Sunteți pe pagina 1din 55

PIPING ENGINEERING CELL

STRESS ANALYSIS
T.N.GOPINATH

Stress Analysis 1
PIPING ENGINEERING CELL

1.0 INTRODUCTION connected equipment (‘b’


above) overrides all other
Stress Analysis is a subject, considerations.
which is more talked about and less
understood. The objective of pipe
stress analysis is to ensure safety 1.2 Classifications of piping
against failure of the Piping System systems
by verifying the structural integrity
against the loading conditions, both The piping systems are
external and internal, expected to mainly classified into three main
occur during the lifetime of the categories and then again
system in the plant. This is to be subcategories. The main categories
undertaken with the most economic are the
considerations. Hence the objectives 1.2.1 Hot Systems.
of stress Analysis could be listed as
1.2.2 Cold Systems.
1.1 Objectives of stress Analysis
are to 1.2.3 Cryogenic Systems.
1.1.1 Ensure that the stresses in the The fundamental reason for
piping components in the this classification is that hot lines and
system are within the cryogenic lines must undergo
allowable limits. Flexibility analysis to determine
thermal forces, displacements and
1.1.2 Solve dynamic problems stresses. These systems are further
developed due to mechanical divided into. .
vibration, acoustic vibration,
fluid hammer, pulsation, relief i) Small bore lines
valves etc.
ii) Large (Big) bore lines.
1.1.3 Solve the problems associated
As a general practice those
due to higher or lower
pipe lines with nominal diameters
operating temperature such as
40mm (1½”) NB and under are
a) Displacement Stress classified as small and 50mm (2”)
range NB and above as large. Further,
piping systems could be classified
b) Nozzle loading on the based on the regulatory codes under
connected equipment which the system is designed. Certain
c) Pipe displacements codes require more stringent analysis
than others.
d) Loads and moments on
the supporting structures 1.3 Hence the steps involved in
the stress analysis can be
When piping is connected listed as.
to strain sensitive equipment,
the flexibility required to 1.3.1. Identify the potential loads
satisfy the acceptable limits of that the piping system would
nozzle loading on the encounter during the life of
the plant.
Stress Analysis 2
PIPING ENGINEERING CELL

1.3.2. Relate each of these loads to required flexibility has been provided
the stresses and strains in the piping layout.
developed.
1.5.1 As per clause 119.7.1/319.4.1
1.3.3. Get the cumulative effect of of the code ASME B 31.1/B 31.3, no
the potential loads in the formal analysis is required in systems
system. which
1.3.4. Decide the allowable limits, i) Are duplicates of
the system can withstand successfully operating
without failure. installations or replacements.
1.3.5. After the system is designed, ii) Can readily be judged
to ensure that the stresses are adequate by comparison with
within the safe limits. previously analyzed systems.
1.4 Types of loads iii) Satisfy equation
specified in clause
All the American code for 119.7.1(A3)/ 319.4.1 (c)
Pressure Piping classify the loads
mainly into three types. 1.5.2 Analyzing the layout by an
approximate method.
1.4.1. Sustained Loads: Those due
Approximate method shall
to forces present during
be applied only if they are used for
normal operation
the range of configuration for which
1.4.2. Occasional Loads: Those adequate accuracy has been
present during rare intervals demonstrated.
of operations 1.5.3 Carrying out a comprehensive
1.4.3. Displacement Loads: Those analysis.
due to displacement of pipe i) Analytical
These are dealt with in detail ii) Model test
in the chapter Pipe Under stress.
iii) Chart method
Hence the content of this chapter is
limited to the details of analysis of
piping system under the sustained
and displacement loads. This analysis 2.0 CODE COMPLIANCE
is most commonly called as the
Let us consider those aspects
Flexibility Analysis. Further those
in the code, which are mandatory
conditions stipulated in the regulatory
requirements for the expansion and
code ASME B 31.1 and ASME B
flexibility of metallic piping. The
31.3 only are considered hereafter.
Piping Specification nominates the
1.5 Conditions of Acceptability code to be used for various aspects in
of Piping System the Piping System. Let us consider
those, which are of importance to the
The Piping Engineer has the Piping Engineer to carry out the
following choices to establish that the flexibility analysis. Every such code
will contain recommendations and

Stress Analysis 3
PIPING ENGINEERING CELL

mandatory requirements on the The system could behave either


following aspects. balanced or unbalanced under such
conditions.
i) Minimum flexibility
require-ments for thermal a) Balanced System
expansion
Displacement strains are well
ii) Allowable stresses for distributed and not excessive at any
various piping materials point. Layout of the system should
aim at such condition, which is
iii) Reinforcement assumed in flexibility analysis
requirements of branch methods provided in the code.
connections
b) Unbalanced System
iv) Support criteria
In an unbalanced system,
It is the responsibility of the Piping stress cannot be considered
Engineer to demonstrate the proportional to displacement strains
compliance with these requirements through out a piping system in which
and achieve the most economical, an excessive amount of strain may
safe and practical layout. occur in a localized portion of the
Let us consider each of the above in system. Unbalance may result from
turn to see how they are dealt with i) Highly stressed small pipe run
by a Piping Engineer. in series with large or
relatively stiff pipe runs.

2.1 Installed And Operating ii) Local reduction in size or wall


Temperature thickness.
Pipe is erected at ambient iii) Line configuration in a system
temperature and that depends on the of uniform size in which the
climatic conditions. 70ºF (21ºC) is expansion or contraction must
the figure commonly used for be absorbed largely in short
calculations. The same piping when offset.
in operation in a Petrochemical Plant
could achieve a temperature in excess Unbalance must be avoided
of 500ºC if it were in a reactor by design and layout of piping
piping system or it could be of the system. If unbalance cannot be
order of -120ºC, if it were associated avoided the designer shall use
with ethylene refrigeration system. appropriate analytical methods as
specified in the code to assure
2.1.1 Displacement stresses. adequate flexibility.
A piping system will undergo Each material has its own coefficient
dimensional changes with any change of thermal expansion. These values
in temperature. If it is constrained are given in Appendix C of the code
from free expansion or contraction, it ASME B 31.3 and Appendix B of the
will be displaced from it unrestrained code ASME B 31.1. If the pipe is of
position causing strain and stresses. carbon steel or of low alloy steel, it

Stress Analysis 4
PIPING ENGINEERING CELL

will expand at the rate of 6 to 7 mm (difference between the temperature


every meter length as the temperature under hot condition and initial cold
rises to 5000C. This means that the condition and also the length of the
pipe running between two pipe). L can be calculated if the
equipments 10m apart may well want coefficient of thermal expansion, an
to expand by 60 to 70 mm or more as important physical property of any
it heats up. The increased length can material, is known. The axial force
be accommodated only by straining generated in the above configuration
the pipe as the ends are not free to can be estimated to be the axial force
move. This straining, if not freely required to compress the pipe back to
allowed, induces stresses in the pipe its original position from its
as well as load at the support points. expanded position.
However, when the line is cooled
during shutdown to ambient In the textbooks, coefficient of linear
temperature the expansion returns to thermal expansion is defined as the
zero, the straining is no longer increase in length of a specimen of
required and hence the load and stress unit length, if it is subjected to a
also disappears. This can be temperature rise of 10C and is often
demonstrated as below. designated as . If a pipe of length L
is heated to a temperature which is
2.2 Magnitude Of Thermal Load
A pipe line (Fig. 2.1), held T0C above its installed temperature,
between two anchors, when heated the increase in length would then be
up, tries to expand against its
restraints resulting in forces, L =  . T. L ….
moments and stresses. (1)
In the codes and many reported
calculations, however,  is used as
inclusive of T (i.e.
.T in above equation) and is called
coefficient of thermal expansion from
installed/ambient to operating
temperature.

The expansion is then written as

L =  . L ….
(2)
This difference in meaning of  used
in equations 1 and 2 should be taken
If the pipe is allowed to expand freely
note of. A typical Table C-1 of total
due to rise in its temperature, it would
thermal expansion from ASME 31.3
expand by L as shown in Fig. 2.2. .
for a specific material is as follows. It
The free expansion will take place
can be used to get  easily for the
when one of the anchors is released.
applicable difference in operating and
L would depend upon the pipe
installed temperature.
length (L), temperature rise

Stress Analysis 5
PIPING ENGINEERING CELL

Table C-1ASME B 31.3


TOTAL THERMAL EXPANSION, US UNITS, FOR METALS
Total Linear Thermal Expansion Between 700F and Indicated
Temperature, in./100 ft

Material
Temp., Carbon Steel... Austenitic Stainless Steel... __ __
Copper and
0
F
Copper Alloys
25 -0.32 -0.46 -- --
-0.50
50 -0.14 -0.21 -- --
-0.22
70 0 0 -- --
0
100 0.23 0.34 -- --
0.34
125 0.42 0.62 -- --
0.63

where A = Area of cross section of


pipe,
In2 (mm2)
If the pipe is to be maintained in the
original position then there will be an E = Modulus of elasticity of material,
axial force P to compress the increase psi (Kpa)
in pipe length of L (Fig. 2.3). P = Compressive force on pipe, lbs
(N)
f = Stress developed, psi (Kpa)
L = Axial compression of pipe, In
(mm)
L = length of pipe, In (mm)
The strain developed in the pipe, , is
then calculated as  = L / L=  To evaluate the magnitude of such
a force, let us consider Carbon
Internal stress developed due to this Steel pipe of 600 mm outside
strain, f = E (Hooke’s Law) = diameter with 10 mm thickness,
E operating at a temperature of
The force required to compress back 3000C.
is
P = Af = AE Referring to ASME B 31.3, Table C6,
E = 26.85 Msi (1.888 x 104
kg/mm2 )

Stress Analysis 6
PIPING ENGINEERING CELL

starts from stress free condition when


Referring to Table C1 cold, gets stressed with stress
 = 3.625 x l0-3 mm/mm reaching maximum at operating
condition and then the stresses get
Area of the pipe reduced to zero when operation stops
A = Pi / 4 [(600)2 - (580)2] and system cools down.
= 18535.4 mm2 This was an oversimplified
picture and this is not what exactly
P = 18535.4 x 1.888 x 104 x 3.625 x happens in practical situations. The
l0-3 piping system can absorb large
= 12,68,563 kg displacement without returning
= 1269 tons exactly to previous configuration.
Relaxation to the sustaining level of
As per most of the design codes, material will tend to establish a
there is no stress in the above condition of stability in few cycles,
configuration, since no bending each cycle lowering the upper limit of
moment is produced in the axial run. hot stress until a state of equilibrium
However, the possibility of buckling is reached in which the system is
due to the development of completely relaxed and capable of
compressive stresses must be maintaining constant level of stress.
considered. In no case, such a design The stress at which the material is
could be accepted. relieved due to relaxation appears as
Alternative to this is the stress in the opposite temperature
piping system with no fixed points, state, with equal intensity but of
allowing the pipe to expand. opposite sign. Thus the system,
However, this is also not feasible, as which originally was stress less,
the equipment cannot float in space. could within a few cycles accumulate
Equipment on wheels is one stress in the cold condition and spring
possibility but seldom practiced. So itself without the application of
every time the plant is in operation external load. This phenomenon is
and during shutdown, the same called “Self springing”. This can be
cycle of events occurs. The pipe demonstrated as follows.

Stress Analysis 7
PIPING ENGINEERING CELL

Consider the piping system start at cold condition, will now have
abc (as installed) as shown in Fig. residual stresses under cold condition
2.4. As it is taken from installed and is said to cold spring. For
condition to hot operating condition, example, if the pipe was cut
let us assume that the leg ‘ab’ would anywhere along its route under this
expand by 2” and leg ‘bc’ by 2”. condition, it would separate into two
This would happen if anchor at ‘c’ segments with violent spring action.
end of the pipe was movable. The The true magnitude of the
pipe would then be in position ab’c’ stress either in hot or cold condition
with the anchor c having moved to c’ cannot be determined by simple
as shown. If both the anchors (at a calculations because the amount of
and c) are rigid, the pipe may still relaxation is unknown and cannot be
expand and attain position ab2c. It judged reliably. It depends in a
may then be compressed to bring b 2 complex way on the metallurgy, pipe
back to position b. Such free route, anchor locations and geometry
expansion is however not allowed etc. However, service failures are
and let us assume that the system related to cyclic rather than static
absorbs this 4” expansion between conditions and it is therefore
anchors at a and c and the resultant permissible to assume that the system
calculated maximum stress is 24,000 will operate satisfactorily if sum of
psi. Supposing the material at the hot and cold stresses is within a
particular operating temperature can stress range, which is considered safe
sustain only 18,000 psi or ¾ of this for an expected number of stress
developed stress, yielding will take reversals. This concept provides a
place and the pipe would be at its logical basis to the design of a piping
sustaining level indicated by ab1c. We system because it takes into
would say that the pipe has absorbed consideration all the stress levels to
stress of 6000 psi by yielding which the system is subjected. The
somewhat. The stresses in the pipe at actual stress intensity at a specific
this stage are only 18000 psi. On stage is of academic interest only.
cooling back to ambient temperature
the system must contract by 4”. It could be seen that the code
Contraction would relieve the recognizes the fact that stresses in
compressive stresses, which were piping system are not necessarily of
developed because the pipe wanted to constant intensity, that the expansion
expand by 4” but was not allowed to stress at elevated temperature may
do so completely. At ¾ of this not be sustained because of
contraction, i.e., at a net contraction relaxation or creep and
of 3”, the system will become consequently will drop to stress level
stressless. Completion of contraction the material can sustain. The
through remaining 1” will result in a phenomenon of yielding in the elastic
stress of 6000 psi in the opposite range or flow in the plastic stage
direction (tensile stress in this case). presents a problem different from that
The system would now be in position encountered in the analysis of
ab3c as shown in Fig. 2.4. The structure which operates at relatively
system, which was stressless at the low temperatures and therefore in a
state of steady stress. In this case, the

Stress Analysis 8
PIPING ENGINEERING CELL

elements are designed to meet the displacement strains are not normally
limiting stress or deformation within considered in the determination of
the elastic limit of the material. In displacement stress range since these
contrast, with stresses from sustained stresses are not significant in typical
loads such as internal pressure or piping layouts. In special cases
weight the displacement stresses can consideration of average axial
cross the elastic limit with stress displacement stress is necessary.
reversal from cold to hot condition Example include buried lines
and still remain safe provided the containing hot fluids, double wall
number of stress reversals remains pipes and parallel lines with different
below the limit to exclude the operating temperatures connected
possibility of failure due to fatigue. together at more than one point.
While stresses resulting from 2.3 Effect of Fatigue on Piping
displacement strains diminish with ARC Markl investigated the
time due to yielding or creep, the phenomenon of fatigue failure of
algebraic difference between strains piping during 1940’s and 1950’s. He
in the extreme displacement tested a number of configurations;
condition and as-installed condition straight pipe, various fittings such as
remains substantially constant during elbow, miter bend, welding tee,
any one cycle of operation. The fabricated tee etc. mostly on
difference in strain produces a 4'' NB size by using cyclic
corresponding stress differential, the displacements to apply alternate
displacement stress range, which is bending stress. Plotting the cycles to
used as the criteria for designing of failure for each applied displacement,
piping flexibility. he found that the results were on the
expected lines and followed the shape
The type of cyclic straining of fatigue curves.
described above, if repeated often Markyl observed that the
enough, will cause the pipe to crack. fatigue failure occurred not in the
The cracking will start at a point or middle of his test spans, but in the
points where the stress is maximum. vicinity of fittings and also at lower
This is what is meant by “Fatigue stress / cycle combinations than for
Failure”. the straight pipes. This lead to what is
Average axial stress (over the called the "Stress Intensification
pipe cross section) due to Factor" which covered under section
longitudinal forces caused by 5.3.

Stress Analysis 9
PIPING ENGINEERING CELL

If an initially applied imposed displacement which is


displacement load causes the pipe to beyond it’s yield strain. When cyclic
yield, it results in plastic deformation, displacement is removed, the piping
producing prestress in the system component, which exceeded the yield
when it is brought back to its original point, will retain the residual
state by withdrawing the load. This distortion equal to imposed strain less
prestress must be overcome by yield strain . This distortion will
subsequent stress applications, induce yield stress, opposite in sign
resulting in lower absolute stresses to that developed during loading and
during the later load cycles. Because equal to difference between the
of this expected system relaxation, calculated stress value and the
the initial thermal stresses are material yield stress (E - Sy).
allowed to exceed the material yield The elastic, and therefore the
stress, with the aim that the system allowable range has been extended by
self springs during the first few the value of this prestress.
cycles and then settles down into
elastic cycling. The reason for this 2.4 Allowable Stresses
allowable over stress is that a The American piping codes
repeatedly applied load which covered under ASME B 31 subscribe
initially forces the pipe into the to the failure of the piping system to
plastic range will, after a few cycles, the basis the ‘Maximum
“shake down” and be reduced to principal stress theory’. The theory
elastic action in the piping system. states that the yielding occurs when
This theory can be understood by the magnitude of any of the three
considering a pipe experiencing an mutually perpendicular stresses

Stress Analysis 10
PIPING ENGINEERING CELL

exceeds the yield strength of the


material. Temperature and pressure
are the significant factors governing
the stresses created in the piping
systems. There are other factors that
influence the stress as well. They are

i) Wind load
ii) Seismic load
iii) Relief valve forces
iv) Fluid hammer Fig 2.9
v) Settlement
vi) Equipment vibration The yield stress is the elastic limit
vii) Weight of attachments and that is the value below which the
viii) Weight of contents stresses are proportional to strain and
when the load is removed, there is no
All these factors contribute to permanent distortion. The ultimate
two distinct forms of stresses. tensile strength is the highest stress
which the specimen can
The sustained stresses – accommodate without failure.
Generated by Pressure, dead The basic allowable material
weight of contents and stress at the hot (operating/design)
attachments, which can be temperature (Sh) is defined by the
expected to be present virtually code as minimum of
at all time of plant operation.
The self-limiting stresses – As per the ASME B 31.1
Generated by thermal effects. i) 1/4 of the ultimate
tensile strength of the material
The allowable stresses for at operating temperature
these two influences are based on ii) 1/4 of the ultimate tensile
different concepts. However, the strength of the material at
allowable stresses specified in code room temperature
are based on the material properties. iii) 5/8 of the yield
They can be classified in two strength of the material at
categories as below. operating temperature
(90% of the yield stress for
austenitic stainless steels)
2.4.1 TIME INDEPENDENT iv) 5/8 of the yield strength of the
STRESSES material at room temperature
Time independent allowable (90% of the yield stress for
stress is based on either yield stress austenitic stainless steel)
or the ultimate tensile strength and
measured in a simple tensile test. v) 100% of the average stress for
a 0.01% creep rate per 1000
hrs.

As per ASME B 31.3

Stress Analysis 11
PIPING ENGINEERING CELL

i) 1/3 of the ultimate tensile The code uses an allowable


strength of the material at stress, which is the smaller of time
operating temperature. dependent, and time independent
ii) 1/3 of the ultimate tensile allowable stress. The time dependent
strength of the material at allowable stress is the smallest of
room temperature. 67% of the average stress to cause
iii) 2/3 of the yield strength of the creep rupture in 1,00,000 hrs, 80% of
material at operating the minimum stress to cause rupture
temperature in 1,00,000 hrs or 100% of the stress
(90% of the yield stress for to give 0.01% of creep rate per hour
austenitic stainless steel) (Ref. 2.4.1)
iv) 2/3 of the yield strength The self limiting stress in
of material at room piping system are essentially cyclic
temperature and the initial hot stresses, if they are
(90 % of the yield stress for of sufficient magnitude, will decrease
austenitic stainless steel) with time because of the plastic
v) 100% of the average strains and will reappear as a stress of
stress for a 0.01% creep rate reverse direction when the pipe cools.
at 1000 hrs This phenomenon forms the basic
vi) 67% of the average difference between the self-limiting
stress for rupture after stresses and the sustained stresses.
1,00,000 hrs The degree of self-springing,
vii) 80% of the minimum as explained earlier, will depend on
stress for rupture after the magnitude of the initial hot
1,00,000 hrs. stresses and the temperature, so that
while the hot stresses will gradually
decrease with time, the sum of hot
and cold stresses will stay the same.
This sum is called the EXPANSION
STRESS RANGE. This leads us to
2.4.2 TIME DEPENDENT the selection of an ALLOWABLE
STRESSES EXPANSION STRESS RANGE.
Self-springing occurs only
Time dependent allowable when the system is subjected to
stress is usually related to the “creep higher temperatures. For the expected
rupture strength” at high temperature. strain (expected expansion per unit
At temperature above 1/3 of the length), if the modulus of elasticity at
melting point, most metals will this high temperature is used to back
exhibit creep in standard tensile test, calculate stress, the stress value will
if the load is kept constant the be lower than when it is calculated
specimen will continue to deform using modulus of elasticity value at
with time. Under constant load, the lower temperature (cold condition).
rate of creep strain will decrease That is, calculated stress value is
initially to a steady state and later higher when material properties in
will increase rapidly until it fails due cold condition are used. This
to creep rupture. provides a built in safety in design.

Stress Analysis 12
PIPING ENGINEERING CELL

Hence the stresses are calculated maximum stress which may be


using the cold modulus of elasticity. developed in a system as temperature
This is a very important point to note. is raised from lowest to highest that
Actual stresses under hot condition will experience in service or when it
would be less than the calculated is shut down.
stresses. This piping “Shakedown” is
When difference in elastic also known as self-springing and can
modulus within a piping system will be represented as shown in the
significantly affect the stress following sketch (Fig. 2.10).
distribution, the resulting
displacement stress shall be
computed based on the actual elastic
modulus and then multiplied by the
ratio of the elastic modulus at the
ambient temperature to the modulus
used in the analysis.
There are other failure modes
that could affect the piping system.
They include buckling, stress
corrosion and brittle fracture. These
topics are not correctly considered in
the piping code. The effects of these Fig. 2.10
must be considered by the Piping
Engineer while selecting the The maximum stress range
materials or restraining the piping may be set to 2 times the yield stress,
system. more accurately the run of hot and
cold yield stresses, in order to ensure
eventual elastic cycling within the
2.4.3 ALLOWABLE STRESS bounds of allowable stress.
RANGE Incorporating a factor of safety, this
The failure modes that the can be represented by the following
piping code addresses are excessive equation
plastic deformation or bursting; SE <= F (SYc + SYh)
plastic instability or incremental where, SE – Expansion stress range
collapse due to cycling in the plastic F – Factor of Safety
range and fatigue which may be SYc – Yield stress at installed
developed in a system as its temperature
temperature is raised from the lowest SYh – Yield stress at operating
to the highest that it will experience temperature
in service or when it is shut down.
Each of this failure, modes is caused For materials below the creep
by a different type of stress and ranges, the allowable stresses are
loading. However ‘Fatigue failure’ is 62.5% of the yield stress, so that a
recognized by the code as the most conservative estimate of the limit of
likely mode of failure of the the bending stress at which plastic
component and place the limit on the flow starts at an elevated temperature
is 1.6 (100/62.5) times the allowable

Stress Analysis 13
PIPING ENGINEERING CELL

stress and by the same reasoning, 1.6 S A = f(1.25 S c + 0.25 S h )


Sc will be the stress at which flow where,
would take place at the minimum S A = Allowable Expansion Stress Range
temperature. Hence the sum of these S c = Basic allowable stress at minimum
stresses represents the MAXIMUM metal temperature during the
STRESS RANGE to which the displacement cycle under analysis
system would be subjected to, S h = Basic allowable stress at maximum
without the flow occurring in either
metal temperature during
hot or cold condition.
the displacement cycle under
Therefore,
Smax = 1.6 Sc + 1.6 Sh = 1.6 (Sc + analysis
Sh) [The value of Sc and Sh are available
in Table A1 of the Code]
But the American design codes f = Stress range reduction factor for
ASME B 31.1 and B 31.3 limit displacement cycle conditions
the stress range to 78% of the yield for the total number of cycles
stress, which gives a total stress range over the expected life
of
The factor ‘f’ has a value of
S Allowable = 1.6 x 0.78 (Sc + Sh) 1.0 for situation where total number
= 1.25 (Sc + Sh) of cycles is 7000 or less. This
represents one cycle per day for
From this total stress range 1 Sh is nearly 20 years, which is a common
reserved for the longitudinal stresses design parameter. Further, if we look
developed due to loading such as at endurance curve for carbon steel
pressure, weight and other sustained and low alloy steel available in the
loading, giving the allowable stress ASME Section VIII Division 2,
range for flexibility as Pressure Vessel Code, it can be seen
that at some point in the vicinity of
S Allowable = 1.25 Sc + 0.25 Sh 7000 cycles, the Sc + Sh limitation
intersects the fatigue curve.
The above value does not consider The code gives the value of
the excessive cyclic conditions. The ‘f’ in the table 302.3.5 (B 31.3) and
code allows it by multiplying by a 102.3.2 (c) (B 31.1) as follows:
stress range reduction factor.
Accordingly, ASME Stress Range Reduction Factor f
B 31.1 in clause 102.3.2(c) and
ASME 31.3 in clause 302.3.5 specify Cycles N Factor
the Allowable Expansion Stress f
Range as : 7,000 or less 1.0
over 7,000 to 14,000 0.9
over 14,000 to 22,000 0.8
over 22,000 to 45,000 0.7
over 45,000 to 1,00,000 0.6

Stress Analysis 14
PIPING ENGINEERING CELL

2.4.4 EFFECT OF SUSTAINED


over 1,00,000 to 2,00,000 0.5 LOADS ON FATIGUE
over 2,00,000 to 7,00,000 0.4 STRENGTH

over 7,00,000 to 20,00,000 0.3 If the alternating stress is


plotted against the cycle to failure, it
can be seen that the mean stress has
This applies essentially to non- an effect on the endurance strength of
corroded piping. Corrosion can the material. As the mean stress
decrease the cycle life. Therefore, increases, the maximum permissible
corrosion resistant material should be absolute stress (Sa + Sm) increases,
considered where large number of while the permissible alternating
stress cycle is anticipated. stress decreases. The relation between
the allowable alternative stress and
f can also be calculated by the the average stress follows the
equation Soderberg line, which correlates
f = 6.0 (N)-0.2  1.0 fairly well with test data of ductile
where, materials.
Alternating Stress Axis
N = equivalent number of full
displacement cycles during the Fig 2.10
expected service life of piping system
[code cautions the designer that the The equation for the
fatigue life of materials operated at Soderberg line is
elevated temperature may be Sa (Allowed) = Sa(for R=1) x (1-
reduced] Sm/SYield)
where,
When the computed stress range
R = Smin / Smax
varies, whether from thermal
Sa = Smax - Smin / 2
expansion or other conditions, SE is
Sm = Smax + Smin / 2
defined as the greater computed
displacement stress range. The value
During the development of
of N is such cases can be calculated
the ASME BPV code section III for
by the equation
nuclear piping analysis, the special
NN   γ N  for i1,2........ncommittee to review the Code Stress
E
5

i i

where, Basis concluded that the required


N  number of cycles of maximum
adjustments to a strain – controlled
E
fatigue data curve based on zero
computed displacement stress
mean stress occur only for the
range, S E number of cycles in the range 50,000
γ S /S
i i E
to 1,00,000 for carbon steels and low
S anycomputed displacementalloy
i
steels and are insignificant to
stress
range smaller than S austenitic stainless steels and Nickel
– Chrome – Iron alloys. Since most
E

N Number of cycles associated with


i
of the plant piping come under this
displacement stress range S i

Stress Analysis 15
PIPING ENGINEERING CELL

material of construction and the


cycles of operation comes much Let us consider a typical example
fewer than 50,000 cycles in the life of for calculation of SA
the plant, the effect of mean stress on
fatigue life are negligible for piping A pipe supplies Dowtherm to
materials with UTS below 1,00,000 the limpet of a reactor, which is
psi. For high strength bolting operated on a batch process with a 4-
materials where UTS is greater than hour cycle every 24 hours. The
1,00,000 psi the mean stress can have Dowtherm temperature is 315ºC
considerable effect on the fatigue (600ºF) and pipe material is ASTM
strength and should be considered A 106 Gr. B. Design life of plant
when performing fatigue analysis. considered 20 years.
For piping analysis the effect Allowable stress at ambient S c = 20,000 psi
of Soderberg line on fatigue Allowable stress at Max. metal temp.
allowable is implemented in a
S h = 17,300 psi
conservative matter. The code 31.1
and 31.3 covers it up in the following 24
Number of cycles =  365 x 20
manner. 4
= 43,800 (total)
When the basic allowable
stress at maximum expected metal The stress range reduction factor = 0.7
temperature hence,
(‘Sh’) is greater than the sum of the S A = f ( 1.25 S c + 0.25 S h )
longitudinal stresses due to pressure, = 0.7 ( 1.25  20,000 + 0.25  17,300)
weight and other sustained loading = 20527 psi (1443 Kg./cm 2 )
(‘SL’) the difference between them
may be added to the term 0.25 Sh in 3.0 LIMITING VALUES OF
the equation for SA. TERMINAL FORCES AND
MOMENTS
In that case the allowable stress range It has been indicated earlier
will be that in case of strain sensitive
S A = f [ 1.25 (S c + S h ) - S L ]
equipment, the main considerations
Code ASME B31.1 covers these for flexibility are the forces and
aspects in clause 102.3.2(c). The moments, which a pipe imposes on
difference between B31.3 and B31.1 the equipments to which it is
is that the stress reduction factor connected. The maximum permitted
indicated in Table 302.3.5 covers values of forces and moments will
upto vary with the type of equipment.
20,00,000 cycles during the span of These are established by the
the regulatory codes based on which
system, whereas Table 102.3.2 (c) in equipment is designed /
B31.1 indicates this value for manufactured. Or else, the
1,00,000 cycles and over. The factors manufacturers of the equipment set
remain the same in both the cases. these values to ensure safe operation
of the equipment. The following are

Stress Analysis 16
PIPING ENGINEERING CELL

the regulatory codes referred by the tanks interaction with the


Piping Engineer. fabrication engineer is
required to establish that the
1. Centrifugal pumps - API 610/
local stress developed due to
ISO 5199
nozzle loadings are within the
acceptable limits. However
2. Positive displacement pumps
the values given under section
- API 676
3.10 can be taken as a guide.
3. Centrifugal compressors
- API 617 3.1 Centrifugal Pumps
The forces and moments
4. Reciprocating compressors acting on the pump flanges due to
- API 618 pipe loads can cause misalignment of
the pump and driver shafts,
5. Steam turbines - NEMA SM deformation and overstressing of
23 pump casing or overstressing of
fixing bolts between the pump and
6. Air cooled heat exchangers the base plate.
- API 661/ISO 13706
The American Petroleum
7. Shell and tube heat
Institute Standard 610 covering
exchangers - Manf. Specific.
“Centrifugal Pumps for General
Refinery Service” is the one, which
8. Fired heaters - Manf.
specifies these limits for pumps. The
Specific.
code specifies the criteria for piping
design as Appendix F. The allowable
9. Flat bottom Welded Storage
external nozzle forces and moments
Tanks – API650
are tabulated in table 2 for nozzle
sizes from 2” NB to 16” NB for
10. For other static equipment
pumps with casing constructed out of
such as Reactors, vessels and
steel or alloy steel.

Table 3.1: NOZZLE LOADING AS PER API 610

FORCE/MOMENT NORMAL SIZE OF NOZZLE FLANGE IN INCHES


2 3 4 6 8 10 12 14 16
Each Top Nozzle

Stress Analysis 17
PIPING ENGINEERING CELL

FX 160 240 320 560 850 1200 1500 1600 1900


FY 200 300 400 700 1100 1500 1800 2000 2300
FZ 130 200 260 460 700 1000 1200 1300 1500
FR 290 430 570 1010 1560 2200 2600 2900 3300
Each Side Nozzle
FX 160 240 320 560 850 1200 1500 1600 1900
FY 130 200 260 460 700 1000 1200 1300 1500
FZ 200 300 400 700 1100 1500 1800 2000 2300
FR 290 430 570 1010 1560 2200 2600 2900 3300
Each End Nozzle
FX 200 300 400 700 1100 1500 1800 2000 2300
FY 130 200 260 460 700 1000 1200 1300 1500
FZ 160 240 320 560 850 1200 1500 1600 1900
FR 290 430 570 1010 1560 2200 2600 2900 3300
Each Nozzle
MX 340 700 980 1700 2600 3700 4500 4700 5400
MY 260 530 740 1300 1900 2800 3400 3500 4000
MZ 170 350 500 870 1300 1800 2200 2300 2700
MR 460 950 1330 2310 3500 5000 6100 6300 7200
F is Force in pounds ; M is Moment in foot pounds ; R is the resultant ; X, Y, Z : Orientation of
Nozzle Loads.
API 610 specifies that the pump casing should be designed to withstand double the forces and
moments as above. The piping configuration that produces loads and moments outside the
above range is also acceptable provided the conditions as specified in Appendix F of the above
code are satisfied. For direction of forces and moments see Fig. 3.1

Stress Analysis 18
PIPING ENGINEERING CELL

Fig.3.1

ISO 5199 – 2002(E) in Annex B with Alloy Steel or Steel Castings at


gives the allowable values of inlet and outlet nozzles as :
forces and moments on the pump
nozzle. The basic values given in
table B3 should be multiplied byFx = 75 D lbs Mx = 125 D ft.lbs
corresponding coefficient given
Fy = 75 D lbs My = 125 D ft lbs
in table B5 or B6. This is based
on the study and tests undertakenFz = 75 D lbs Mz = 125 D ft lbs
within EUROPUMP (European
Association of Pump where D isinthe
USCSnominal diameter of
Manufacturers) together with the nozzle in inches.
support of piping specialists.
3.3 Centrifugal Compressors
3.2 Positive Displacement API 617 “Centrifugal
Pumps Compressors for General Refinery
Service” has been specifying that the
compressors shall be designed to
withstand external forces and
Y
Z moments on each nozzle at least 1.85
times the value calculated in
X accordance with NEMA - SM 23.
Experience has shown that there has
be not been uniform interpretation of
The American Petroleum “1.85 times NEMA”. Therefore API
Institute Standard 676 specifies in 617 has modified the equations and
clause 2.4 the limiting values for the attached in Appendix G the formulae
Rotary Positive Displacement Pumps

Stress Analysis 19
PIPING ENGINEERING CELL

for the calculation of the acceptable Fc = Combined resultanat of inlet side


forces and moments. and discharge forces in
pounds/Newtons
The forces and moments acting on
compressor(s) due to the inlet pipe Mc = Combined resultant of inlet side
and discharge pipe connections are: and discharge moments resulting
from forces in ft lbs/Newton meters
1. The total resultant force and total
resultant moment imposed on the Dc = Diameter (in inches) of a
Compressor at any connection must circular opening equal to the total
not exceed the following: area of inlet side and discharge
opening upto a value of nine
3F + M  927D or (9)inches (230 mm) in diameter. For
values beyond this use value of

Dc = (18 + Equivalent Diameter)


inches
F + 1.09 M <= 5.41 D in SI units 3
F = Resultant force (lbs/Newton) = 460 + Equivalent Diameter in
M = Resultant Moment mm.
(ft.lbs/Newton meters) 3
D = Pipe size of the connection
(IPS) in inches/millimeters The components of these resultants
upto (8) inches shall not exceed
(200 mm) in diameter. in USCS
For sizes greate than this, use a value Fy = 231 Dc My = 231 Dc
of D equal to (16+IPS)/3 inches or
(400+D)/3 in mm. Fz = 185 Dc Mz = 231 Dc
Fx = 92 Dc Mx = 462 Dc
2. The combined resultant of the in MKS units
forces and moments of the inlet side
and discharge connections resolved Fy = 40.5 Dc My = 12.3 Dc
at the centre line of the discharge Fz = 32.4 Dc Mz = 12.3 Dc
connection must not exceed the Fx = 16.1 Dc Mx = 24.6 Dc
following two conditions.
where,
These resultants must not exceed Fy = Vertical component of Fc
462D c M c Fz = Horizontal components of Fc
Fc  in USCS at right angles to compressor
2
shaft
Fc = 40.4Dc – 1.64Mc in SI Units Fx = Horizontal component of Fc
Parallel to compressor shaft

Stress Analysis 20
PIPING ENGINEERING CELL

Mx = Component of Mc in a
hoizontal plane parallel to
Compressor shaft
My = Component of Mc in a vertical
plane
Mz= Component of Mc in a
horizontal plane at right
angles to the compressor shaft

Stress Analysis 21
PIPING ENGINEERING CELL

+y R ig h t a n g le t o
c o m p resso r sh aft

Fy
+M y

Fx
+x P a r a ll e l t o
c o m p resso r sh aft
+M x

Fz +M z

+z

F ig . 3 . 2 : C o m p o n e n t s o f f o r c e s a n d m o m e n t s
o n c o m p r e s s o r c o n n e c t io n

Stress Analysis 22
PIPING ENGINEERING CELL

3. These values of allowable forces 3.4 Reciprocating Compressors


and moments pertain to the API 618 “Reciprocating
compressor structure only. They do Compressor for General Refinery
not pertain to the forces and moments Service” do not specify the limit on
in the connecting piping flanges and the allowable forces and moments in
flange bolting which should not the code. The vendor shall specify
exceed the allowable stress as the forces and moments for each
defined by applicable codes and nozzle in the tabular form. However,
regulatory bodies. the values as per API 610 can be
considered for guidance.
Forces on inlet connections
are to be transferred along with
3.5 Steam Turbines
moments to discharge connection to
NEMA – SM 23 requires that
analyze the compressor for resultant
the forces and moments acting on
forces and moments. But, the
steam turbines due to the steam inlet,
transfer of force will generate
extraction, and exhaust connections
additional transfer moments, which
should be evaluated by simple set of
are added to the total of moments to
force/moment calculation similar to
give resultant moments.
centrifugal compressors. These
computations shall be done as below.

1. The total resultant force and


total resultant moment imposed on
the turbine at any connection should
not exceed the values calculated as
per the following equation.

500D -
M

Stress Analysis 23
PIPING ENGINEERING CELL

3 F + M  500 D or F  Fc = Combined resultant of inlet,


3 extraction, and exhaust forces,
where, in lbs.
Mc = Combined resultant of inlet,
F = Resultant force (lbs) including extraction, and exhaust
pressure forces where moments, and moments
unrestrained expansion joints resulting from forces, in ft lbs.
are used at the connection Dc = Diameter (in inches) of a
except on vertical exhausts. circular opening equal to the
Full vacuum load is allowed on total areas of the inlet,
vertical down exhaust flanges. extraction, and exhaust
openings up to a value of nine
F= Fx2 + Fy2 + Fz2 inches in diameter.

M = Resultant moment in foot- For values beyond this, use a value of


pounds, Dc (in inches) equal to:

M= Mx2 + My2 + Mz2 (18 + Equivalent diameter in


inches)
D = Nominal pipe size of the
connection in inches up 3
to 8 inches in diameter.

For sizes greater than this, use a value The components of these resultants
of should not exceed:
(16 + IPS) Inches Fx = 50 Dc Mx = 250 Dc
D (in inches) = Fy = 125Dc My = 125 Dc
3 Fz = 100 Dc Mz = 125 Dc
2. The combined resultants of The components are as follows:
the forces and moments of the inlet,
extraction, and exhaust connections, Fx = Horizontal components of Fc
resolved at the centerline of the parallel to the turbine shaft.
exhaust connection should not exceed Fy = Vertical component of Fc
the following two conditions. Fz = Horizontal component of Fc at
right angles to the turbine
These resultants shall not exceed: shaft.
Mx = Component of Mc around the
250 Dc - Mc horizontal axis parallel to the
Fc = 2 turbine shaft
My = Component of Mc around the
where, vertical axis
Mz = Component of Mc around the
horizontal axis at right angles
to the turbine shaft.

Stress Analysis 24
PIPING ENGINEERING CELL

+y R ig h t a n g le t o
t u r b in e s h a f t

Fy
+M y

Fx
+x P a r a lle l t o
t u r b in e s h a f t
+M x

Fz +M z

+z

F ig . 3 . 8 : C o m p o n e n t s o f f o r c e s a n d m o m e n t s
o n t u r b in e c o n n e c t io n

3. For installation of turbines pressure loading) of 15 ½ times the


with a vertical exhaust and an exhaust
unrestrained expansion joint at the area in square inches.
exhaust, an additional amount of 4. These values of allowable
force caused by pressure loading is force and moment pertain to
allowed. (This additional force is the turbine structure only.
perpendicular to the face of the They do not pertain to the
exhaust flange and is deemed to act at forces and moments in the
its centre), For this tye of connecting piping, flange,
application, calculate the vertical and flange bolting, which
force component on the exhaust should not exceed the
connection excluding pressure allowable stress as defined by
loading. Compare this with one sixth applicable codes and
of the pressure loading on the explanatory notes.
exhaust. Use the larger of these two
numbers for vertical force component
on the exhaust connection in making 3.6 Air Cooled Heat Exchangers
calculations outlined in 1 and 2. The American Petroleum
The force caused by the Institute Standard 661 for “Air
pressure loading on the exhaust is Cooled Heat Exchangers for General
allowed in addition to the values Refinery Services”
established by the foregoing up to a covers the allowable loads on the
maximum value of vertical force in vertical, colinear nozzles found in
pounds on the exhaust connection most single multibundled air-cooled
(including heat Exchangers.

Stress Analysis 25
PIPING ENGINEERING CELL

3.6.1 Each nozzle in corroded


Fig.3.9: Direction of Forces & condition shall be capable of
Moments on the Nozzle withstanding the
simulataneous application of
API 661 has the following two the following moments and
requirements. forces.

Table 3.2: Nozzle loading as per API 661 / ISO 13706

Nozzle size Forces in lbs (N) Moments in ft lbs (Nm)


NB mm (In) Fx Fy Fz Mx My Mz
40(1.5) 150(670) 230(1020) 150(670) 80(110) 110(150) 80(110)
50(2) 230(1020) 300(1330) 230(1020) 110(150) 180(240) 110(150)
80(3) 450(2000) 380(1690) 450(2000) 300(410) 450(610) 300(410)
100(4) 750(3340) 600(2670) 750(3340) 600(810) 900(1220) 600(810)
150(6) 900(4000) 1130(5030) 1130(5030) 1580(2140) 2250(3050) 1200(1630)
200(8) 1280(5690) 3000(13340) 1800(8010) 2250(3050) 4500(6100) 1650(2240)
250(10) 1500(6670) 3000(13340) 2250(10010) 3000(4070) 4500(6100) 1880(2550)
300(12) 1880(8360) 3000(13340) 3000(13340) 3750(5080) 4500(6100) 2250(3050)

350(14) 2250(10010) 3750(16680) 3750(16680) 4500(6100) 5250(7120) 2630(3570)

Stress Analysis 26
3.6.2 The design of each fixed or The limiting values for forces
floating header, the design of fixed and moments should be laid down by
headers to side frames, and the design the manufacturer. Restrictions are
of other support members shall applied on nozzle rotations also in
ensure that the simultaneous this case to take care of the
application (sum) of all nozzle clearances between the tube and
loadings on a single header will cause refractory lining. The thumb rule
no damage. The components of the used is :
nozzle loadings on a single header Forces = 200 to 300 lb/in.
shall not exceed the following values. nominal bore of
nozzle
Mx = 4500 ft lb (6100 Nm) Moments - Equivalent to Sh/4
My = 6000 ft lb (8130 Nm) Nozzle Rotation - From 1/2º
Mz = 3000 ft lb (4070 Nm) to 1º
Fx = 2250 lb (10010 Nm)
Fy = 4500 lb (20020 Nm) 3.9 Flat bottom Welded Storage
Fz = 3750 lb (16680 Nm) Tanks
to API 650
Note :- The application of the The design of the piping
moments and forces shown in table system connected to thin walled,
will cause movement that will tend to large diameter cylindrical vertical flat
reduce the loads to the values given. bottom storage tanks pose a problem
in the analysis of the interface
3.7 Shell & Tube Type Heat between the piping system and the
Exchangers tank nozzle. The parameters to be
The designer has to set the considered are the stiffness of the
limiting values or to check the vessel tank shell, the radial deflection and
connections for the meridional rotation of the shell
the nozzle loading imposed by the opening at the nozzle connection
connected piping. resulting from the static head,
pressure and uniform or differential
temperature between the shell and the

The rough guide generally followed Y M C = M Y

is: - R
F L = F Y

Resultant : 200 lb./in NB of nozzle


Maximum
Force
2a

X
Bending : Equivalent to bending F R = F X M r = M x

Moment stress in standard schedule


A
X

pipe between 4000 to Z


c
L = X

t
5000lbs./in2
B
X

3.8 Fired Heaters bottom. Although three primary


forces and moments may be imparted
by the piping on to the shell
connection, only the radial thrust (FR) the piping connection at the lower
and two moments i.e. the longitudinal half of the bottom shell course.
moment (ML) and the circumferential The following are the steps involved
moment (MC) are significant causes in the determination of allowable
of shell deformation. loads.
a) Determine the non dimensional
Fig.3.10: Storage Tank Nozzle quantities XA/(Rt)0.5 , XB/(Rt)0.5
Details and XC/(Rt)0.5 for the nozzle
connection under consideration.
a = Outside radius of the nozzle in b) Layout two sets of orthogonal
mm(in) axes on graph paper and label
FR = Radial Thrust applied in N (lbf) abscissas and ordinates as
FP = Pressure head at the nozzle a2P specified in ‘e’ below.
L = Vertical distance from centre line c) Layout two sets of orthogonal
of axes on graph paper and label
nozzle to tank bottom in mm(in) abscissas and ordinates as
t = shell thickness at the nozzle specified in ‘f’ below.
connection d) Using the values of FR, ML and
in mm(in) MC obtained from the piping
P = Pressure from liquid head in Mpa analysis determine the factors
(lbf/in2) (/2YF)(FR/FP)
XA = L + a in mm (in) (/aYL)(ML/FP)
XB = L - a in mm (in) and (/aYC)(MC/FP)
XC = L in mm (in) e) Plot point (/2YF)(FR/FP) ,
YC = Coefficient determined from (/aYL)(ML/FP) on the nomogram
Fig. P-4B with X axis and Y axis
(API650) respectively.
YF , YL = Coefficient determined from
Fig. P - 4A (API650) f) Plot point (/2YF)(FR/FP) ,
 = a/(Rt)0.5 (/aYC)(MC/FP) on the nomogram
MC = Circumferential moment at the with X axis and Y axis
nozzle respectively.
from the piping system in N-mm
(in-lbf) g) Construct boundaries as lines at
ML = Longitudinal moment at the 45 degree angles between
nozzle abscissa and ordinates passing
from the piping system in N-mm through the calculated value.
(in-lbf)
h) The external piping loads FR, ML
Appendix P of API 650 and MC are acceptable if both
establishes minimum points determined from b and c
recommendations for the design of above lie within boundaries of the
storage tank opening subjected to nomogram constructed for that
external piping loads. This is particular nozzle.
recommended only for tanks larger
than 36M (120 ft) in diameter and is
considered as an accepted practice for
3.10 Static Equipment such as
columns, reactors, tanks and
vessels

Each nozzle, 2” NB and larger, for


columns, drums and shell & tube heat
exchangers constructed out of steel or
alloy is recommended to be designed
to withstand forces and moments
from the thermal expansion and
N O ZZLE

F A

M T

M b
M L
F L F
F R
ER
AN G
XCH
H E AT E
R
EL O
V ESS
sustained loading from the piping as
per the following criteria. These
forces and moments shall be
considered to be acting at the
intersection of nozzle and shell in he
corroded condition. A total of 7000
full temperature cycles shall be 3.11 Orientation of Forces &
considered for the expected life of the Moments on Vertical Equipment
equipment.

NOZZLE Fig 3.12 Orientation of Forces and


Moments on Horizontal Equipment
FA

MT

MZ
9.1 Nozzle to Shell or channels
MX a) Moments
FZ
i) Longitudinal bending moment
FR FX
ML =  130 D2 Nm
ii) Circumferential bending
moment
FL M =  100 D2 Nm
M iii) Resultant bending moment
FR
ML
F A
LE
N OZZ
MT F
VESSEL
Mb = (ML 2 + M 2) ½
=  164 D2 Nm
iv) Torsional & moment
Flange Rating  Valve
Mt =  150 D2 Nm ANSI DIN Heat Columns and
Class Exchangers Drums
b) Forces 150 # PN 10 & 16 0.75 0.6
i) Axial force in plane of Flange 300 # PN 25 & 40 0.75 0.7
FL = 2000D N 600 # PN 64 & 100 1.25 0.8
900 # PN 160 3.00 1.8
ii) Tangential force in plane of
1500 # PN 250 & 320 4.00 3.0
Flange 2500 # PN 400 5.60 3.3
F= 1500D N
Table 3.3
iii) Resultant shear force
FR = (FL 2 + F 2) ½
=  2500D N 3.10.3 The local stress intensity at
iv) Radial Tensile or the nozzle connection due to
compressive Force attachment of piping can be
FA =  2000D N computed using the welding
9.2 Nozzles to Formed Heads research council bulletin 107
a) Moment setting the limitations as.
i) Resultant Bending moment
a) The local sustained stress
Mb =  164 D2 Nm
intensity at the nozzle connection
Where Mb is the resultant of should be less than 0.5 sm
the components MX and MZ B) The sum of local sustained stress
ii) Torsional Moment intensity and the local expansion
Mt =  150 D2 Nm stress
b) Forces intensity at the nozzle connection
i) Resultant shear force must
FR = 2500D N where FA is be less than 2 sm
the resultant of the
components FX and FZ where, sm is the allowable stress
ii) Radial Tensile or compressive intensity for the material at the
force operating temperature.
FA = 2000D N
Where  is as per the table 3.3 4.0 Data required for flexibility
and D is the nominal calculations
diameter in inches. The
orientation of the forces and The following data will be
moments shall be as per Fig required for the flexibility
3.11 & 3.12. These loadings calculations if it is carried out
shall be considered as being manually or by the use of software. It
caused by 67 % thermal and is therefore prudent to have this ready
33 % dead weight load. before starting.
The direction of coordinates are fixed 16. Young’s Modulus at
as below: Flexibility
Temperature
17. Thermal Expansion at
Flexibility
Temperature
18. Allowable stress at
Ambient/ Installation
temperature
19. Allowable stress at
flexibility temperature
20. Bend radius and type
of bend
1. Code of Practice
21. Branch connection
2. Basic Material of type
Construction of Pipe
22. Weight of attachments
3. Ambient / Installation - Valves and
temperature Specialties
4. Number of Thermal 23. Terminal
Cases movements with
5. Flexibility directions
Temperature (See Note: The Code states that the design
Note) temperature shall be assumed to be
6. Design Pressure same as the fluid temperature unless
calculations or test supports the use
7. Outside diameter of
of other data.
Pipe
8. Type of construction
of pipe
9. Nominal Thickness of
Pipe 5.0 METHODS OF
10. Manufacturing FLEXIBILITY ANALYSIS
tolerance 5.1 Check As Per Clause
11. Corrosion allowance 119.7.1/ 319.4.1 of the Code
Clause 119.7.1(A3)/
12. Pipe Weight
319.4.1(c) specifies that no formal
13. Insulation Weight analysis is required in systems which
14. Specific Gravity of are of uniform size, have no more
Contents than two points of fixation, no
intermediate restraints and fall within
15. Young’s Modulus at the empirical equation.
Ambient/Installation
Temperature
DY
K 1
U = Anchor distance, straight line
 L U  2
between anchors in ft. or (m)
where, K1 = 30 SA/Ea in USCS
D = the outside diameter of pipe
In inch (or mm) = 208300 SA/Ea in SI units

Y = Resultant of total displacement – As per B31.3


strains K1 = 0.03 for USCS
In inch (or mm) to be absorbed by = 208.3 for SI units.
the
- As per B31.1
Piping system.
Where Ea is the Modulus of Elasticity
L = Developed length between the at the installation temperature
anchors
and SA is the allowable stress range
In ft. or (m)
Example: Let us apply the above check for the following system.

e2.7"/100ft
Pipe - 6" (150 mm NB) Sch. 40
2.735
carbon steel to ASTM A106 Δx 
100
0

Gr. B Δy
2.730
0
Design Temperature - 400 F 100
2.745
Δz 1
(2040C) 100
Y Δx 2
Δy
Step 1  0.945 2 
1.739"
To establish the anchor to anchor
distance U Step 4
Total length in X direction = 35’
DY
K
Total length in Y direction = 30’  L U  2
Total length in Z direction = 25’ + 6.6251.739

20’ 11064.42 2
= 45’
0.0055
2
U= x 2 + y2 + z 0.03
 35 2  30 2  45 2 64.42'
since K  K 1 the configuration is safe
Step 2 Please note that no general proof
To determine value of L. can be offered that this equation
L = x  y + z = 35 + 30 + 45 = 110 ft. will yield accurate and
conservative results. It is not
applicable to systems used under
Step 3
severe cyclic conditions. There is
To calculate resultant total no assurance that the terminal
displacement Y reactions will be acceptably low,
From Appendix C, ASME B 31.3 even if the system satisfies the
above equation.
Linear Expansion between 70 o F
and 400F.
5.2 Guided Cantilever Method vessels are of large
Suppose that we have two diameter and, therefore,
vessels T1 and T2 say 20 m apart thick and the pipe is small.
and we have to run the pipe from
T1 and T2 between two nozzles at
the same elevation. Obviously the
most economical way of doing this,
purely from the performance
aspect, is to join them by a straight
pipe as shown in Fig. 5.2.
However, if the equipment
is laid out differently, it will be
possible to run the pipe in two
different sections at right angles to
each other as shown in Fig. 5.5.

Now further, suppose that


everything is in carbon steel and the
vessel T1 has its temperature raised
to 200C. When the valve is
opened, there will be expansion in
the connecting pipeline, which can
be calculated as below.
Expansion of carbon steel from
210C to 2000C = 2.2 mm/ m

(Refer Appendix C ASME B 31.3)


With this configuration of
Total Expansion = 20 x 2.2 = 44 piping, as the point B moves out to
mm B’, it is able to bend the leg BC to
To absorb this expansion, one of the position B’C.
following two things can happen. It is simple to calculate
1. As the pipe expands it will length L of BC which will allow
dent the sides of the vessel expansion "  " to be absorbed
as in Fig. 5.3 while the stresses are restricted to a
given value and this is the simplest
concept of all in the field of
flexibility analysis, namely
"Minimum Length".

2. The pipe will buckle as


shown in Fig. 5.4 if the
If L is length of BC in ft. ( l = 12 L)
3
144PL
δ=
EI
Hence,
EIδ
P  above
144L3
Maximum bending moment at Bor C
 M ft.lbs.
M Y12
Maximum bending stress f  l
I
ODof pipe
Y
2
12M Y
f 
I
when the pipe bends as shown by 12 P L D
the dotted line, i.e. B’C, it is  
I 2 2
referred as "Guided cantilever “ EIδ
Substituting P 
As per Elastic Theory, 144L3
12 EIδ L D
Pl 3 f    
δ I 144L3 2 2
12EI
where, DEδ
i.e. f 
δMovement in inches 48L2
P Force required to bend BC in lbs DEδ
L 
lLength of BC in inches 48f
E Young' s Modulus in lbs/in 2
By putting the value of stress range
IMoment of inertia about bending axis in in 4
calculated, as discussed earlier and
the modulus of elasticity of the
material, we can arrive at the
"Minimum Leg Length".
e.g. :- In the previous layout if we
restrict the stress at 16,000 psi and
consider modulus of elasticity of
carbon steel as 29.5 x 106 psi and
assume the pipe size as 6" NB
(6.625" OD)
Expansion of piping between T1
and T2,
δ  0.8
D
L 
4
6

14.

This indicates that the length BC


should not be less than 4.6 m.
We can also calculate the stress the stress developed can be
developed in such a system of calculated as
known dimensions of leg BC by the
same method. l = AB = BC = 10 m = 394
inches
Pl3 E = 29.5 x 106 lbs/ in2
 = R = 6.625/2 inches
12EI  = 1.73/2 inches

12 EI 6 x 29.5 x l06 x 6.625 x 1.73


Hence P = SE =
l3
(394)2 x 2 x 2
Pl
M = = 3267 psi
2

12 E I  x l 5.3 Piping Elements - Their


6 EI Individual Effects
= =
3 Let us analyze each of the piping
l 2
elements to see how it contributes
l2
to flexibility.
M 6 EI  5.3.1 STRAIGHT PIPE :
F = = FLEXIBILITY IN TORSION
Z l2 Z
I
L
I
R = ;Z =
Z M
M
R
6ER
Solving for SE = 
2
l
Where R = Outer radius of pipe,
inches Fig 5.7
I = Moment of inertia of
cross section, in4
Z = Section modulus, in3
E = Modulus of elasticity,
lbs/in2
l = Length, inches
In Fig.5.5 if the vessels are
arranged in such a way that AB and
BC are equal and 10 M each, then
G = Modulus of rigidity, lbs/in2
If a bending moment M is applied (KPa)
to the end of a straight piece of
pipe, it behaves as a uniform beam J = Polar moment of inertia, in4
and exhibits change of slope from (mm4)
end to end, as given by the It can be shown that for metals
expression.
G = E / 2.6 and
J = 2 x I for circular cross
section
ML
  Hence,
EI
TxL
Where,  =
 = Angle, radians
E / 2.6 x 2I
M = Bending moment, in lbs (mm –
N) TL
= 1.3
E = Young’s Modulus, lbs/in2 (KPa)
I = Moment of Inertia, in4 (mm4) EI

L = Length, inches (mm) This result is important as far as


piping routings in three planes are

Fig 5.8 concerned. It shows that a given


length of pipe can give 30% more
If the same pipe is subjected to a rotation if the moment from the
constant twisting moment, the adjacent legs produces torsion
rotation of one end relative to the instead of bending. This can create
other end is given by : moments, which should also be
TL kept in limits.
 
GJ There are piping
components other than straight
Where, pipe, which are required to make a
complete piping system. These are
 = Angle of twist in radians elbows, tees, reducers, valves, etc.
The knowledge of individual effects
T = Torsion moment in – lbs (mm on the flexibility and the stresses in
– N) each element is essential to analyze
L = Length, inches (mm) a piping system close to its true
nature of behavior.
5.3.2 ELBOWS:
Early calculation on the
flexibility analysis containing
elbows proved that the structural
engineering theory and the results
of experiments did not agree well.
Practical piping system was more
flexible than predicted and the
discrepancy was due to the
flexibility of elbows.
The first theoretical analysis
of the behavior of pipe bends when
Fig. 5.10
subjected to a bending moment was
made by Theodore Von Karman, Let us consider how flattening of
who showed that when a curved the cross section occurs
pipe is subjected to a bending
moment in its own plane, the
circular section gets flattened and
this results in increased flexibility.
This was further developed by
Hovgarrd, Berkins, Vigners &
Markel.

Flexibility Factor
The ratio of the flexibility of
a bend to that of a straight pipe
having the same length and cross
section is known as its "Flexibility
Factor", usually denoted by the
alphabet "k". Fig. 5.11
When bending moment M is
applied as shown, tensile stresses
are developed on the outer fibres
and compressive stresses on the
inner ones.

ML
=
EI
Fig. 5.12 of pipe wall and which is
circumferential in direction. If we
Let us consider a thin slice taken consider one half of the ring, the
between two radial planes "XX" stress system gets illustrated as
and "YY". (see Fig. 5.12). above.
The resultant effect of the These circumferential
tensile load "T" in the outer fibres stresses due to bending moment M
is an inward radial load on the can be many times the value of
element. Similarly the resultant of MY/I obtained by bending theory
the compressive loads "C" in the of structural members. The factor
inner fibres is an inward radial load by which the circumferential
on the element. stresses exceed the longitudinal
stresses in the bend is called the
If we view the slice as a "Stress Intensification Factor"
cross section, and draw a loading called S.I.F. It can be defined as the
diagram for the ring, we arrive at ratio of the actual bending stress for
the situation shown below. Under a moment applied to nominal
the applied loading the ring flattens section.
into an ellipse with its major axis
horizontal (see Fig. 5.13.a) The effect of the existence
of these circumferential stresses is
that when elbow is subjected to
repeated "in-plane" bending, it
ultimately develops a fatigue crack
along its sides.
When we take additional
benefit by Flexibility factor due to
flattening of elbows, consideration
If the bending moment is should be given to the induced
reversed, the tensile and circumferential stresses by
compressive forces will also get multiplying the stresses at the bends
reversed and cross section gets due to overall bending moment by
elongated instead of getting the appropriate "Stress
flattened. (See Fig. 5.13.b) Intensification Factor".
Appendix D of ASME B
Fig. 5.14 Circumferential Stress 31.1 & 31.3 tabulates the
in Pipe wall expressions to be used for
calculating the Flexibility Factor
If the element is analyzed in more and Stress Intensification Factor.
detail, it is seen that the flattening The parameter used for the
produces bending moments in the calculation of these factors is called
rings, which are maximum at the the "Flexibility Characteristic"
ends. These moments produce a denoted by letter "h"
stress which varies from tension to
compression through the thickness
In case of Mitre Bends an
equivalent bend radius is used in
the equation to calculate 'h'. The
equivalent bend radius (Re) is
estimated by
Re  r2 (1  0.5s / r2 cot  )
for closely

Fig. 5.15 Re  r2 (1  cot  )


TR1 spaced mitres
Flexibility Characteristic h =
 r2  2 for widely spaced
T  Wall thickness, inches (mm) mitres
where,
R 1 Mean radius of bend, inches (mm)
r2 Mean radius of pipe, inches (mm) S = mitre spacing at centre
line,
Using this parameter, code
indicates that inches (mm)

The flexibility factor = k = 1.65/h  = One half of angle


Inplane S.I.F. = ii = 0.9/h between cuts
Outplane S.I.F. = Io = 0.75/h r2 = mean radius of pipe,
When any problem is analyzed, the inches
following considerations are made. (mm)
a) The Flexibility Factor applies to
bending in any plane.
5.3.3 TEES
b)The stress intensification factor
is greater for "inplane" As far as branch off is
bending than for "outplane" concerned the flexibility factor is 1
bending. ASME B 31.3 and the stress intensification factor
permits the use of inplane SIF can be calculated based on the
for any plane whereas B 31. 1 specific formula adapted by code
does not separate out these ASME B31.1 / B31.3 and given in
two. Appendix D. These vary depending
on the type of branch connection.
Virtual Length The unreinforced fabricated
The product of length of arc centre tee is modelled using same formula
line and Flexibility factor is for widely spaced miter bend with
referred as "Virtual length" of a single miter i.e. half angle of 45o.
bend and these are considerable This produces the flexibility
while analyzing thin walled large characteristic of
diameter pipes. h  T / r2
Mitre Bends
For buttweld tees, Markyl
adapted bend equation with
equivalent radius (Re) and factor occurs when the ratio of
equivalent thickness (Te). branch to header is 0.7 at which the
non-conservation is of the order of
two.
h  c(Te Re/ r2 2 )
5.3.4 FLANGES
where, For flanges also the
flexibility factor is 1 and the
c = ratio of tee to pipe section various types of flanges are
modulii. considered to have the following
= (Te/T)3/2 as recommended by Stress Intensification Factors.
ARC Markyl. S I F for Flanges:
Te = Equivalent pipe wall thickness
inches (mm) Weld neck flange 1.0
= 1.60T as recommended by ARC Slip-on flange 1.2
Markyl Socket weld flange 1.3
Lap joint flange 1.6
Re = Equivalent bending radius Threaded flange 2.3
inches (mm)
= 1.35 r2 as recommended by
ARC Markyl The flange when attached to the
Substituting these values in the bend exerts a severe restraint to the
expression flattening of the cross section due
for h to its heavy construction. Hence
attachment of the flange to an
2
h  (Te / T ) 3 / 2 (Te1.35r2 / r2 ) elbow or a mitre bend reduces the
flexibility as well as the stress
h  4.4T / r2 intensification factor. Flange at
both ends of the elbow reduces
As far as the stub these factors further.
connections are concerned, the ASME B 31.3 indicates these
major problem is in the out of plane correction (reduction) factor as:
bending moment on the header.
Stresses due to these moments can C1 = h1/6 for one end flanged
never be predicted from the size on C1 = h1/3 for both ends flanged
size tests. Errors due to these
moments can be non-conservative Flanges are designed to remain
as much as a factor of two to three. leakfree under hydrostatic test
Further when r2 / R ratio is very pressure when cold and under
small the branch connection has operating pressure when hot. The
little impact on the header and the design of flanges does not take into
calculated stress could be account bending moment in the
unreasonably large by using large pipe. However, the flanges transmit
SIF. It has been pointed out by R. some bending moment before they
W. Schneider of Bonney Forge that ‘open up’, the value being very
the highest stress intensification small. This generates the wire
drawing effect on the mating 5.3.7 EFFECT OF PRESSURE ON
surface of the flange. Hence, SIF AND FLEXIBILITY
additional flexibility is to be FACTOR
considered where a flange joint is In large diameter thin
located near a point of high bending walled elbows and bends, pressure
moment. can significantly affect the
Flexibility Factor `k’ and Stress
5.3.5 REDUCERS Intensification Factor `i’. Hence
The reducer could be either the correction factor as below
eccentric or concentric and in both should be applied on the values
the cases the stress intensification available from the table.
factor and flexibility factor are 1.
Divide ‘k’ by
The overall length is very small
compared to the piping and hence   P  r 
7/3
 R1 
1/3

1  6  2    
the effect of the same is neglected   E  T   r2  
in ASME B31.3. ASME B31.1
indicates the SIF as 2 max for a Divide ‘i’ by
concentric reducer as per ASME  5/2 2/3

 P  r   R 
B16.9. 1  3.25  2   1  
  E  T   r2  
5.3.6 VALVES
where
Valve is normally
considered as a short length of very T = Nominal wall thickness of
thick pipe. The effect of fitting,
temperature when the valve is inches (mm)
closed is more significant in the r2 = Mean radius of the matching
analysis. end,
inches (mm)
P = Guage pressure psi (KPa)
E = Modulus of Elasticity psi (KPa)
Rl = Bend radius, inches (mm)

This is called Bourden effect and


this stiffening effect of pressure on
bends are not considered in ASME
B 31.1.

6.0 CODE STRESS


EQUATIONS
Fig. 5.16 The stress equations specified in the
code are derived from the
theoretical and substantiated by
investigative work. To make the
calculation simpler, the code
calculates the stress intensity only MA = resultant moment due to
for expansion stress, since this load weight and sustained loads, in-lb
case contains no hoop or radial (mm – N)
components.
= Mx2 +My2 + Mz2
6.1 ASME B 31.1
ASME B 31.1 specifies under Z = Section Modulus, in3 (mm3)
clause 104.8 that to validate a tn = Nominal Thickness, in (mm)
design under the rules of this Sh = Basic allowable stress at the
clause, the complete piping system operating temp., psi (KPa)
must be analyzed between anchors
for the effects of thermal expansion,
weight, other sustained loads and 6.1.2 THERMAL EXPANSION
other occasional loads. STRESS RANGE
The effects of thermal expansion
6.1.1 STRESS DUE TO must meet the requirements of the
SUSTAINED LOADS following equation.
The effects of pressure, weight and i Mc
other sustained mechanical load SE =  SA + f
must meet the requirements of the (Sh –SL)
following equation. Z

P Do 0.75 i MA In USCS units


SL = +  Sh 1000 i Mc
4 tn Z SE =  SA + f (Sh –SL)
Z
in USCS units
In SI unit
P Do 1000(0.75 i) MA Where,
SL = + SE = Expansion stress range psi
 Sh (KPa)
4 tn Z Mc = Range of resultant moments
due to thermal expansion,
in SI units in - lb
(mm – N)
where
SL = Sum of the longitudinal = Mx2 +My2 + Mz2
stresses due to pressure,
weight and other sustained SA = Allowable stress range (Ref
loads, psi (KPa) 2.4.3 above) psi (KPa)
i = Stress intensification factor
(ref. Appendix D-1)
The product 0.75i shall
never be taken as less than 1. 6.1.3 STRESS DUE TO
OCCASIONAL LOADS.
The effects of pressure, 6.2 ASME B 31.3
weights, other sustained loads and
occasional loads including 6.2.1 ASME B 31.3 does not
earthquake must meet the provide an explicit equation for
requirements of the following sustained stress calculation, but
equation. requires that Piping Engineer
should compute the longitudinal
P Do 0.75 i MA 0.75 i MB stresses due to weight; pressure and
+ + other sustained loading and ensure
 KSh that these do not exceed Sh. The
4 tn Z Z thickness of pipe used in
calculating SL shall be the nominal
in USCS units. thickness less the erosion and
corrosion allowance. This is
P Do 1000(0.75 i) MA calculated by looking at Clause
+ + 302.3.5 (c)
4 tn Z
Fax (iiM i) 2 + (ioM o) 2
1000(0.75 i) MB PDo
=  SL = + +
KSh Am Z
Z 4t
In SI units.  Sh in
USCS
Where units
K = 1.15 for occasional loads
acting less
than 10% of any 24 hr.
operating Fax 1000 [(iiMi) 2+
2 1/2
period. (ioM o) ]
K = 1.2 for occasional loads SL = +
acting less
than 1% of any 24 hr. Am Z
operating  Sh in SI units
period.
MB = Resultant moment loading on where,
cross section due to occasional SL = Sum of longitudinal stress due
loads. If calculation of moments to pressure weight and other
due to earthquake is required, use sustained loading, psi (KPa)
only one half of the earthquake
moment range. Effect of anchor Fax = Axial force due to sustained
displacement due to earthquake (primary) loading, lbs (kg)
may be excluded from the equation Am = Metal cross sectional area,
if they are covered in Thermal in2 (mm2)
Expansion stress range calculation.
Mi = In-plane bending moment due  i i M i  2  i 0 M 0  2
to sustained (primary) Sb 
Z
Loading, in-lb (mm–N)
Mo = Out-plane bending moment 1000  i i M i  2  i 0 M 0  2
Sb 
due to sustained (primary) Z
Loading, in-lb (mm–N) where
ii,io = In-plane and out-plane stress i i in  planeSIF
intensification factors i o out  planeSIF
Sh = Basic allowable stress at the
M i in  planeBendingMoment
operating temp., psi (KPa)
M o out  planeBendingMoment
6.2.2 THERMAL EXPANSION ZSectionModulusofPipe
STRESS RANGE

The computed displacement stress


range shall be done as below (Ref.
Clause 319.4.4).

(a) The range of bending and


torsional stresses shall be computed
using the as installed Modulus of
Elasticity ‘Ea’ and then combined as
below to determine the computed
stress range.
2 2
S E  S b  4S t Fig. 6.1 – Moments in Bends
where
S b Resultant bending stress, psi (KPa)
S t Torsional Stress = Mt/ 2z, psi
1000Mt/ 2z, KPa
M t Torsional moment, in - lb (mm - N)
ZSection Modulus of Pipe, in 3 (mm 3 )

(b) The resultant bending stress to


be used in the above equation for
elbows and full size branch
connection shall be calculated as
follows Fig. 6.2 – Moments in Tees

For Reducing outlet branch


connections, the equation shall be
as follows.
For Header
 i i M i  2  i 0 M 0  2 sustained and occasional loads shall
Sb  not exceed 1.33 times the basic
Z
allowable stress. The method
in USCS units. adopted is to calculate the sustained
and occasional stresses
1000 [ (iiMi)2 +(ioMo)2 ] 1/2 independently and to then add them
Sb = absolutely.
Z
6.3 COMPARISON OF CODES
In SI units.
Based on the above, we can
identify the difference in
approach between these two
codes
For Branch
6.3.1 Stress increase due to
i M   i
2

M 
occasional
2

loads are
S  i i 0 0

Z different in each code.


b

in USCS units
1000  i M 
6.3.2 Allowable
2

 i M 
stresses2 are
 1/ 2

S  i i
different for each code.
0 0

Z
b

in SI units6.3.3
ASME B 31.3 include Fax
where in the stress calculation due
to sustained load where
S Resultant bending
b
stressas
ASME B 31.1
Z effective section modulus
e
neglects all
= πr T forces 2
2

r mean branch cross - sectio


2
6.3.4 ASME B 31.3 calculation
T effective branch
s
wal
methods are undefined l thicn
for
lesser of T h sustained
andand (ioccasional
)( T b) i

T hThickness of
loads pipe
whereas ASME B
matchin
31.1 is explicit for the same.
or header exclusive of renf
T bThickness of pipe matchin
6.3.5 For calculation of stresses
For branch connection, the resultant
due to sustained loads
bending stress needs special care as
ASME B31.3 neglects
section modulus Z of header and
torsion where as ASME
branch is different.
B31.1 includes it.
6.3.6 ASME B31.1 intensifies
6.2.3 STRESS DUE TO torsion where as ASME B
OCCASIONAL LOADS 31.3 does not.
ASME B 31.3 do not specifically
define the equation for calculating
7.0 MEANS OF
the stresses due to occasional loads.
INCREASING
The code, under clause 302.3.6
FLEXIBILITY
only states that the sum of
longitudinal stresses due to
The pipe thickness has no
significant effect on bending E, y, L remaining the same, P = k I
stress due to thermal 3E y
expansion but it affects the where k =
end reactions in direct ratio. L3
So overstress cannot be
nullified by increasing the For 6”(150 mm) NB Sch. 40 pipe
thickness; on the contrary,
this makes the matter worse I = 1170 cm4
by increasing the end Z = 139 cm3
reactions. This is For 6” (150 mm) NB Sch. 80 pipe
demonstrated in the I = 1686 cm4
following example. Z = 200 cm3
Therefore,
Let us consider two simple
cantilever arrangements having
Sch. 40 Sch.
the same deflection, pipe size
80
and length but with varying
Load P 1170 k 1686
thickness.
k
Moment M 1170 k L 1686
kL
Stress = M/Z 8.4 k L 8.4 k
L

Form the above it can be seen that


although the stress remains the
same, the forces and moments
increase with the increase in
thickness of the pipe.

Where the piping system


encounters large end reactions or
detrimental over strain and it lacks
built in changes in the direction to
absorb the same, the Piping
For simple cantilever, the deflection
Engineer should consider adding
is given by the formula
flexibility by one or more of the
P L3 following means; addition of
y = bends, loops or offsets, swivel
joints, corrugated pipes, expansion
3EI joints or other means permitting
angular, rotational or axial
3EyI movements. Suitable anchors shall
Hence P = be provided to resist the end
L3 forces.
Ea = Modulus of elasticity at
8.0 COLD SPRING installation temperature.

Piping Engineer may also Em = Modulus of elasticity at design


provide cold cut or cold spring to temperature.
limit the value of stress and C = Coldspring factor from 0 for no
moment. coldspring to 1.0 for 100%
Cold spring is defined by coldspring.
the code ASME B 31.3 under
clause 319.2.4 as the intentional The factor 2/3 is based on
deformation of piping during experience, which shows that
assembly to produce a desired specified cold spring cannot be
initial displacement and stress. fully assured even with elaborate
precautions.
No credit for cold spring is
permitted in the stress range The value of reaction forces
calculation since the service life of or moments at the temperature at
a system is affected more by the
range of stress variation than by S h Ea
C1  1 
magnitude of stress at a given time. S E Em
ASME B 31.3 gives the which the piping is assembled is
formula for calculation of given by:
maximum reaction or moment Ra = CR or C1R which ever is
when cold spring is applied to a two greater
anchor piping system in clause
319.5.1 as below. Ra = Estimated instantaneous
reaction or moments at the

R m  R 1 2
3
c  EE m installation temperature.
a
SE = computed displacement stress
where range

R = Estimated instantaneous Sh = Maximum allowable stress at


maximum reaction force or design temperature
moment at maximum or minimum ASME B 31.1 deals with these
metal temperature. factors under the clause 119.9 and
R = Range of reaction force or 119.10.
moments derived from flexibility The reaction thus computed
analysis corresponding to the full shall not exceed the limits, which
displacement stress range and based the attached equipment can safely
on E. sustain.

9.0 SELECTED CHART 9.1 The book on Flexibility


SOLUTIONS analysis "Piping Stress
Calculations Simplified", by
S. W. Spielvogel deals with comprising of three legs, each of
a number of shapes which is at right angles to the other
frequently encountered in two.
practice. This can be used
for a quick check on the
terminal forces and
moments. This method
neglects the effect of
flexibility of the bends and
to that extent the values are
over estimated.

One of the shapes considered is the


"Three Dimensional 90 Turns"

AB = L1 = 25’ The book gives the following formula


BC = L2 = 10’
CD = L3 = 5’ The stresses are evaluated from the
L1/L3 = m = 25/5 = 5 equation,
L2/L3 = n = 10/5 = 2 S = KC D/L 32 lbs/ sq. inch
By referring to chart, we get :
The forces are evaluated from the
equation,
Kb = 8.61 Kt = 4.35
F = KC I/L 3 2 lbs
Moments are evaluated from the
Kx = 1.60 Ky = 0.09 Kz = 0.57equation,

Kxy = 1.40 Kxz = 1.50 Kyz = 0.40 M = KC I/L 3


ft/lbs

where C is the expansion factor


calculated from the expression
EMBED Equation.3
Expansion in inches/100'Ec I
C  Reaction F KC p

1728 100
L
x x 2
3

1.6  510  13.3


Expansion of C.S. @ 450 F 
= 3.16 inch / 100 ft. 5 5
 434 lbs.
Ec = Cold Modulus of Elasticity EMBED Equation.3
= 27.9 x 106 lbs/sq.in. I
Hence, Reaction F  K C p

L
y y 2

3.1627.910 6 3

C  510
1728100 0.09  510  13.3

Ip for 4" NB Sch 160 pipe = 13.3 in . 4
5 5
 24 lbs.
Allowable stress at installation temp EMBED Equation.3
Sc = 20,000 psi I
Allowable stress at design temp Reaction F  K C p

L
z z 2

Sh = 19,450 psi
3

0.57  510  13.3



Bending Stress = K b C
D 5 5
L3  155 lbs.
8.615104.5 EMBED Equation.3

5 Ip
= 3952 lbs./sq.inch Moment M xy  K xy C
L3
EMBED Equation.3
1.4  510  13.3
D 
Torsional Stress  K C 5
L
t

3
 1899ft.lbs.
4.35  510  4.5 EMBED Equation.3

5 Ip
 1997lbs./sq.inch Moment M xz  K xz C
L3
EMBED Equation.3 1.5  510  13.3

Expansion Stress Range 5
 SE Sb  4St 2 2
 2035ft.lbs.
 (3952)  (1997)
2 2

EMBED Equation.3
 5619 lbs./sq.inch Ip
Allowable stress range = SA Moment M yz  K yz C
= f ( 1.25 SC + 0.25 Sh)
L3
= 1 (1.25 x 20,000 + 0.25 x 0.4  510  13.3

19,450) 5
= 29,862 psi  543ft.lbs.
SE < SA.
9.2 TUBE TURN METHOD fe fs
f
The above method (Guided f1
Cantilever Analysis) does not give where,
allowance for the elbow flexibility. f = Expansion Factor allowing for the
e

Chart solutions incorporating temperatu re and material of pipe


the flexibility of elbows for certain given in table.
single plain configurations were fs = Shape factor allowing for the
produced by M/s. Tube Turns Inc. ratios of lengths.
Charts were developed which f1 = Factor for effective diameter length,
unfortunately limit the application to allowing for the excess virtual length
Z, L, U and symmetric expansion (EVL) of the elbow.
loops. Nothing is available for 3 These EVL are tabulated in the catalogue.
dimensional configurations.
In this method, the expansion stress is For example
calculated by the expression.

Pipe size = 4" std. wt. ASTM A106Gr.B


Operating Temp. = 450 F
From table
Effective Elbow Diameter Dr = 8.78 inch
Effective Elbow Length Lr = 4.7 feet
Length of Short Vertical Leg : h = 5'
Length of Long Vertical Leg : H = 25'
Length of horizontal offset : W = 5'
h/H = 5/25 = 0.20
H/W = 25/5 = 5.00

From Chart, shape factor fs = 1.66


Square corner length = h + H + W = l = 35.0'
Sum of elbow lengths = 2 Lr = 2 x 4.7 = 9.4'
Total effective length = l + 2 Lr = L = 44.4'
Effective Diameter length = L/Dr = f1 = 44.4/8.78 = 5.06
From Table 1 [This can also be calculated by the
Expansion factor fe for 450 F = formula Expansion inches/inch x
73,000 psi Young’s Modulus i.e
f e  Ec x  L problems. Some of the software
For C.S. Ec = 27.9 x 10 6 available are: -
Expansion in inches per 100 ft. for
1. ADL PIPE
C.S. = 3.16 from Appendix C. 2. AUTOPIPE
ASME B31.3 3. CAESAR II
Hence,
4. CAEPIPE
3.1627.910 6
fe  5.
73,470 psi PIPEPLUS
10012 6. TRIFLEX
Computed Stress Range 7. Q-FLEX
fe fs The pipeline geometry is fed into
SE  the system along with all the
f1 parameters such as design
730001.66 temperatures, pipe sizes; bend radii,
 23948psi type of branch connections,
5.06 locations of anchor points and
S A f 1.25Sc0.25Sh  restraints. This is termed as
ForCS to A106 Gr.B, `Modeling’ the problem. The model
can be generated by anybody who
S c 20,000 psi knows how to prepare the input.
S h 19,450 psi at 450F The programme executes the
solutions. First computer analysis
S A 11.2520,0000.2519450
was done in the year 1957.
29,862 psi
The analysis of the solutions is the
S E S A real engineering and is the job of a
Piping Engineer.
10.0 COMPUTER ANALYSIS
So far the consideration was 11.0 ANALYSIS OF REBOILER
given only for the calculation of CONNECTION
stresses in a pipeline with uniform
diameter and supported at two fixed Let us analyze the
points. This does not represent a requirements of the flexibility of
life size problem. In a real case, the connecting piping for the vertical
pipe routings will be connecting to thermosyphon reboiler, which is
various equipments, guides or most commonly used. The analysis
restraints will be provided in is restricted to the fixed tube sheet
between anchors and the pipes will type of heat exchanger for
be of different diameters and simplicity. Being fixed tube sheet
thicknesses. Further one section of type of heat exchanger, there will
the piping may be cold while the be a bellow type expansion joint
other part is subjected to high provided on the shell. This is to
temperatures. accommodate the differential
expansion between the reboiler
There is computer software
shell and the tubes. This has
available to handle such complex
nothing to do with the flexibility of
the piping connection.
Let us first look into the the differential expansion in vertical
functioning of the system before plane between the nozzle on the
analyzing the various arrangements column and those on the reboiler
of the reboiler supports. The will be negligible in the normal
heating fluid in the reboiler shell operating conditions.
provides latent heat to the liquid
inside the tubes, thereby vaporizing When the tubes and the
a part of it. The mixture of vapor column shell have different material
and liquid returns to the column via of construction, the piping leg ‘cd’
vapor return connection. This in Arrangement-1 has to
mixture will be at the same accommodate the differential
temperature as the liquid drawn expansion. For accommodating
from the bottom of the column. expansion in the horizontal plane,
When the shell of the column and the reboiler is supported on low
the tubes of the reboiler are of the friction slide plates, which allow
same material of construction, then the same to move.
the vertical expansion between the If an operating condition
nozzles on the column will be occurs in such a way that the
practically same as the expansion temperature of the reboiler and/or
between two corresponding nozzles the column varies, then the
on the reboiler differential expansion will have to
There can be different ways be accommodated by the flexibility
of attachment of the reboiler to the in the connecting piping. This
column. First we will consider the condition could occur if the heating
simplest system to analyze as of reboiler is started before
illustrated in the Arrangement-1. establishing the level in the column
or with the result of a leaky valve in
11.1 Arrangement-1 (See Fig. the heating circuit.
11.1)
As the top nozzle is rigidly
Here the reboiler is attached and normally will be of
supported on its shell from the higher sizes designed for two phase
bottom part of the column as near flow, the flexibility can be achieved
to the elevation of the vapor return only by the liquid inlet piping.
nozzle as possible. In this case the Hence the analysis has to be done
support lugs on the reboiler shell is for the upset condition as above and
arranged as near to the top tube not the normal operation.
sheet as possible. This minimizes
the differential expansion between
the centre line of the vapor return
nozzle on the column and the
corresponding reboiler nozzle i.e.
between ab and a’b’.
As indicated earlier, when
the tubes are of the same material
as that of the shell of the column,
between the reboiler outlet and the
column, as they are close coupled.
The only solution to this problem is
to support the reboiler on the spring
supports with the addition of low
friction slide plates to allow the
reboiler support to slide. This is as
illustrated in the Arrangement - 2.

Fig. 11.1
11.2 Arrangement-2 (See Fig.
11.2)
If it so happens that the
column shell will not permit the
support of reboiler on brackets
attached to the shell just below the
vapor return nozzle, then the
Fig. 11.2
brackets are to be attached to the
vessel skirt. This will result in the 11.3 Arrangement-3 (See Fig.
support bracket at a larger distance 11.3)
from the level of the vapor return
nozzle. When the reboiler is too heavy to
support from the column or the
In all reboiler analysis skirt as illustrated
problems, the centre line of the
vapor return nozzle is considered as above, an independent support has
the datum and the movements are to be provided for support of the
assessed first on the column side reboiler. In this case there will be a
and then on the reboiler side. considerable offset between the
Applying this to the present column and the reboiler centrelines.
problem, the expansion of the The movements of vapour return
reboiler shell between the vapor nozzle will be at fixed elevation.
return nozzle and the support The spring support is the only
bracket will be more as the shell is solution and will be as illustrated in
at a higher temperature and the Arrangement-3 (See Fig. 11.3)
tubes and column shell are at the
same temperature. This differential
will not be possible to be
accommodated in the connection
Fig. 11.3

xxxxxxxxxxxxxxxxxxxxxx

xxxxxxxxxxxxxxxxxxxxx

S-ar putea să vă placă și