainless steels);

.\. 5)

, .

100% of the average stress for a 0.01 % creep rate per 100 hours;

6) 60% of the average stress for rupture after 100,000 hours; and

7) 80% of the minimum stress for rupture after 100,000 hours.

Expansion:

SE = i Me / Z < f ( 1.25 Se + 0.25 Sh ) + Sh . 8L

Where:

Me = resultant range of moments on the cross-section due to variations in loading (usually due to thermal effects)

= [Mx2 + My2 + Mz2Jlf2, in-lb

SL = Sip + 0.75 i Ma / Z (where 0.75 i >= 1.0)

Occasional:

The occasional stress equations are:

For Service Level C (Emergency):

Soec = Bl x Slpmax + B2 (M; + Mb) / Z < 1.8 Sh <= 1.5 Sy

For Service Level D (Upset):

Socc = Bl x S!pmax + B2 (Ma + Mb) / Z < 2.4 Sh

Where:

Slpmax

= pressure stress due to the peak pressure, psi

resultant moment on the cross-section due to occasional (primary) loads [Mx2 + My2 _+ Mz2J1l2, in-lh

=

=

Sy

yield stress of material at temperature. psi

=

..

0
0
0'
C·
l
":":;":":
.-
0-
0
t""
,.,
•
•
(1
(I
~!
\
l..-
•
•
-- •
:~+
.-
.-.
•
•
•
•
(/)
v ,
0
0
0
•
•
Q
'J
Q
•
•
• e 8

-~

• • o

e

• • • e

o

eCJ

• • • • • • • • • .0

o (J

• • • o

o • • • o


1.5.6 831.4 Fuel Gas Piping

The B3L4 piping code requires that the engineer calculate and check the sustained, expansion, and operating stress.

Sustained:

SL = Sip + Sb < 0.75 x 0.72 x SYield

Where:

SIp = the longitudinal pressure stress, psi

Sb = bending stress due to sustained loads, psi

Expansion:

s, = (Sb2 + 4 St2 )1/2"< 0.72 SYield

"Where:

Sb = range of bending stress due to varying loads, psi

St = range of torsional stress due to varyingloads, psi

SYield = specified minimum yield stress of material , psi.

Operating:

Sope = FIE a dT - v SH I + Se + SL( I-F) < 0.9 SYield

"Where:

F = % of pipe axial restraint (long buried pipelines are considered to be fully axially restrained, i.e. F = 1; while pipelines above ground on slide plates are not axially restrained, i.e. F = 0

E = modulus of elasticity of pipe material, psi

a = thermal expansion coefficient of pipe material, in/inl°F

dT = temperature change of pipe from ambient, of

v = Poisson's ratio

1-51

_ ... ~~~~~£Vf;~{i;j.fJ;,."j~!~ ... ·e·'·"'-k!i~··y"rtg~t:tI;i;&~~.~';;."""~Uir.~~~tnt:~"'t·'i"~ ":


SH = hoop stress, psi

= Pdo / 2t.

Occasional:

Socc =

SIp + Sb < 0.75 * 0.72 * SYield * k

Where: .

;\

SIp '\, =

Sb = k

longitudinal pressure stress

resultant moment due to occasional loads

occasional load factor

=

1.5.7 831.8 Gas Transmission and Distribution Piping Code

Like the B 31. 4 code, the B 31. 8 piping code requires that the engineer calculate and check the sustained, expansion, and operating stress.

Sustained:

SL = SIp + Sb < .75 S x F x T

Where:

Sip = longitudinal stress due to pressure, psi

= P dj2 / (<102• dj2)

Sb = bending stress due to sustained loads, psi

St = torsional stress due to sustained loads, psi

S = specified minimum yield strength of pipe material, psi

F = Construction Type

1·52

e e

•

o·

•

(J ,,', o o

• •

~"
'J
(:~
•
•
•
~-, •
•
•
•
•
cj
0
0
0
•
•
C
I)
0
•
•
• Q
0
,.
•
•
0
f)
•
•
•
0
0
ti~)
•
•
•
•
•
•
•
•
•
.C)
e
Q
•
•
•
0
0
--
•
•
•
G COADE Pipe Stress Analysis Seminar Notes

Factor F

Wasteland. Deserts, Mountains. Grazing Land. Farmland. Sparsely populated areas and Offshore_ Fringe areas around c t t te s , Industrial areas, Ranch or Country Estates.

Suburban Housing Developments,

0.72 0.60 0.50

A B C

Shopping Centers. Residential Areas. Multi-Story Buildings are prevalent, Traffic is heavy and where there may be numerous other utilities underground.

o

0.40

T = Temperature Derating Factor T Pipe Temperature deg. F.
1.0 250 or less
0.967 300
0.933 350
0.9 400
0.B67 450 Expansion:

Se == (8b2 + 4 8t2 )112 < 0.728

Where:

Sb == bending stress due to varying loads, psi

8t = range of torsional stress due to varying loads, psi

Operating:

Sape = 8e + 8L < S

Where:

terms are as defined previously.

Occasional:

80cc = S1 -+- Se < 0.75 * SYie1d * F * T * K

Where:

K = occasional load factor

all others as defined previously

1-53


o

1.5.8 Canadian Z1831Z184 Oil/Gas Pipeline Systems

Sustained:

SL = 0.5 * Shoop + SB ::;; S * F * L * J :I< T

Where:

Shoop = hoop stress

I = Pdl2tn \

'""

= resultant bending stress

G • • ~

SB = specified minimum yield strength
F = design factor
L = location factor
J == join t factor
T == temperature derating factor 0
e
C),
-; ..
•
•
•
" _ _ '. Expansion:

SE == (Sb2 + 4St2 )112::;; 0.72 :;. S * T

1-54

• • • •

"".

CO

"0

o e

• • o

o \j

• • •

Where:-

Sb = resultant bending stress

St = torsional stress

Occasional:

Soee = Fax I A + 0.5 :;. Shoop + SB S; S :;. F * L * J * T * K

Where:

Fax = axial force due to sustained and occasional loads

.'

K = occasional load factor

A = cross sectional area of the pipe

SB = resultant bending stress due to sustained and occasional loads

e


e

-. 1.5.9 RCC-M C

•

• o

Sustained:

SL = Pd, 14tn + 0.75 * i * MA I Z :;:; Sh

• • •

o

e .c)

• • • • • • • •

Where:
p = design pressure
do = outside diameter of pipe
tn = nominal wall thickness
= stress intensification factor
MA = resultant moment
= <Mx2 + My2 + Mz2)112
Z = section modulus
Sh = material allowable at design temperature MB = resultant moment due to occasional loads

Expansion:

SE = i Me I Z::; fCl.25Se + .25Sh )_ + s, - SL

'Where:

Me = range of resultant moments due to expansion loads

Se = material allowable at room temperature

• Occasional:

.C) Sace = Pmax do / 4tn +-0.75 * i * (MA + MB) / Z $1.2 * Sh

0
e
•
•
•
0
0
.
•
•
•
0 Where:

Pmax = maximum pressure occurring

1-55


/

1.5.10 Stoomwezen

Sustained:

SL == P (De· d ) / 4d + 0.75 * i >!< MA / Z < f

Where:
p = design pressure
Dot = outside diameter
d " == formula wall thickness
= stress intensification factor
MA == resultant moment Z = section modulus

f = sustained allowable, the minimum offive equations (see code for details)

Expansion:

SE == i * MB I Z < fe

Where:

MB = resultant moment due to expansion loads

fe = expansion allowable, the minimum of two equations (see code for details)

Occasional:

Socc == SL + 0.75 * i * (MA + MB) / Z < 1.2f

Where:

MB == resultant moment due to occasional loads

1.5.11 Special Considerations of Code Compliance

1 Many of the non-power codes separate the in-plane and the out-or-plane stress intensification factors (and do not intensify torsion). For the power codes the SIF's can be computed for in-plane, out-of-plane, and torsional moments using SIF :;: 0.9/ h2l3. For the petrochemical and other non-power codes:

13
•
•
A·
,
tt~;_, ./i
•
I)
Q
0
•
•
~
()
\-!
•
•
•
"
-.
•
•
•
•
<~
-,
0
CD
0
•
•
0
,J
u
•
•
• 9 COADE Pipe Stress Analysis Seminar Notes

e

,.

• •

o 0'

• • • o

o

e:_)

• • • • • • • • •

tf_)

Bends: 10 = 0.751h213 ii = 0.91h213
Miters: io = 0.91h213 Ii = 0.91h213
Welding tees: io = 0.91h213 Ii = 0.75io + 0.25
Reinforced tees: 10 = 0.91h213 li = 0.75io + 0.25
Unreinforced tees: 10 = 0.91h213 Ii = 0.75io + 0.25 2 - The power codes do not recognize the extruded welding tee, the sweepolet, or the weldclet.The non-power codes do, andifany of these fittings are used in a power application, the engineer must consider the validity of using the stress intensi-

fication factors from the chemical codes. -

3 - The power codes explicitly define the equation to use for the sustained stresses.

The non-power codes do not. The non-power codes do however tell the user to compute the longitudinal stresses due to sustained loads. and B31.3 Interpretation 4-10 issued May 8,1985 instructed the user to include the axial force term in this longitudinal stress. The power codes explicitly omit this axial force term from the definition of the sustained stress calculation.

1-57

4 - Power codes do not include pressure stiffening effects on bends, while the petrochemical and related-codes do.

5 - Note that the power codes use the term 0.75i in the sustained stress equation. while the non-power codes historically have not. In Interpretation 1-34 issued February 23. 1981 the B31.3 code permitted its users to employ the 0;75 i stress intensification term for sustained and occasional loads. (CAESAR n provides this as an option.) In Interpretation 6-03 issued December 14,1987, the B31.3 code permitted its users to ignore the stress intensification term. for sustained loads. It is recommended that this latest interpretation be ignored and that i or 0.75i be used as the stress intensification factor for sustained and occasional loads.

o

6 - Power codes provide special formulas for reduced branch connections. Nuclear and fossil codes have not come together on their interpretations as of this time, however. These rules come into effect whenever the branch diameter is less than 0.5 times the header diameter.

o • • o

o o

• • • Q

7 - Class 1 pi ping rules also allow flexibility coefficients to be computed and inserted in to reduced branch intersections in order to consider the flexibility of the branch relative to the header. No other piping code at this time includes this as an

,

option.

S - B31.3 was the first piping code to instruct the user to remove corrosion allowance from the section modulus before making sustained and occasional stress calculations. Other piping codes simply warned of the deleterious effect of corrosion when joined with cyclic loadings. Arguments are that B31.3 is


instilling a false sense of security when dealing with corrosion in this manner, implying thatB31.3 is leading the user to believe that he is properly considering the effects of corrosion. Other arguments state that B31.3 considers corrosion in the wrong stress calculation, and that if anything it should have been incorporated into the expansion stress calculation.

9 ~ Some of the piping codes include longitudinal weldjoint efficiency factors in their allowable stress tables. In the majority of the codes where these factors are included, the codes also instruct the user to divide the allowable stress by the joint efficiency before using the table value for flexibility calculations, thus increasing the allowable stress. Note that post~1980 B31.3 codes DO NOT

~, include the longitudinal weld joint efficiencies in the stress tables.

-:

10 ~ The piping codes are unanimously sllenton the point of Bourdon pressure effects.

If included, the pressure will cause-some distortion of the piping system. If excluded there will be no displacements due to pressure.

11 ~ European piping codes for the most part are formulated differently than the U.S. codes. The Europeans use an effective stress calculation and com pare the resul ts directly to an allowable without emphasizing the concern for fatigue. The Swedish piping code does have a provision that allows its users to employ the AS:ME B31.1~1977 code providing the- Swedish allowables are used.

12 ~ Almost all piping codes allow the exact expression for pressure stress to be used in place ofPdJ4 t in the sustained stress calculations. The exact pressure stress value is:

P df2 / ( do2 ~ di2)

13 ~ Most piping codes also allow the use of a increased section modulus for the stress calculations at the branch end of a reduced intersection. The reduced section modulus is calculated as:

z = (p) (r2) te
Where:
r =
te =
tnh =
tnb = mean radius of the branch, in

lesser of tnh or i tnb. in

nominal wall thickness of header, in

. "

nominal wall thickness of branch, in

14 ~ Note that the B31.3 and related piping codes do not intensify the torsional stress term in the expansion or sustained stress equations, while the power codes do. This is considered an oversight on the part of the code as Markl's tests clearly indicate that the torsional moment should be intensified. This was confirmed by the research documented in WRC 330 .

0
G
e
(j.
,
JI .... ...;
•
0
a
0
•
•
~
G
e
C~
.'
•
-- •
.~~ . .,.
•
•
•
•
•
{J
e
C3
0
•
•
0
o
0
•
•
• 6 (I

.~

• • o

o

• • • o

e

.:J

• • •

'.,.

• • • • •

rJ_)

@ Q

• • .,

o o

• • • o


15 - Stress indices are used in ASME Section III, Class 1, 2 and 3 piping codes. There are three different indices:

B

Related to gross plastic deformation (sustained stress intensification factor).

c

Gives the magnitude of the primary plus secondary stress range (elastic shakedown stress intensification factor),

K

When used with C, gives the magnitude of the primary plus secondary plus peak stress range (C*K is the fatigue stress intensification factor).

There are three subscripts used with the stress indices:

1

Used for pressure loadings

2

Used for moment loadings

3

Thermal gradient loadings

Asa rule of thumb, 2i = C2 * K2, where i is the stress intensification factor for the B31 codes (as discussed above).

16 - In almos t all cases, the cold modulus of elasticity and nominal dimensions are to be used in the flexibility analysis of piping systems. Using the cold modulus produces larger, and therefore more conservative stresses. BS 806 and ASME Class 1 codes have provisions for using rna terial properties in the hot condi tion. liEMA SM23 also provides for using the hot modulus of elasticity for eval ua ting loads on turoine nozzles.

1.5.12 Evaluation of Multiple Expansion Range Cases

It is often the cases that the temperature of the piping system is not constant throughout the operating cycle,or there is potentially more than one operating cycle (i.e., pump A on, pump B on, both pumps on, both pumps am,

In these cases it is common to find that the temperature rises on some occasion to a maximum value, say T e; then, as events occur during the normal course of operation the temperature varies between T e and other lower temperature states, say T1, T2, ...• Tn. In these cases the piping codes have provided a simplified method by which the cumulative damage due \0 the various thermal cycles may be evaluated by converting reduced thermal expansion cycles into equivalent full temperature cycles. The user will find that cumulative damage rules usually become important only either the number of thermal variations is large, or when the magnitude of the temperature variation is a large percentage of the maximum design temperature expected. The following rules should be followed when evaluating systems with multiple temperature variation cycles:

1-59

.. ,..=n::rt:CiiIIdP_ ...... ill

::::::::w:::a::s ..... ,

fA'! "fWrlfiif'i.

... --:::::::zt:lU_;n::w=;::w::::::a:4_ J. , 4.;


• T e should be selected as the highest operating temperature of the piping system, even if the startup cycle does not go directly to this temperature.

• The expansion allowable stress should be based on T e. i.e. SA should be calculated

from Sh for temperature T e· '

• The range dTe is determined as the difference between Te and the ambient temperature. N e should be estimated as the total number of times during the life of the unit that the temperature will be expected to vary from ambient to T e.

• \ The temperature ranges between T e and each of the other reduced temperature \ states should be calculated, i.e.:

Te-T2,.··

dTn =

• The number of cycles associated with each operating mode should be estimated as:

Temperature change dT1 occur N 1 times,

Temperature change dT2 occur N2 times, ... ,

Temperature change dTn occurs Nn times

• The total number of equivalent full temperature cycles that these partial cycles represent can be calculated as:

• The cyclic reduction factor f should be selected based upon the number of equivalent cycles, N, while other components of SA and SE should be based upon

temperature Te. :

Example:

A particular process line varies in temperature as the quality of the catalyst varies. The particulars of the operation are outlined below:

Ambient = 700F

Startup goes to 560°F

It is estimated that the maximum temperature ever required will be 6500F and the minimum temperature required during operation will be 4300F. The temperature 'Will fluctuate between 5600F and 6500F perhaps 10 times per day, and between 5600F and 4500F perhaps 5 times per day. The design life of the unit

1-60

o

• •

n
()
I)
r~
1......
•
•
.- •
{: ••••
•
•
•
•
0
0
0
0
• • e

'.)

o

• • •

...,..

• COADE Pipe Stress Analysis Seminar Notes o

Ne = 1 shutdown/month x 12 monthsiyr x 12 yr = 144

• • • o

o

• • • c

G)

.c)

is 12 years, and it is estimated that the unit will be shut down at least once each month for maintenance.

•

T e should be selected as the highest operating tem perature of the piping system. In this case, it is equal to 6500F.

The range dTe is determined as the difference .between 6500 and the ambient temperature of700F, so dTe = 5800F. The estimate of Ns, the total number of times that the temperature will be expected to vary through this range is:

•

•

The temperature ranges between T e and each of the other reduced temperature states are:

N = 144 + (90/580)5 x 43800 + (200/580)5 x 21900 = 255

Tl = 560oF; so dTl = 6500F - 5600F = 900F

•

The number of cycles associated with each operating mode are:

• • • • • • • • • .C)

Nl = 10 times/day x 365 days/yr x 12 yr = 43800 N2 = 5 times/day x 365 days/yr x 12 yr = 21900

•

The total number of equivalent full temperature cycles is:

•

The cyclic reduction factor fis selected based upon 255 cycles, so f = 1.0 (for fewer than 7000 cycles). As noted, the material allowable stresses SA and Sh are based . upon 650oF, and the expansion stress. SEt is calculated for the system operating at 650oF.

o

Warning: These cumulative damage rules don't fully address those cases where one part of the piping system stays at T e while another part of the piping system undergoes a temperature fluctuation. In these cases it is common to simply analyze each case separately. The ASME Section III. Subsection NB (Nuclear Class 1 Piping) Code provides rules which may be followed should the user be concerned about the cumulative damage where different parts of the piping system cycle through different temperature states. The requirements are described below:

\

Q •

• 41

CI o

• •

Cumulative Damage: If there are two or more types of stress cycles which produce significant stresses, their cumulative effect shall be evaluated as stipulated in Steps 1 through 6 below:

1 ~ Designate the specified number of times each type of stress cycle of types 1, 2. 3, ... , n, will be repeated during the life of the component as nl.nZ. n3 .... , nn, res pectively. In determining n 1, nZ. n3 ...• nn consideration shall be given to the

•

1-61

;Lab

_ #.$OC ..... m .. n_ ...

~UL .... E

;... QJttw:z4=wL = ==


superposition of cycles of various origins which produce the greatest total alternating stress range. For example, if one type of stress cycle produces 1000 cycles ofa stress variation from zero to +60,000 psi and another type of stress cycle produces 10,000 cycles ofastress variation from zero to"50,OOO psi, the two cycles to be considered are shown below:

(a) Cycle type 1: nl=1000; and Saltl=(60000+50000)/2; (b) Cycle type 2, n2=9000; and Salt2=(50000+0)/2

2 '! For each type of stress cycle, determine the alternating stress intensity Salt, "i· which for our application is one half of the range between the expansion stress cycles (as shown above). These alternating stress intensities are designated as

Saltl, Salc2, ... , Saltn'

3 On the applicable design fatigue curve find the permissible number of cycles for each Salt computed. These are designated as N 1, N 2. , .. , N n'

4 For each stress cycle calculate the usage factors U 1, U 2, .. " Un, where U 1 = nIl

NI, U2 = n2!N2, ... , Un = nnlNn·

5 Calculate the cumulative usage factor U as U = U 1 + U2 + ... + Un·

6 The cumulative usage factor shall not exceed 1.0,

e
rl·
'-.~.J
•
()
0
,)
•
•
0
<')
(}
4
•
•
•
~.
',:i •
•
•
•
•
•
C6
0
'0
•
•
0
: ....
.._,
V
•
•
• •
0
0
I·
I
I.
I.
I
:0
I
0
•
.• 0
•
I·
1.
t
).
'.
•
) .
•
I.
•
C •
••
, ('
· )
).
0
Ow
z w
.~ :::;
w ~
I-
ez w
w
u _J
<.:l
OLtJ Z
Z -c
o~ ®
\ ::>
, I-
q_ -r
• o

N

x

>- ------(

'-to/ f
9
o
•
•
"
()
'-
0
•
•
•
0
0
J)
•
•
•
•
•
•
•
•
•
0
•
0
0
•
•
"
0
0 ~
•
•
• CAESAR II Ver 3.24 Job: SNFAIL

1

DYNAMIC ANALYSIS INPUT DATA

Analysis Type (HARMONIC/SPECTRUM/MODES/RANGE/TIMEHIST)

Static Load Case for Nonlinear Restraint Status Stiffness Factor for Friction (D.D-Not used) Max. No. of Eigenvalues calculated (O-Not used) Frequency cutoff (HZ)

T ? <-----
? <-----
0.0 <-----
0 <-----
55 33 <----..,
1.8 0.1 <-----
1.0 20 <-----
.02 0.03 <-----
20 0.5 <----- Closely spaced Mode Criteria/Time History Time Step (ms) Load Duration (Time History or DSRSS method) (sec.) Damping (Time History or DSRSS) (ratio of critical)

ZPA (Reg. Guide 1.60 -- g's)/# Time History output Cases

N <----- Re-use Last Eigensolution (Frequencies and Mode Shapes)

MODAL <-----
SRSS <-----
GROUP <-----
Y <-----
Y <-----
ASS -<-...,---
SRSS <-----
ASS <----- spatial or Modal Combination first spatial Combination Method (SRSS/ABS)

Modal Combination Method (GROUP/lO%/OSRSS/ASS/SRSS) Include Pseudostatic (Anchor Movement) Components (Y/N) Include Missing Mass Components (YIN)

Pseudostatic (Anchor Movement) Comb. Method (SRSS/ABS) Missing Mass Combination M~thod (SRss/ABS)

Directional combination Method (SRSS/ABS)

Y <----- Sturm Sequence check on computed eigenvalues (YIN)

6 <----- Estimated no. of significant figures in eigenvalues

lE-12 <----- Jacobi Sweep Tolerance

lEla <----~ Decomposition Singularity Tolerance

Subspace size (O-Not Used)

No. to converge before Shift Allowed (O-Not Used) No. of iterations per shift (O-Pgm computed)

% of iterations per shift before orthogonalization Force orthogonalization after convergence (Y/N) Use out-of-core eigensolver (Y/~)

Frequency Array Spaces

a -<--_-_-
2 <-----
0 <-----
a <-----
N <-----
N <-----
100 <----- ***** NOTES *~*** NOTES *-~** NOTES **~** NOTES ***** NOTES **~** NOTES *--**

The frequency cutoff is set so that the maximum contribution of each of the four components of this event are included. The Pulse spectrum generator may be used to calculate the maximum Dynamic Load Factor (DLF) for each component. Examining the results shows that the fourth component is the last to peak at a value estimated at 55 Hz.

For accurate analysis, there should be ten time steps (analyses) over the highest mode of vibration used in the analysis. Here that means 10 time steps over 1/55 seconds; where 1/55 is the period of the highest frequency of S5 Hz. one tenth of 1/55 seconds is 1.8 milliseconds.

Minimum load dUration should be greater than event time plus one quarter of the fjrst period of vibration to ensure that the system response from each mode is fully developed. For this system, the first peri9d is 3.134 secpnds, therefore, the minimum analysis duration is 123 ms + 784 ms or 907 ms.

Round up to 1.0 seconds.

~wenty output cases are chosen as that produces a report every 50 ms.

Missing mass is automatically includeg to account for the system response 'which is not contained in the calculated modes of vibration. The remaining ~rnass" (in this case, force) which is not part of the modes below'55 Hz uses a DLF of 1 or a DLF-equal to the DLF of the last extracted mode depending

on the missing mass parameter set in the proqram configuration -- SPECTRUM or EXTRACTED, respectively.

Page:

CAESAR II Ver 3.24 Job: SNFAIL

•

•

•

SPECTRUM DEFINITIONS

3

G 1 1 o

* FOUR SHOCKS 'ARE SPECIFIED FOR THIS STEAM HAMMER ANALYSIS * TH1 MODELS THE FORCE IMBALANCE ON THE VALVE (30) AS THE, * PRESSURE WAVE TRAVELS TO THE FIRST ELBOW (100)

* TH2, TH3 & TH4 MODELS THE LOADS ON THE THREE UPSTREAM ELBOWS * 100, 110 & 120

o 10 66 76

TH1,T,E',LIN,LIN

o * RISE TIME IS 10MS

*IT TAKES 66 MS FOR THE

o * WAVE TO MOVE FROM * 30 TO 100

o

.0

•

•

•

•

•

•

•

•

•

,J:;\

•

•

TH2,T,E',LIN,LIN

* EVENT AT NODE 100 *-STARTS AT T·= 66 MS * EVENT LASTS 26 MS

* WITH 10 MS RISE TIME

o 66 76 92

102

o o 1 1 o

TH3,T,E',LIN,LIN

* EVENT AT NODE 110

* STARTS AT T = 92 MS * EVENT LASTS 15 MS


.'

o o 1 1 o

•

•

•

o

92 102 107 117

TH4,T,E',LIN,LIN

* EVENT AT NODE 120

* STARTS AT T = 107 MS * EVENT LASTS 6 MS


* (PEAK or 1 NOT REACHED)

o 107 113 117 123

o

o 0.6 0.6 o

•

***** NOTES ***** NOTES ***** NOTES ***** NOTES ***** NOTES ***** NOTES *****

These four time histories are each run through the pulse spectrum generator to exarnione the frequency at which the responses peak (attached). The highest of these four frequencies may be used to conservatively set the frequency cutoff for the analysis. This value is also used in calculating the time step -- Time Step = 1 / (10 * highest frequency).

o

•

•

•

9 a

• • •

1633

e::?

.c

.""D

co

-.3

o

o

to

•

•

•

•

• 4433

•

c o

- •

• c

• • •

-I-

Imbalance on leg 1, 2, 3, 4
I I I i o

50

100

150

Time (ms)

- 0.6(1433)

Imbalance on leq.t,

o

50

100

150

Time (ms)

0
~~lIn
4) S)"'NN
lSI .........
... 111
• I 'ISI~ISI
~~"'C\l
M ...... ooi
• :0:+++
•
0
61
•
•
•
~
1-1
0 ~<I
C/'J~
0 ZZ
O(,/)
Ilo.
.0 C/'JZ
~O
~
• p:
:>141
U::C
• Z::C
~<I
:=.::t
• 13'
WI:
p:a:
• ~4l
rot
(,I)
•
•
•
•
•
~)
0
0
•
•
•
e
()
•
•
• i

~

.. , .... , .... , .... , .... , .... en

.... +.+~ ~ ~ ~ ~ ~ r"".~ ••

. . .

, .••.•...••••..•....••••••..•• lSI

Ol

... -: .... -: .... -: .... -: ..... :. .... :. ... -:. ... -: ..... :. .... ~

• . ISl

~ .. ~ , ~ .. ,.~ .. ,,~~~ ~ I.D

.-

.. -: .... -: .... -: ..... :. .... :. .... ; ... -: .... -: ..... ~

. . .+~: +~; ~ : ~ :

.. . ~ .

• + + • • •

. • • . •.. ' ••••• '. • • : • •• .•.••...•.•••••••. IS)

""'

..••••.• , .•..•..•.•...•••••. lSI

C\l

N

!';~#;~ff~W~;i~~~'1i'[:;j~;!i,f,~tM;~,?,~il'.:,.~~3#~~~~_7;'~~:;!j·,~~{~W\;;~~·~\:b:~;~:';~;;~";~~';;;\i:i'~'"jiM',,;ij\-;,:''.;~;,;~j~j;,!':;[:2"

,.:.": -

Establishing Time History Input - An Example

' ..

~roductjan

~ure I illustrates the layout of the piping system named S'NFAIL. This job examines the dynamic load on a snubber i.,alled to reduce the system response to a steam hammer traveling through the system. The snubber has a rated load cePacity of 2.500 pounds and the analysis will predict whether or not the snubber will fail (ergo the name, ~FAIL). In the past, this situation was investigated using the force response spectrum method of analysis. Since this .tem has a specific definition of the dynamic load timing, it is a good candidate for time history analysis. Events with • andom duration and content such as earthquakes and relief valve loads are best suited for the general analysis letnd ill' response .spectrum methods. More precise load definition yields better results through time history analysis. Gsponse spectrum analysis combines the maximum response for each mode' of vibration to generate a total aponse without respect to time. Time history follows the ~'.~\ responses t.1'Irough each time step to produce a system ~nse with respect to time. Time history analysis also allows the coordination of separate events in the system. In ~FAJL, the steam hammer causes a pressure imbalance in ii_ch upstream leg one after the other. Force response ~ectrum analysis cannot maintain this sequence of events ~Iile time history can. The load characteristics of SNFAIL "!'I'ake it an excellent candidate for time history analysis.

tere are three basic categories of dynamic analysis input il,r CAESAR II - modification to the static analysis model, ?efiniiion of the dynamic load, and specification of the ,alysis parameters. The model developed. for a static analysis needs only the snubber definition for the dynamic iilalysis. This iaformation is entered through Item 2 - S'NUBBERS of the- Dynamic Analysis Input Menu. A one aillion lbf.lin. stiffness is entered in the X direction at node 1M. (The input is entered as: 1 E6. X. _I 00.)

~(' Definition

~~./analysis examines the effect a sudden valve closure at e= 30. Steam is flowing through the line at 45 fps and suddenly stops. To make matters simple, it is assumes that <lie valve closes instantaneously. This assumption will allow a conservative guest as to the pressure rise in the line. At .e instant the flow stops at node 30, all the upstream vapor is still moving at 45 fps towards node 30. Each molecule of .por will keep on moving towards the valve face and only stop once it is acted upon by the building mass of stopped .por. The inertia of the moving vapor is converted into the higher density and pressure of the stopped vapor. So as the Olpor is moving down the line it is stopped by a pressure wave traveling upstream. This pressure wave is traveling ~stream at the speed of sound in the line. The CAESAR II analysis will follow this pressure wave up the line.

•

•

•

At steady state conditions, the pipmg system is in

. equilibrium. One component of the load On the system is the pressure effects on the straight runs of pipe. One was "of visualizing this pressure load is to crawl inside the pipe and look upstream and downstream. The surfaces you can see are the surfaces on which 'the line pressure ad. Again. at constant upstream and downstream pressure, the loads are .in equilibrium - (Pup) • (Aup) = (P dn) * (Adn): Pup = Pdn = P & Aup "" Adn• But as the pressure wave travels up the line into your straight run of pipe, that run of pipe is no longer in equilibrium. The downstream elbow (or valve face) is now loaded with P "'- 6.P where 6.P is the pressure rise due to the instantaneous valve closure. This imbalance (6.P) will continue until the pressure wave passes the upstream surface' at which time the run is again in balance (at the higher pressure).

Snubber Failure (SNFAIl)

SNFAIL.OOC

~~~~i[~1d~~~~~~~~'~~~~~4~~'i.¥~~~f:~~I,~~~lt;*~$!'l!t~~~~~~~

r ~. ....

•

~ .

1 ,

,""

12 in. Oia .• Sch 80 Lew Carbon Steel T" 7S0F

? .. 5oo psla In.s.ul,, 3.0 in. CS Use 631.1

Sc" 17000 psi Sh .. 10950 p!li

S1eam Flowrale .. 45 Ips

~

Tbe designer of !he

orig ina I svstem seled.e Ii a snubber at r.ode 100 rated for a 2SOO til. load.

Wi H thIs be sulfici&lllif the angle valve (alSO) is shvI instantly?

Figure 1

Establishing Time History Input - An Example

3

" •

• •

Set #2 (definition) TH2. T. F. LIN. LIN

Time

Load

(data)

o 66 76 92 102

o o 1 I o

o

Set #3 (definition) TH3, T. F. LIN. LIN

• •

(data) Time Load
0 0
92 0
102
107 1
117 0 •

o

Set #4 (definition) TH4, T. F. LIN, LIN

o

.0

(data) Time Load
0 0
107 0
113 . 6
117 .6
123 0 •

•
Item 7 - SPECTRUMITIME HISTORY FORCE SETS
•
Load Direction Node Force Set
•
-1433 X 35 I
• -1433 Y 100 2
1433 Z 110 3
• 1433 X 120 4 .em 9 - SPECTRUM,ITIME HISTORY LOAD CASES
• Time Hisr. Factor Dir. Force Set #
• THI X 1
TH2 Y 2
.C) TH3 z 3
TH4 X 4 e

One note on the magnitudes or factors specified above.

oThere are three CAESAR II components which combine (Q form the total load used in the time history analysis:

•

a) the force in the time history definition (lor .6 in the time history definitions);

b) the force in the force set definition (1433 in the time history force sets); and

c) the scale in the dynamic load case definition (I in the lime history load cases).

•

•

CWith the total load calculated as ( a * b * c ), us.e any convenient way to enter these data, for the three terms.

•

•

•

Analysis Parameters

There is only one more Dynamic Analysis Input item required to run this time history analysis in CAESAR 11- B - CONTROL PARAMETERS. While some of the settings are obvious (e.g, analysis rype > Time History), several analysis parameters have a range of possible values - some having a major impact on the accuracy of the analysis. What is the "correct" entry for frequency cutoff. time step and load duration'? Should missing mass be used. and if so. how should it be included? How important is damping? Guidelines for these data are provided here.

FREQUENCY CLrrOFF ., The frequency cutoff establishes [he number of modes of vibration included in the analysis. With each additional mode included. the accuracy of the analysis improves. But there are diminishing returns with the higher modes - the analysis takes much. much longer with very little change in the results. So how many modes to include? A conservative approach is to examine the frequency content of the event and analyze the system at least to the point where the frequency response is maximum . The system frequencies not included in the analysis will be approximated by the missing mass correction (see M.E.N. VoL 16 - "Use the Missing Mass Correction Option in Spectral Analysis"). Fortunately the Response Spectrum Generator in CAESAR II (hem 4 in the Dynamic Analysis Input) can be used to convert the four time histories into their frequency response curves. The response curve data. are shown in Figure 3. This plot was generated using.the CAESAR II 2-D Plotting package (Item F on the Main Menu).

-~

~ U h _~ d ~ U ~ _ r.-'""un.t't

Figure 3

The response curve for the first event (30 - I 00) has the highest response and it peaks at the lowest frequency. This is reasonable as this is the longest duration event and therefore can "activate" lower modes of vibration. The first

SNFAIL.DOC

!!;;:"igN&';i;~~;f,[,~,j';;:;i';iiiif"ifS,";,;~,!;~j~~¥:!:~t~~f~pt<f~~!lJ~~~)i,1;\(''';!l;;~:.;i~,,~J:;i,!;i'~;~~4~f"'I!!t{~~~;:iU;\~:~!7K};;~;~:;;jl~~~;!~f,

, . ":: '·i~'"

Q
0
•
•
•
0
0
•
•
•
0
('I
C)
•
•
•
I
•
•
•
•
•
•
0
0
0
•
•
•
e
0
.
•
• SECTION 2

\

\

•

it
e
•
•
•
o
0
•
•
•
0
0
-:J
•
•
•
•
•
•
•
•
•
Ci
0
0
•
•
•
0
G
.'
0
•
•
o COADE Pipe Stress Analysis SemjnarNotes Section 2

Table of Contents

2.0 Piping Design For Loading Types 1

2.1 Designing For Sustained Loads - Pressure 2

2.1.1 Minimum Wall Thiclmess Requirements 2

2.1.2 Pressure Design of Elbows and Miters 4

2.1.3 Pressure Design of Flanges and Blanks 5

2.1.4 Pressure Design of Branch Connections 6

2.1.5 Restraint of Unbalanced Expansion Joint Pressure Loads 8-9

2.2 Designing For Sustained Loads - Weight 10

2.2.1 - Calculation of Weight Stresses 10

2.2.2 Use of Standard Weight Spans 13

2.2.3 Consideration of Nozzle Loads 19

2.3 Designing For Expansion Loads 22

. .

2.3.1 Magnitude of Thermal Load : 22

2.3.2 Guided Cantilever Method ' 24

2.3.3 Refining the Model Through the Use of Restraint Stiffnesses 26

2.3.4 Use of Expansion Loops 27

2.3.5 Simplified Expansion Stress Check ; 29-30

2.3.6 Stress Reduction through Use of Expansion Joints 30

2.3.7 Expansion Stress - Other Solutions 33-33

2.4 Hanger Design 34

2.4.1 Variable Spring Hanger Design Basics 35

2.4.2 wad Variation 37

2.4.3 Hanger Selection Table 37

2.4.4 Hanger Design Process - Restrained Weight, Free Thermal, and More .. 39

2.4.5 Restraint Placement Using Distance to First Rigid Criteria 40

2.4.6 Notes on Hanger Design 43

2.4.7 CAESAR II Hanger Design Control and Options 45:,49

\

2.5 Designing For Occasional Loads (Static Equivalent of Dynamic Loads) 50

2.5.1 Wind Loading 50

2.5.2 Earthquake Loading 54-56

2.5.3 Quickly Applied Loads 56

'"
0
•
•
•
()
e
•
•
•
0
0
(J
•
•
•
•
•
•
•
•
•
0
(3
Ii)
•
•
•
0
0
, .
8
•
•
0 COADE Pipe Stress Analysis Seminar Notes

2.0 Piping Design For L.oading Types

As described in Section 1.0, the pipe stress analyst is concerned with two types of loads - primary and secondary. Not only are the causes and the failure modes of these two loading types quite different, but not surprisingly, the solutions to these two types of loading are usually different as well. In fact, the solution to a problem caused by one of the loading types often ca uses a problem with the other loading type. Therefore, a compromise must often be reached in order to find the solution to these two types ofloading.

Note that primary loads are usually classified further, according to their duration ofloading. Those primary loads which are nearly always present throughout operation are called sustained loads, while those which occur less frequently are called occasional loads. The methods of resisting these two types ofloads are similar, with the main difference being found in the use of a higher allowable stress for occasional loads (as seen in Section 1).

2-1

. -.; . .;; ' .. ,. "r' ~-. • • . . , - ,. •


2.1 Designing For Sustained Loads - Pressure

All piping systems must be designed to withstand sustained loadings. Sustained loads are classified as those caused by mechanical forces which are present throughout the normal operation of the piping system. Therefore these loads:

• Are force driven, as opposed to displacement driven, and

• Are present for relatively extended periods of time, as opposed to those which change dynamically.

Typical sustained loads consist of:

• Pressure -loads due to operating (or design) pressure,

• Weight - uniform loa ds due to the weight of the pipe, fluid, and insulation, and concentrated loads due to the weight of in-line components (such as valves, flanges, etc.), and

• Spring hanger pre-loads and other applied forces.

2.1.1 Minimum Wall Thickness Requirements

Since hoop pressure stresses are approximately twice as large as longitudinal pressure stresses, pipe wall thicknesses are initially sized for hoop stresses. Because ofthis, pressure design of components is usually done far before, and therefore in isolation, from the pipe stress analysis phase of piping design. Because of this, pipe stress software such as CAESAR IT does not normally handle this part of the design effort. A discussion of pressure design of components is included here for the sake of completeness, and is based upon an amalgam of the requirements of various codes. Note that pressure design of piping components must be done according to the requirements of the user's specific code, not to the rules described here!

Because the pipe wallis sized for the large hoop stress, this usually provides sufficientmargin between the allowable stress and the longitudinal pressure stress to accommodate the weight stresses. The requirement for the minimum pipe component wall thickness is:

tm == t + c

Where:

tm == minimum wall thickness, in

t == minimum wall thickness required for pressure, in

c == sum of allowances for thread or groove depth, corrosion, erosion, and

manufacturer's tolerance, in

2-2

0
•
•
O·
-~,~. •
'"
0
0
•
•
..
(')
Q
o.
•
•
•
'.
•
•
•
•
•
O.
~
0
•
•
0
0
iJ
•
•
• £)
e
~
•
•
0
e
•
•
•
0
0
(J
•
•
•
1
•
•
•
•
•
•
0
G
G
•
•
•
0
0
•
•
•

For thin wall (t < D/6), straight pipe under internal pressure, t may normally be calculated, through various approximations of Lame's equation, as:

t = PD I 2(SE + PY), or:

t = PD I 2SE, or:

t = (DI2) x (1 - [(SE - P) I (SE + P)J1I2), or:

t = P (D, + 2c) I [2(SE - P(l·y)l

""'here:

P = design pressure, psig

D = outside diameter, in

Di = inside diameter, in

S = basic allowable stress at design temperature, psi

E = casting or longitudinal weld quality factor (typically ranges from 0.8 to

1.0)

,Y == material coefficient, with a value (depending upon the specific code

requirements) to be interpolated from:

Temperature. _oF
Materia1 <"'900 950 1000 1050 1l0O 1150 1200 >1250
Ferritic 0.4 0.5 0.7 0.7 0.7 0.7 0.7 0.7
Austeniti c 0.4 0.4 0.4 0·.4 0.5 0.7 0.7 0.7
Ni del Alloys 0.4 0.4 0.4 0.4 0.4 0.4 0.5 0.7
Other ductile 0.4·- 0.4 0.4 0.4 0.4 0.4 0.4 0.4
Cast iron O. Re quirements for pressure design of other piping components are described in the foll~wing sections. (For B31.3 y = 0.0, for B31.1 y =.4. The CAESAR n check uses 0.4 for all codes

except B31.3, where y = 0.0.) 1

2·3

.J.t 4tl)&WDlBt.5 US # tt M,.:i '&4hLa.&CAGa~='aJ4#bL2

C)S$!l

2Q =:e za ne.xesu.


/

10 • •

2.1.2 Pressure Design of Elbows and Miters

o·

• n

e Q

• • fIJ

When using elbows, the minimum wall thickness after bending shall not fall below that calculated for straight pipe.

For mitered elbows, the maximum allowable pressure is calculated differently depending on whether the angle of the miter cut is less than or greater than 22.5°.

For 0 < 22.50, the allowable maximum pressure is the lesser of:

Pm I = [SECT w C)/r2J x [(T . c) I (T w c + 0.643 tan 0 (r2(Twc))1I2)]

or:

For 0 >= 22.50, the allowable maximum pressure is:

Pm Where:

= [SE(T w C)/r2J x [(T . c) / (T . c + 1.25 tan 0 (r2(T-c))lf2)]

o O.

• • •

••

• • • • •

08

Pm = maximum allowable internal pressure for miter, psig

T

= minimum pipe wall thickness, in

r2 =:: mean pipe radius, in

. 0

= effective radius of miter bend (defined as the shortest distance from pipe center line to the intersection of the planes of adjacent miter joints - see Figure 2·1), in

= angle of miter cut (see Figure 2·1), degrees

e Q

• • • o

u • • •

"

e
0
•
•
•
Q
0
•
•
•
0
o
Cl
J
•
•
•
•
•
•
•
•
•
0
0
0
•
•
•
G
(3
--
•
•
•
e COADE Pipe Stress Analysis Seminar Notes

2.1.3 Pressure Design of Flanges and Blanks

Pressure design offlanges is a complex task, requiring consideration of the configuration and materials of the flange, bolts, and gasket. Potential causes of failure are bending stresses in the flange, localized stress concentrations in the hub, yielding of the bolts, or unloading of the gasket, causing leakage. Design offlanges is covered in detail in Section vm of the .4..SME Boiler an Pressure Vessel Code; however, due to the complexity, it is rarely done by the pipe stress engineer. Instead, the most common piping codes endorse the use of flanges conforming to recognized standards such as ANSI B1G.5 "Pipe Flanges and Flanged Fittings". This standard designates standard pressure classes of flanges, which are recognized by the codes to be acceptable for the following combinations of pressure and temperature:

Design Pressures (psig) for Flange Pressure Classes
Pressure Class
Temperature. OF 150 v 300 400 600 900 1500
100 275 720 960- 1440 2160 3600
150 255 710 945 1420 2130 3550
200 240 700 930 1400 2100 3500
250 225 690 920 1380 2070 3450
300 210 680 910 ' 1365 2050 3415
350 195 675 900 1350 2025 3375
400 180 665 890 1330 2000 3330
. 450 165 650 870 1305 1955 3255
500 150 625 835 1250 1875 3125
550 140 590 790 1180 1775 2955
600 130 555 740 1110 1660 2770
650 120 515 690 1030 1550 2580
700 110 '; 470 635 940 1410 2350
750 100 425 575 850 1275 2125
800 92 365 490 730 1100 1830
850 82 300 400 600 900 1500
\.
900 70 225 280 445 670 1115
.,
950 55 155 220 310 465 770
1000 40 85 160 170 255 430 2·5

;' COADE Pipe Stress Analysis Seminar Notes

A more detailed discussion of flange analysis, with specific regard to determination of leakage under load, is provided in Section 3 of these seminar notes.

Blanks are designed based upon formulas for the calculation of bending' stresses for plates under pressure loading. The minim urn thickness for a blank is calculated as:

t -

m

= dg [3P I 16SE]112 + c

Where:

dg = inside diameter of gasket for raised or flat face flanges, or gasket pitch

. diameter for ring joint and fully retained gasketed flanges, in

2.1.4 Pressure Design of Branch Connections

A pipe having a branch connection is weakened by the opening that is cut in it, so it may be necessary to provide reinforcement to replace the metal removed from the wall thickness at the opening. A typical fabricated tee is shown in Figure 2-2.

Limiu.or

! ------tP.,.

k,

------p--------- ---

Figure 2~2

For fabricated tees, with the angle between branch and header of at least 45°, the area required to replace the area of the opening is calculated as:

0
t'I
•
n·
~,~,: ... )~
•
"
@
0
•
•
0
0
e
4
•
•
-. - •
",
~;I": •
-.
•
•
•
•
•
q
Q
~
•
•
0
0
0
•
•
• Al = area required to be replaced, in2

e
~
-.
•
•
0
0
•
•
•
o
0
(J
•
•
•
,
•
•
•
•
•
•
0
0
(I
•
•
•
0
0
.
•
•
•

Where:

th = pressure design thickness of header pipe, in

dl = effective length of pipe wall removed from header at intersection, in

~ = smaller angle between axes of branch and run, degrees

This required area must be exceeded by the total available reinforcement area, or:

Where:

= (Tb· c) + (Th - c) + d1/2, but_ not less than dl

A2 = area resulting from excess thickness of header pipe, in2

d2 = half-width of reinforcement zone, in

Th = minimum wall thickness of header, in

Tb = minimum wall thickness of branch, in

Aa = area resulting from excess thickness of branch pipe, in2

L4 = height of reinforcement zone outside of header. in

= lesser of 2. 5(Th - c) or 2.5(Tb - c) + T r

tb = pressure design thickness of branch pipe, in

T r = minimum thickness of reinforcing ring or saddle, if any, in

~ = area of welds and reinforcement provided for the intersection wi thin the area

of reinforcement as defined as a parallelogram extending a distance of d2 on either side of the centerline of the branch, and from the inner wall of.the header pipe to a distance of La along the axis of the branch, measured from

the outside of the header pipe. "

2-7

".,:: .. j-


2.1.5 Restraint of Unbalanced Expansion Joint Pressure Loads

Pressure usually only creates stress in the pipe, rather than loadings on supports/restraints, because pressure loads are neutralized at the_cross-section by the tension in the pipe wall. One exception to this is when the pipe is not continuous from' anchor to anchor, such that tension is not present in the pipe wall at all locations of the system. (Note that a second exception occurs when the Bourdon effects of pressure are considered. The, Bourdon effect is due to the axial extension of pipes either under high pressure or in long runs, causing displacements which must be absorbed by the piping system. Since this is a displacement load, it is a secondary load, and therefore is not considered here.)

\

" .

Tension "In the pipe wall is not continuous when there are expansion joints or slip joints present in the system. These types of components are too flexible in the axial direction to transmit the pressure force, therefore the unbalanced pressure load must be handled by ~either tie rods or external pipe restraints. The unbalanced pressure load is calculated as:

Fp = P Ae

Where:

Fp = Pressure force, lb

P = Design pressure, psig

Ae = Effective area of expansion joint

De = Effective diameter of expansion joint, in

= internal diameter of pipe + depth of one corrugation, in

When using restraints to absorb the unbalanced pressure load, it is recommended that guides be located on the adjacent pipe runs in order to reduce the danger of buckling. The Expansion Joint Manufacturers Association recommends that the first guide be placed a distance no further than 4 pipe diameters from the expansion joint, with the second guide placed no further than 14 pipe diameters from the first.

Figure 2-3 shows some typical piping layouts using expansion joints.

0
•
•
n·
)
'. ;., ,~~
•
__ ,0
0
0
•
•
f!t
0
0
C:.
•
•
•
--
•
•
•
•
•
•
q.
0
g
•
•
0
0
\)
•
•
• 9
€)
•
•
•
0
e
•
•
•
C
0
()
•
•
•
,
•
•
•
•
•
•
0
Ci
0
•
•
0
0
0
- -
0
•
• CO/illE Pipe Stress Analysis Seminar Notes

• D=Pipe 0.0.

~ ISf: (~.58m)

, 15f~ (IS.25m) i -j

/, A I • , Verticol Supporl

2 X

~\ B '/ Pipe

\ Exponsion! I

Anchor Joint ~ C

~

~ (

I I •

-";40 ~'4D"-:

I : I

, .

~~~IJ--; -rifJ

inchor \ 2nd Guide \

lsi Guide All Olner Guides

Figure 2-3

More information on the use of expansion join ts is found in Section 2.3.6 and Section 3 of these

seminar notes. -,

\

2-9

@

, ~ ,. :_'

COADE Pipe Stress Analysis Seminar Notes i

2.2 Designing For Sustained Loads - Weight

2.2.1 Calculation of Weight Stresses

Stresses due to weight loads acting on a supported pipe can be estimated through the use of beam theory. The simplest method of estimating pipe stresses due to weight is to first consider the pipe as being a con tin uous run, with supports loea ted at constant intervals (this is a somewhat accurate model of piping traveling horizontally. mounted on racks. and with a minimum. of in-line components):

5 JE 1 1 1 )
:.--.-. e e t-'

Figure 2-4 Elementary beam theory can be used to determine stresses in a member due to loading on that member. Normally the member considered is one-dimensional. homogenous with respect to cross-sectional and material parameters, and restrained in a number of degreesof-freedom at one or both ends. This model can only be used if the effects examined are limited to two adjacent support points and the straight run of pipe between those support points. The question is what beam stress equation should be used?

Beam theory states that if both support points are pinned (free to rotate):

w

jJltttttIJfJI f{I II t IJIJIlltftl

Figure 2-5

The maxim urn moment in the beam is in the center of the span, and has a value of:

Mmax = w12/8

where:

Mmax = maximum moment in the beam, in-lb

w = uniform weight of pipe, fluid, insulation, etc., Ib/in

I = length of hearn, in

2-10

0
G
•
0:
(')
0
a
•
•
fit
(')
0
l).
•
•
•
~.
!I ,.
"
•
•
•
•
•
()O
Q
0
•
•
0
0
I,,)
•
•

.. If both ends are fixed, or rigid (restrained against rotation):

•

• o

o

• •

~ II iii i I f I i I j 1I11I11I111 f IIIl ~

Figure 2·6

The maximum moment is at the ends of the span, and has a value of:

• Mma = w12/12

o o

tJ

• • •

'. •

• • • •

C)

o e

• • •

o

G • • o

Which formula is more appropriate? Examining a typical pipe support detail:

Figure 2-7

The clamp/pin/rod hardware allows rotation of the pipe" therefore simulating a pinned connection. However, if all spans are of identical length and loading, the reaction of the adjacent pipe span prevents rotation at the support, therefore simulating a fixed connection. The true condition is somewhere in between, so a compromise approximation is reached:

Mmax = w12/10

, with the location of the maxim um moment being somewhere between the ends and the center (i.e., anywhere) on the span.

2-11


Of course, there sometimes are concentrated loads (valves, flanges, etc.) in the pi ping system. The effect of these on the pipe stresses can be estimated as well. For pinned connections:

p

f----- 0 ---'-+j-'- b I

..c..1'<"------[>!<]r---~A

~------e-----~~

The maximum moment is located at the point ofloading, and has a value of:

Mmax = Pabll

Where:

a = longer portion of span, in

b = shorter portion of span, in

For fixed connections:

~~---------e----------~

Figure 2-9

The maximum moment is located at the end nearer to the load, and has a value of:

Mmax = pa2b/l2

In either case (or actually some case in between). the additional stress (M/Z) due to concentrated loads should be added to the stress from the uniform load in order to determine the total stress in the pipe. Or. examining the formulas above. it is evident that, as the shorter portion of the span (b) approaches zero length, the moment, and therefore the stress, approach zero as well. This points to an important rule of design - if supports are placed as near as possible to concentrated loads, the effect of these loads from a stress point of view may be neglected. (They must still be considered for the support loads. of course.)

0
•
•
0.-
'< .. .i'" •
t'JI
0
C
•
•
8
0
0
C).
-
•
" ....... ~ •
""
•
•
•
•
•
•
.0,
Q
o
•
•
G
0
G
•
•
• o
()
•
•
•
0
0
•
•
•
0
0
CI
•
•
•
!
•
•
•
•
•
•
C)
0
u
•
•
•
G
0
.-
0
•
•

2.2.2 Use of Standard Weight Spans

Implementation of the preceding analysis provides a simple way to design for weight loading. The engineer may first support all concentrated loads in the system as closely as possible, reducing the stresses due to those loads to near zero. Next, converting the formula Mmax = w12/10 into its corollary:

Lall = (10 Z SaIl / w)1I2

Where:

Lall = allowable pipe span for weight loading, in

Z ::: section modulus of pipe, in3

SaIl = approximate allowable stress of piping material for weight stresses (Sh, less

pressure stresses, divided by intensification factor. for example). psi

If the pi ping system is then supported, such that no straight span exceeds Lall. the engineer can be sure that allowable weight stresses are not exceeded in the system, and no analysis per se need be done.

In order to save even the brief. time required to calculate Lall, the Manufacturer Standardization Society of the Valve and Fitting Industry has calculated allowable piping spans for various piping configurations, arid published them in their standard MSS SF·69 (Figure 2·10). They have calculated the maximum allowable piping weight spans based upon the following criteria:

1 - the pipe is assumed to have standard wail, with insulation,

2 the maximum moment is calculated as Mmax = wI2/10.

3 - no concentrated loads are present,

4 ~ there are no changes of direction in the spans, which are assumed to run in the horizontal plane,

5 . the maximum allowable stress is assumed tobe 1500 psi, combined bending and shear.

6 - maximum deflection of the span under load is limited to 0.1", and

7 stress intensification factors of components are not considered.

Due to the low allowable stress value used. there is sufficient factor of safety that this standard span may be applied to a wide range of piping configurations.

If the engineer supports a- piping system such that no span in the system exceeds the standard spans listed in the table, it is virtually certain that the system is adequately supporte~ for weight loading. However, it is rare that a piping system has no concentrated


loads, consists of only horizontal runs with minimal changes in direction, etc. Therefore, standard practice dictates that standard spans be applied subject to the following four

caveats:

1 Supports should be located as close as possible to concentrated weights .... The theoretically best location for a support is directly on the concentrated load; however, . this is usually impracticaL

2 A developed length of 3/4 of the standard span or less should be used when the

piping run changes direction in the horizontal direction, in order to minimize eccentric moments. The theoretically best location for a support is on an elbow; , however, this is not recommended due to the bend stiffening and increased local stresses associated with attachments on a bend.

'\

13·
I iii
C
0-
.2
•
"
~
0
•
•
"
0
0
()a
•
• •

.. ,

'i-

•••

•
••
•
•
•
'At
\
0
0
•
•
0
0
I,)
~
•
• •

:~;~tf~~;g~~~~~~~\~~~~~~~~~~f~~~S~~~~'¥~.~

~ ~ .~

(d)
(9
•
•
•
0
0
•
•
•
0
0
0
•
•
•
r
•
•
•
•
•
•
0
G
.
0
•
•
•
Q
""-
'oJ
--
S
•
•
'-9 COADE Pipe Stress Analysis Seminar Notes

CJ Z o

f

en

~ a: o Q. Co :=I CI)

Q Z < a: au CJ Z < :r:

w Co

c::

...I

~

Z o t::!

cc 0' :r.

:E :;:)

:! x c(

:e

MSS

STANDARD PRAcnCE

SP069

VI
~ Q g
... • :;.J
- 1.:='" fOLWW PIPf WANUI'ACTURU'S RECONMfNOAnoNS FOIt SI'ACN(; AND '"
- "",,, SERVICE CONDITION. ::
",.::0
III "- i
lO:
~ '"
... FOI..1..O'lt' PI'E WANUF ACTURER'S RECOMWfNDAnoNS FOIt MATWAL AND i5
2 VI
::i SER VICE TENPEllA TURE.. II.l
.... ~
VI ..
..... 8 rt 1l-<4I11) MAXS'ACING. F01.LOW MANUFACTURER'S AECOJOIENDATIONS. 1S~
Ct s
SEf. SEcnON 18. <ct
o ~~
I '"
c ... ~'"
... , ~~
"'w rcuow PIPE MANUF ACTURER'S RECOMMENDATIONS.
"" "'::t:
I =w ~g
~v
~-
... z_. 10 it Cl.Om) MAX SP .... CING MIN OF ONE rn ltANGD. PER PiPE SECTION CLOSE 8~
.... "'0- TO JOINT ON THE BARREt. AlSO AT CHANGE. OF 01 RECI'ION AND BRANCH ~!i
<",0
v_'" CONN EcnONS.
...:ii!
~o; ; 12 ft f3.1m) WAXSPACINC MIN OF ONE (l)HANGD. PER PIPE SEO'ION CJ..QSE _ :!5
'" !!~5 TO JOINT ON THE BARREL. ALSO AT CHANCE OF DIIlECIlON_AND BRANCH ";!5~
j§!!f CONNECf IONS. ~~~
z ::i!!6
W • 0 FOLLOW REQUIRnlENTS OF THE NA nOHAl FIRE PROTECTION ASSOClA- wiw
'" $~6 .!;j iii: .:
~ ....... nON. SEESECTJON 14. • CI !i
I- §w%
Q-
W e "'l ~ ~ .. ... C!; -.: CI ... '" Ct:: ... c oc ... ~~tl
=u N N ... ~ ~ .; .,; .0 ~ r: .0 iiii:!~
os: - - - ... ... ..
..
III ....'" ~...:-
11:1 <w ....
~ >", .:: '" '" ... ~ 0- 2: - ... ... ... '" <0 ~ ... ... 110 \I,IUlU
- - - - - - ... ~ .... ~Za:
...
- ~ !-Qo
... ~IW ~ ... :
t E "'l "'l "'l "'l ~ .. "I; r: C! -.: ... C! ... OJ: .... <0 _;SS
8 o..lU - - - - - N ... N ... ... ... ..; ... ~ .. vi .,;
... ... :s: ""'~
<'" ~~ -
:tW .: .... .... '" '" oc ... ... ... .... CI - ... ... ::! 'CI ... .... g<t3
..... - - - - - - - uS
UJ e ~ ... -.: r-: ... .... r-: C! ., "! ~ .... ... ... ., ~ "'! r: ., ~ ... .,_ 8~~
'" a:u "" N .... ..... N N ... ... .. ... .. ...; ... ..:i ,'" ... ~ .... 2 - - N ... %.~~
c... 2:> - - - - ;;~ts
... 0:
<a: ;~~
-I >w .: III> ... <0 .... '" '" N ~ .. .... ~ ... ~ ::i. .. '" ~ ... ... .... .... ... ;:
... '" - - - - N ... ... ... ... ... ..
... ~:;-I
- ...
til ><
~ a:W E -: - -: -: .... CI -.: '"': co "'! Ct:: N co - C! "! "! "'! ... .,,-1>
W::! .... N ... N N '" M ... ... .., ..' ... ... ,..; ,..; ~ r- r- Ill> III> ~ ~ g !i t -
- Q ... > <::3
I- <iii: ..., US3
." a:L:J .:: ... ... r- ... .... ... 0- 2 - N ... ... oc ... 0- N ... '" .... "" ~ ....
." - - - - - - - N ... N ... ... ,.., .., ::z
:;f31>o
w f8~
c...
t:: w ..... _
N .:!. .., t:!. .:!. .:!. .!:! !::. ~
~ct:;;:; - _N
- ... - '" - - - - .........
<ow - - - N ... ... ... ... '" '" "" 0 = .. \D ~ 0 ... eo Iii
Z III - - - - ... M ...
- - E
~ -
o I'" Z
Z Figure 2·10


3 - The standard span doesn't apply on risers, since no moment (and thus no stress) develops regardless of the riser length. The number and location of supports should be determined by the location and strength of building steel. However, it is preferable to locate supports above the center of gravity oflong risers in order to prevent toppling.

4 - Support locations should be selected as close to building steel as possible in order

to simplify support configuration.

The steps involved in supporting a piping system for sustained loads can be illustrated with an example. In Figure 2-11, the system consists of a 12" diameter, standard schedule steel pipe filled with water, with a design pressure of 150 psi, and a design tern perature of 3 50oF, which runs between two equipment nozzles.

The engineer first must determine the standard span for the system. For 12" diameter, water filled pipe, the standard span is shown in MSS SP-69 to be 23 feet. For changes of direction, 3/4 of this span is 17 feet-4 inches.

Next, the engineer locates supports. The first concern is to locate them near concentrated loads - supports should be located as close as possible to the two valves (for example, near node points 20 and 70). The first of these is optional, depending on whether the nozzle at node point 10 is assumed to act as an anchor, and whether it is desirable to minimize the

nozzle loads on the equipment. -

The next concern is the placement of supports on the riser. Assume that the capacity of the building steel dictates that the weight of the riser be split between two supports. It is recommended that one of these be placed above the center of gravity of the riser (for exam ple, 15 feet below the top of the riser).

Now supports can be located elsewhere in the system, starting at the nozzle at node point 10. A support was located near node point 20 earlier; we now want to locate the next one downstream within the standard span. It is evident that pipe changes direction within 23 feet, so the developed length to the next support should be maintained as less than 17 feet- 4 inches. The next run of pipe accommodates a full 23 foot run, so two supports can be located between node points 30 and 40. The line of action of the supports on the riser provide support to the end of the horizontal 30M40 run, so no additional support is required at node point 40. Support locations can be continued to be selected in this manner until all locations meet the selection criteria; one solution is shown in the Figure 2-12.

Once completed, what does this accomplish? By using the standard span criteria, the engineer can assume that the maximum stress in the piping system due to weight loading does not exceed 1500 psi. Therefore, substituting this value for the weight component of the stress equation:

Ssus = PAj/Am + 1500 = 150(113.1)/14.58 + 1500 = 2664 psi < 20,000 psi

2-16

t!)!
•
C
O·
'.~ .
f:)
rd
0
•
•
"
()
"
('"J.
'-_
•
•
•
i.) •
.,
•
•
•
•
•
C_),
t)
o
•
•
0
0
0
•
•
.. o

1 2· 01,1., - STO SCH PIPE MAT'L - Al05 GR 8 FLUtD - wATER

PRESSURE - 150 PSt TEMP - 350 DEGREES F' ELBOWS - ~DNG RADtUS INSU\.A'l'tOt>: - 2" CS

VALV::S - 1506 GATE VALV::S (vr. .. S25(f) NOZZLES (ANCHOR POIN"'5) 0 1 a &: 90 se .. 2C,OOO Fl51

Sh ., 20,000 PSI

THERMAL CYCLE:S <:7000


".

•

•

o

•

•

•

o

o

•

I !

•

•

Figure 2·11

•

o

'.2" CiA - STD SCH ""I"::

MAT'l - Ai06 GR :3 f"LVIO - WATER PRESSURE - 1 50 PSI TEMP - 350 DEGREES , ElBOWS - L.ONG RADiUS INSULATION - 2" CS

VALVES - '50# GATE VALVES (111'"=825#) N CZZLES (ANe t-fO R POINTS) e 1 0 & 90 Sc =- 20,000 PSI

Sh = 20,000 PSI

THERMAL CVCLES <7000

•

•

•

•

•

()

G

""' ,

i®

)0(

•

PROPOSED HANGER LCCATION

•

o

Figure 2-12

o

•

•

•

2-17


Piping sag is not a problem, since displacement is limited to 0.1 inches.

Therefore the engineer has demonstrated that this piping system meets the sustained stress criteria, without having to do any actual "work".

This can be confirmed by actually doing an analysis of the supported system. The results in Figure 2-13 show that the maximum sustained stress actually calculated for the configuration shown in Figure 2-12 by CAESAR II is 2418 psi, showing that the shortcut analysis is reasonably accurate, yet conservative. The CAESAR II analysis also shows a maximum vertical displacement under weight of 0.0046", which is also conservative.

\

\.

LICENSED TO: TETRACOM SERVICES IOiF 13269 Ver 3.18 PAGE:1
CAESAR I I DISPLACEMENT REPORT FILE:SUPT01
CASE 2 (SUS) W+PI DATE:NOV 4,1992
-----Transl ations( In. )-- -- --- --Rotations{deg. )--
NODE OX OY OZ RX RY RZ
10 .0000 .0000 .0000 .0000 .0000 .0000
20 .0000 .0000 .0000 .0000 .0000 .0000
22 .0000 .0000 .0000 .0010 .0000 -.0017
28 .0000 -.0040 .0000 .0024 .0000 -.0044
29 .0000 -.0046 .0000 .0034 -.0001 -.0031
30 .0001 '.0039 .0000 .0052 -.0002 -.0024
34 .0005 .0000 .0000 - .0065 _ -.0003 -.0022
36 .0024 .0000 .0000 .0075 - .0004 -.0015
38 .0034 -.0012 .0000 .0004 -.0004 -.0012
39 .0034 - .0008 -.0002 .0030 -.0005 -.0011
40 .0032 -.0005 -.0009 .0027 -.0005 - .0010
44 .0009 .0000 .. 0067 .0013 - .0005 -.0007
46 -.0014 .0000 -.0073 -.0012 -.0005 -.0004
48 -.0024 -.0005 -.0012 -.0032 -.0004 -.0003
49 -.0024 -.0009 -.0004 -.0041 -.0004 -.0002
50 -.0023 -.0016 .0000 -.0017 - .0004 -.0002
55 - .0015 .0000 .0000 -.0010 -.0004 -.0001
60 -.0006 -.0024 .0000 .0042 -.0003 .0000
70 -.0005 -.0010 .0000 .0042 -.0003 .0000
72 -.0004 .0000 .0000 .0029 -.0003 .0000
78 -.0001 -.0001 .0000 -.0003 -.0002 .0000
79 .0000 .0000 .0000 .0007 -.0001 .0003
80 .0000 .0001 .0000 .0004 . 0000 - .0010
85 .0000 .0000 .0000 .0004 .0000 -.0012
90 .0000 .0000 .0000 .0000 .0000 .0000 0
•
•
n·
_". :7
'\,- .. '~
•
C
0
o
•
•
8
0
0
C).
•
•
- •
}- •
•
•
•
•
•
() .
Q
'"
•
•
0
0
Q
•
•
• 9 COADE Pipe Stress Analysis Seminar Notes

6

•

•

• C

e

• • • o

c

()

• • • • • • • • •

c.

o • • •

o • • •

LICENSED TO: TETRACOM SERVICES CAESAR II STRESS SUMMARY CASE 2 (SUS) W+Pl

10# 13269 Ver 3.18 PAGE: 10 FI LE: SUPTOl

OATE:NOV 4.1992

*****CODE STRESS CHECK PASSED

PIPING CODE: 831.3 (1990) HIGHEST STRESSES: (lb./sq.in.)

ALLOWABLE:

20000.

CODE STRESS %: 2418. @NQOE 34
BENDING STRESS: 1255. @NODE 34
TORSIONAL STRESS: 54. @NOOE 22
AXIAL STRESS: 1300. @NOOE 46
I 3D MAX INTENSITY: 2631. @NOOE 22
Figure 2·13 A further implication of this approach is that in order to eliminate a stress or deflection problem due to weight loadings, the best solution is usually to reduce the unsupported span of the piping - i.e., add more supports.

2.2.3 Consideration of Nozzle Loads

The previous discussion has primarily concerned the effect of supports on system stresses. The engineer is also interested in determining loads on 'supports and nozzles, in order to select the appropriate support hardware, to check the overloading of'equipment, or to calculate vessel stresses.

A review of the restraint loads shows that the hanger loads are on the order 0[2000 to 3000 pounds. These loads would be used as an upper limit for the selection of the support hardware - for example, the rods, clamps, brackets, supporting steel, etc. must be capable of resisting these loads at a minimum.

A review of the weight load (Yforce) on the nozzle at node point 10 (see Figure 2~14) shows a relatively small load, of only 237 pounds, which should be acceptable for most types of equipment. However, closer inspection shows that the sign is positive, indicating that the piping system is pushing up on the support, rather than down. This seems unnatural for a gravity load, and in fact is due to the unbalanced elbow at node point 30 pivoting about the hanger at node point 22. Therefore, even though the nozzle load is low, this is tiot an

optimally supported system. '

.0

...


LICENSEO TO: TETRACOM SERVICES lOll 13269 Ver 3.18 PAGE: 5
CAESAR II RESTRAINT REPORT FILE:SUPTOI
CASE 2 (SUS) W+P1 DATE: NOV 4.1992
Forces(lb.) - -- Moments(ft.lb.)
NODE- FX FY FZ MX MY MZ TYPE
10 O. 237. -1. 397. - 3. -100. Rigid ANC
22 O. - 2024. O. o. O. O. Rigid +Y
34 O. -2300. O. O. O. O. Rigid +Y
36 \ O. -2190. O. O. o. O. Rigid +Y
44 i O. -3014. O. o. o. O. Ri gi d +Y
46 O. -3054. O. o. o. O. Rigid +Y
55 O. -1606. O. O. o. O. Rigid +Y
72 O. -2044. O. O. O. O. Rigid +Y
85 O. -803. O. o. O. O. Rigid +Y
90 O. -604. l. 31- - 3. 1110. Rigid ANC
Figure 2·14 0
•
•
(j.
'--
•
t,)
0
o
•
•
e
0
0
C)e
•
•
~- •
~(~ ,
::t- .
. , .'.
-,
•
•
•
•
•
()O
c
0
•
•
•
0
·9
•
•
• The system support can probably be Unproved by moving the hanger a bit closer to the elbow to reduce pivoting - but how close is enough? Figure 2-15 shows the restraint loads for a configuration with the restraint at node point 22 moved 2'-0 closer to the elbow (i.e., 3'-0 from the end of the valve). The sign is now correct (indicating a reasonably balanced system), but the load on the nozzle is now 495 pounds, larger than before. It is not certain that this is an improvement.

LI CENS ED TO: TETRACOM SERVICES IOiF 13269 Ver 3.18 PAGE: 1
CAESAR II RESTRAINT REPORT FILE:SUPTOI
CASE 2 (SUS) W+P1 OATE:NOV 4.1992
---. Forces(lb.) ... - -- - - Moments(ft.lb.) - -
NODE FX FY FZ MX MY MZ TYPE
10 O. -495. -I. 239. 4. 17l. Rigid ANC
22 O. -1348. O. O. O. O. Rigid +Y
34 O. -2224. O. O. o. O. Rigid +y
36 O. -2219. O. - o. O. O. Rigid +Y
44 O. -3005. O. o. O. O. Rigid +Y
46 O. -3055. O. o. O. O. Rigid +Y
55 o. ·1606. O. o. O. O. Rigid +y
72 O. -2045. O. O. O. O. Rigid +Y
85 O. -804. O. O. O. O. Rigid +Y
90 O. -603. 1- 30. 5. 1109. Ri gi d ANC Figure 2·15

2-20

Q I:J

•• • •

,0

o

• •

• G

o

I()

• • •

r:

•

• • •

• •

.:)

o

• • • o

o

• • • e


However, this exercise demonstrates that support and nozzle loads may be tailored by adjusting the locations of the supports. The best location for the hanger may be estimated by interpolating between the two results, in order to minimize the load acting on the nozzle. This shows that the best (where "best" is defined as minjrn;zing Ydirection weight force on the nozzle at node point 10) location for the hanger is:

d

= 1.0 - (3.0 - 1.0) :x 2 x (237) 1(-495 - 237) ;:: 1.648 ft = 1'-7-3/4"

\Vhere:

d

;:: distance of hanger from valve, ft

Analyzed Case #1

237#

Location (1"-7-3jL")

/

Interpolation Line

-495#

\

Analyze~ Case #1

Tuning nozzle loads may also be done by varying the support loads, rather than the support locations. This is done by refusing to allow the system weight to settle on its own, but rather by forcing weight unbalance at certain support loea tions. In this way, if the support at node point 22 is underloaded, the system is less likely to push up on the support. For example, if the support at node point 22 only takes -1725 pounds, the shortfall will be split up between the nozzle at node point 10 and the support at node point 32, with the bulk of the shortfall going to the nozzle, w hich is closer. This shortfall, of approximately -300 pounds, will reduce the upward load at node point 10 by approximately 225 pounds (with the support at node point 34beirig reduced by the remaining 75 pounds), down to approximately zero pounds. (Proof ofthis is left to the reader.)

The load at selected supports can be forced to be unbalanced through the use of pre-loaded springs (i.e., the loads are set to something other than the naturally distributed weight load), thus influencing the resulting loads on the nozzles. This is most easily done by releasing degrees-of-freedom at anchor points during the restrained weight phase of hanger design, as discussed in Section 2.4 of these seminar notes.


2.3 Designing For Expansion Loads

Note: It must always be remembered that the engineer must consider the RANGE of expansion stresses (between the cold and hot conditions of the pipe, for example) when considering expansion loadings. The absolute stress value is not a particularly meaningful parameter when discussing expansion stresses, due to the shakedown (self-springing) effect.

2.3.1 Magnitude of Thermal Load

A piping system, when heating up, normally tries to expand against its restraints, resulting in internal forces, moments and stresses:

p

Figure 2-17

The axial force generated in the above configuration can be estimated to be the axial force required to' compress the pipe back to its original length after it has been allowed to grow freely. Its free growth is:

Figure 2-18

D. = cd

Where:

6. = thermal axial extension of unrestrained pipe, in

ex. = linear thermal expansion of material from ambient to operating tempera

ture, in/in

I = length of pipe, in

2-22

0
•
0
O·
•
(,)
0
0
•
•
e
0
G
C).
•
•
•
i ,:' .
.. ~.
•
•
•
•
•
{)
•
()
0
•
•
0
f':;'
0
•
•
• fa e

• • • o

G

• •

COADE Pipe Stress ~alysis Seminar Notes

The axial force required to compress that growth. is:

p

·r---------e--------~I

Figure 2-19

= Pl! AE

• Where:

o

o .()

• •

• •

•
•
•
•
•
ee=;
0-
{)
•
•
•
0
0
. .
•
•
•
C = axial compression of pipe under load, in

P A

= compressive load on pipe, lb

= cross-sectional.area of pipe, in2

E = modulus of elasticity of pipe material, psi

Equating the deltas, the axial force can be estimated as:

cd

= PlJAE, or: P = AEex

Considering a rather benign operation - a 12-inch diameter, standard wall pipe (A = 14.58 square inches, E = 29E6 psi) operating at 3500F (ex = 1.88E-3 in/in) - the axial load is calculated as:

P = 14.58 x 29E6 x 1.88E-3 = 800,000 pounds

From the point of view of most piping codes, there is no stress, since no moment is produced in the axial run (although the codes do state that the possibility of buckling must be considered); however, this is not a good design.

AIl alternate is no restraint at one end, allowing the pipe to grow unimpeded; therefore no load develops. However this is not good design either, since the pipe must normally attach to some relatively fixed piece of equipment, and cannot usually be floating in space.

What is the sol u ti on to this problem? It is necessary to have some restraint on the system. but too much may cause excessive forces, moments, and stresses. Looking at the examples above. allowing no movement produces a force of about 800,000 pounds. Allowing 100% of the pipe's desired free movement causes no force. Interpolating, if we could devise a means by which the piping system remained intact, yet allowed 99.8% of the pipe's desired free movement, the developed force would be approximately:

(1.0 - 0.998) x 800,000 = 1600 pounds

This is a much more manageable situation.

2-23

r- 10'-0 -j

I I

\ --r

12" o 5TO :...C5 @350'F

10'-0

0'
0
•
O·
.
.
. ~HJ,
•
0
0
0
•
•
8
0
Q
C).
•
•
•
, •
•
•
•
•
•
o,
e
0
•
•
•
0
0
•
•

2.3.2 Guided Cantilever Method

One proposed means of allowing nearly all of the pipe's free movement (while still holding the system together) is to provide adj acent, perpendicular legs to absorb the thermal growth through bending, as shown in Figure 2·20.

t::. = PI-3 / 12E! =:: <X I

Each leg can be modeled as a guided cantilever. According to beam theory:

~ = PI3 / 12EI = c l

M = PI / 2

= PI / 2

2·24

CJ
e
•
•
•
0
G
•
•
•
0
0
eel
•
•
•
•
•
•
•
•
•
c
0
0
•
•
0
0
0
0
•

Solving for P: P:: 12EI all3 Solving for M: M = 6EI 11112

Solving for SE: SE = 6EI t.112Z:: 6ER A 112 'Where:

I = moment of inertia of pipe cross- section, in4-

I = length of leg absorbing thermal growth, in

Z = section modulus of pipe cross-section, in3

I/R

R = outer radius of pipe, in

Note that the calculated expansion stress range SE is independent of the wall thickness of the pipe (on a system-wide basis). Therefore, increasing (or decreasing) the pipe wall is usually not an adequate sol ution to an expansion stress problem. This equation also points out that the stress range decreases with the square of the length of the absorbing leg, so the longer the leg absorbing the displacement, the lower the stress range.

For the configuration shown in Figure 2-20:

A = 1.BBE-3 x (10 x 12);:; 0.23"

SE = 6 x 29E6 x 6.375 x 0.23 I (10 x 12)2 :: 17,700 psi

An expansion stress range of 17,700 psi is normally not a problem .. however it must be remembered that this equation 'did not take into account the stress intensification factor (SIF) at the elbow at the top of the leg. Considering an in-plane stress intensification factor for a long radius bend oftypical SIF value of2.8, this would result in a stress range of about 49,600 psi, which is probably excessive for typical low carbon steel a pplications, (Note that this value is actually conservative, since the guided cantilever model does not take into account the fact that thejunction of the two legs will rotate some under the load, and further neglects the additional flexibility of the elbow.)

Against what do we compare the 49,600 psi stress range? We compare it against SA. which is:

SA ;:; f [1.25 (8e + SH)· 81). or, conservatively: = f(1.25 Se + 0.25SH)

,

For a typical low carbon steel (A106 Grade B, for instance) and a typical piping code (B31.3, for instance), 8e and SH are both 20,000 psi, giving a conservative value for SA of 3 0, 000 psi (the non-conservative value of'S A cannot be calculated without knowing the sustained stress S} at the point of interest).

The expansion stress range can be approximated for any run of pipe using the guided cantilever equation shown above. as long as the displacements to be absorbed are known.

-- f'

¥ .4:;: ,... & a:=:u=:m::r::aAI U.... . n=


2.3.3 Refining the Model Through the Use of Restraint Stiffnesses

What if the calculated expansion stress range is too high? How can we reduce it?

Consider the initial example, with the pipe fully anchored at its ends. What would happen in real life? The restraints would probably bend some under the enormous load, allowing some piping expansion, which would then reduce the internal load (since expansion loads are self limiting). What happens if the bending of a support - i.e., its flexibility - is explicitly consi dered in the analysis? (Normally a pipe stress program by de fa ul t considers a restraint to be "infinitely rigid". For example, CAESAR II's default restraint stiffness is in the range of IE 12 pounds per inch.)

If the res train t actually has a lateral stiffness of 10,000 pounds per inch (instead of IE 12), the thermal growth is partially absorbed by the pipe and partially absorbed by the restraint:

--- L

due to pipe leg

due to support

6 = PL~3/ 1 2EI -To P /10000

P = 6/(L~3/ 1 2EI ... ,1/10000)

M = 6. L/2(L---3/1 2EI + 1/10000)

S:- = D. L/2(L---3/ 1 2EI ..:.. 1/1 OOOO)/Z

)

Lateral Stiffness

of Support = 10000Ib/ir.

Figure 2 .. 22

For a 12" diameter standard wall pipe, the calculation is as follows:

I = 279.3 in4 Z = 43.8 in3

SE = 0.23 x 1201 [2(1203 1 (12 x 29E6 x 279.3) + 1/10000) x 43.8]

= 2675 psi

2-26

4 !IJ#aii="" =
0
•
e
CJ·
",; __ )
•
,0
Q
0
•
•
0
0
a
C).
•
•
__ • .,:,.-o.._ •
:-
~' •
,
-
•
•
•
•
•
'J
\'
•
(1)
Q
•
•
0
0
.~,) ~
•
•
• o o


• This significantly reduces the stress range (from the previous value of 17.700 psi) - not

• through any actual modification. but simply through a refinement of the model. From this we can gain two insights:

• o

G o

• • o

e

• •

• • • • • • •

-=.;) o

Q

• • •

o

G

Q

• •

o

1 - It is sometimes a good idea to provide actual restraint (and nozzle) stiffnesses in the model- the closer to reality the model becomes, the more accurate are the results. Refinement of the model may save the cost of modifying piping systems which initially appear to be over stressed,

2 - If a system really is over stressed, a potential fix may be the introduction of flexibility at the restraints, either by removing restraint or by providing less than

infinitely-rigid restraints (or gaps), '

Restraint stiffnesses may be calculated through any means and then entered by hand. or sim ulated in the pi ping model through the use of structural or piping elements. Vessel nozzle stiffnesses may be calculated manually using Welding Research Council Bulletin 297 or some equivalent. Modeling ofrestraints using CAESAR IT's structural modeler and use of \VRC Bulletin 297 are discussed in Section 3 of these course notes,

- Note that it is best not to selectively enter flexibilities for some restraints and not for others, This will result in the inaccurate distribution ofloads, resulting in nonconservative results.

2.3.4 Use of Expansion Loops

In the event that model refmement is not sufficient to solve the problem (i.e .• there is a real problem, and not just one on paper), something must be done. Re-examining the equation for the guided cantilever model..

= 6ER b./12

It is evident that the stress analyst cannot easily change the terms 6, E, R, or delta. This leaves only 1, the length of the leg absorbing the thermal growth. This can be done through the addition of an expansion loop, In this case, the thermal growth is partially absorbed by each of the legs running orthogonally to the thermal growth:

= P(120)~ / 12EI + P(240)3 / 12EI

= P (1203 + 2403) / 12EI

P

= 12EI b. / (1203 + 2403) = 6ER 11.1/ (1203 + 2403)

L·~.'

Pipe Stress Analysis SEMINAR COADE

Transféré par

Informations du document

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Pipe Stress Analysis SEMINAR COADE

Transféré par

Droits d'auteur :

Formats disponibles

'Of

Vous aimerez peut-être aussi

Pipe Stress Analysis SEMINAR COADE

Transféré par

Informations du document

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Pipe Stress Analysis SEMINAR COADE

Transféré par

Droits d'auteur :

Formats disponibles

﻿'Of

Vous aimerez peut-être aussi

'Of