Professional Documents
Culture Documents
Mechanical Design of Process Systems (Vol. 1) Piping and Pressure Vessels (PDFDrive)
Mechanical Design of Process Systems (Vol. 1) Piping and Pressure Vessels (PDFDrive)
Design
of Process Systems
Volumel
Piping and
Pressure Vessels
A.Keith Escoe
IY
Contents
vl
Foreword
The engineer who understands the impact of process proper respect in two excellent chapters on fluid me-
design decisions on mechanical design details is in a po- chanics and the engineering mechanics of piping.
sition to save his client or his company a lot of money. The chapter on heat transfer in vessels and piping il-
That is because the test of any process design is in how lustrates lucidly the interrelationship between process
cost-effectively it yields the desired product, and how and mechanical design. Every engineer working with in-
"cost" generally translates to "equipment": How much dustrial process systems will benefit from reading this
will the process require? How long will it last? How chaDter.
much energy will it consume per unit of product? Although the author has made a herculean effort in
In this two-volume work on Mechanical Design of covering the mechanical design of pressure vessels, heat
Process Systems,A. K. Escoe has performed a monu- exchangers, rotating equipment, and bins, silos and
mental service for mechanical design engineers and stacks, it is true that there are omissions. It is hoped that,
chemical process engineers alike. It is presented in such as the author hints in his preface, a future volume might
a manner that even the neophyte engineer can grasp its be added covering multiphase flow, specific cogenera-
full value. He has produced an in-depth review of the tion processes, turbines, and detailed piping dynamics.
way in which process design specifications are inter- Still, at this writing these two volumes comprise an
preted into precise equipment designs. Perhaps most outstanding practical reference for chemical and me-
valuable of all are the extensive worked examples chanical engineers and a detailed instructional manual
throughout the text, of actual designs that have been suc- for students.
cessfully executed in the field. I recommend these volumes highly for each design en-
The piping system is the central nervous system of a gineer's professional library.
fluid flow orocess. and the author has treated this with
John J. McKexa. Ph.D., P.E.
Joe C. Waher Professor of Chemical Engineering
UniversitY of Texas ' Austin
vii
Dedication
v|ll
d{ ry,'
heface to Volume I
This book's purpose is to show how to apply mechani- felt that this book is a valuable supplement to any stan-
cal engineering concepts to process system design. Pro- dard or code used.
cess systems are common to a wide variety of industries The book is slanted toward the practices of the ASME
including petrochemical processing, food and pharma- vessel and piping codes. In one area of vessel design the
ceutical manufacturing, power generation (including co- British Standard is favored because it nrovides excellent
generation), ship building, and even the aerospace indus- technical information on Zick rings. The book is written
try. The book is based on years of proven, successful to be useful regardless of which code or standard is used.
practice, and almost all of the examples described are The intent is not to be heavily prejudiced toward any
from process systems now in operation. standard, but to discuss the issue-engineering. If one
While practicality is probably its key asset, this first feels that a certain standard or code should be men-
volume contains a unique collection ofvaluable informa- tione.d, please keep in mind that there are others who
tion, such as velocity head data; comparison ofthe flexi- may be using different standards and it is impossible to
bility and stiffness methods of pipe stress analyses; anal- discuss all of them.
ysis of heat transfer through pipe supports and vessel The reader's academic level is assumed to be a bache-
skirts; a comprehensive method on the design of hori- lor of science degree in mechanical engineering, but en-
zontal vessel saddles as well as a method to determine gineers with bachelor of science degrees in civil, chemi-
when wear plates are required; detailed static and dy- cal, electrical, or other engineering disciplines should
namic methods of tower design considering wind gusts, have little difficulty with the book, provided, of course,
vortex-induced vibration and seismic analysis of towers; that they have received adequate academic training or
and a comparative synopsis of the various national wind experience.
cooes. Junior or senior undergraduate engineering students
Topics include.d in the text are considered to be those should find the book a useful introduction to the applica-
typically encountered in engineering practice. There- tion of mechanical engineering to process systems. Pro-
fore, because most mechanical systems involve single- fessors should find the book a helpful reference (and a
phase flow, two-phase flow is not covered. Because of source for potential exam problems), as well as a practi-
its ubiquitous coverage in the literature, flange design is cal textbook for junior-, senior-, or graduateJevel
also excluded in this presentation. Since all of the major courses in the mechanical, civil, or chemical engineering
pressure vessel codes thoroughly discuss and illustrate fields. The book can also be used to supplement an intro-
the phenomenon of external pressure, this subject is only ductory level textbook.
mentioned briefly. The French philosopher Voltaire once said, "Common
This book is not intended to be a substitute or a re- sense is not very common," and unfortunately, this is
placement of any accepted code or standard. The reader sometimes the case in engineering. Common sense is of-
is strongly encouraged to consult and be knowledgeable ten the by-product of experience, and while both are es-
of any accepted standard or code that may govern. It is sential to sound engineering practice, neither can be
ix
--*
learned from books alone. It is one ofthis book's eoats to and words of encouragement. I also wish to thank other
unite these three elements of "book learning," c6mmon engineering faculty members at the University of Texas
sense, and experience to give the novice a better grasp of at Austin for their comments. I must exDress thanks to
engineering principles and procedures, and serve as a Larry D. Briggs for reviewing some ialculations in
practical design reference for the veteran engineer. Chapter 4; and last, but certainly not least, I wish to ex-
Finally, I wish to thank Dr. John J. McKetta, professor press gratitude to William J. Lowe and Timothy W. Calk
of chemical engineering at the University of Texas at of Gulf Publishing Company, whose hard work and pa-
Austin, who had many helpful comments, suggestions, tience made this book oossible.
The study of fluid energy in piping systems is a com- Pr - Pz = V,t=- vrt + (y. _yr;€1p (r-2)
prehensive subject that could in itself fill countless vol- p 28" c"
umes. This chapter is primarily concerned witl fluid en-
ergy dissipated as friction resulting in a head loss. where subscripts I and 2 refer to flow upstream (after
Although this topic is popularly known in industry as the flow process) and downstream (before the flow pro-
"hydraulics," the term "piping fluid mechanics" is used cess), respectively, and
here to avoid confusion.
Pt - Pz : change in pressure head
p
BASIC EOUATIONS
Vt^- V' : change in velocity head (kinetic energy)
29"
The basic equation of fluid mechanics, originally de-
rived by Daniel Bernoulli in 1738, evolved from the :dz
principle of conservation of energy:
(Yr - Yr) I = change in static head (potential energy)
,]V r ,{E -r- ,llr. + ,1ll- (l-l)
p 29" g" ^
cm (kg)
F : friction 1o* in !JlQ,
where density, lb./ft3 or g./cm3
P: pressure, lb/ft2 or kg/cm2
velocity, ftlsec or cm/sec
The following are expressions of the Bernoulli equa-
8": conversion constant, 32. 17 (ft-lb./sec2lbr)
: tion when applied to various incompressible and com-
gravitational acceleration 32.2 fllsecz,
cm/sec2; g/9" : 1.0
pressible flow conditions:
height above datum, ft, cm
dY: differential between height above datum and Incompressible flow-
reference point, ft, cm
p, - P. v,2
F: head loss, friction loss, or frictional pressure - v.2
drop, ft-lbr/Ib., cm-kg6/g. P zE" gc
He: energy added by mechanical devices, e.g.
pumps, ft-lb/Ib., cm-kg/g.
HE: energy extracted by mechanical devices, e.g. Compre s sib le -i s othermal f low -
turbines, ftlb6/1b., cm-kg/g.
: _[*l + (zz - z,
Rewriting Equation l-1 along a fluid streamline between
points 1 and 2 with steady, incompressible flow and no
FJn H X[ [*l]
mechanical energy added or extracted results in +F+HA+HE
2 Mechanical Design of Process Systems
Compre s s ib le -adiabati c f low Forcing a fluid through a pipe component requires en-
- ergy. This energy is expended by shear forces that de-
H [1 [' - (,*J'.-"'] : xl' -FJ^ [*J] velop between the pipe wall and the fluid, and to a lesser
extent among the fluid elements themselves. These shear
forces are opposed to fluid flow and require excess en-
+(22-z)+F ergy to overcome. Figure 1-l shows a simple version of
this phenomenon and illustrates how shear stresses in-
+HA+HE crease in the radial direction away from the pipe center
line and are maximum within the boundary layer next to
. /o \* /p\ the wall. Friction energy loss is a resuit of these shear
where l- | : l:l : general gas law stresses next to the pipe wall. Excess loss in energy oc-
\Prl \rrl
curs because of local turbulence and changes in the di-
k : .specific heat ratio (adiabatic coefficient), rection and speed of flow. As a fluid changes direction,
t- lt-
energy is expended because of a change in momentum.
Cp : sPecific heat at constant pressure,
The methods used to determine energy loss caused by
Btu/lb.-'F
C, : specific heat at constant volume, Btu/lb--"F
wall friction are essentially the same, where the pipe
component is treated as a straight piece of pipe. How-
Equation 1-2 is the analytical expression that states a ever, the methods used to determine energy loss caused
pressure loss is caused by a change in velocity head, by change in momentum differ, and a couple are de-
static head, and ftiction head. The most cofirmon units scribed as follows.
are "feet of head." lb. and lbr do not cancel out and the
expression is exactly "energy (ft-lb) per pound of Equivalent Length
mass."
In most industrial fluid problems, Equation 1-2 is In this approach to determining energy loss caused by
cumbersome to use, because the friction loss is the pa- a change in fluid momentum, a piping component is ex-
rameter most often desired. The friction loss is the work tended a theoretical length that would yield the same en-
done by the fluid in overcoming viscous resistance. This ergy loss as the actual component. This length is called
friction loss can only rarely be analytically derived and is the "equivalent length" because it is that length required
determined by empirical data developed through experi- to obtain the same amount of friction pressure drop as
mental testins . the piping component alone. The major problem with
dv
oy
x+c
Figure 1-1. Shear stresses in fully developed flow. Shown here are imaginary fluid elements "slipping" over one another.
Piping Fluid Mechanics 3
this method is that the equivalent length for a pipe com- into the piping system, the factor F in Equation 1-2 be-
ponent varies with the Reynolds number, roughness, comes the desired parameter. This friction loss is the
size, and geometry of the pipe. All these par.rmeters work done by the fluid in overcoming viscous resistance
must be analyzed in using this method. and loss attributed to turbulence. The parameter F is
composed of two components, pipe wall friction and
Velocity Head llethod losses for the various pipe fittings, pipe entrances, pipe
exits, and fluid obstructions that contribute to a loss in
Since the excess head loss is mostly attributed to fluid fluid energy. These latter losses are described in terms of
turbulence, the velocity head method is widely accepted velocity heads, K;. In solving for F in Equation 1-2, we
and is replacing the equivalent length method in fluid first obtain pressure loss attributed to pipe wall friction,
calculations. Throughout this book, the velocity head ap- represented by
proach will be used.
The velocity head is the amount of kinetic energy in a
fluid, Y2l2g". This quantity may be represented by the -AP.' =.: eyll]
2e. \d/
(1-3)
thors use 4f1 in place of f. If such factors are used, par- factor data can be obtained and better understood
ticular attention should be paid to the specific friction through use of new methods for measuring roughness.
factor chart used. Figure 1-3 is broken into three flow regimes-
The friction factor f is dependent upon the dimension- laminar, transition from laminar to turbulent, and turbu-
less term expressing the roughness of the pipe (E/D, lent. The Reynolds numbers establishing these zones are
where E is the depth of the pipe) and the dimensionless 2,100 for laminar, 2,100 to 3,000 for transition zone,
Reynolds number Nr" : dpV/M, where l1, is the absolute and 3,000 or more for turbulent
viscosity of the fluid, inJb1-sec/ftz. The Reynolds num- The basis for Figure 1-3 is the classic Colebrook equa-
ber is the single most important parameter in fluid me- tron
chanics because it establishes flow regimes and dynamic
sirnilarity. The relationship between the friction factor f, | : -.^.to8ro Idd 2.51 I
(l -6a)
the pipe roughness, and the Reynolds number is shown r1r, [- " **,rpi
in the classic relationship given by Moody in Figure 1-3.
Figure l-3 may be presented in a more convenient for (3,000 to 4,000) < NR" < 108
form as shown in Figure 1-4, where the relative rough- For laminar flow the friction factor is determined by the
ness of the pipe is based on a single value of roughness. simple expression
This value of roughness must be an average value esti-
mated to simplii/ the problem. The figures presented "64 (1-6b)
herein are the best available until more reliable friction Nn.
.09
.08
.07 .05
.04
.06
.01
.o?
.0t5
.04
^
: .01
.008
.006
.03 a
=-
oo4 :
003 :
002
.0015 :
^^,
-0008 -'
.0006
.01
.009
.008
? 3 4 56 I 2 3 4 56 Blo5 2 3 4 56 € to7 \2
z J 4 56 8 rot -q-s9l
r, -If* , o i' n., ,' ir *4r =
R?ynotds Nunber n"
= f
'. ff
Figure 1-3. Moody friction factors. (Repdnted from Pipe Friction Manual, @ 1954 by Hydraulic Institute. Data from L. F.
Moody, Frioion Faaors for Pipe Flow, permission of ASME.)
#( -8u
,=
Equation 1-6a, which describes the friction factor for Dimensional forms of Equation 1-4 are presented in
turbulent flow in pipe of any roughness, is a simple addi Table 1-1 [1], where the equation is conveniently shown
tion of the Prandtl solution for smooth pipe and the von in various units that are used to solve fluid pressure loss
Karman solution for rough pipe. The relationship holds problems.
for the transition between rough and smooth pipe.
To solve Equation 1-6a for the friction factor f an itera-
tive analysis is required because the function is nonho- NON.NEWTONIAN FLUIDS
mogeneous and inseparable. There are several empirical
relations of f expressed as an independent separate func- The Colebrook equation holds for fluids whose flow
tion of f G/d, NR.), but with today's micro-computers properties are dependent on the fluid viscosity. These
Equation l-6 can be solved more accurately and expe- fluids consist of all gases, liquids, and solutions of low
diently with iteration. molecular weieht and are known as Newonian fluids. In
-rll
Mechanical Design of Process Systems
Table 1-1
Dimensional Forms Used With Equation 1-4
[11
Plessure Ol?p, ne r, ana
= rate -APr or pHr L
Row
g IrNr">2,ooo'r:[2 loglqQ27etD+(?/NR"o)],'
w {*-r^,.i
\ rr 'l *
pt)- +
pD
nvz
pD2 e
'l uoQ' cQP dQ pv'
/!& * ",. D4 pD
"- \ D ' -L\J D2 .
Units and constants
Conventlonal units Metric units
Newtonian fluids the viscosity alone defines the rheolos_ Time-independent fluids that are purely viscous are
_
ical behavior. classified as. pseudoplastic, dilatant, Bingham, and yield-
Non-Newtonian fluids are those in which the viscosity pseudoplastic fluids. ln pseudoplastic
alone does not define their rheological behavior. Sucir fluids an intinites-
imal shear stress will initiate motion and the ratio of
fluids are solutions composed of solid particles that ex_ shear stress with velocity decreases with increasins ve_
pand. Clay and very dense slurries are examples of non_ locity gradient. This type of fluid is encountered in iolu_
)iewronian fluids. The flow properties of suih fluids are tions or- suspensions of fine particles that form loosely
a function of the particle characteristics, e.g., size and bounded aggregates that can break down or reform witL
flexibility and thermal expansion. an increase or decrease in shear rate. Such solutions are
Purely viscous non-Newtonian fluids are classified aqueous dispersions of polyvinyl acetate and of an acrv_
into dree categories: time-dependent and time-indepen_ clic copolymer: aqueous solutions of sodium carboxy_
dent and viscoelastic. A time-dependent fluid displays methyl cellulose, and of ammonium polymethacrylatl;
slo*' changes in rheological properties, such as thixbtr-o_ and an aqueous suspension of limestone.
pic fluids that exhibit reversible structural chanses. Sev- In dilatant fluids an infinitesimal shear stress will start
eral ty,pes ofcrude oil fit inro this category. Anoiher rype motion and the ratio of shear stress to velocity increases
of tinre{ependent non-Newtonian fiuid is rheooectic as the velocity is increased. A dilatant fluid ij character-
fluids- Under constant sustained shear, these fluidi. rate ized by an increase in volume of a fixed amount of dis-
of srrucrural deformation exceeds the rate of structural persion, such as wet sand, when subiected to a deforma_
decav. One such category of fluids is polvester. tion that alters the interparticli distances oI its
Rheqectic fluids are less common than thixotrooic flu- constituents from their minimum-size confisuration.
ids. Such fluids are titanium dioxide particles in waier or su-
':bl&,,
crose solution. Dilatant fluids are much rarer than ( 100,000 the following empirical relations can be used
pseudoplastic fluids. for determinins the friction factor:
ln Bingham fluids a finite shearing stress is required to
initiate motion and there is a linear relationship between
the shearing stress-after motion impends-and the ve- (Ni") b"
locity gradient. Such fluids include thickened hydrocar-
bon greases, certain asphalts, water suspensions of clay, where bn : 0.0019498 (n)-45"
fly ash, finely divided minerals, quartz, sewage sludge, n= (7.8958 x l0-7) (a") 182.1321
and point systems.
Yield-pseudoplastic fluids are similar to Bingham flu- Typical values for "y and n are given in Table 1-2 [3].
ids, but the relationship between the excess shearing Values for "y and n not available in literature must be de-
stress after motion impends and velocity gradient is non- termined by viscosimeter measurements.
linear. Fluids in this category are defined by their rheo- Figure 1-5 shows the rheological classification of non-
grams, where relationships between the shear stress and Newtonian mixtures that behave as single-phase flow.
rate of shear exhibit a geometric convexity to the shear The reader is urged to refer to Govier [4] for further in-
stress axis. Such fluids are many clay-water and similar formation on non-Newtonian fluid or other complex
suspensions and aqueous solutions of carboxypoly- mixtures. Usually, the mechanical design of process sys-
methylene (carbopol). tems does not involve non-Newtonian fluids, but knowl-
Viscoelastic fluids make up the last category of non- edge of them and their peculiarities is a must if the need
Newtonian fluids. The term "viscoelastic fluid" is ap- anses.
plied to the most general of fluids-those that exhibit the
characteristic of partial elastic recovery of the fluid
structure. Whenever a viscoelastic fluid is subiected to a
rapid change in deformation, elastic recoil oi stress re-
laxation occurs. Many solutions exhibit viscoelastic
properties under appropriate conditions-molten poly-
mers, which are highly elastic; and solutions of long-
charged molecules, such as polyethylene oxide and poly-
MULTI.PHASE
acrylamides. Processes such as coagulation, oil-well SINGLE PHASE
fracturing, and high-capacity pipelines rely on polymeric
additives to cause pressure drops. Viscoelastic fluids ex-
hibit the "Weissenberg effect," which is caused by nor-
TRUE HOMOGENEOUS
mal stresses and produces unusual phenomena, such as
the tendency of the fluid to climb up a shaft rotating in
the fluid.
For any time-independent non-Newtonian fluid, Met-
zer and Reed [2] have developed the following general-
ized Reynolds number fraction:
=
_ D' U2-np
N*" (1-7)
"l PLASTIC C OILAIAI.II 9
where D : pipe ID, ft
U : average bulk velocity, ftlsec
p : density, lb,/fC I
^l : generalized viscosity coefficient, lb./ft
sec :gc c 8n-r (see Table 1-1)
: experimentally determined flow constant, //g"
for a Newtonian fluid
n: empirical constant that is a function of
non-Newtonian behavior (flow behavior
index), 1.0 for Newtonian fluids
For n: 1.0 and C : p/g", Equation 1-7 reduces to Figure 1-5. Rheological classification of complex mixtures
Np" = Du p/p for Newtonian fluids. For 2,100 < NR" that behave as single phase fluids [4].
Mechanical Design of Proces: Slstems
Tabte 1-2
Rheological Constants tor Some Typical Non-Newtonian Fluids* 13I
Rheological Constants Rheological Constanis
ol Fluid n of Fluid n
23.3% Illinois yellow clay in water 0.229 0.863 18.6% solids, Mississippi clay in
VELOCITY HEADS tered by the flow. In a screwed elbow there are abrupt
changes in the wall causing local turbulence and henie
Returning to Equation 1-4, let's focus on the term increased velocity head loss.
EKi. This term represents the excess velocity heads lost Analytical determination of velocity heads can only be
in fluid motion due to fluid turbulence caused by local accomplished in a few simplified cases. The values for
turbulence at the pipe wall and change in flow direction. velocity heads must be determined and verified empiri-
The latter is the greatest contributor to the DKi term. cally. Comprehensive listings of such velocity head (K)
When a fluid strikes a surface and chanses flow direc- values are given in Figures 1-7 t5l, 1-8 [5], 1-9 [6], and
tion. it loses momentum and. therelore. Jnergy. Consid- 1-10. Using these values in Equation 1-4, you can ana-
ering the 90' elbow in Figure l-6, we see that the fluid lyze most cases of friction pressure drop for pipe under
changes direction from the x to the y direction and im- 24 inches in diameter. For pipe with diameter greater
parts reactions Fx and Fy, each a function of the pressure than 24 inches, an additional analysis must be made in
and velocity of the fluid. End conditions of the elbow de- solving for the velocity head term. This method, pre-
termine some of the velocity head loss, that is, where the sented by Hooper [7] is called the "two-K method."
fitting is a "smooth elbow" or a "screwed elbow." A
smooth elbow is one that is either flansed or welded to TWO.K METHOD
the pipe such that a smooth internal srirface is encoun-
As explained previously, the value of K does not de-
pend on the roughness of the fitting or the fitting size,
but rather on the Reynolds number and the geometry of
the fitting. The published data for single K values apply
to fully-developed turbulent flow and K is independent
of N*. when N^. is well into the turbulent zone. As Nq.
approaches 1,000, the value of K increases. When Na"
< 1,000, the value ofK becomes inversely proportional
to NR". In large diameter pipe ( > 24 in.) the value of NRe
must be carefully considered if values of 1,000 or less
are encountered. The two-K method accounts for this de-
pendency in the following equatron:
Nominol Size t/^n 3/q" I Y4" 1Y2" 2V2.3" 8.10" 12-16" t8-24"
Friclion .o27 .o25 .o23 .o22 .021 .0r9 .018 .o17 .01 6 .01 5 .014 .01 3 .ot2
Focfor ( fr.)
o.s(in9(, - o'r K.
: Kr= - Formula z , Formula
rz=
t/\2/^1
Ba tr
O
4
lA
,, K,- o r !sin i(r - trt + (r - E )2
tJ'|
Kz= "iG-p)\f"# a4
Kr
lf
k.
Kr= j.n - 0(Formulaz 'Formula+) uhen d = r8o"
/.\
6(sin+(I - P)'?
2
K,
K, _ ,__]____184 = E K,+O [o : (, - g') + (t - 9')']
Kz=
|, a2\2 Kr
R4
^ d,r
lJ \2
Kr=SO *Formr.rla I + Formula l az_\d,J
12\ -
1I
_au
"
Subscript I dennes dimensions
and coefncients with reference to
Kr +sin3[o.a 0 - P\ +2.6 (t - 02)2) the smaller diameter.
K"= Subscript 2 refers to the larger
E;l
Et#
JLI I.-
-+ -ffa-r-
TL€
r-L-r
fNr,
FI-/f F
- K: rcof7 K:sof,
lf . ts = r,0 = o. ........... . K' : 8 -/r Minimum pipe velocity Minimum pipe velocity
P< r and 0 < 45o ........K2: Formula 5
(fps) for full disc lift (fps) for full disc lift
B< r and 45"<0< r8oo...Kz - Formula 6 =)5vv -a8!V
r
E
If: B: r...Kr=6oo/z
9. r.. .K, = Irormula 7
Minimum pipc r tlocitr itp.; ior full .lisc Iifr
If: B:r...Kt=l+ofr : F p2 \,/ v
lf: r...K,:
tr
9= ;s fr
B< r. . . K, = Irormula 7
lf A-t.. Kr=sjfr Minimum pipe velocicy (fps) for fr-rll disc lift
: t4o B|V V
l--4-lV
l'- I z++ll l-
F Ftr-IF
L
li:il- -:::i::
8t'-n*"
40
+.[
< 45. ora ( 22.50
lf d
l 830
:E
.4
llo
20
. td,-
d=arcslnl- d,ll=- 0
\21 /2
Figure 1-8. Calculated Crane K-values for concentric conical Figure 1-9. Correction factor.
diffusers are tabulated in Table l-6.
T./1
Dl ( ,l
_l_ \,-,/
TWO.MITERED ELBOW
1.0
0
Figure 1-10A. Velocity heads for change of flow [6].
'*
12 Mechanical Design of Process Systems
ot
1Eo
n =number miters
or segments
1
mitered ell
644+
I
smooth ell
.5 t.o 1.5
R/o
2<.- <*3
-V-
+s"(o<go.
an az/og
=
Figure l-10C. Velocity heads for change of flow [6].
'4
Mechanical Design of Process Systems
on= 9/og
Figure 1-10D. Velocity heads for change of flow [6].
Piping Fluid Mechanics 15
\sri
\7 tAl
| |I Ai= A2: A3
aR = or
/ag
Figure 1-10E. Velocity heads for change of flow [6].
16 Mechanical Design of Process Systems
on or/ag
=
OR: O1/o3
'*
18 Mechanical Desisn of Process Svstems
o*= 02/o3
2-----> -_)> 3
--v'.
.
---)t -->2
OR= 01/O3
' 'drF'
Mechanical Design of Process Systems
-llt-
2--+ ----)3
Kzg
on= oz/o,
Figure 1-10J. Velocity heads for change of flow [6].
it - 'I[||,,'
Small pipe fittings have more surface roughness and Table |-3
abrupt changes in cross sections, making Kl insignificant Constants for the Two-K Method I7l
at values of Nr" ) 10,000. For this reason, the new
Crane method is recommended for pipe diameters 24 in. Filting Type Kl K-
and less. Comparison of the methods for elbows is de- Standard (R/D : l),
screwed 800 0.40
picted in Figure 1-11. Table 1-3 lists Kr and K- values. :
Standard (R/D 1),
The two-K method is preferred over the equivalent flanged/welded 800 0.25
length method because in large, multi-alloy sysiems the Long-radius
equivalent length method could predict losses 300% too (R/D : 1.5), all
high, resulting in oversized pumps and equipment. With typqs 800 0.20
laminar flow, the equivalent length method predicts head 90" 1-Weld
losses too low. Also, in the equivalent length method, (90") I,000 1.15
Mitered 2-Weld
every equivalent length has a specific friction factor as- elbows (45') 800 0.35
sociated with it, because the equivalent leneths are de- (R/D : l 5) 3-Weld
rived from the expression L. = K D/t. (30') 800 0.30
The Hydraulic Institute's widely used K-factors are 4-Weld
good for l-in. to 8-in. pipe, but result in errors in larger (22t/2") 800 o.27
5-Weld
piping. (18") 800 0.25
Elbows
The disadvantage of the two-K method is it is limited
Standard (R/D : l),
to the number of values of K1 and K- available, shown all types 500 0.20
in Thble 1-3. For other fittings, approximations must be Long-radius
made from data in Table 1-3. (R/D : 1.5), all types 500 0.15
Mitered, 1 weld, 45" 500 0.25
Mitered, 2 weld, 22rlz" 500 0. 15
Standard (R/D : 1),
screwed 1,000 0.60
180' Standard (R/D : l),
flanged/welded 1,000 0.3s
PIPE FLOW GEOMETRIES Long radius
(R/D : 1.s), all
In using Equation 1-4 the geometry of the flow area
types 1,000 0.30
must be considered if the area is noncircular. In calculat- Standard, screwed 500 0.70
ing the Reynolds number and the diameter for a noncir-
Used Long-radius,screwed 800 0.40
as Standard, flansed or
cular cross section, the hydraulic radius is applied: etbow welded 800 0.80
Tees
Slub-in-type branch 1,000 1.00
R11 : hydraulic radius : cross-sectional flow area Run- Screwed 200 0.10
wetted perimeter through Flanged or welded 150 0.50
tee Stub-in-type branch 100 0.00
This relationship applies to noncircular cross sections Gate, Full line size,
flowing full or partially full, oval, rectangulat etc., but
p: 1.0 300 0.10
ball, Reduced trim,
not to extremely narrow shapes, such as annular or elon- p:0.e 500 0.15
gated openings, where the width is small relative to the plug Reduced trim,
length. In such cases the value of Rs is approximately p=0.8 1.000 0.25
one-half the width of the passage. Globe, standard 1,500 4.00
The value of 4RH is substituted for d in Equation 1-4. Globe, angle or Y-type 1,000 2.00
Diaphragm, dam type 1,000 2.00
Butterfly 800
**" : r,1*, Lift 2,000
0.25
10.00
Check Swing 1,500 1.50
Tilting-disk 1,000 0.50
Thble l-4 provides hydraulic radii for various cross :
Note: Use R/D = 1.5 values for R/D 5 pipe bends, 45' ro 180'.
sections. Use appropriate tee values for flow through crosses.
< - -8l
D K1
rfi-I|l
0.8
10
6
0.3
H
r\
Long
radius
K
I 10
T
Branch
}J flow
Flanged
0.3 1 4
D
FlangEd tee
K o.2
0.1
0.3 0.6 I
D
Long
radius
o.4
t\\_t Branch
flow 1
20 p!"..ffi
Contraction
o'?FidG,i\
0
*Hl--
p 0.6
X 0.4
K=o.b 0.3
ffi'=.err-r,rf(fia-r] 0.3 0.5
Inward ptojocting pipe
Enlargem€nt
Figure 1-11. Velocity head values for common piping components [1]. (Reprinted by special permission from Chemical Engi-
neering, @ 1978, by McGraw-Hill, Inc., New York.)
'd
24 Mechanical Design of Process Systems
E F nninqfricrlon lactor for fllribl. m6.tl hc. Il4l
v
Swing chock valve
n tLtLtt- _
Screwed
4 flilj-ll-lH;E
Tvoictl dimh'ioo3
0.
9 o.12
.T Nomimr LDr LE 0.08
tize {in.l {in.l u tL
1/2 0.520 0.250
,I 0.6 '1 24
314 0.750 0.275 0.04t04
r t.000 0.187 106
r1n 1.500 0.200 Feynolds number, /VF"
2 2.O@ 0,161
3 2.962 0.143
K
4 4.000 0.081
Head loss in conical diffuse6 1.2
Flanged
4 6 10 tr9l
1.0
5 o.e
Angle valve
Scre$red
6l-
-
4l-
Ir
,t- P, des.
,I llIl
0.3 0.6 1
147l
63
Flanged 1
K= 1.2 (1 -cos0
5" o.24
0.05
to' 0.29 to' o.52
20" 1.54
2o' r.56
+0" 10.8
lo" 11.3
I18.0
60" 206.0 6o'
IN/
E+=
0.153 Di Compressibility effects of a fluid are small when the
fluid velocity is small, compared to the fluid sonic ve-
locity. If V1/al is equal to 0.3, the error in the velocity is
less than I % when using the incompressible assumption.
For ambient air, this limitation corresponds to a velocity
of 300 ftlsec without causing significant error.
The phenomenon ofnonsteady flow is somewhat more
complex than that of steady flow. The acceleration or de-
celeration of liquid particles immersed in a two-phase
solution is one such example. The time required for the
nonsteady phenomenon to occur is compared to the time
reouired for a sound wave to traverse the flow in which
substantial differences in velocity occur. If the time dif-
ferences are small, then the incompressible Bernoulli
equation (Equation 1-2) may be applied.
Table 1-5 Thus, ifone has a fluid such as a fuel oil (see Example
Reasonable Velocities 6-1), which for a given temperature has an absolute vis-
Liquids Feet per Second cosity of 139.53 centipoise, we calculate the kinematic
Service water mains 2to5 viscosity, z, in the English system of units as follows:
General service water piping 4 to l0
Boiler feed water piping 6 to 13 p= 139.53 cp at 90'F
Heat medium oils 2to6 w : 54.725lb^/tr
Lubricating oils 2to6 lb,-rT, ij!-.
Gases (rry )r)cP /o.oooozol\
|1 r ., /= :2. iz,
lbrsecr
Low-pressure steam heating
and process piping 15 to 70
Low-pressure steam mains 70 to 165 (s4.72s) E
fP
High-pressure steam mains 165 to 400
Steam engine and pump piping 100 to 150
Steam turbine piping 150 to 330 r: 0.0017 :ft2
sec
Natural gas 100
Air, 0 to 30 psig 70 ft2
Forced draft ducts 40 to 60 0.0017 -::-
Induced-draft flues 30 to 50 centistoke : 159.261 centistokes
Chimneys and stacks
Ventilating ducts
35
7 0.00001076e il
sec
SSF to centistokes with COz in steam. The properties of the mixture are as
follows:
:: Seconds Saybolt Furol
p= 0.01322 cp
lR4 p= 0.085 lb/ft3 P : 10 PSig
Rrr 25 ( t( 39, 2.24r - -:- : v e: 0.0015 in.
t
60
5rt>40, 2.16t - -:--
t
= v The velocity head approach is used in determining the
friction pressure drop. The line shown in Figure 1-12 is
coming off two nozzles on top of the reboiler and merg-
ing at a tee before entering the amine still tower. The di-
EXAMPLE l.l: FRIGTION PRESSURE DROP mensions shown are identical with both sides coming off
FOR A HYDROCARBON GAS.STEAM the reboiler nozzles (exemplified by the word "TYR"
MIXTURE IN A PIPE meaning for both sides). To solve this problem, we must
apply Equation l-4. To use this equation, we divide the
An amine still reboiler boils off a hydrocarbon gas- connecting pipe into three components (see Figures
{eam mixture that flows in a 3M ss line connecting the 1-13-l-15)-an 18-in. f portion with W = 25,291 lbl
:eboiler with an amine still tower. It is desired to deter- hr a24-in. d portion with W : 25,291lblhr; and a24-
rine the maximum pressure drop in the line as shown in in. d with W : 50,582 lb/hr, Equation 1-4 is applied to
Fieure 1-12. The sas is a small tract of amine immersed each portion and the pressure drop for each is added to
d
Mechanical Design of Process Systems
SERVICE ov
iE;G'dr;BrE-vEEcrw REYNoLDS
E No= t
Ki= No oF ver-ocrri ems; x=.ov/,
FoR sERvtcE = DEpENDENT ptpE FRtcrtoN = rt6
K VALUES ILD = 5OK, D=rNStDE D|AMFTEFtfrll lL= o,ot32? cp
FOR COMPONENTS: p
y== o,O85 Lb/cu tl
1, I L'l I'
)*, € = O,OOI5 in
PIPE ENTRANCE . o,7 80
?1.x ta" u FFUSER (cs/{rRrc). o,079 f=
t6" t R 90' ELL = o, oza D: 17' 50 in.
..r-
- '
raszgr\E(#c'"J
("
- 21,++2 *
i,,''
""n #n"cr.z\ Gfi],,)
flow-gas or liquid.
Figure'l-14. Fluid analysis for single phase-gpm
J
Mechanical Design of Process Systems
lt Z''tt" = 2,9l.1 k
Sh
= <2."
"4q
rs'srz'tiF (=za!=ai
= r7,s84 *
fo,oa,iS h*l\ru.(ffi
--gpm
lr1:50)r, 1+r.+tzr a
secro.oasl k
Sch
=DVP: \ tz I rr- 80
3lc x
3lt 0.742 0.423 1.500 6.104 0.126
p tlz 0.742 0.546 1.500 3.746 0.036
(8.884 x 10-6) ;lb'
n-sec Sch
80 4s 0.957 0.423 2.000 7.672 0.225
: 690,491.450 1X 1lz 0.957 0.546 2.000 5.898 Q.121
3lc 0.957 0.'742 2.000 3.081 0.022
,fl-05 - _2 ron,^ [{g
"- * _?r_] ( 1-6a) Sch
[3'7 NR"(f)"'l 80 tlz 1.278 0.546 2.000 10.545 0.318
Let f: 0.014 I r/4 x r.278 0.742 2.000 7.701 0.153
1.278 0.957 2.000 4.603 0.040
'0.014)-05
: 8.452 : -2 logr0 [(2.317 x 10 5)
Sch
+ (3.072 x :
10-5)l -8.537 80 1.500 0.546 2.500 10.999 0.373
f- 0.014 3lq 1.500 0.742 2.500 8.720 0.225
1tl2 x
1.500 0.957 2.5W 6.235 0.099
lPf = (?. r'.l# (l -4)
Sch
lUq 1.500 1.278 2.500 2.545 0.009
fr lh
H*) 40
Itlq
2.469 0.957
2.469 1.278
3.500 12.474
3.500 9.796
0.406
0.237
ztlz x
S€C'lD6
lt/z 2.469 1.500 3.500 7.957 0.143
2.469 2.067 3.500 3.292 0.013
lP1, : 9-929 O.t Sch
40 3.068 0.957
1
fr lh
S€C'lD1
c--,-..,&
32 Mechanical Design of Process Systems
't:
(0.013x2.917)
rP.- I + 2.201
(23.s0)
(o.oss) (s4.884F
H
-i1Fl",ooJ
r,.r r' ft lb.
-\J-.-/ --7-a-
S€C'lD1 Figure 1-16. Process surge tank. (Courtesy of Tranter, Inc.)
34 Mechanical Design of Process Systems
oil system in Figures l-17, l-18, and 1-19. It is desired Bottom Head Hot Oil Supply
to determine how much frictional pressure drop will be
incurred for the entire tank so that pump sizes may be I Hot Oil Entrance from the 2-in. Header and Flow
selected. Through Station 1. (Figure 1-U):
The tank is divided into two systems-the hot oil sup-
ply system and the hot oil return system. Each system
connects to the three components-the four internal coils
inside the tank, the outside shell jacket coils, and the Yt,, z.s'
jacket coils connected to the bottom head-and each of 6 gpm ,l/ 30 spm
the three components must be analyzed separately.
"1,
t
Bottom llead System I
g=11h"x31a"
C=11h"x1,
D=2"xEa"
E=2" x1"
F=2'x1112"
___.1
rl
tl t^
JI r3-
q- 0l Il-
q.\
{t
{t-\| \t I
7sm \l
|\l
NI
t\l
| \l
t\
tt\
_l I
-@
n L: 4.0 ft
osD* { I rc \ltgi
_DVp_ ?8\o"*" .io\
'--'
sar
rnin \7.479.9aU \60 sec/
Nn"
P
"I x lo )
v_ : 2.869 ft/sec
(t.oo8 -tD'
n-sec
ilrt I
(J.JJbtrn.'t-......_l
\144 in.'/l
:3,452,9\0
l{q!q'!) _\
12
67lr,,z.sost
/
a,sa.;, I
I 2.51 I
DVp 12 sec
(1o s -z roe,o
ffi +
a,qsz,srol(D\ (1.008 x 10 4)
--l!'
n-sec
K-factors [/o.oora\ I
1 l\-0.
-2 logp 3r -
i z.st rzz I
or,^ _ *,.orol
_ rzr '' [ro.o+orr+.orrtzt
I Q.067)
(2.067) *
[to.o+oxz.zoelt ,.r6al I
[ I
_^ ^-^.. fc I I ft,
.-^-. lb {r.6oe]' I
rh rr2 / rfi2 \ ()d. /) , l.fil
/5R?\::11 M)\2:L | -" I =
n' sec' \r+4 ln.'/
'-' ,ftr.- -. ss62 \144 in.2/
lrql rr ft lb'
-'--'-'sec2
,rrta j:b:l'- sec'lDf
lbr
APr, = 0.126 n.t
I
Mechanical Design of Process Systems
[/ooore\ I
1ft
l\ r:+ 1* z.st ,,^, / I ft3 \/1 min\
1
-z
a I t"' ca1
(flo
[- rnrsvltf 'oi" \2.+zs s"il \60 '*/ =
s
'oc'o 4.457 ftlsec
1
/ t e'z \
(0.864)in.,tfr1n}/
:0.M2
K-values
A
o 2-in. x ltlz-in. LR ell Nn. = t#)o,0.0", or.a *
(1.008 x 1o-) .lb'
ft-sec
t' - e'i],u,
. - "'
o
[(9 : 226,889.525
1u9'
Ir rr o\2
= l- -^"1 = 0.607
lb''4
/
l\ o0s7 2.51
I
|
1
-r rocro * |
[ 3i tz26seo]flri
\2.M7 5
(o0
= P4 : 0.368
:0.051
* _ 0.8 sin [(5.423)(l - 0.607)] _ o.o'
(0.368)
1)
Q*:18-:0.667-K:0.53 . lr/2-in. x f-in. reducer
p,)] _
--- [c(l -
,. _ 0.8 sin 0.8 sin [(6.23s)(0.593)1
^R - - - r.v
et- - 0.166
6 gpm
*r:I (0.051x1.0x12)
(r .049)
+ r. 1811
,l
t!' h
r1 l2x't
I
I
,<o rr lb r,,, ,.t', fC / t t' \
---A -spm | +6 -oFm
sec \t+a 1"31
= O.222 psr
12 fr-lh
I
@ S€C'-lD6
Piping Fluid Mechanics 39
/4\ sal /
1ft3 t-l \/r -i"\ @ Flow from statioo 2 to station 3 0.063 psi
\"/ min t-t
- \7.479 ga| \60 sec/ _ 3.612 ftlser @ Flow from station 3 to station 4 0.222 psi
@ Flow from statioD 4 to exit B O.zlO5 psi
(0.533)in.'(+) 1.002 psi
a rsa.zr'tt'I
{o't'lo,r.utl'sec' Dot, = 1.002 psi = Total frictional pressure drop
\12l : A4.449 from entrance A to exit B
1^ Station 1-
(o0.5
= 0.055
35spm
I-values 2
o.7
o" =:.r : 42 : 0.167
. l-in. x 3/a-in. reducer Qr
Kr = -0.032
Header entrance = K = 0.78
_____7-_
,. _ 0.8 sin [a(l - F1] _ 0.8 sin [(3.081X1 - 0.601I station 1 : K = -0.03
(0.361) K : O.75
Q: : 42
: 0.0'[8
epm
'L = 10ft;/ = 0.15cp;d = 2.067in.
. p:58.7 lb/ft3;6 = 9.9613
3-90' LR ells V = 4.016 ff/sec; f : 0.020; Nr, : 402,829
APr : 9.195 *'
K : 3(0.025X30) : 2.25O
Station 2-
Pi1r exit + 1q : 1.0
&r, _
- l(0.055x2.0)(12) +
28 gpm
t to^szat
3.2esl
40 Mechanical Design of Process Systems
n. 1 Station 5-
Qr: ::] = JJ.-:0.200
Qr
Kaz = -0'03
L = 5in. = 0.417 ft;d = 2.O67 in.
Qa : 35 epm
Y : 3.346 tusec; f : 0.020; Nn : 335,691
AP, : g.gg1 n"t
Station 3-
9:1:o.soo
Qr 14
L: 10ft
Krz : 0.015
v = 1.339 ft/sec;,f : 0.021
Nn":134,276
APs : 9.914 O.t
Q3 = 28 gPm
Q:.128 :o.zs
Qr
L = 10ft
d : 2.O67 rn.
v : 2.677 ft/sec; K32 : -0.036
Nr":268,553
f : 0.020
APr : 6.952 n.'
Station 4-
21 gpm
o^:Qt=ro
I Qr
&r = 1.28
Tspm For 2-in. x 1tlz-in. reducer,K = O-129
K : 0.311
I
14gpm lrlz-in. xl-in. reduceq
I l-in. x 3/a-in. reducer, K : 0.048
2 r/+-in. plug valve, K : 18(0.025) : 0.450
2-l1lz-in.90" LR ells, K = 2(30)(0.021\ : 1.2@
1-1-in. 90" LR ell, K : (30X0.023) = 0.690
1-3l+-in. 90. LR ell, K : (30X0.025) : 0.750
Q:1:o.rgr
2l Exit into coil, K : 1.0
Q:
&z = -0.030 Q = 7 gPm; L : 7 -25 f7
L:5in.:0.417ft; d = 2.067 ln. F.- = 5.168
!K
Qr : 21 epm :
. For 2-in. { pipe, d 2.067 in.
V = 2.008 ff/sec; f : 0.021 L:7in.:0.583ft
Nn' = 201'415 K: 1.049
AP+ : g.ggt O.t V= 0.669 fl:/sec
Piping Fluid Mechanics 4'l
Nr":67,138 n.
f:0.023 QR::r
\J3
= 0.333
AP : 0.0M psi Kr: -0.030
o For lrlz-in. d Pipe, d : !.610 in. L: 10ft
L:3ft Qr : 21 gpm
K : 1.571 p:58.7 lblff; p :0.015 cp; e:0.0018;
V : 1.103 ftlsec; Nx" : 86,195; d : 2.067 \n.
f : 0.023 V = 2.008 tusec; Nr" : 201,415t f : 0.021
: AP1 : g.g3g O.'
AP 0.016 psi
o For 1-in. { pipe, d : 1.049 in.
Station
L:Zft 2
-
K = 0.738
V : 2.599 ftlsec; Nq" : 132,292',
f : 0.024
AP : 0.055 psi
. For 3/a-in. { pipe, d : 0.824 in
L :2ft
K : 2.2O
Y : 4.Zll ft/sec; Nx" : 168'416;
f:0.025
AP = 0.330 pst
Toral frictional pressure drop from station 1 to bottom shell o, : Q' : o.5ool L = roft
Qr
.nil is
K:z = 0.015
!s\
Q3 : 14 gpm
llP = 0.195 psi + 0.001 psi
i-!-Z
+ 0.052 psi +
\-!-
0.001 psi
--/- v : 1.339 ft/sec; N3" : 134,2'16: f: 0.021
\--
3 :
AP 0.014 psi
station 1 station 2 station station 4
+ 0.014 psi + (0.004 + 0.016 + 0.055 Friction pressure drop from station 2 to coil entrance:
+ 0.330) psi
station 5 station 6
3
n.
+ 21spm o" : lll
r)^
= 1.0: K,, = 1.28
For 2-in. x 1-in. reducer, K : 2.538
;'il] I
1-in. x :/q-in. reducer, K = 0.048
3/4-in. plug valve, K : 0.450
I Exit into coil, K : 1.0
r 14 gpm For l-in. 90" LR ell, K = 0.690
2 For l-in. { pipe, d : 1.049 in.
d
Mechanical Design of Process Systems
Q : : 4fr,
7 gPm, L EXAIIPLE l-3: FRIGTION PBESSURE DROP
:
K 0.738 FOR A WASTE HEAT RECOVERY SYSTEM
V = 2.599f1/sec; f : 0.024
Ns.:132'292
A gas turbine manufacturer specifies that the maxi-
AP : 0.079 mum back pressure on the unit used in this system be 10
in. of water pressure, therefore, the waste heat recovery
system should be designed so that the frictional pressure
For 2-i'].0 pipe:Q:7gpm,L: llft drop does not exceed 10 in. of water. The system is
K:2.538
V : 0.669 tusec shown in Figure l-20.
Nn":67'138.184
f : 0.023 Z Turbine exhaust dntq for outside air ot 6l)'F
Ap : 0.0t1psi
Temp. ofexhaust gas : 795"F; V= 131 fusec
For3A-in. opipe:d : 0.8241 L: 2 ft: K = 1.450
V: th .n
4.211 tusec tt : 0.0759 .':. \0.4132) lb/n-hr
It-hr --l= = 0.03 t cp
Nn":168'416
f : 0.025
Ap : 0.245 psi L: l2O ft; : (commercial steel)
O.OOOO+Z
D
Total frictional pressure drop from station 1 to bottom shell
coil is e= 0.00015
sr
l-r AP 0.030 psi + 0.014
= \-!-\-.-/ psi
a = 0.0759 lb {--!t
: VDP, '|
N""
station 1 station 2 t' ft-hr \3,600 sec/
B{SS STAC i(
42'6 670 9a
4zV srD
, : log,o I n o!goo38j : :
10.00001I + :-:
-:* -2 K.*r 0.770 + 3.161 + 0.340 4.271
ri t rttot / From equation l-4,
f:0.0130
op = ILL*
\d -
rr)ey
I2e,
D= 41.25 in. = 3.438 ft
fL (0.0130)(120) (r 3r .oo,12
il
.. _ , "-,1
to.oiU rt"
-19,
sec'
D 3.438 ^,r..4 ur = [(0.0130)(I20)
t- 1- +,zt1l
L (3.44) I
2\32.2)
fr
". (144)
fr2
It- sec'
\:lr nKr K- nK-
o, - [<o.ot:ot<+.ot
| 5.714
* o u,rl
es and Fittings Kr I :fr
2132.2)
fr2
( _
qso
-" | (0.75) /ll I _lr\ = 0.770 in
'::::::L
H,O ar 62.F
662.224 \ 4t.251 AP = 0.005 osi t27.912t
' t psi
: 0.140 in. H2O
AP thru heating coils : 2 in. H:O
Itirer K-values
Total AP = 7.564 tn. + 0.140 in. + 2.000 in.
i.De entrance at turbine nozzle l .000 OI
rl-in.dx30-in.d 0.161
.lrste heat recovery unit entry duct 1.000 AP : 9.704 in. HrO < l0 in. allowed
l-rck exit 1.000
3. 161
Entrance, K: 0.78
Dr = o.oso+o.l2o = 0.570
AP = 5.935 psi; NRe =22.494.325
V :83.973 ftlsec: = 0.01913f Flgure 1-22. Relief valve piping system.
The Engineering Mechanics of Piping 45
Static and dynamic analyses require clear and precise These forces and moments are controlled by structural
c.efinition of terms-their misuse can often lead to mis- supports attached to the piping using pipe supporrs ro
:-;nderstandings, a problem the engineer greatly appreci- control forces and moments in the pipe and attaching
The application of engineering mechanics to piping components bring up two fundamental concepts-stiff-
"tes.
:s mainly referred to in industry as "pipe stress analy- ness and f lexibility-which are discussed later in this
-.rs." However, the term is not comprehensive enough chaDter.
recause engineers are usually more concerned about
:orces and moments exerted on equipment than stress.
Cerrainly, stress is a concern and is discussed along with
rtier phenomena in the chapter.
-{nother popular term used in industry is "piping flex- PIPING CRITERIA
-:ility analysis." The word flexibility can pose a prob-
-em because in the stiffness method of analysis it is actu- In analyzing piping mechanics, the following parame-
-:lh the structural stiffness of pipe supports, rather than ters must be considered:
:ieribility, that is important. For this reason the term
-'piping 1. The appropriate code that applies to the system.
flexibility analysis" is avoided.
-\ piping component is any constituent part of a piping 2. The design pressure and temperature.
.-, stem, of any finite length of pipe-a valve, flange, el- 3. The type of material.
:\.\\\'. pump, or anything else within the piping system. 4. The pipe size and wall thickness of each pipe com-
llping is supported for various reasons-an obvious one ponent.
5. The piping geometry including movements of an-
-rng to counteract the force of gravity-and to begin to chors and restraints.
-:rderstand the applications we must start with some ba-
a.: concepts. 6. The allowable stresses for the desisn conditions set
Consider a piping component as shown in Figure 2-1. by the appropriate code.
i{:re we have a three-dimensional axis system with the 7. Limitations of forces and moments on equipment
: rmponent-a short length of straight pipe-subjected to nozzles set by API, NEMA, or the equipment man-
: rrces and moments about each axis. The forces and mo- ulacturers.
:.nts are considered as vector quantities and often ex- 8. Metallurgical considerations, such as protecting
::essed in terms of resultant vectors. For convenience material from critical temperatures, like carbon
:: u'ill express resultant vectors in terms of a resultant steel below its transition temperature.
' :.ror operator defined as follows:
For any piping system, these criteria must be consid-
ered and satisfied. While it is sufficient to analyze a
pip-
; \.\.2) : ,tll*-Z (2-1) ing system, it is not always necessary. For example, a
system having only two terminal points and pipe of uni
-je resultant force and moments change in magnitude form size does not require a formal analysis if the fol-
-J direction along the length of the piping system. lowing approximate criterion is satisfied:
47
Mechanical Design of Process Systems
inro the plastic region. For complete plastic deformation imen is unloaded, residual stresses are developed that
to occur, the entire area ofthe pipe wall must exceed the cause reverse yielding when the material exceeds the
minimum yield point. This would not be acceptable in compressive yield point. This is shown graphically in
practice because of permanent deformation and the pos- Figure 2-5. The specimen is loaded to point A and an ex-
sibility of rupture. cessive load deforms it to point B. At point B, part of the
There are acceptable cases where the loads will fail be- material is in the plastic range and the other portion is in
tween Figure 2-2 and Figve 2-3. This condition is the elastic range. When the specimen is unloaded, the
shown on Figtre 2-4, where part of the pipe wall is in stresses in the material go into compression shown at
the elastic range and the other part is on the plastic re- point C. Residual stresses caused by the combination of
gion. For cases where the portion in the plastic range is material in the elastic and plastic regions make part of
small compared with the portion in the elastic range, the the material exceed the compressive yield point and the
amount of permanent plastic deformation is impercepti- specimen deforms from point C to point D. Upon appli-
ble. For this reason, the distance between points A and B cation of the same initial tensile load, the material is
m Figure 2-4 is small compared to Figure 2-3 because loaded to point E. Point E is larger in value and, thus, to
the portion of material in the elastic range limits the the right of point A, because the initial loading of part of
amount of permanent deformation . Thus , when the spec- the specimen into the plastic range causes strain harden-
B
I
tl
ll
,l
,l
STBAIN =>
ing and, thus, increases the minimum yield point of the This analysis indicates that the allowable stress should
material. As excessive loads are applied, the minimum be based on the yield point rather than ultimate strength.
yield point E is exceeded and the material deforms to The material's ability to revert into compression and
point F. As the material is unloaded again the initial pro- limit itself to the amount of permanent plastic deforma-
cess repeats itself and the stresses in the material move to tion is termed "shake down." The material "shaking
point G and then to point H as the compressive yield down" limits the amount of deformation and, thus, has
point is exceeded. an elastic response.
Point Q represents the stress in the loaded condition af- From this discussion, we see tlat there is a range of
ter several loading cycles, and point P represents the allowable stresses available. Direct membrane stresses
stress in the unloaded condition. It is possible that no sig- are limited by oy, bending stress is limited by l.5oy, and
nificant plastic deformation will occur after many load a limited, one-time permanent deformation from A to B
cycles. However, should stress values of Q and P exceed occurring from secondary stresses is limited by 2oy.
the fatigue limit of the material, small cracks will propa- Table 2-l gives recommended values for design allow-
gate throughout the strain-hardened material. After the able stresses. As shown in ASME Section VIII, Division
small cracks appear, further cyclic loading will result in I, paragraph UA-5e, different stress levels for different
brittle fracture failure. The stress magnitude P results stress categories are acceptable.
from the specimen being unloaded when the load condi-
tion, point Q, is reached. Thus, since Q is the tensile
stress opposite to the compressive stress P in the paral- FLEXIBILITY AND STIFFNESS OF PIPING
lelogram OB'QR the sides OB' and QP are equal in SYSTEMS
iength. Therefore, Q : 0.5 B'. Fracture by strain hard-
ening will not occur if the theoretical tensile stress B' There are two basic approaches to piping mechanics-
does not exceed twice the minimum yield stress of point
flexibility and stiffness. The former approach is more
A, and the magnitude of Q does not exceed the ultimate common and easier to understand. Piping mechanics
yield strength of point A. (more popularly known as "pipe stress") is often re-
When a ductile material, that is a material with a de- ferred to as "flexibility analysis," but it will become ob-
fined minirnum yield point, is subjected to repeated load- vious in the following discussion that such a term is not
ing, a certain behavior occurs. When a component, such complete.
as a nozzle on a header pipe, is repeatedly loaded and In the flexibility approach, the piping configuration is
unloaded, the strain hardening makes the material made more flexible by using loops that allow the pipe to
stronger from load cycle to load cycle. As the material
becomes harder, it is better able to resist yield. However,
the maximum point at which this repeated loading cycle
Table 2-1
can occur is 2oyp. The stress o : 2ovp is the limit ofthe Allowable Stresses'
maximum stress range. This process is called elastic
shakedown. that is. the material "shakes down" to an Pressure Component Design Conditions
elastic response, and undergoes deformations or strains
l. Internal pressure . ....... oA
2. Internal Dressure
induced by loads beyond the minimum yield point of the plus therinal loading . ... . 1.25 (oa * op)
material. 3. Temporary mechanical
It must be noted that at elevated temperatures the value overload . ..... l.33oa < oy
of 2oyp can be altered by hydrogen embrittlement. Car- 4. Hydrotest . . ... oo X hydrotest factor
bon steel exposed to hydrogen at elevated temperatures Non-pressure Components Design Conditions
1. Pipe supports and
can fail during elastic shakedown because the hydrogen connections other
combines with the carbon causing embrittlement. than bolts .. ... 1.330a
The relationship between the maximum stress range 2. Bolting ....... Per AISC Manual of Steel
and the initial yield point can be expressed as Construction considerable
savinss in material can be
incuried if high strength
bols are utilized, such as
o1,s 3 Zoyp (2-1,4) SA-193-87. Followins
AISC guidelines in n6n
where MR : maximum local stress range not producing pressure components will
fatigue failure, psi result in prudent
YP : initial yield point of the matedal at the economical desisn.
operatrng temperature, psl ' Courtes) of American Socier) o[ Mechanica] Engineers
The Engineering Mechanics of Piping 53
displace itself, resulting in lower stresses, forces, and A piping element has six degrees of freedom, three in
moments in the system. This method is often the most translation and three in rotation, as shown in Figure 2-6.
desirable when relatively inexpensive piping material is The amount of force or moment required to produce unit
used (pipe elbows can be very expensive in alloy piping) displacement in each degree of freedom at points all
and space is available for the loop(s). along the piping element is described mathematically as
However, the stiffness method becomes quite impor- the stiffness matrix. K. which is defined as
tant when the flexibility method is neither practicai nor
economical. When limited space reduces piping flexibil- P:KU
ity or makes it irnpossible or undesirable to use flexibil-
ity loops, restraining the piping using the stiffness of where we have an elastic element subjected to a set of n
pipe supports becomes the alternative. This approach is forces and moments
gaining popularity with the increased use of modular de-
(2-ts)
signs of petrochemical plants, offshore platforms, and
other industrial facilities.
the corresponding displacement of each P1 is described
The following is a summary of the advantages of both
by the matrix
methods:
l. Requires less pipe fittings and is thus more eco- Therefore, the stiffness matrix can be expressed as
nomical than flexibility method, because pipe re-
straints required are far less expensive than the p
number of fittings they replace. In alloy piping (2-17)
U
these savings are enormous.
2. Requires far less space for piping, such as in modu- which can be in pounds per inch or foot pounds per de-
Iar skid-mounted plants, offshore platforms, and gree. The relationship
ships.
3. Method is safer because in case ofa failure, such as II
a leak in a weld crack, the pipe restraints can (and (2-18)
P
have) kept systems from blowing apart.
4. Piping and system is more resistant to dynamic
loads, such as vibration and seismic shock loads. is defined as the compliance or flexibility matrix and can
be in inches per pounds or degrees per foot-pounds.
Flexibility Method Advantages Thus, the stiffness K ofa system is the inverse of the sys-
tem compliance or flexibility, C, that is, the piping sys-
1. Utilizes simpler pipe supports, and requires less tem becomes more flexible, or less stiff than its initial
piping engineering skill. configuration .
2. Is more desirable in noncritical systems, e.g. ex- The system stiffness matrix, K, is made up of elements
haust and flare lines. that are either direct stiffness or indirect stiffness compo-
3. Many solutions do not require a computer. The nents. The direct stiffness component K;; is the value of
problems can be solved manually. stiffness at the point i when the displacement U1 is pro-
duced by a force or moment P acting in the direction of
To better understand these two methods of piping me- U1. The indirect stiffness Kij is the value of stiffness at
-'hanics, it is necessary to examine some basics of struc- the point j, with the displacement Uj acting in the direc-
::rral analysis. tion ofj, due to a force or moment at another point i in
Stiffness is the amount of force or moment reouired to the direction of i. The indirect stiffness can also be
:ioduce unit displacement. either translational or angu- thought of as relative stiffness-those stiffness values in-
-.lr movement. The simplest concept of stiffness is to duced by forces and moments in the system other than
::nagine using X pounds to compress a spring one inch. the point in consideration. It is the combined grouping of
Thus, the spring stiffness is in terms of pounds per inch. the complete direct and indirect stiffness values that
This simple example illustrates translational stiffness. form what is called the "stiffness matrix." Each direct
Rotational stiffness can be thought of in a similar manner and indirect stiffness is considered in the matrix when all
as a spring that resists rotational movement, foot-pounds other matrix components are zero. Such as the system
rer unit degree of movement. described in the followins:
Mechanical Design of Process Systems
lH ft: e ft [: t_3
io-pon"nt, ur. utilized in practical applications'.we will The force required to move the 4-in. pipe
l/+ in' is
consider each type of force or moment rnduclng ols-
olacements, thai is, each component of the P matrix cor- : lh
iesponding to each value of the U matrix' Table 2-2 lists
F (5.687.66t .11 (0.25) in. = 1.421.92 lb
Table 2-2
Stittness Properties ot Piping Elements
*"t
t"
-(*--u
rl'>'*- ->x
f A\l
l,k
r\ \.t
AF
Ktt=Kr:? Kr::Kzs:0 Kla:K2a:0
r,,:r,,:4EL _- =
or, tzEl ,. l2Er
il-+ e)L3 ^44:(l+o)L:
,, : -lzBl ,. : - tzEr
K:r:IQr:K5r=Iqr=0 ^or 11 1oy rr ^" rJi) IJ-
(
K.. f.r
/P t"t
--(.4-=-Y/^-
-""/ l--------9)'
\-1._________J /, .lffil
,/
TT
^.-_____________
Tu o-------4
"(\YI
f-l T/ t<----_T+
Krs=K:r=illo)U
-6FI - lzBr
GALI
and
--
K = cJ lorsronal sunness
-:L
56 Mechanical Design of Process Systems
g
,r,
FOR ONE END PINNED AND THE OTHER FIXED 10Q SCHEDULE 40
K1o )) K4
K,, = -3!.L- \
' l1+olL'
a"q scHeoure
M
a: (-1.75)do,o'ft = -o.o70in'
The force required to move a 4-in schedule 40 pipe
0.070 in. is
th Figure 2-8. Pipe size makes a significant difference in nozzle
Fq: (5.687.66) I tO.OUOr in. = 398.14 lb'
loadins.s.
The Engineering Mechanics of Piping
l ielding a moment of
vro: (8,854.82)(4) : 35,4r9.27tt-lbl2 : l7,7o9.64ft-tb
\'r1 : (3e8.14)
! : na.ze uv
at nozzles A and B.
It is clear that the 10-in. pipe would exert moments
u ell above the allowable moments for most rotating and
stationary equipment. To reduce the loading at the noz-
zle, the engineer is faced with two options-make the
piping configuration more flexible or restrain the piping.
To fabricate the piping configuration to within a toler-
ance of 0.070 in. would be well beyond the practical
range of any fabricating shop. Exchanger
First, we will analyze a case where space is premium
and there is not enough room to make the piping more
flexible. This requires using piping restraints to transfer
loads from the pipe to structural steel or concrete. Con-
sider the piping system in Figure 2-9, where two alumi-
num heat exchangers are piped parallel to one another.
Here we use the fewest 90' elbows needed to give the
svstem enough flexibility to stay within the maximum al-
iowable stress range for the material at the given temper-
ature. Piping restraints are then placed close to the heat
3\changers to transfer loads from the pipe to the steel in-
stead of the nozzle of the exchanger.
Now, we analyze the component that makes the system
$ork-the pipe restraint at the equipment nozzle. The
:estraint's function is to transfer forces and moments ex- Figure 2-9. An MRS support-restraint designed to reduce
erted by the pipe to the structural steel below, simultane- forces and moments on an aluminum olate-fin heat exchanser.
ously allowing the equipment to move freely. This re-
quires a more careful design of the piping restraint, as
.\e are expecting it to do more.
In this example the piping restraints must allow the ex-
Jhangers to move upward as shown in Figure 2-9. A re-
straint that resists moments by transferring the moments simple MRS restraint is the anchor, where the pipe itself
:iom the pipe to the steel is termed a moment restraining or a pipe attachment is welded down to structural steel or
support (MRS). Different types of MRS supports are immersed in concrete. In that case, it is resisting three
shown in Figure 2-10. An MRS can vary from a boiled degrees of freedom in translation and three degrees of
plate connection shown in Figure 2-10A to a sophisti- freedom in rotation. In most applications, the moments
iated type in Figure 2-10C. MRS restraints' sophistica- at nozzle connections can become excessive, and it is of-
:ion is a function of how much rotation is resisted and ten desirable to resist rotation in one. two- or tlree axes
iow much translational movement is allowed. The most while allowins translational movement. Resistine rota-
58 Mechanical Design of Process Systems
Nozzle flange
_> Hequrres
D Requires
meters 8-in. and smaller, attaching an MRS on one side equipment, it is often more economical and desirable to
is sufficient for most modular construction. design the piping to be flexible enough to reduce load-
Circumferential shear flow, on the other hand, is not a ings on supports and equipment nozzles. For pipe racks,
lactor in most installations because torsion is very effec- long headers, etc. this method is the only practical ap-
lively transferred to the structural steel by the MRS re- proach to solving piping mechanics problems. Tools used
sralnt. in this approach include such well known devices and
Using piping restraints to transfer loads to structural techniques as piping loops, cut short and cut 1ong, and
iteel or concrete to lower loads at equipment nozzles is expansion joints.
'becoming quite popular
and more widespread because it
is more economical in modular skid design. Also, where
erpensive piping materials are used, the stiffness method
PIPE LOOPS
can help reduce the number ofelbows used for flexibility
end, thus, reduce the cost of the job because restraints
The most common types of pipe loops used today are
and supports are far cheaper than piping elbows. "U" shapes, "2" shapes, and "L" shapes. Curves for
these shapes showing stresses plotted against the loop di-
FLEXIBILITY IIETHOD OF PIPING mensions are shown in Figures 2-12 and the equations
HECHANICS are as follows:
rol
t
,I I
8l
,l
Ry
it
tof
I
"I
I
"l
_l
Ry
6
1o
Ry
ro
Ry
'|
to 12 14 16 1A 20 22 24
Ay
Figure 2-128. Uloop with one leg twice the other leg.
Mechanical Design of Process Systems
to
RY
Figure 2-128 (continued). UJoop with one leg twice the other leg'
10
6
Ry
180 ^ z&
Figure 2-12C. UJoop with one leg three times the other leg'
The Engineering Mechanics of Piping
6
Ry
"* o,
ooo
Fv
10 t2 tO t"o, 22242a303234
Figwe 2-12C (continued). U-loop with one leg three times the other leg.
Mechanical Design of Process Systems
to
I
I
7
Rv
tao oo 22o
10
I
a
6
Ry
15 ,o
2O o,
1, 25
Figure 2-12D. Uloop with one leg four times the other leg.
The Engineering Mechanics of Piping
'to
Figure 2-12D (continued). UJoop with one leg four times the other leg.
a =
-i1 n=*
Ry
Ab
Ry
10 20 30 40 50 60 m 80 gOAv IOO llo l2O l3O lr|o 15O 160 17O t8O
Ry
4
40o A, so
"=E
U e
- 180
-l-l [DNkr I
Circumferential :H: , MD"IC. ..
" - {fadl?llSl
where K : KRX or KRY, ft-lb/deg EI
M = moment, ft.lb
e : angle of rotation, deg where C1 : 0'09 for in Plane bending
F : modulus of elasticity of vessel metal at C. : O.2'l for out of Plane bending
ambient temperature, Psl D= diameter of vessel or pipe header' in.
I : moment of inertia of vessel rLozzle, in.a Dg = diameter of branch' in.
Dy = diameter of vessel nozzle, in. E = modulus ofelasticity. lb/in 2
: Figure 2-16. Pipe lug support for a pipe with internal pres-
Circumferential R" stresses must be added'
suie-primary and secondary
The Engineering Mechanics of Piping
^C^L,RU
' 2Rna 2Ru L
(B,B)os Q-20)
rr
pq
< | .u : [, - i(- -
'ff),, ",]
(Btp)o: (2-21)
:< t
: r+l/ 6w/ \0.65
I (2-22)
Table 2-4
Recommended Minimum Weld Sizes for Plates
\5
Thickness t ol Thicker Minimum Size, w' ot Fillet
:rd the concentration factor for bendine stress is deter- Plare welded (in.) weld (in.)
Sned by t<Y2 3/ro
tlz1t13/c
=,*(-t'\" (2-23) 3l+<t1lt/z 51rc
100
+ :H (9) :
'"'r"'o''"r
bending stress
,, N" ,,
o"T-^o 6M,
(2-2s)
*
The longitudinal stress and circumferential stress rep-
resent the secondary stresses in the pipe wall. These
stresses must be added to the primary stress which, in the
case of internal pressure in the pipe, is the pressure
stress. The pressure stress is determined by
I
€tE
z l:\ P"GD)
-
'2t
OD: . DSI Q-26)
oT = q6+ op Q-2',1)
oT:qx+op Q-28)
Mx
T
.ol
.30 .35 .40 45 .50 Figure 2-21. Bending moment, M"/P, induced by radial
p load P [3].
Mx
-T
spring that will provide the same spring rate for any moves upward. The amount of deflection, A, relates to
force great enough to cause initial deflection. Constant the amount of excessive force as
springs are used in critical installations where forces or
deflections induced on the piping system are critical. F" : AK' lb (2-29)
These springs are considerably more expensive than the where K: spdng constant of spring, lb/in.
variable types and are usually avoided by piping engi- A= deflection, in.
neers when not needed.
Constant springs provide constant supporting force for
It is common practice to calibratethe hanger in such a
the pipe throughout its full range of contraction and ex-
manner that when the piping is at its operating (hot or
pansion. As shown in Figwe 2-23, this constant support cold) condition, the supporting force of the spring is
mechanism consists of a helical coil spring working in equal to the weight of the pipe. This means that the maxi-
conjunction with a bell crank lever in such a manner that mum variation in supporting force occurs when the pipe
rhe spring force times its distance to the lever pivot is al-
is in the down condition, when primary stresses are non-
ways equal to the pipe load times its distance to the lower existent because of no internal pressure. Therefore, in
pivot. Thus, the constant spring is used where it is not the cold position, the suppo ing force of the spring is
desirable for piping loads to be transferred to connecting
equipment or other supports.
F:F"+WP (2-30)
--F _
.eB 9
sg !,i
to
s sss
g$
E
f; egfi E$$3m$Ege$$$$8$gE
bE89 EE$f;9Ffr;S
FN&SRREhR s8E8
83RE $EHTEESgEP ?bEr
9383
e$pEIgEE$EENFEFFEE*
5
I
SHHEESFPEFFFFe$es$ErE HHDF E
3EE8
FF FF gBE g gBfi Bg EE BEEEFE NShE 8 t5 a5
ao
It)
ctt F$$ffg$fi gfi FEFFFgEgHE g PbEF
E 8 I
(t
E r58PS9gN HFH€i a I
It 8888 gEFgggFBi igE9Fggi 89R I I e
dE 3[8i ri
3KtF838
dl.N Fxmt5H$fiRHR$ I
ao
.(g HEHiHS;gEEEPR:
Fc Ett
889: PFFEEF:9EEEEFEFEEEEE 33I
U'
l! ;gEEEEEEE-FFEEEEgggEE EEEg
'6. FNF[g$EIHHEHEESTFf Ff R NKNS 8 8
F $$$MEEfr EEdFFFPFH$JgEH
i€$$$s$eH;$sg*HHEE$E$ 3Rf,N E
$HESHShgBSg;8s5$$t$$$
N&RNRREhE&
EFE$FSRHHF*fr $TRSft $*8ft
RN633$g333EFRR33EbEBp
EsEpp:::$FFSppi33$EEq
EESF FPRESbS$8588839:P:&ft &
83SE33633bEFFrRR8$$E3
j's s
9E
S ss
g8 (,
F
F
()
The Engineering Mechanics of Piping 77
It is common practice to utilize the smallest spring size l/a in.) will cause such a spring to jam, as shown in Fig-
possible. rre 2-26A. To avoid jamming, a guided load column is
In critical and large systems, spring loadings should be used to prevent such a problem (Figure 2-268).
eYaluated by computer analysis. Often, in large systems, Springs are often used to support equipment to reduce
piping movements are not intuitively obvious and errors nozzle loadings, which are discussed in Chapter 6.
can result because the entire system must be evaluated if
a correct analysis is to be obtained. In most systems,
hand calculations are far too cumbersome.
Occasionally, springs are used as moment resisting de- EXPA]ISION JOINTS
vices, as shown in Figure 2-25. In such an application,
the spring preloads the pipe in a specific direction. As These devices accommodate movement in piping
the pipe expands or contracts, the spring counters the caused by temperature changes. Such items range from
rbrce created by the movement and, thus, reduces the special slip joints that only allow movement in the axial
moment at an end connection. Such a system in normal direction to corrugated bellows joints that can be de-
practice usually works in the operating mode but when signed to accommodate movement in several directions.
ihe system shuts down the spring overloads the piece of It is the latter type that we will concentrate on, as they
equipment protected in the operating condition. Thus, if are by far the most numerous and complicated of expan-
such a scheme is used, care must be taken to ensure that sion joints.
Cre protected items are safe in both the operating and Corrugated bellows expansion joints have a bad repu-
Jown conditions. These schemes can be avoided by use tation with some users because of ignorance. Many bel-
..'i MRS devices where space does not warrant piping lows expansion joints have been incorrecdy specified
t-lexibility. and the consequences attributed to the device itself. This
lrcation of spring supports is of critical importance. is unfortunate because this device is invaluable when ei-
\\-hile springs should be placed where they will be most ther re-routing the line is impossible or cold spring or
3fficient, often such locations are undesirable from tle other alternatives are not available. The surest way to
itructural engineer's viewpoint. The piping engineer avoid problems with bellows expansion joints is to have
:hould always be cognizant of available structural steel the piping (stress) engineer specify the unit and to pro-
Lrr concrete and loads to be placed on structures. Most cure the unit from a reputable manufacturer.
prings are supported from above at either mid spans or The bellows expansion joint is like the MRS device
at elbows. Many times it is desirable to support the pipe discussed earlier because the more the unit is required to
tiom below. When using this type of spring, one must be accomplish the more complex is its configuration. The
.autious of pipe movement, as excessive movement ( > simplest corrugated bellows expansion joint is the single
= momenl generated bY
lvla
\ movement at Pl A
)L2Nozzle
Direction oI Pt A movement
Figure 2-25. Utilizing a spring to counter a moment generated by piping is appropriate only when the spring movement, Ms, does
ilot overload the nozzle or overstress the piping system when the latter is in the down condition and there is no movement at A.
This condition is required after the operating condition is met.
Mechanical Design of Process Systems
:
:
Figure 2-26'. (A) Enough piping movement will cock load flange andjam spring. Note: arrows indicate
direction ofmovement.
(B) A guide load column shown here will prevent situation in
1a--;. ttreie arJvarious designs for guide load columns, but for pipe
movement greater than t/+" one should consider a column with rollers or Teflon on top;f the ioad flange.
bellows type shown in Figtxe 2-27 A. This specific joint not support its own weight so this joint would not be de-
is shown with flanges welded on each end, but is avail- sirable where each end exceeds the maximum amount of
able from manufacturers with pipe spool pieces on each pipe span shown, as calculated by the following equa-
end to enable the unit to be directly welded into a line. tion:
The piping engineer should try to utilize this type ofjoint
whenever possible because of economy and simplicity of
operation. The single bellows is free to move in all de-
L:0.131
grees of freedom except about the longitudinal cen-
terline. In fact. no bellows expansion joint can accom-
modate torsion and any tendency for the pipe to exert a E : modulus of elasticity of pipe material, psi
high torsional moment could seriously damage thejoint.
I = moment of inertia of pipe, in.a
P : design pressure (psig)
External restraints are placed on the joint to restrict A : bellows effective area. in.2
movement in one or more degrees of freedom. Such de- K : axial bellows stiffness (KTZ in F\gure 2-21)
vices are tie rods or hinges that restrict movement or
pressure thrust. Figve 2-278-E are examples of joints The maximum length of unsupported pipe implies that
that are so restricted. Following the same nomenclature the unit itself is within this length. Preferably, the joinr is
shown in Figure 2-10, we consider each joint in a three- close to one support or nozzle to avoid excessive deflec-
dimensional axis system. KTX, KTY, and KTZ are the tron.
translational stiffness values lb/in., about the X. Y. and Thejoint in Figure 2-278 has values ofKTX and KRy
Z axes, respectively. KRX, KRY, and KRZ are the and KTY : KTZ : 10e lbs/in. and KRX = KRZ = 10e
rotational stiffness values, ft-lb/deg, about the X, Y, and lbs/deg. This means that the joint is free to translate in
Z axes, respectively. For all bellows joints, the X-direction and free to rotate in the y-direction and is
KRZ: (1.0 x 10') ft-lb/deg, as previously srared. rigid in all other directions. This type ofjoint is called a
ln Figure2-27 A, we have finite values for KTX, KTy, "hinged" joint and is self-supporting in the y-direction
KTZ, KRX, and KRI the joint is free to translate about shown in Figure 2-27 . Placing high vertical loads on a
three axes and rotate about two axes. The bellows does joint must be approved by the manufacturer.
The Engineering Mechanics of Piping 79
The joint in Figure 2-27C has values of KRX and KRY Pressure thrust is the amount of force generated by in-
irut absolute values of KRZ, KTX, KTY, and KTZ. Nor- ternal pressure and is simply internal pressure times
nally, these joints are used in pairs to allow rotation minimum bellows radius area (PA), lbs. This force can
lbout two axes similar to swivel or ball joints and not become quite high as the pipe size and the internal pres-
ellowing any translation. This action is seen in Figure 2- sure increase. In many applications, the piping itself is
t8. anchored and the joint is allowed to compress when the
The unit in Figure 2-27D is a pressure balanced uni- thermal compression force exceeds the pressure thrust
r ersal joint. It is free to move about all degrees of move- force. As seen in Figure 2-28, when movement in the
:nent except KRZ and is restricted by tie rods that bal- form of lateral translation is desired (KTZ and KTY), tie
.rnce pressure thrust. This type of joint is very common rods are used to restrain the joint in the axial direction
in engine exhaust systems. (KTZ: @). If tie rods are being used to overcome
Figrre 2-278 depicts one of the most complicated ex- pressure thrust, then any equipment flanged to the joint
pansion designs-an in{ine pressure balanced expansion should be able to withstand the load reouired to over-
'oint. This joint eliminates pressure thrust, is self-sup- come pressure thrust. Generally. tie rods are only used to
rorting, and does not require a change in the piping sys- permit lateral movement.
aa :em to install. It is desirable where structural supports Bellows expansion joints can be restrained and com-
.ire not available and a joint is needed because flexibility bined in pairs or trios to perform certain tasks. It must be
rs required of the piping. emphasized that just because a joint is free to move in
)
A
{:,
Ftgure 2-27. Types of bellows expansion joints: (A) flanged-flanged end simple bellows joint; (B) hinge bellows expansion
::nt: (C) gimbal bellows expansion joint; (D) pressure balanced bellows expansion joint; (E) "inJine" pressure balanced self-
'-:oorting bellows expansion joint. (Courtesy of Pathway Bellows, Inc.)
80 Mechanical Design of Process Systems
lA: lnt€rmedlaleAnchor
MOVEMENT HOT PG: Planar Guid6
-LATERAL TUEJ: Tied Univorsal Expansion Joint
Figure 2-28. Generally the use of tie rods is to allow only lateral movement. (Courtesy of Pathway Bellows, Inc.)
directions KTX, KTY, KTZ, KRX, and KRY does not PRESTRESSED PIPING
mean that the corresponding stiffness values are small.
As internal pressure and pipe size increase the values of Piping systems are sometimes prestressed to reduce
KTX, KTY, KTZ, KRX, and KRY increase, because the anchor and restraint forces and moments. This prestress-
bellows wall thickness increases to resist increased inter- ing of the pipe is best known as cold springing, but is
nal pressure. The bellows can be a single wall construc- also called "cut short," meaning that the pipe is cut short
tion (single ply) or multiple wall construction (multi-ply) a percentage of the amount of thermal expansion ex-
and the stiffness values vary with each manufacturer. pected. The opposite is true in cold systems where the
Some people erroneously think that the purpose of us- pipe contracts, so the pipe is fabricated extra long, with
ing bellows expansion joints is to make the pipe stress the extra length being a percentage of the amount of ther-
analysis unnecessary. Such is definitely not the case, be- mal conEaction expected. This procedure is best known
cause values of stiffness in each direction must be en- as "cut long." Some refer to cut long as "hot springing,"
tered in each computer stress run so that it can be vei- which may cause confusion because it is not as popular
fied that the displacement and piping loads are not as the term cold springing and to some it means hot
excessive to the equipment nozzles. As shown in Figure forming, which hds nothing to do with fabricating the
2-28 a pipe can either be properly guided or anchored, pipe extra long.
and such restrictions should be modeled into the com- "Credit" may not be taken for prestressing the pipe in
puter stress analysis. computing the stress in the piping system. Several piping
The piping engineer is encouraged to refer to the Stan- codes are specific about this and, if the piping is over the
dards of the Expansion Joint Manufacturers Association allowable stress range, one cannot cut short or long to
(EJMA) t4l in accessing piping layouts when using bel- lower the stress. However, credit may be taken for an-
lows expansion joints. Also, it is desirable to specify the chor and restraint reaction forces and moments.
joint such that the manufacturer is required to meet The procedure of cutting short or long involves a per-
EJMA requirements. One should follow EJMA guide- centage of thermal movement. The whole purpose of the
lines and requirements, and include modeling restraints prestressing process is to balance the forces and move-
and stiffness values in computer stress analysis to verify ments between the down and operating conditions. Thus,
that attached equipment is protected. Expansion joints cutting short or long 1007o (i.e., cutting short or long the
are not cataloged items to be bought at random but rather exact amount of thermal movement) is normally not
sophisticated pieces of equipment that must be engi- done. Exceeding 100% is not recommended and doesn't
neered into the piping system. With this approach, the make good sense. Normally, the amount cut is 50% and
user should not expedence any problems with bellows should not exceed 66% of the thermal movement' The
expansion joints. reactions, R6 and Rp in the operating and down condi-
The Engineering Mechanics of Piping 81
tions, respectively, are obtained from the reactions R de- Bellows expansion joints should be avoided if a more
rived from calculalions based on the modulus of elastic- economical and practical method is available for provid-
ity at ambient temperature, 8". The relationships are as ing flexibility oi restraint to the pipe. ln many ipplica-
follows: tions, only the bellows expansionjoint will suffice, e.g.,
movement and vibration in straight runs of pipe at ele-
n":lr-?*l&
- \ 3 lEo
vated temperatures between different pieces of equip-
ment can only be compensated by bellows joints. How-
ever, as the joint becomes more sophisticated and
Ro: XR thereby more expensive, other alternatives should be
considered. Such alternatives lie in either the flexibility
ot or stiffness methods Dreviouslv discussed.
I
R" -t: 11 - ].
whichever is greater, and with the additional condition FLUID FORCES EXERTED ON PIPING
that SYSTEilS
b
When fluids move in a piping system, they import en-
---: < 1.0
ergy to the system when they are forced to change direc-
4
where X: cold (or hot) spring factor ranging from zero tion by the pipe. In other words, it requires energy to
to one, one being 100% cold or hot spring change the direction of a moving f luid . This fundamental
:E computed expansion stress. psi fact is known as the impulse-momentum principle, ex-
Ep : modulus of elasticity in the down condition, nressed as:
psi
Ee : modulus of elasticity in the operating
condition, psi l)l ph : Mv, - Mv, (2-31)
R: maximum reaction for full thermal movement \-]
based on Ep which is the most severe This states that the change in momentum in a system re-
condition. lb- or in.-lb mains constant during the exchange of momentum be-
Rp : maximum reaction in down condition, lb or
tween two or more masses of the system. Applying the
in.lb
Ro : maximum reaction in operating condition, lb equation to that of a pipe elbow shown in Figwe 2-29,
or in.-lb we apply the principle to obtain:
cv: (rJ05 = 0.95 to 0.98 cause the process is considered adiabatic making the to-
nozzle adiabatic efficiency tal temperature constant. Thus,
nozzle exit pressure, psia
ambient pressure : 14.7 psia t)l
nozzle exit area, in.2 rc = l-l+
r,. oF
nozzle throat area, in.2 \K r,t
and
critical pressure : P, l---- I ' /"_,..-,f 5
\K -f l/
vc=V.=(skRrJo5=|\#',f (2-36)
100
90
80
70
60
o"A,
10
8
7
ject of design is to avoid resonance. If the inverse of the The response of a right circular cylinder at resonance
Shouhal number < 1, where f is the natural frequency of with vortex shedding is a function of the following:
the structure, then resonances with vortex shedding from
the first, second, and third harmonics are avoided. This Damping : (2zs)2 6.
can be accomplished by adding mass, such as insulation,
and putting pipe support spacing at uneven intervals. If where 6. : reduced damping _ 2m(2rl) (z-38)
pipe supports are spanned evenly, periodic wave motions pW
can form, resulting in resonance.
and
-W
e' IYi?Y,",li,t
L^YE
lt6%i!13i*,,*.*[i'"
voirEx srREtr rs.PP.REnt
BAS UIiOEROOXE
Figure 2-33. Strouhal-Reynolds number function for circular The analysis of wind-induced vibrations on tall verti-
cylinders [7]. cal vessels is discussed in Chapter 4.
Mechanical Design of Process Systems
Table 2-6
Natural Frequency ot vibration ot Beam Elements
Concentraled Load on Relatlvely Uniform Load on Beam Supported Unitorm Load on Cantilever Beam
Light Beam or Spring at Ends
ffi
. / \0.5
r - t l9l f: f : (3.89) (D)
'- t\-Di (3.55XD)-0 5 0.5
or r,rer r.re
= r stn r3l s2/p#
Figure 2-34. Damping, d (dimensionless), versus amplitude, Ay/D (dimensionless).
88 Mechanical Design of Process Systems
EXAilPLE 2.1: APPLYIilc THE STIFFNESS guish between the types of equipment. The heat
]UIETHOD TO A IIODULAH SKID.IIOUIITED exchangers HE-A and B shown in Figures 2-35 and2-36
GAS LIQUEFAGTION FACILITY are aluminum plate exchangers, and the cold separator
and power gas volume tank are made of reasonably
thick-walled stainless steel. Thus, the critical items are
the aluminum heat exchangers. The line between points
Figure 2-35 depicts the preliminary piping design of a l0 and 25 in Figure 2-35 must be cut because the relative
gas liquifaction plant mounted on a skid module. Space Z-movement between these points overloads the nozzles
is severely limited, as the equipment and piping are lim- at points 5 and 30, creating a very high y-momenr and Z-
ited by the structural steel skid supports, so such devices moment, because the pipe wants to move in the -Z and
as piping loops are unthinkable. Expansion joints are not *Y directions. These movements can be accommodated
allowed by the client, because high-pressure hydrocar- by using certain structural devices, such as shown in
bon gas is highly combustible and an expansion joint fail- Figure 2-37. Even though flexibility has been added to
ure would mean certain gross property damage and pos- the system to get the piping within the allowable stress
sible loss of human lives. Therefore, the piping engineer range, the equipment nozzles are still overloaded by ex-
must utilize the stiffness of pipe supports to transfer cessive moments above the X, Y, and Z axes-M;q, My,
loads from the piping to the structural steel rather than to and Mz.
the equipment nozzles. This transfer of loads is not total, To counter the movements of the piping at the nozzles
but enough to guarantee that the equipment nozzles load- numbered 5 and 30 at HE-B and A, respectively, variable
ings will not exceed allowable levels. springs are placed to support the pipe while allowing the
For the stiffness method to work, the piping configura- pipe to move at the same time. One spring is placed at
tion must be flexible enough for the piping itself to be point 20 with a simple Y support added at point 56.
within allowable stress limits set by the applicable code. These additional supports help reduce the moments at
This is the first significant criterion, because if the pip- nozzles 5 and 30, but not enough. So, we must add MRS
ing exceeds the allowable stress range in any part of the restraints (see Figures 2-9 ard 2-10) in pipe members 5-
geometry, the system design is faulty. Conversely, the l0 and 30-35. Each MRS is designed to allow nozzles 5
piping system can be well within the stress range and the and 30 to move upward but to transfer moments M;,
equipment nozzles still be overloaded. Thus, the piping Mv, and Mz from the pipe to the structual steel below.
itself must have a certain amount of flexibility to be Also, each MRS allows pipe members 5-10 and 30-35
within code allowables. The piprng supports must be freedom to nove along the axis so that we have the fol-
stiff enough to protect equipment nozzles from excessive lowing restraints at each MRS: KTZ, KRX, KRY, and
loads. Here our case has been stated; adding additional KRZ of Figure 2-38 (see Figure 2-10). Thus, we have
flexibility is not acceptable. one translational and three rotational restraints, each
From computer calculations the original configuration with a stifftress value K in lb/in. or ft-lb/deg. The pipe
in Figure 2-35 is found to be overstressed and the expan- and exchanger are free to translate along the X and Y
sion stress exceeds the ASME 831.3 allowable stress axes.
range provided in Equation 2-4 for 3O4 SS pipe. There- One can readily see that the MRS restraints must allow
fore, the piping must be changed to bring the maximum nozzles ar points 5 and 30 to move upward, as the ex-
stress within the accepted stress range. This analysis in- changers are bolted down to structural steel higher up on
cludes the nozzle movements shown in the figure. Each the units. Restraining the nozzles from moving upward
nozzle is considered as an anchor. would anchor the unit at the nozzles and at the support
Figure 2-36 shows the final configuration after several point causing the exchangers to rupture. Pipe members
iterations are made to determine what configuration 5-10 and 30-35 must be allowed to move along the x-axis
would best suit the structural limitations set by the mod- for thermal expansion.
ule skid. This configuration is found to have a maximum We now have the conceptual model of what the solu-
allowable well within the stress range of ASME B3 1 . 3 . tion looks like and the next step is to finalize the details.
To achieve this acceptable stess, a limited amount of The MRS restraints are resisting forces and moments
flexibility must be added to the system. Thus, regardless shown in Figure 2-38. It is necessary to design the re-
which method is used-flexibility or stifftiess-a certain straints such that each has enough stiffness to transfer the
amount of flexibility is required to make the piping sys- loads to the steel and protect the nozzles at points 5 and
tem operate properly. 30. We will now compute the support stifftress values
Once we have obtained the minimum flexible configu- KTZ, KRX, KRY, and KRZ. Once these values are de-
ration required, we now focus our attention to the equip- termined, they can be input back into the computer run
ment nozzles. To consider this question, we must distin- and verified to be sufficient for the nozzles.
The Engineering Mechanics of Piping 89
\,zr/
\/W
\\
\\\
\\\
\ \\
\ \'"
\
\\--l \p't 'a2
,r^
o
ao
o
'4
z
o_
tsJ -4
7
"3
2:1 7^
, 4e.
t 2l/o
o.,
e-:x $-
90 Mechanical Design of Process Systems
o
o
',',
ao
ct
N
o
.D
IL
9
z
The Engineering Mechanics of Piping 91
TO ENSURE PFOIEC'IoN
OF CAF8ON STEE L COSPONENTq
\\ rtrslrL^r
---
--' +
-tb
)o lb
ioo ft-rb
For Torsion
T : 12,800 ft-lb
For Shear
(12,800)ft-rb (r.,J (*,
._ _ 4(R2 + Rr +
3(R + r)
12)
. I"igit'1 :"it jotffi''. ,r:a: Tc
---- 4oJ0 in.o
)
T-- $-Q.?
q .'_-* r, Shear Distribution
At Point A-
O : 22.83"
For a circular .thin-walled cross-sectron'
rl4lc, (r,roonb (ry) in.'?(r.e75) in.
Q : 2R'zt cos o
"=+:+i:
: psi-max at neutral axis
(40.49) in.a (0.432) in.
--'\-n 16'
625 + 5.76 I
4
(0.432) cos (22 83") = 7 '635
rs 521-648
1,100) (3.215) _
' _ (40.49)
(
: 10,308 ftib (0.432)
202.182 psl
/.^.rn'l\ c-
10,308 ft-lb lY At Point
M \ln/
o =; = ----r,j/ti.-= = 1o,114Psi A:?)R'lo=r= 480.143 psi
The Engineering Mechanics of Piping
(l.100) ( l 1.039)
= 13,043.143 psi
' (12.5) (0.322\
'
* lf--lf o= 7.358 psi + I43 psi = 7.501 psi
{_ lt e __-f o:1oJ14 -143
<--ll - ll ------) O=seflpsi r = rs + z1 : 9,138 psi + 520 psi : 9,658 psi
Torsion
or KRX : KRZ = 19.467 .s92.6
-:*
fr-lh
in qeg
2,800)ft-lb (12)
It^
(| t4.313)in.
' (72.5)in.3
= 9.138 psi 2(29.0 x 106) (72.5) in.4
ln-'
KRY = --1
(36.0) in.
Tensile Stress
: 116,805,55.6 in.Jb/deg
: l,200 lb :
o r+r
870GJ Psr
or KRY : 9,133,196.3 ft-lbldeg
94 Mechanical Design of process Systems
Entering these stiffness values into the computer run. therefore making such a unit sensitive to external loads.
we see thar lhe nozzle loads fall very sharply it points 5 Always be careful when subjecting rotating equipment or
and 30. Further reduction in loads can be obtained bv vessels made of light material to excessive nozzle loads.
adding springs abo',e the MRS restraints to counter ; In the final analysis the pipe loadings transferred by
negative moment above the Z-axis. Using springs above -
the MRS to the steel must be considered by the structural
these supports is not always necessary, but in this case engineer. who must design the loundation accordingly.
they are required because of the large vertical movement Sometimes it is necessary to model the stiffness of the
of points 5 and 30. A weight run should be made to ver- steel foundation members when nozzle loadines become
ify that the springs do not ovedoad the nozzles durins critical.
shur-down.
The MRS restraints vary in design and are conceptu-
ally shown in Figure 2-10 and Figure 2-39. These iup-
ports are made ol interlocking sliding plates wirh eaih
sliding surface coated with high-strength Teflon. The
precise details of such supports vary and are customized
for each application.
Looking to other parts of the piping system, we notice
that nozzle 75 on the cold separator has a high moment
about the negative x-axis. This moment is attributed to
the aluminum exchangers (HE-A and B) moving upward
and the cold separator shrinking downward. Because
space is premium and we are "locked-in" and can't add
any more flexible piping, we add a spring at elbow 65
pulling downward to counter the exces5ive neqative x-
moment at nozzle 15. The spring is sized ro b6 accept-
able for operating and shut-down modes.
Table 2-7 lists the forces and moments at each equip-
ment nozzle.
Upon reviewing Table 2-7 , you will notice the dispar-
ity in nozzle loadings. The aluminum heat exchangers,
HE-A and B, have lower loads, especially moments,
than does the cold separator or power gas volume tank.
This is because each has acceptable loadings that are dif-
ferent. The cold separator is made of 23la-in. plate stain-
less sreel. which makes rhe loads shown easilv acceot-
able. {The method of determining whether such-loads ire
acceptable on pressure vessels is discussed in Chapter 8.)
Such loads would be very unacceptable lor the aluminum
heat exchangers because aluminum cannol withstand Figure 2-39. The BIAX-MMRS installed and in operation ar
nearly as great a load as steel and is not very elastic, olant facilitv.
Table 2-7
Equipment Nozzle Forces and Moments
EXAIIPLE 2-2: APPLYING THE along the other axes are moderate and the problem of
FLEXIBILITY IIETHOD TO A STEAiI alignment must be solved. The extremely high y-moment
TURBINE EXHAUST LINE is caused by the thermal expansion of the pipe member
extending along the z-axis from point 95 to point 145 al-
most Ze in. With this expansion along the positive z-axis,
A client has added a steam turbine to a chemical plant the pipe rotates about the positive y-axis from point 20
and has piped up the turbine with make-shift parts and through the expansion joint at point 45 to the elbow at
existing pipe, plus a newly purchased bellows expansion ooint 75. This torsion is transmitted to the turbine nozzle
joint. When the turbine technicians determine they can- it point 5. Thus, the adjustable base elbow support at
not cold align the turbine with the exhaust piping, the cli- point 31 is entirely useless in resisting this vertical mo-
ent decides that the piping must be rerouted, but requests ment and the expansion joint at point 45 transmits all of
an evaluation of the system, which is shown in Figure the torsion motion to the turbine nozzle at point 5.
240. An earlier section discussed the fact that these joints
The system is modeled with a computer software pack- are totally rigid in torsion-a moment about the axis is
age, and the results indicate that a moment about the y- parallel to the longitudinal axis, which in this example is
axis in the magnitude of 31,000 ft/lbs is exerted on the the y-axis. In fact, with the vertical moment as great as
turbine exhaust nozzle under operating conditions. Such 31,000 ftlbs the expansion joint at point 45 will either
a load is well above any turbine allowable. The reactions be destroyed or have a short service life because the bel-
li
lri
ii
i
d-o.^* .u""'"'
Figure 2-40. Original piping configuration of 20-in. 0 steam line for turbine exhaust: temperature : 300"4 pressure : 16 Hg
psia.
96 Mechanical Design of Process Systems
lows are not designed to resist such high torsional mo- Turbine nozzle (Point 5)-
ments. Thus, the diagnosis is to avoid the high torsion : : : 343'9 lb, Fr
and stop the .8-in. movement at point 135. To do this- ec-
Fx 46.51bs, Fy -530.8Ib, Fz
: 634.l lb
onomically with minimum alteration to the piping, a bel- Mx : 1,198.4 ftlb, MY :
1,978'2 ft-lb' Mz
lows expansion joint is added at point 123 and.the shoe = 745.3 ft-lb, Mr :
2,430.0 ftlb
on the dummy leg is stopped in the *z direction (i'e''
movement in the 1z direction is stopped, and the vessel Vessel nozzle (Point 85)-
nozzle at point 85 is protected by the joint at point 123 '
An expansion joint is sized based on the manufacturer's F" = -46.4Ib, Fv = -3,311.8 lb' Fz
standard dimensions for a 20-in. pipe and the joint stiff- : -3,348.5 lb, Fr : 4,709 9lb
ness values are as follows: Mx : 5,968.7 ft-lb, MY : 9,742 0 ft-lb Mz
'
: 5,0?6.0 ft-lb, MR = 12,501.9 ft-lb
KTX : KTY - 1,500 lb/in.
KRX = KRY : 200 in.-lb/deg The loadings at the turbine nozzle are acceptable' (The
KTZ : l2O lblin. basis for conaluding this is discussed in Chapter 6') The
reactions at point 85 seem excessive and would be for a
These values are provided by the joint manufacturer' steam turbin;, but considedng the vessel is five feet in
The problem of turbine alignment is directly related to diameter and made of 3-in. plate, these loads are not ex-
the inabilitv of the turbine technicians to adjust the pipe cessive. Pressure vessel nozzle loading analysis is cov-
because of the pipe's inflexibility, which is caused by the ered in Chapter 4, but one can deduce that pressure ves-
suided base elbow at point 3l . The base elbow support is sel nozzles tan withstand much greater loads than most
ieplaced by a spring depicted in Figure 2-41 and mod- tvDes of equipment.
''The
eled into the compuier siress program. This mn is made svstim is implemented and in two days the turbine
with the added ixpansion joint at point 123 and. the will be fired up and operating well. The concluding re-
spring at point 3. ihe following results were obtained marks are that the expansion joint at point 45 is accom-
plishing nothing and the capital expended for its pur-
from the computer run:
riu. wa$;d' ln fact, it would not hurt to move the
"hu."
unit, but this is not necessiuy since the high torsional
moment has been eliminated.
The expansion joint at point 123 was specified and
ourchased for those stiffniss values previously listed'
the final configuration is shown in Figve 2-42
*f--ttt t
I
,-tt^\.
,"_'-wJ_\
tN:z Dlf,Ectro{s
- = A,B [q]
on : tzzol toss.at tr.sol
\L/ 30.0
r^ : rr7.704.9\lL-'l
' \L/
112.3 fr
: 7.5 supports
15 ft (between supports)
Mechanical Design of Process Systems
Figure 2-43. (A) Initial piping configuration; (B) final piping configuration installed and operating.
Therefore, place a loop 6 ft x 6 ft (arbitrary dimen- t = corroded pipe wall thickness. rn.
sions) every seven supports. One could increase L' by
making the loop larger (increasing Ia and La'1, but space c
: L :
-'2 0.188 in.; cr l.z) ln.
limitations in this application prevent it. See Figure ,
2-43. The shess intensification factors (SID in the code
were made equal to one because computer stress runs
have verified that the curves are conservative enoush to
n :
ft> t,u [' - i(,! -')rr - r,r] rB,B,ro,
make SIF : 1.0.
R5?5+?qRt
R.: ---- ; --'= 4.t52in. Md -. Fisure 2-19
P-
The Engineering Mechanics of Piping
-i- :_ lrnrOl
or,ao
t-P l[-i,l
: tu," t:t
lotl _ ,",
A.yzt"
{6) (2,0q0)
= E.*,.5.j2 Total stress : o-t: oO + or : 3g,5g6.Ot O.t
od : (1.38) (2,842.30) + (1.36) (rs,o4s.72) Thus, a lug with these dimensions is acceptable.
: 24,309.65 lb Bolt shearing stress : 78
..
x': t, + [-1l" = t-.lz1o.zzztlo" = r'36 The distance from the lug hole centerline to the lug
lr,+w-l l*rcrr:l edge is to be a minimum of AISC Table 1.16.5.1, p.
t-ril 5-51.
Nx -. -' P
f,:-: -'-"
2.000 : 1.600
Pi R-
Fizure 2-18 ZvL 2(0.25) (2.50)
!1r f*:0.707Eoe:
P- -
Fisure 2-2r
E = Joint efficiency
f* :
*.:llolEl =,r' (2,000)
14,140.0 psi
6M-.
-.-^ [ur,-l
r..^r r-. | [rp] ta\t
/n ,....-r'2'000)
tz - tPltt']l - \0.322),
EXAIIPLE 2.5: RELIEF VALVE PIPING
= 14,467.03 SYSTEM
-. NY ..
ox:K"itooti 6M"
Examine Example 1-4 relief valve system for external
loadings induced by valve discharge. The gas properties
qx : are as follows:
(1.38) (3,739.87) + (1.36) (14,467.03)
:24,836.19 k:c1c":1.451
N: 243,755 lb/hr T": 294"F = 754"R
Primary or pressur" ,1r"a, : o. = I
2t Ar = 28.89 in.2 M: 170.9
100 Mechanical Design of Process Systems
[r"ln'
l__:l
I : 0.15 damping factor for small pipe (4-in. d<)
IMl
I : 0.025 for large pipe (>6-in. d)
I l ^ \2451105
0.1443 (1.15) 't lrr.+srt| ' l* | A o.o7 c
[^ .^ o.7z lo'5 (2-39)
Cc: v.45u I D (6. + 1.9)52 [-'-- (6, + 1.9)Sl
p-: G 61 710 =
' ccA,- (0.055) (28.89)
42.6t3 = 50.40 lb
6 : 12.569 in.
D =14
From Table 2-5,
F : (0.99X1.1s)(0.98)(28.89)(42.613)
r- = 3.55
(12'569)u r
: l.trut cvcles/sec
[zrr.+sr,l_r
[ 0.45r v.4stl I
)#[, _ l_ryq)l-,",,1"
\42.613/ I I
. 4mzf (2-38)
+(28.89) (4) 'pD-
F : 2,385.879 lb
For air at 60'F,
Reaction moment at the vessel nozzle is
p: 0.076 lb./ft3
MR : (2,385.879) lb (8.5) ft : 20,279.9'12 ft-lb
0.140 i-th
The reactions at the vessel nozzle are discussed in Chap-
.= tn :0.004&
ter 4, along with external loadings on vessels. ft ll.
'1)
-- )- _:L
sec2
EXAIIPLE 2.6: WIND-INDUGED Air velocity under investigation : 25 milhr : 36.65 ttl
VIBRATIONS OF PIPING sec
Values from Figure 2-34 indicate we are close to reso- RD : reaction in down (non-operat-
ftrnce, as we are within an L/D ratio of 5 and L/D = 30. ing) condition, lb
Thus, we should experience resonance at 25 mph for the R(x, y, z) : vector resultant operator
l-in. S Sch 40 bare pipe. The line should have more sup- fi inside radius, in.
ports added at uneven intervals closer than 30 ft and the ro outside radius, in.
previous analysis repeated for a range of wind velocities . T torsion, ftlb
Such a problem can be approached with a computer pro- t thickness, in.
gram based on experimental data. U displacement, in.
As is obvious, vortex shedding vibrations is still a sub- w weight of fluid, lb.
_jective phenomenon based on empirical data, but this ex- weld size, in.
ample should assist one in protecting piping surrounded Z section modulus, in.3
by vortices. z^ section modulus of mean sec-
tion radius, in.3
OR bending stress, psi
I{OTATION oc circumferential stress, psi
OL longitudinal stress, psi
A: area, in.2 o" pressure stress, psi
C= compliance, in./lb or deg/ft-lb radial stress, psi
D:
OR
diameter, in. torsional stress, psi
Ep :
OT
modulus of elasticity in down oy yield stress, psi
condition, psi T shear stress, psi
E" : modulus of elasticity in operat- F,M + X, +Y,+Z forces or moments acting only
ing condition, psi in +X, +Y, or *Z direction,
F = force, lbs respectively
G
I
modulus of rigidity, psi
moment of inertia, ft'
4M-X, -Y,-Z : forces or moments acting on.ly
in -X, -l or -Z direction,
J polar moment of inertia, fta respectively
K stiffness, either translational
(lb/in.) or rotational (ft-lb/deg)
K= stress concentration factor for
bending
K: stress concentration factor for
pure tension or compression
REFERENCES
KTX : translational stiffness along X- l. Faires, V. M., Design of Machine Elements, The
axis, lb/in. Macmillan Company, New York, 1965.
KRX : rotational stiffness about X- 2. Przemieniecki, I.5., Theory of Matrix Structural
axis, ft-lb/deg Analysis, McGraw-Hill Book Co., New York, 1968.
KTY = translational stiffness along Y- Wichman, K. R., Hopper, A. G., Mershon, J. L.,
axis, lb/in. Welding Research Council Bulletin 107, Local
KRY : rotational stiffness about Y- Stresses in Spherical and Cylindrical Shells Due to
axis, ft-lb/deg External Loadings, Welding Research Council, New
KTZ : translational stiffness along Z- York, 1979.
axis, lb/in. 1
Expansion Joint Manufacturers Association, Inc.,
KRZ : rotational stiffness abovt Z- Standards of the Expansion Joint Manufacturers As-
axis, ft-lb/deg sociation, Inc., New York.
L: length, in. 5. Hesse, W. J., Mumford, Jr., N. V. 5., Jet Propul-
M= moment, ft-lb sion for Aerospace Applications , Second Edition, Pit-
P= force (lb) or moment (ft-lb) in man Publishing Corporation, New York, 1964.
stiffness matrix 6. Blevins, R. D., Flow Induced Vibration, van Nos-
Pi= internal pressure, psig trand Reinhold Company, New York, 1977.
P"= external pressure, psig 7. Lienhard, J. H., "Synopsis of Lift, Drag and Vortex
Pn: internal pressure evaluated at Frequency Data For Rigid Circular Cylinders,"
radius R, psig Washington State University, College of Engineering
R: reaction, lb Research Division Bulletin 300, 1966.
Heat Transfer in Piping and Equipment
Providing thermal energy to process systems and 5,000 centipoises or more. Such high-viscosity fluids are
maintaining desired temperatures are key responsibilities quite common with coating mixes used in manufacturing
of mechanical design. Although they border on chemical roofing tiles. Tracing such viscous mixtures with several
engineering, the concern here is with the mechanical as- tracers has proven to be so inferior to jacketed pipe that
p€cts of process systems, and not with the processes the disadvantages ofjacketed systems are offset. With a
themselves. (Chapters 2 and 4 illustrate how mechanical viscosify of 4,000 centipoises, one should consider jack-
design borders civil engineering in a similar manner.) eted pipe.
Process systems require thermal energy for various Most jacketed pipe is limited in commercially avail-
reasons, and the most common are to accelerate chemi- able sizes. Normally 8-in. by 10-in. is the largest size
cal reactions; to heat products and services so the prod-
ucts remain liquid and do not clog piping or equipment,
such as with asphalt and roofing materials, viscous fuel
oils, and syrups; and to cool products and services, for
example to protect epoxy from polymerizing.
In piping systems there are three ways to transfer heat
to the process service-tubular tracers mounted exter-
nally to the pipe, jacketing the process pipe with a larger
pipe forming an annulus in which the heat transfer fluid
flows, and electrically tracing the pipe. We will discuss
the first two types of transfer systems.
103
104 Mechanical Design of Process Systems
DIMENSIONS
COMMON TO ALL 150 LB.' 3OO LB:
Holes Holeg
I o T
stzE tPs tPs u NPI .oD No. Dia. BC RF K OD No. Dla, BC RF K
Y2t1Y1 Y2 1r/a 2.56 4.25 0.62 3.12 2.00 0.75 4.88 0.75 3.50 2.OO 0.88
3/tt1Yz 1Y2 2.56 5.00 0.62 3.88 2.84 0.75 6.12 0.E8 4.50 2.88 0.88
,l
1t2 Vq 6.00 0.75 4.75 3.62 0.75 6.50 0.75 5.00 3.62 0.6E
1Yarz '| \/+ 3.44 3.62 0.88
2 % 6.00 0.75 3.62 0.62 6 0.75 5.00
1l2x2l2 1L/2 , L/^ 7.00 0.75 4.12 0.69 7.50 0.8E 5.EE 4.12 1.00
2x3 2 3 4.69 1 7.50 0.75 6.00 s.00 0.75 8.25 I 0.68 6.62 5.00 1.12
3x4 3 4 4.44 1 9.00 0.75 7.50 6.19 0.94 '10.00 8 0.8E 7.88 6.'l9 1.25
4x6 4 4.31 1 11.00 I 0.88 9.50 8.50 1.00 12.50 12 0.88 10.62 8.50 1.44
6xE 6 I 4.31 1 '13.50 0.88 11.75 10.62 1.12 15.00 12 'L00 13.00 10.62 1.62
8x10 8 10 4.88 1Y2 16.00 1.00 14.25 12.75 1.19 17.50 16 '15.25 12.75 1.88
.Flanges of higher pressure class and other facings available.
Figure 3-1C. Expansion joints for jacketed pipe. (Courtesy of Parks-Cramer Co.)
106 Mechanical Design of Process Systems
DIMENSIONS
150 LB., DUCTILE IRON. STEEL
FLANGE DIMENSIOIiISi
OD Holea A T
ID
srzE tPs lns. mm. No. Dla- BC RF x lns. mm. U TIPT
1Y1r2 1Y4 6.00 152 4 o.75 4.75 3.62 24.OO 610 3.44 3/t
1t/2f,Y2 1Y2 7.@ 178 0.75 4.12 o.75 25.00 635 3/q
3x4 9.00 28 8 0.75 7.50 6.19 1.00 26.00 660 4.44 '|
4x6 11.00 279 6 0.88 9.50 8.50 't,12 26.50 673 4.31
,l
8x10 I 16.00 406 12 1.00 14.25 12.75 1.38 u.25 470 4.8 1Y1
All dimensions in inches (ins.) unless otherwise noted.
'Flanges ol higher pressure class and other facings available.
carried in stock, but larger sizes can be specially fabri- what simpler than steam-traced systems, because steam
cated. When a jacketed system is selected, a careful traps and condensate return lines are unnecessary. How-
stress analysis should be made to ensure that the system ever, hot oil can be expensive and if there is ample auxil-
is not overstressed. (Chapter 2 covers such stress analy- iary steam available for tracing, steam is favorable for
ses.) moderate- to low-temperature systems. When there is
much piping to be traced, steam at the available tempera-
ture and pressure may condense into hot water before
TRACIilG PIPING SYSTEMS tracing the entire system. For these situations, only hot
oil can be used. Thus, hot oil is used in tracing applica-
When process fluids have low to moderate values of tions where steam is either not practical or not available.
viscosity 1g 4,500 cp), it is best to trace them with tubes There are many types of hot oils marketed by various
containing hot or cold fluids. The tracing can be done chemical companies as heat transfer fluids.
with or without heat transfer cement around the tracer It is most desirable and should be mandatory to use
tubes (Figure 3-2). We will consider two methods for heat transfer cement in tracing tubes on process piping,
analyzing both systems. because it provides more heat transfer area. Heat trans-
Usually, steam or hot oil is used to trace systems. Hot fer cements are available in all major countries and in
oil is used when the fluid to be traced is hotter than satu- some of the larger Third World countries. However
rated steam at typical operating pressures, which would there are times of expediency in which traced systems
be about 350'F and above. Hot oil systems are some- must be installed without the cement.
Heat Transfer in Piping and Equipment 1O7
1 T]
(3-2)
u. kt h.
- L)
(v,xA")(t" (3-3)
a: (hJ(AJ(n)(tt
Q: ) (r)(Qr)
- t") (34)
(3-5)
Figure 3-2, Various traced pipe configurations: (A) single where Ao : outside insulation surface area, ft2lft
traced pipe, with tracer under pipe, with heat transfer cement A, : outside surface area of tracer tube, ftzlft
(HTC); (B) process pipe with two tracers with HTC; (C) one h, : convection film coefficient from tracer or heat
tracer on top ofprocess pipe with HTC; (D) process pipe with transfer cement (HTC) to air space,
three tracers with HTC; (E) jacketed pipe. Btu/hr-ft -'F
108 Mechanical Design of Process Systems
b 1.o
.c o.9
o.a
o.7
9
a:
7
35710 152030 50
OUTSIOE DIAMETER OF INSULATION IINCHESI
Figure 3-5. Heat transfer outside
horizontal pipes. ho=COMBINED OIJTSIDE HEAT TRANSFER FILM COEFFTCIENT
Heat Transfer in Piping and Equipment 109
Qr : heat transfer per lineal foot from air space to vessel wall temperature. The contribution of radiation
atrnosphere. Btu/hr-ft from the tracer and pipe or vessel to the inside wall ofthe
Qz : heat transfer per lineal foot from tracer to air insulation is negligible, as is the film resistance to heat
space, Btu/hr-ft transfer on the inside insulation wall.
L : temperature of outside air, oFoF The procedure for tracer design with heat transfer ce-
ti = temperature of tracer fluid, ment is as follows:
"y : safety factor; 1.3 for piping systems without
HTC, 2.0 for piping systems with HTC, 1.5
for vessels without HTC, 2.5 for vessels with 1. Determine the scheme of tracers to be applied us-
HTC ing Figure 3-7.
2. Calculate the metal wall area (equals wall thick-
lf ta > ti, then the system is adequate. The maximum ness) A*; the outside surface area of insulation, Ao,
spacing of tracer tubes for cylindrical vessels is calcu- the outside surface area of pipe, Ao, and the outside
lated in the same manner except that a flat plate approxi- surface area of tracer tube or heat tfansfer cement
mation (T. = t) is used to compute the heat losses, or Q (Hrc).
values. 3 .
Assume a value of the minimum pipe wall tempera-
ture, to, equal to or greater than the minimum pro-
cess f luid temp€rature.
Traced Piping Wlth Heat Transfer Cement 4. Assume a value of air space temperature, ta.
5. Estimate the natural convection coefficient, h",
One mode of heat transfer in a system with heat trans- from the HTC to air space.
fer cement is conduction from the tracer tube through the 6. Calculate T" using Equation 3-1.
pipe or vessel wall to the point of the wall most distant 7. Determine the outside film coefficient of the insu-
from the tracer. The thermal distribution of such a sys- lation to the atmosphere, h., from Figure 3-5 and
tem is shown in Figure 3-6. The other mode of heat calculate Uo from Equation 3-2.
transfer is the natural convection from the tracel and the 8. Calculate the average pipe wall temperature tp and
pipe or vessel wall to the air space. Thus, the air space estimate t}le natural convection coefficient from the
temperature is lower than the minimum process pipe or pipe or vessel to air space, ho, from Figure 3-4.
9. Formulate a heat balance around the pipe or vessel temperature of air space, 'F
wall and air space and perform an iteration analysis ambient temperature,'F
solving for t" and te with the following steps: length of heat flow through metal, ft
pipe temperature at point nearest tracer, 'F
Qr: (u.)(,\)(r" - t") (3-6)
pipe temperature at point farthest from tracer.
.F
Qz: (hJ(At)(tt
- t") (3-7)
(hPxApxtp * t") (3-8)
Likewise for traced systems with HTC, for traced ves-
sels, the maximum tracer tube spacing for traced vessels
Qn : (2Xq) (I*)o.u,, * ,,,' (3-e) is calculated by the same procedure, except that the flat
plate approximation (Te : t) is used to compute the heat
Q+) (r)(Q:) (3-10) losses, or Q values.
Qz+ Q:)Qr (3-l l)
: Condensate Return
where Am cross-sectional pipe wall area (equals pipe
thickness), ft,/ft Steam differs from hot oil in that condensate is formed
Ap = outside surface area of pipe, ft /ft
by loss of heat energy. During energy shortages, the use
ho = convection film coefficient from pipe to air
of condensate return lines is normally justified. Consid-
space, Btu/hr-ft2-oF
k : thermal conductivity of vessel shell material, ering the use of 1/2-in. tracers, normally a l-in. conden-
Btu/hr-ft2-'F sate subheader will handle condensate from 2-8 tracers,
n, : number of tracers, dimensionless a ltlz-in. header from 9-20 tracers, and a 2-in. sub-
Q: = heat transf€r per lineal foot from pipe to air header from 21-50 tracers. With a condensate collection
space, Btu/hr-ft and return system the steam supply pressure should be at
Q4 : heat transfer per lineal foot from tracer to least 100 psig. Even though these rules of thumb are well
pipe, Btu/hr-ft tested in field practice, the reader is encouraged to calcu-
Heat Transfer in Piping and Equipment 11 1
late the condensate load for his particular needs. Con- where hr8 : enthalpy of vaporization (also called latent
sider the following analysis: heat of vaporization), Btu/lb
Total heat loss from steam tracer:Qr The steam in the tracer is assumed to transfer energy
For systems with HTC (by adding Equations 3-'7 and 3-9), as heat for a given mass of steam under constant pres-
sure. A typical condensate return system is shown in
Qr: Qq+Qz (3-12) Figure 3-8.
When collecting condensate, care must be taken to
For systems without HTc, prevent water harnmer caused by the mixing of conden-
sate at different temperatures and pressures. To prevent
Qr:Q: (34)
water hammer in condensate systems, spargers and
The steam in the tracer is assumed to enter the system steam separation kegs should be considered.
as saturated steam at an initial temperature or pressure. To size the condensate return lines, as well as the trac-
ers themselves, use the methods presented in Chapter I
Considering the amount ofheat loss over a given temper-
ature range, the condensate load from n tracers on a for line sizing. In systems where a large quantity of con-
given process pipe is densate is formed by steam flashing, a condensate return
pump may be required. Normally, condensate return
,ir : $,
nnfc
rum (3-13)
pumps are the horizontal centrifugal type. Pumps and
their applications are presented in Chapter 6.
STEAM SUB
HOR.
E
COND.
HOR.
SEEDETAIL A(TYP.)
SEE FIGURE 3
COND HOR.
SEE
OETAIL A
(wP.)
OETAIL-
Figure 3-9 illustrates details ofjacketed pipe. Forjack- For an annulus, the hydraulic radius, Rs,
eted systems, it is customary to assume a temperature
drop over a given length of pipe for hot oil. In applica, Dr-i - d.o
Rs (3-18)
tions of hot oil heating a viscous fluid such as asphalt,
100'F drop per 100 ft, or I 'F per foot, is quite common.
If one is not familiar with a given service, then a heat in which
balance must be made, like those done for tracers, How-
ever, using a temperature drop over a given lenglh of D:4Rs (3-19)
pipe simplifies the analysis and has been proven in prac-
tice, because all examples cited are from actual, success- ._ NC"
(3-20)
ful operating systems. The following steps illustrate one
such method of designing jacketed pipe:
l. Compute the overall heat transfer coefficient, U, and thus
by the following relation:
Nru"k
' .D- (3-2r)
u = Er*r+ln(ry'r'* ll-r (3-14)
kzr
[h': h
9d LR
PE(rcESS IINE
v2t. ss n BING
JACKETIJI
PFOCESS LITIE
rrNE srzE prJ li.L
ll
The LMTD is solved using the following formula: To facilitate manual calculations refer to Figure 3-10 [1].
The concept of the logarithmic mean temperature differ-
_ (GTTp) - (UrTp)
/1-r1r ence is widely explained in most basic engineering text-
, /crro\
]n l-l books, so its explanation will not be presented here. The
\rjrTD / reader is referred to Kern [2] or Ludwig [3] for a formal
description of the significance of the LMTD.
where LMTD : logarithmic mean temperarure difference
GTTD : greater terminal temperature difference 3. Once the amount of heat transferred is determined
LTTD : lesser terminal temDerature difference from Equation 3-22. assuming a given temperature
to,4 40
i5
tll E
o F
t
E
Figure 3-10. LMTD chart. (O 1978 by Tubular Exchanger Manufacturers Association. Repdnted by permission.)
Heat Transfer in Piping and Equipment 1 15
drop, the amount of flow rate ofthe heating fluid is jacket (outside the tank) and exits through another side
determined by heating the vessel's contents. This can be seen in Figure
)-tz-
q : rirCoAt (3-24) Before we analyze in detail these various components,
we must first look at the overall heat reouirements of the
where rir : hot fluid flow rate, lb/hr
vessel to determine how much heating surface is re-
: specific heat of hot fluid, Btu/lb-"F
At : hot fluid temperature drop
quired. The controlling criterion in determining the
amount of heating panel surface area of a vessel is the
From this formulation we determine the flow rate re- transient state, i.e., how much surface area is required to
quired. heat a given mass of fluid of specified properties to a
specified temperature within a specified time. Figure
4. Using Figure 3-11, the amount of pressure drop in 3-13 illustrates a control mass inside a oressure vessel.
the annulus is determined and added to the Dressure Consider two transient boundary conditions in the ves-
drop in the whole sysrem (which includes the pip- sel-the fluid resting at steady state and the fluid moving
ing connecting the annuli). The pressure drop for through the tank at a given mass flow rate. Thus, the fol-
the piping other than the annuli is determined by lowing two criteria must be established before the heat
using the methods presented in Chapter 1. Chapter transfer area required for a process vessel can be deter-
6 shows how to select and size the pumps to handle mmeo:
fluids that usually require jacketed services, such
as hot oil. 1. A vessel shown in Figure 3-13(a) contains a static
fluid of X gallons at an initial temperature, Y'F.
Once the flow rate is determined, the hydraulic analy- How many degrees of temperature per hour will
sis made, and the pressure drop judged adequate for the the fluid mass rise for a given surface area of
size ofpumps selected, the jacketed system details can be clamped-on jacketed coils?
designed. 2. Using clamped-on jacketed coils shown in Figure
Typical jacketed piping components are depicted in 3-13@), how many degrees of temperature per
Figure 3-9. In extensively jacketed systems, valves can hour will be transferred to a given mass of fluid of
be procured that have jacketed spaces built in. These defined properties flowing through the vessel at a
types of valves are recommended for services where constant mass flow rate with an initial temoerature
jacketed pipe is required (p > 5,000 cp). Some of these of Y'F
valves are shown in Fieure 3-9.
These two criteria are established bv considerins the
following relationships:
and
Systems that require piping to be either traced or jack-
eted likewise require similar components for vessels.
Q : UA(LMTD) (3-26)
The complexity of traced components depends on the
viscosity of the process fluids being handled. For high- Equating Equation 3-25 to 3-26, we obtain
viscosity, non-Newtonian fluids special items must be
added to vessels, such as agitators that are composed of UA(LMTD)
blades and usually powered by electric motors. There At : (3-27)
mCp
are many reasons to use agitators, and one of the most
common is to keep suspended particles in a non-Newto- The U value, or overall heat transfer coefficient, is
nian fluid evenly distributed to prevent particle settle- calculated on the basis of whether the panel of heat trac-
ment on the tank bottoms. ing tubes are clamped on outside the vessel or located in-
There are two basic types of heating and cooling de- side the vessel. These overall heat transfer U values are
vices used for vessels-internal and external iackets that determined through extensive laboratory tests and accu-
fit on the inside and outside of the vessel, respectively. mulated field experience. The U value used in calcula-
These jacket types are shown in Figure 3-12. The hot tions should be that recommended by the heat transfer
fluid (normally steam or hot oil) enters one side of the panel manufacturer, as various panel designs are avail-
coil and flows through the baffle (inside the tank) or able and the calculation of the U value analvticallv can
116 Mechanical Design of Process Systems
tt;
A P-inchesoluater I
Flgure 3-11A. Engineering data pressure drop through jacket. (Courtesy of Parks-Cramer Co.)
't
dl
o
=
Figure 3-118. Engineering data pressure drop through jacket. (Courtesy of Parks-Cramer Co.)
I
1 18 Mechanical Design of Process Systems
f,
fi
{.
1" x 2" Sch.40 Jackeled Pipe
curyes based on 100 leel ol jack€ied pipe (livs 20{"
rengrhs) and rive t4" o.D. x .065" wall lubrng jump_
'nclude and exii losses. water @ 60'F- {16'c.}
overs plus entrance
J
I
o
=
J
g A P- inches of water
Figure 3-1|C. Engineering data pressure drop through jacket. (Courtesy of Parks-Cramer Co')
Heat Transfer in Piping and Equipment 119
d A P-psig <'
ut
o
=
A P - inches ol water 3
Figure 3-11D. Engineering data pressure drop through jacket. (Courtesy of Parks-Cramer Co.)
120 Mechanical Design of Process Systems
AP-psig
t
'to
=
J
Figure 3-11E. Engineering data pressure drop through jacket. (Courtesy of Parks-Cramer Co.)
Heat Transfer in Piping and Equipment 12'l
lD
I
Figure 3-11F. Engineering data pressure drop through jacket. (Courtesy of Parks-Cramer Co.)
122 Mechanical Design of Process Svstems
=
J
A P - inches ot water
Figure 3-11G. Engineering data pressure drop through jacket. (Courtesy of Parks-Cramer Co.)
Heat Transfer in Piping and Equipment 123
d A P - psig o
Figure 3-11H. Engineering data pressure drop through jacket. (Courtesy of Parks-Cramer Co.)
124 Mechanical Desisn of Process Svstems
o
=
J
A P - inches ol waler q
Figure 3-1 11. Engineering data pressure drop through jacket. (Courtesy of Parks-Cramer Co.)
Heat Transfer in Piping and Equipment 125
o
=
A P - inches ol wate. q
Figure 3-11J. Engineering data pressure drop through jacket. (Courtesy of parks-Cramer Co.)
Mechanical Desisn of Process Svstems
AP-psig
t
-
,LENT 7
)aa:
-:,-: j: o(
7.,
o(
E,
o-
(ra
lo
a
=, =
7 '::
AM tN/ \R
::::l::::l: ::::l::.1:.::1.::t,:
Park :s -l ,fal ne r:
JACKETET NG SYST :l\4S t COMPONENTS
A P - inches ol water q
Figure 3-11K. Engineering data pressure drop through jacket. (Courtesy of Parks-Cramer Co.)
Heat Transfer in Piping and Equipment 127
AP-psig
-/;;
!r:::r'
JT T FTOW i,? ,/..:
i:,1-,:
.1-=: F=
h-0b, t1=
/. / I l::::
F= t.-
t:
lv c
=,
(r. )1
@
o,
= 7/. o
=
/
ul
L AMINAR
A I
:l ':r:i::::l :::l't
Pe k s-l fame I I
,l
i. coMP ONEI {TS
!l :
:;::l ::l :::.[]
.. .t.._
...1..1 . . .1.-.
A P - inches ol water q
Figure 3-1 1 L. Engineering data pressure drop through jacket. (Courtesy of parks_Cramer Co. )
128 Mechanical Design of Process Systems
Figure 3-12A. Heat transfer internal plate or panel baffles inside a vessel. (Courtesy of Tranter, Inc.)
Heat Transt'er in Piping and Equipment 129
,/ w*\ \ \
/ _-.--\r\1
'ffiFt*
,(?,''l
\t\
\-*'/
e
C=2%" MlN.
l[ [ l\ l\
I I
D ----v
3RD. S€T OF CHANNETS
HEADER SIZE
USED WHEN B DIM.
EXCEEDS 7T'
OVER,AI.I WIDTH
I(NO. OF PLATECOTL-|) (CJ.3',]
Figure 3-128. Schematic depicting how heat transfer panel plates heat up or cool down process fluid in tank. (Courtesy of Tran-
ter. Inc.)
ffi N
Figure 3-12C. Heat transfer panel plates designed to fit on
curved surfaces. (Courtesy of Tranter, Inc.)
Mechanical Design of Process Systems
Figure 3-12D. Vessels with typical external heat transfer plate panels. (Courtesy of Tranter, Inc.)
Heat Transfer in Piping and Equipment 131
-'-- -*-i
fluid
x gallons
v'F
oF/min
Figure 3-13. Two schemes in which the heat transferred must be considered: (A) conrol rnass scheme; (B) control volume
scheme.
132 Mechanical Design of Process Systems
HEAT TRANSFER IN RESIDUAL SVSTETyIS and convection are the controlling modes of heat trans-
fer. The convection can either be considered as natural
or free convection, or in the case of wind, forced con-
lleat Transfer Through Gylindrical Shells vection. It has been found that using the free convection
coefficient is the most desirable in many cases, since
Heat transfer through pipe supports, vessel skirts, and vessels are normally surrounded by other equipment and
empty branch piping connections to hot or cold headers structures, making free convection more applicable.
can cpuse critical stress problems as well as damage to Assume that the temperature inside the skirt is the
equipment. Excessive thermal deflections can result in same as ambient temperature and wind chill factors are
unacceptable loads on rotating equipment and vessel not present. Air seepage under the skirt and open aper-
nozzles. In cryogenic service, vessel skirts can fail by tures on the shell allow for equilibrium to be established
brittle fracture if the transition temperature point be- with the outside temperature.
tween alloy steels and carbon steel is not considered. The first step is to determine the free convection film
This section discusses the analysis procedures for ana- coefficient for the outside surface of the oressure vessel
lyzing heat transfer in such residual components as ves- skirt insulation. In normal conditions. the air temoera-
sel skirts and pipe supports. The methods used have been ture inside the vessel skirt. ti. is assumed five degrees
tested with empirical data and have been used for several lower than the outside ambient, ts. The free convection
years in design practice. For derivations to the following film coefficient is found by iteration using the following
method on heat transfer through cylinders, the reader is equatrons:
referred to the author's paper [4].
Vessel skirts are normally insulated on the inside and ,, - [r, ln(r2lrr) -, qlnG!lt2) ,
@l = insulation
flll = metal
.f
Figure 3-14. Vessel skirt insulation detail. Sometimes the in-
side insulation is left off.
Heat Transfer in Piping and Equipment 133
^- /2"r,.,\ [ r l I
tltt tr"'rrl -
r 1-3'll
" \U,q,/ l" (t ,ttj
^/ \[
z - ll:+l lroho,ttn . t5) - k2',
l\KmAny' [
[ ,, - (t/'i)lll
,
tl" {r./rr) I"
l]L (3-34)
z : ztQ (3-35)
proportional to the insulation thickness. The air around Figure3-15. (A) Stub piece used as header support: (B) pro-
the outside insulation surface is in a state of local turbu- cess line is connected to a turboexpander. The line is sup-
lence and for this reason the variance of the Grashof ported by a short section of pipe welded to a base plate; (C)
number down the outside insulation wall is insignificant. branch line from a header (hot or cold) connected through a
Experimental measurements confirm this fact. The shut-off valve to a ftagile piece of process equipment.
reader will see in Examples 3-7 and 3-8 how to apply this
method to vessel skirts.
Piping that is supported by piping sections is treated in header. If the process header is in cryogenic service, the
a similar manner to vessel skirts. Such piping supports stub piece must be analyzed to assure the design engineer
are shown in Figure 3-15 in which the pipe supports and that the carbon steel structural members are adequately
branch lines are subject to thermal gradients from a hot protected from temperatures below the transition tem-
or cold process header. Figure 3-15a shows a stub piece perature.
used as a piping header support. The temperature gradi- Shown in Figure 3-l5b is a common situation in which
ent through the stub piece must be analyzed to determine a process line connected to a turboexpander is supported
if the Teflon slide beneath the base plate will be pro- by a section of pipe welded to a base plate. If the pipe
tected from the elevated temperature inside the process stub deflects enough (shown by 61), the thermal deflec-
134 Mechanical Design of Process Systems
tion could induce a sufficient bending moment on the where c and m are determined as previously for skirts
turbine to cause serious mechanical damage.
Figure 3-15c shows a branch line running from a hot h,j = (k"r.Nr")/d (3-32)
or cold pipe header to a fragile piece of equipment. Even
though the valve on the branch line is closed, the residual tj : - t.) + t"
(u3/h;) (ti (3-29)
temperature distribution through the branch line may be
enough to cause the pipe to deflect and damage the
Atj : t3 - ti < 2'F (3-39)
equlpment.
Once Atj criterion is met, we can proceed with the fi-
Referring to Figure 3-16, the procedure for determin-
nal iterative value for the film coefficient, h.. With this
ing the temperature distribution through the empty
final value. we solve lor the parameters Q, Z. and Z as
branch pipe or pipe support is similar to the case of a
follows:
vessel skirt. First, solve for the free convection film co-
efficient on the exterior surface of the pipe insulation. To 2nk1
do this, use the equation for the overall heat transfer co- ^ (3-40)
efficient:
;"{$
," (;) ,, ," (,:) t / r- \ r
- kJ,
t't - ---tT
ur: 1,,
- r( -n;/]
,
(3-37) (3-41)
\ k- l--^J lrrtr"tt:
rn lll
\r,
t
(2.496 x ro ,z'L
t0-
(6.536,l0-?)
(109 - zr
{rr,,
't---
..
-------l1
[sech(LQos.; tanh(LQo + arctan [sinhtLQo'y]l
- 2f s
- 87r, / e:Loo i
,a"-- \l + etLao
5i
5
(106)Qo Figure 3-17. Pipe supported on a shoe.
136 Mechanical Design of Process Systems
u.
rr i-1 n-l
t___
I
olo I I
ol I
o
3
H_-
'{"1 ltll!'1-
*---l
'-
A-^ : o 216
12
: o.ol8 fr: t' - j1
2n,
EXAIIIPLE 3-2: HOT OIL TRACING DESIGN 0'690 ft :
- 2(1.0)
0.345 ft
A 3-in. schedule header contains asphalt which is to be
maintained at least to 445'F. The 3-in. header is to be
traced with hot oil (Ce : 0.50 Btu/lb-'R p = 58.7 lbl hr" = 0.33[ffiu'.*r,,J: r't,
ft3 at 475"F). Determine the size and number of hot oil
tracers required to maintain the asphalt at a minimum tr" : hr" + eh,:3.992 + (0.90)(1.185) : 5.059
temperature of 450'F. For asphalt, Cp : 0.368 Btu/lb-
"F at 500"F. 1-0203+ I = u^:0.449
For most applications, l/z-in. copper tubing is the stan- u" 0.1 5.059
dard size for tracing operations. We select a l/z-in. 18
BWG gauge steel tube, At : 0.131 ftlft, k^ : 27.5 Now performing a heat balance we have
Btu-ft/hr-ft2-'n First we will try one tracer,
t, = 350'F and tn, : 490"F. Using Equations 3-6 through
Di : 3.50 + 0.50 : 4.00 in. : 0.333 ft 3-11 with 70'F ambient,
138 Mechanical Design of Process Systems
t)1 \\
h,":033[sffiffi,,t:,,,,
q^ = t2) l-::-:l (0.018)(550 - 450) = 286.957 Btu/hr-ft
\u.J4)/ h- : hr. + eh, : 1.996 + (0.90X1.185) : 3.063
/rt.\
r,": o.s({q::g)'" = 1.383 9a = t2\Qt l '1j". | (0.018X550
1d/
- 490) : 544.954 Btu,/h.:
\u.z
Heat Transfer in Piping and Equipment 139
t1'7 al
q. - (2x2) l^'-:j" l(0.018)(550 - 500) = 454.128 Btu/hr-ft
^, _
VDP
\u.l r6/
qr>2qandq2+q3>qr (j.781) I t4.026)in. Ull,n.7 '- lb 160h,secl
Nn"
sec 1,, '" ""ttr \ /
Therefore, the system is balanced.
For 100 ft of pipe, b'rL\
9r : (100)f(454.128) Btu/hr-ft = 45,412.8 Btu/hr
{e38 08)ce l\ ill
*p i
or, Nn" = 3.242 << 2,100 + Laminar flow
q : 22,706.422 Btu/hr for each tracer For laminar flow, we have [2]
ofChapter 1 and size pumps to handle the hot oil. (Chap- nr Btu
ter 6 discusses how to select the pumps required to dis- -^ hr-fr-'F
tribute the oil in the system.) : 7 ,654.733
/ rr'\
EXAMPLE 3-3: JACKETED PIPE DESIGN D - 4 026 = 0 336: L= loo rt
'" t,-rJ
A jacketed pipe shown in Figure 3-1b is to be ana- '
r (7.654.733)r
lyzed. The process fluid to be heated is a film coating hr., = 1.86(3.242)r ' (H)
mixture used in the manufacture of roofins tiles.
1.0
'(#)(10)#.F
ra ln (ry'rr)
f',, = 2.415-T-
ra , 1 nr-rt'- - f
ht z kz,: h: +
For hot oil in annulus, Di/D. : 0.664 > 0.2, and Perry
For film coating in inner tube, [6] recommends
: : q : 0.75t Rr''
"* ( 7.80)ftr(72.135).F
D 4Rn 1.566 in. = 0.131 ft "ts
It-nr-
''
Nr" : (4o.ro7r ]9:99:
(0.1s)( L566)
- ),0n.24r q : ricpat
Err'
(29.121){0.071) ntu th
1^ _ NN,k hr-ft'/-"F/ft
D 0.131 ft ftr
: 1 l-'
-, 15r$l
(3.033) ln (2.2so/2.0r3)
rr
- |13.033) + lh Rr"
L 2.4rs I- = (235.428) _: (0.5) _= (t00)oF
hr lb-'F '
Ri,r or
ft']-hr-'F q: I1,771.400 Btu/hr
A: heat transfer area : outside surface area of inner Thus, 0.5 gpm is a sufficient flow rate to transfer the re-
tube quired heat to the film coating mixture.
A- 1.178 fPlft : 117.800 ft, for 100 ft ofpipe
(9.52x35.667)(LMTD) + (4.0X377.078XLMTD)
-'^, _
- (?6,110 ooxo 34l
at:7.410'F/hr
Referring to Figure 3-19 we can now determine how
COATING MIXTUR€ AT
TEMPERATURE t
many degrees per hour the fill coating mix will rise using
external clamp-on jacketing on surfaces of the flanged
Figure 3-19. Coating surge tank. and dished head, the vessel shell, and four internal pan-
els just considered:
min oal lb
Q
- : (60) :hr" (150)'min
":' (12.82)'gal
-
coating mixture is flowing through the tank at 150 gpm x 0.34 Btu (t - 360)'F
ar 360"F.
lb-'F'
From Figure 3-10, we have Q : 39,229.20(t - 360) Btu/hr
Now,
LMTD : (550-360)-(450-4oo) :104.869'F
, [sso - :ool UA(LMTD) : 39,229.20(t - 360)
'" t4so - 4ool
with Q: mceAt (9.s2)(35.667)(LMrD) + (4.0)(377 .U 8)(LMrD)
: 39,229.20(t' - 360)
and Q : UA(LMTD)
1,847.862(LMTD) = 39,229.2O(t' - 360)
, UA(LMTD)
LMTD :
DLp
2L.230(t' - 360)
Now,
th
m - {6.000)gal(12.82) : 76.920.00 lb
gal
= (550-360)-(450-r') 21.23(t' - 360)
Using heat transfer panels shown in Figure 3-12 we 1550 -
|'"t450-tl 3601
-
142 Mechanical Design of Process Systems
Solving for t,
wP: wetted Perimeter : (--.Jo*t*
(21.23r' - 1 ,642.80)
ln (450 - r')
(110.394t', - 39,841.956)
WP:
* ,642.80)
Letting y _
(21.23t' "t
(l 10.394t'
- 39,841.956) A : cross-sectional f low area : 0,68 I in.'? : 4a(W)
OI in the plates)
K23: 29 BTU/hr-t12-"F/fl
From Kern [],
ri :0.44in =O.O37 tt thn
- : = ro.o27r ([) ,t-.,., ,N,,,' ' (uJ"'" r. =0.545 in = O.O45 tt
Btu
I /tl-
'hr-ft2-'F/f
r. = {0.027)
ro.aozrin. llrt
\l2 in.i
I
rtt = -
U.U9
i r\ir-:r.r.o\."
l-ll "I /c-J''
l!l Approximated as Circular Tubes
\DJ \ p / \k/ Figure 3-20. Cross section of panel plate tube approximated
as a cylinder whose surface area is equivalent to the heat trans-
p: 95.909 lb/ft3
k= 0.1 Btu/hr-ft-'F
k: 0.071 Btu/hr-ft2-.F/ft or
r-) +:r't:'log.o
h,q - 0 0e (oo
['s
'qs
eoqJ'^'
lT -
where W : werred perimeter (WP) The baffle plate area for all four baffles is determined
from the baffle plate manufacturer's dimensions, as fol-
sr lows:
LJ | = '2 D" + r' . n = number of flow circuits
Length of channel per baffle = 51.'123 ft
To account for the residual heat transferred through the
plate connecting the hexagon tubes we consider to be r' A= surface area - 2210.431)i" {.:-tf) rsl.72t rr)
the equivalent radius of a cylinder that is the total surface \rzln./
area of the baffle plate separating the hexagon tubes.
Now,
A: lt.672 fe
For all four baffle plates,
D, : ,o,u, distance between channels (Frgure 3-21)
A= 4(tr.672) ft2 : 46.690 ft2
Zor' = Dt
It Heat Duty of lnternal Baffle Plates
2tr
For hot oil the anticipated temperature through each
From above, plate is 100'R as stated earlier, thus
\.r q:
q= UA. {f I - t1) - 2trlJ lzr 1t, - ta1 UAAT
2T(tr Rr',
- ta) n q - (29.492J, "-1:1, (46.690)ft, (100).F
r" r" ln (r./r) I nr-It'- -f
"-T-
finl-2 Kt I L
--f
q : 137,698,1a8 f.!
nr
in which
(0.360) lL
tt-hr
Nn" = 138,150.85
w= 1.375 in
1.,1o. : PCo
= 2.535
k
w = eflective heat transfer area
For approximalion, analyze the ligure as a tube with a
h,, : (0.027)
(H1.ol ,rrr,rro r5)0s (2.535),/, (1.0)
circumference = w
2Tr j = 'l.375
in which, Btu
rz = O.219 in., or D2 = 0.438 in. h',,- = 77.260
hr-ft2-'F
Equivalent llow,
Perry [6] gives the correlation for heat transfer for jack-
13 1341/|) 111
= 6.448 fusec eted walls to the agitated liquid as follows:
14,400 lb/fi-hr
speciflc heat of coating mix
.-
u=l t
(0.703 )
+
(0.73) ln (0.328/0.219)
substituting above values into Equation 3-49 we have the A' : 257 .249 ftl
following:
Ar = 100"F for hot oil-coaring mix servicc
Rr'
(s.42t(3,360.0X9s.909 q - {8.14lr .nr-It'- '
:i-_r25t.24o1ft (100)'F
,*,(,*)[ t -
(14,400) l- or
, :ltr+,+ool]"'
,o.ur',o,o
lto Rr,'
q= 209 .414.44 --:
Btu nr
h,,.
' = 14.060
hr-fC-'F
Thus,
II _
Heat Duty of Jacket Plates Clamped to
Bottom Vessel Head
The bottom head panel sections are depicted in Figure
l-17. In Chapter 1, Example 1-2 we analyzed the hy-
draulics for the hot oil flow through the panels mounted
on the tank. From this analysis we determined the fol-
lowing velocities required to obtain l0 psi pressure drop
Table 3-1 through the panels:
Values of Constants for Equation 3-49
innerpanel : V: 7.315 fi/sec
Agitator
Feynolds Number
Range
outerpanel : V: 5.237 ft/sec
Disk, flat-blade
turbine 0.54 2lz 0.14 40<NR.<3x105 Heat Duty for Bottom Head Inner Panels
Pitched-blade
turblne 0.53 2lt 0.24 80<NR"<200 Similar to the shell panel plates above, we must com-
Propeller 0.54 ,/3 0.14 Nn" = 2000 pute the equivalent tube diameter and equivalent veloc-
Paddle 0.36 2h 0.21 3oo<NR"<3 ity. As determined above the equivalent radius is
x 105
Helical ribbon 0.633 Vz 0. l8 8<NR.<105 rr : 0.219 in.
Heat Transfer in Piping and Equipment 147
S.lce the bottom baffles have the same flow area as the
.rell plates, the cross flow area of the equivalent tube is
q' - uA at = (8.s90). lr)--
hr-ftr-'F
1r.+r+)n, lroo;"r
(7.315) -:'
sec
(0.31)in.2
(5.237rt0.31
-quivalent velocity = Y. =
0.151
tcutvatent \ etoclt\ - r
lu. /f I lt/sec
in.'z 0. 151
/r\ r58.7 rr3.600
{ 10.75l,(0.438r l-l )
Nn.
0.160
\- 322,453.78
/^ ^--\
h,2
'- - {0.027) lfffl rz.lo..lss.:41)0b r2.s35) '(1.01
\0.438/
-.. .: (0.027) l9J11l r:zz.+s:.28)0" {2.535,r 'r { 1.0)
\u.4J6/ h,, = 1t6.303 ,nr-lt'-'f
:tu^=
Btu
:..:152.2ll Similarly as for inner panels,
hr-fP-"F
The vessel-side film coefficient is the same as for the f'n, : .l.l+,SlO
--!$-
panels, nr-rI'--t
'hell
Thus,
-. : t+.gt+hr-ftr-'F
Btu
i.r.
Thus, r (t.trr:, rn lo
34'l
,,
" | 0.703 _ \0.219/
OI
1 = (18$.231) in. (1.375) in. (2) : 1,067.635 in.2
OI A= 38.'144 ft2 for all four outer panels
Total Heat Duty of Tank Thus, the minimum hot oil flow rate in pipe header sup-
plying the total hot oil to surge tank is 10 gpm, the actual
At maximum flow rates the total heat duty is as fol- flow rate is 16 gpm.
lows:
q= Btu Btu
137,698.148
hr
,
209,414.44
hr
EXAIIPLE 3.6: TRANSIENT AND STATIC
HEAT TRANSFER DESIGN
internal shell side
panels panels Roofing shingles are made by adding asphalt, filler
material, granules, talc and adhesives to a plastic-glass
Btu
+ 6,368.11 + 24,772.333
Btu sheet, which is the basic component of the roof shingle.
hr hr The process is shown schematically in Figure 3-23.
Granules are added last, after adhesives and talc. The
two nner four outer sheet must be cooled so that workers can handle it with
panels on panels on
gloves. Cooling is accomplished by water sprays, circu-
bottom head bottom head
lating water through the rollers, and using radiant heat
transfer to the surroundings. The sheet, once cooled to
q= 378,253.631 P!! the desired temperature, is cut into specified dimensions
hr
by mechanical cutters and then packaged into boxes for
From Example 3-4 the total heat duty required is shipment.
There are two aspects to this problem-static heat
transfer and transient heat transfer. First, we solve the
q-t = 23g,832.rc28:
nr static conditions and then the transient case to determine
how fast the sheet can be cooled with the coolins svstem
Now, designed in the static case.
7,s65.073 P
4.", : nr /r.+a gur\ Table 3-2
]
|I
FINAL
COMPONENT
For convection, For 600 shingles/hr (or 144,000 Btu/hr) the heat removal
would be : (144,000X0.9375)(O.299)(313.63 - 125)
Q: hAAt
: 7.614 mm Btu/hr
For flowing air, h,in = 2 Btu/hrlft2/.F
h."^ : 50 Btu/hr/ftrl.F
removal>
Thus. Total heat Heat removal
requtled
Use h. = 23 Btu/hrtfP l"F
and the cooling system is adequate.
/6R {r
";:-'(l)
A = s.708 fr2 For vaporization, q: 29 Btu/ft,
t2
104.40 g
mtn
h, : 1.857 Btu/hr/ft,/"F
gpm =
th
= 12.518 gpm
8.34:
gal
Thus, the water pump to be used is to be sized for 13
gpm at a terminal exit pressure of 200 psi.
Total Heat Removal
hr : :
h" + h. + Transient Heat Transfer Analysas
23.000 1.857 = 24.857 Btuthttfet"F
Q = h1A(At) : (24.8s7X5.708) (223.63) = 3r,729.464 This method is based on the Fourier analysis of un-
Btu/hr steady-state heat conduction. The following assumptions
are made:
AI
Qv = 104,400.00 Btu/min Q = -k-dX
Qv : 6,264,000.00 Btu/hr
The heat transfer across x = 0 is zero and at the mid-
Total heat removal : 6,264,000 + 31,'729.464 plane of x : 0 the sheet behaves as a perfect insulator-
+ 3,000,000 an adiabatic surface. Consequently, the solution to this
Q= 9,295 ,729 .464 Btulhr problem applies to a slab that is perfectly insulated at one
Heat Transfer in Piping and Equipment 151
m/L2
Figure 3-24. Heisler's main chart for the infinite slab [7].
152 Mechanical Design of Process Systems
(o oo4l7j( | o7o' lyzed, -200"F, - 100'F, and -50'F. The skirt is made
(Nr o),"o
'a (0.094)'
0.505
of Type 3(X stainless steel and is insulated on the inside
and outside as shown in Figure 3-28. The insulation is
For a Fourier number of 0.505, sized for the most extreme process temperature that the
vessel will be exposed to, -200"F. Data used in the ex-
(*).,,-,, : oo,o ample are given in Figure 3-25.
First, determine the natural convection film coeffi-
cient for the skirt. The temperature inside the skirt, ti, is
Let t" : lequirad water temperature assumed to be five degrees lower than the ambient tem-
peratute, t5.
l)5 - r
Jl+ - t,\
uo : .,. O*4, * u'i,r:.',n) *
[', 'L(no,r iJ-'
in which t* = -11.86"F and is well below the freezing
temperature of water.
=
I +
r l-,
[7.r1s -hJ
Thus, for a cooling section of 150 feet long, the sheet
moving at 600 ftlmin cannot be cooled to 125'F since Assume h4,5 = 0.275
the theoretical value of t* is below freezing.
u4:0.093
At V, = 499 Lr-tn,
r= O.3"75 min : 0.0063 hr
(0.0063X 1.070)
{Nro),.u.r = .: ^: --j--- = 0.7568
(u.u94r
ln\
l;l = 0.180
lts - r
314 - t*
and
,. =
[*J
(tr - ts) + ts #[",('[4 - 'f5)
- -, - t, ll
:l /o.osl\l{-))+bU
- . ['-h '(;-.,Jll
\0.275i
: ra-1nl.2gsr
63sx0 27sx-1 69) - 0 14
s8.31'F l'3
t
L-t5:58.31 - 60 = -1.69"F
I sa.: r 55.00
\o, = [d37,gB( Atl)(3,6o0)2]/p,
| ,,1. Iturr\-
|ln t-t ln t-l
l(7 .27)3(0.07 6T)' Q2.2)(0.00r923)(r.69)(3,6tJ0)21 [ \3.13s/ \3.073/
(0.04339f
Z:65.1O2"F1ft2
Na,:1,613,'720,723
N", : C(Nq,Np,)-
(-515.56)(2.89)-+
= 0.129(l,148,969,155)r/3 : t. _ 57.78.1
135. 11 1 + (2.89y'
(fqi, (0.014661(l35.ll)
n.1 I.") :
' = d - 7.27
0.2125 Similarly for:
IOOqII
t, : - 100'F,
ti - l#l (-s) -60: s8.29"F
t." _
(
-515.56X2.89)- + 57.781
r + (2.89)"
: 58.31 - 58.29 = 0.02 and for t, : -50"F
:4.275 r _ (-215.56x2.89).
+ 5? ?Rl
l + (2.89t^
Figure 3-26 shows these distribution curves.
The axial deflection of the skirt will now be calculated
For a cylinder with insulation on both sides,
using the first three terms of the stainless steel deflection
equation (Equation 3-45). The hyperbolic terms in the
equation are not necessary when the steel temperature is
.:[ffH][dil.*-.19] greater than -300'F or less than 1000'F,
^
I
= t_t"ro.
z I t_ r+r I r ! rl (2.055 x
[(8Xl.2oe, li . /z.sz:\ t -f
100
l0-3\z2L
Ut : 0 '0137
4 +'74,115,250.451
Z : zlQ : 9.587'F
tx = t2 (i. - Zt.'oo1lt 1-.:roo51; + Z ^- 6". : 0.0155 ft : 0.1860 in. axial deflection
: iZ t,, - 9.587)e''0 "'7'o 5/11 + e2r'02?210s)l
+ 9.587
This example shows that residual heat through a closed
For t, : 900'F, branch line can be significant enough to cause thermal
movements, which can result in high stresses. These
|'' : I.780.83 (1.313)'r q^ \87
-. thermal deflections are particularly important when
tl + 1.313)'?1
( space is limited and the piping system has little flexibil-
Ity.
For t. : 699'P
-
r,' = 1.180.83 0.313)-
. + 9.587
lr + (l.3l3y'l
For t, : 399'P
580.83 (l.3l3l
+ 9.587
ll+ (1.313),.1 -'
Curves depicting t, are shown in Figure 3-27. Unlike
Example 3-7, the slopes of the curves change much less,
almost approaching straight lines.
Axial temperature gradients along a section of piping
produce thermal deflections. The pipe support will now
be analyzed for thermal deflections.
-
E
The surface temperature, ts, of the branch pipe at the E
point of the contact with the header is 600'F. The aver-
E
age temperature inside the pipe may be calculated from
the 600"F curve in Figure 3-27 which shows a tempera-
ture at a distance x of five feet to be 294'F.
t; = (600 +294)t2:447"F
Through the process of iteration, h. : 0.68 Btuihr-
ft-'F at the average internal temperature of 447"F. This
was obtained using the natural convection iteration tech-
nique described in Example 3-7.
Using the same techniques, Q = 0.2719 ft 2, Z :
66.7916'Flftz, and Z : 245.6476"F.
To calcuiate the axial deflection, substitute these val-
ues into the expanded thermal deflection equation for
X distanc€, ll
carbon steel, Equation 3-39. Note: Values for the arctan
used in the equation must be calculated in radians. Cal- Figure 3-27. Temperature of a branch pipe connected to a
culate the arctan m degrees and convert to radians in header through a closed valve plotted from the pipe to valve
which the relationship is 2zr radians : 360 degrees. connection every six inches for a distance of five feet in Figure
Using equation 3-39, 3-15C.
156 Mechanical Design of Process Systems
L:90'F
NOTATION
\5' = corrected convection coefficient, Btu/hr-ft2- X: distance of plotted temperature points along
the vessel skirt or piping, ft
ki = insulation conductivity, Btu/hr-ft-'F Z: heat transfer factor, 'F/ft2
ki = thermal conductivity ofair inside empty pipe, Z= heat transfer factor, ZiQ, 'F
Btu/hr-ft-"F
k, : thermal conductivity of vessel skirt or pipe,
Btu/hr-ft-'F Greek Symbols
L= length of branch pipe, ft
:
N6. Grashof number, dimensionless 0: volumeric coefficient of thermal expansion,
N51. : Nusselt number, dimensionless ,IK
Np. : Prandl number, dimensionless 6.,, 6,, : axial deflection of carbon or stainless steel
Np" : Reynolds number, dimensionless skirt or pipe, in.
Q: heat transfer factor, ft 2 ?: safety factor for traced pipe
Qr : heat transfer from air space to atmosphere, p: absolute viscosity, lbift-hr
Btuihr p: densiry, lb/ft3
Qz : heat transfer from tracer to air space, Btu/hr
Q: : heat transfer from pipe to air space, Btu/hr
Q+ : heat transfer from tracer to pipe, Btu/hr REFERENCES
t" : air space temperature, oF
t; : process fluid temperature, 'F 1. Tubular Exchanger Manufacturers Association, Stan-
tj : air temperature inside the vessel skirt, pipe dards of the Tubular Manufacturers Association
support or branch pipe, 'F QEMA), sixth edition, New York, N.Y, 1978.
t, : surface temperature of the branch pipe at con- 2. Kern, Donald, Process Heat Tiansfer, McGraw-Hill
tact point with the header, or operating tem- Book Company, 1950.
perature in a pressure vessel, "F 3. Ludwig, Ernest E., Applied Process Design for
tj( : temperalure at distance x along the vessel Chemical and Petochemical Plazls, volume 3, sec-
skirt, pipe support or branch pipe, 'F ond edition, Gulf Publishing Company, Houston,
to, t5 = ambient temperature, oF Texas, 1983.
t3, t3' : temperature and corrected temperature at OD 4. Escoe, A. Keith, "Heat Transfer in Vessels and Pip-
of the pressure vessel insulation, 'F ing," Hydrocarbon Processing," January, 1983, vol.
t4, t4' : temperature and corrected temperature at OD 62, no. l, Gulf Publishing Company, Houston,
of the pressure vessel insulation, 'F Texas.
:
At tr - t in piping example, ta - t5 in vessel 5. Chapman, Allen 8., Heat Transfer, third edition,
skirt exarnple, 'F Macmillan Publishing Company, New York, 1974.
At' : t4 - ta' in vessel skirt example, "F 6. Perry, Robert H. and Don Green, Perry's Chemical
Atj' : t3 - t3' in piping example, "F Engineers' Handbook, sixth edition, McGraw-Hill
U3 : overall heat transfer coefficient at OD of pipe Book Company, New York, 1984.
insulation, Btu/hr-ft'?-'F 7. Heisler, M. P., "Temperature Charts for Induction
U+ = overall heat transfer coefficient at 14. Btu/hr- and Constant Temperature Heating," Transactions of
ftr-'F the A.S.M.E., vol. 69 (1947), pp.227-236.
The Engineering Mechanics of
Pressure Vessels
The specifuing, design, and construction of pressure The first problem you face in designing a vessel con-
containing vessels varies all over the globe. Each taining pressure is how to physically make the compo-
adopted code that has been used for any significant nents and assemble them. In the petroleum refining in-
length of time has proven to be workable because its use dustry (CPl-Chemical Process Industry) and allied
has resulted in safe, economic designs. The main differ- industries, the most practical and economical method is
ences in codes are the theories of yield that are used for welding. We will refer to welding later in more detail,
determining maximum allowable stresses, material spec- but first we will look at the vessel from a pure engineer-
ifications. and basic procedures. ing viewpoint assuming perfect welds with given effi-
With increasing international competition and cyclic ciencies. Some have proposed bonding pressure vessels
economic conditions, there is a growing need to empha- together with glue, as is done with aircraft components.
size economics and familiarity of foreign codes, and The main disadvantages to bonding are
avoid unnecessary overdesign that relies on only one set
of codes and standards. This chapter emphasizes the op- 1. Clean surfaces are required for assembly.
timization of economics and safety. If you choose to be 2. Glues that exhibit high tensile and compressive
conservative in your design, you can be; however, if you strengths are very expensive.
are bidding in a highly competitive market, you can use 3. Chemical bonding, especially in thick-walled ves-
these methods to produce a safe, economical design. sels, takes much longer than any welding process.
International competition and economic condltions
have caused engineers to restructure their thinking that a Another form of assembly that has been even more se-
good design uses only enough material that produces a riously considered than bonding is threading components
safe and economical product. Thus, this chapter's philos- and screwing them in place. Even though this may ap-
ophy is to optimize engineering design within code rules, pear to be simple, the process becomes enormously ex-
whatever the code. Overly conservative design that re- pensive with large diameters. Thus, welding is the most
sults in excessive material use becomes unproductive practical and economical means of assembling pressure
and expensive when one is competing in the world mar- vessels for the foreseeable future.
ket today.
A thorough treatment of vessel engineering and its
concomitant aspects of static and dynamic phenomena
would fill several volumes. To present this broad subject
with clarity. various physical phenomena are briefly dis- DESIGNING FOR INTERNAL PRESSURE
cussed and references are made to sources that give de-
tailed theoretical explanations. lt is not this boo-k's pur- The two factors that must be considered in the desisn
pose to give a trearise of static and dynamic problems. for internaf pressure are crr??ponent thickness and quatiry
but rather descriptions of proven practices. The theory of weds. Before either of these two factors can be ad-
of these problems is always available, but proven solu- dressed, you must know what the vessel is to contain.
tlons are not-hence, the reason for this book. This chapter only considers gases and liquids. Vessels,
159
160 Mechanical Design of Process Systems
silos and bins containing solids are discussed in Chapter design temperature and pressure. The value of the
MAWP at the two conditions gives the exact range of
In the design for liquids under pressure, the most se- temperature and pressure that the vessel can withstand if
vere condition of coincident pressure and temperature the owner decides to use it in another application. The
expected in operation must be considered in computing reader is cautioned to consult his respective code on the
shell thickness. This is fairly universal in codes through- practice of using a vessel for another application. The
out the world. The intent ofthe statement is that the most following example illustrates how the MAVr'P is applied:
frequently occurring liquid level should be considered. An ASME Section VIII Division I vessel is made of
For example, if a vessel is filled to a certain level "A' SA 240-304 SS, design pressure : 500 psig, design
75% of the time and a higher level *8" 25% of the time, temperature : 150'F. The vessel has a shell thickness
level "A' should be used for design purposes. The nor- of 1.00 in. and a ioint efficiencv of 1.0.
mal liquid level to be used for vessel design and its quan-
titative value should be determined by the process engi- MAWP (NAC) =
(18,800) x (1.00) x (1.00.)
neer. For upset conditions each code allows an increase (21.00)+(0.6)x(1.00)
in allowable stresses under temporary conditions, and
you should consult whichever code is to be used for exact 870.4 psig
amounts allowed. It is recommended that a value of 30
psig or 10% be added to the operating pressure for de-
MAWP (Design) : (18,300) x (1.00) x (1.00)
sign pressure. This practice varies with each company (21.00)+(0.6)x(1.00)
throughout the industry.
Once the internal pressure is determined it must be de- 847.2 psig
cided how the vessel is to be welded. The factors affect-
ing this decision are as follows: The 18,300 psi is obtained by linear interpolation of
the allowable stress values in Table UHA-23 of the
l. Size of vessel-whether rolled plate or seamless
ASME Code.
pipe is used.
The vessel owner knows the maximum allowable pres-
2. The toxic nature of the fluid to be contained.
sure for the shell at the new and cold condition as well as
3. The economics of fabrication as to whether a full the design condition. It is a common practice to limit the
joint efficiency is necessary.
MAWP by the head or shell and not by the flanges or
openings, only the MAVr'P is determined by the flanges
One can appreciate the degree of types of welds re- or openings when the vessel is to be reapplied in another
quired for a vessel. A slug catchel which acts as a scrub- application or a design oversight is made.
ber handling a non-toxic substance, does not require the Finally, in computing the minimum thickness of the
same caution as a vessel containing cyanide gas. shell or head, mechanical allowances must be consid-
The quality of a weld joint is determined by a radio- ered. In the manufacture of heads, the metal is thinned
graphic inspection. Full radiography includes a complete on forming the section (a forgery process). This forming
X-ray inspection (1OO% for butt weld and 907o for sin- allowance must be considered when the nominal thick-
gle-welded butt joint) and spot radiography implies 85 % ness is specified. When a minimum thickness is specified
for buttjoints. See Thble 4-1 for maximum allowable ef- to the head manufacturer, the forming allowance is not
ficiencies for arc and gas welded joints. The reader is considered because it is the manufacturer's responsibility
strongly urged to consult whatever code happens to gov- to ensure the minimum thickness.
ern. Listed in Thble 4-2 are the joint efficiencies for the
various welded combinations for pressure vessels under
ASME Section VIII, Division I[1]. DESIGNING FOR EXTERNAL PRESSURE
Any discussion on designing for internal pressure
must include maximum allowable working pressure, The design for external pressure of vessels is fairly
which is the maximum gauge pressure permissibie at the standard in the ASME and codes of other nations. The
top of the completed vessel in its operating position for a procedures for determining minimum shell thickness,
designated temperature. This pressure, MAVr'P, is nor- spacing, and section properties of stiffening rings are
mally specified on two conditions*new and cold (ambi- straightforward and simple. Because there is much pub-
ent) (NAC), and design. "New and cold" implies the lished material on external pressure design, the subject is
MAWP for a new vessel (non-corroded) at atmospheric not discussed here. The reader is ursed to consult the
condition, and "design" implies the vessel corroded at oressure vessel code to be used.
The Engineering Mechanics of Pressure Vessels 161
Table 4-1
Maximum Allowable Joint Efficiencies for Arc and Gas Welded Joints [11
Degree ot Examination
(a)
Fully (b) (c)
Type ol Joinl Radio" Spot Not Spot
No. Description Limitations graphed Examined Examined
(l) Butt joints as attained by dou- None 1.00 0.85 0.70
ble-welding or by other means
which will obtain the same
quality of deposited weld metal
on the inside and outside weld
surfaces to agree with the re-
quirements of UW-35. Welds
using metal backing strips
which remain in place are ex-
cluded.
(2) Single-welded butt joint with (a) None except as in (b) below 0.90 0.80 0.65
backing strip other than those (b) Butt weld with one plate off-
included under (l). set-for circumferential joints only,
(3) Single-welded butt joint without Circumferential joints only, not 0.60
use of backing strip. over 5/a in. thick and not over 24
in. outside diameter
(4) Double tull fillet lap joint l-ongitudinal joints not over 3/8 in. 0.55
thick. Circumferential joints not
over s/r in. thick
(s) Single firll fillet lap joins with (a) Circumferential joints for at- 0.50
plug welds conforming to UW- tachment of heads not over 24 in.
t7 outside diameter to shells not over
t/2 in. thick
(b) Circumferential joints for the at-
tachment to shells ofjackets not
over s/a in. in nominal thickness
where the distance from the center
of the plug weld to the edge of the
plate is not less than 1r/2 times the
diameter of the hole for the plug.
(6) Single tull fillet lap joints with- (a) For the attachment of heads 0.45
out plug welds convex to pressure to shells not
over s/e in. required thickness. only
with use of fillet weld on inside of
shell; or (b) for attachment of heads
having pressure on either side, to
shells not oyer 24 in. inside diame-
ter and not over t/+ in. required
thickness with fillet weld on outside
of head flange only.
162 Mechanical Design of Process Systems
Table 4-2
Joint Elficiencies for Arc and Gas Welded Joints per ASME
Joint Types H, C, and L
T1 = Type 1 Joinl (ASME UW-12)
T2 = Type 2 Joint (ASME UW-12)
Asterisk (+) denotes which joint type governs. Illustration of weld joint locations Typical of Categories A, B,
C. and D-see ASME Section VIII Division I.
L00
Spot
None
Spot
Spot Spot 0.85
Spot None
None Full I
I
None Spot I
I
None None I
I
Spot Spot 0.85 I
I
None I
1.00 I
I
Spot Spot I
I
Spot Spot Spot 0.85 I
spot Spot I
I
None None I
None Full I
I
None Part 1.00 V
*L Weld soverns in circumferential stress calculations.
rrI E-^r-^^-r-^
'FL^ LrrSrrr!!rur6 i'[echanics of Pressure
rr. Vessels 163
:------------- _--------=--:- |
r
spot Spot tu i
Spot Spot Spot 0..85 0.80 100 0.85 0.80 100 0.85 0.80 100
Spot Spot None
Spot None Full
Spot None Spot
None Full Full
None Full Spot
None Full None
None Spot Full
None Spot Spot 0.85 0.80 0.80
None Spot None
None None Full
None None
None None None 0.70
164 Mechanical Design of Process Svsterrrs
Spot Full
I
I
r00
Spot Part 0.80
Spot Spot
I
I
85 i
Spot None
I
I
I
None Full
None Part
I
None Spot I t
None None 0.65 0. 70 0. 65
At Saddle
01 : Eoan + op
The allowable stress for compression is based on the ac- o1 < Bl2
cepted formula for buckling of short cylindrical col-
umns, which is
where r = radius of cylindrical shell, in.
thickness of cylindrical shell, in.
/"\i,\[, - (,1,'*,(i)]
\-rt \;i f
modulus of elasticity of shell, psi
D B factor in the ASME vessel code, psi
The Engineering Mechanics of Pressure Vessels
Referring to Figure 4-3, oy occurs at either lBl6 + 0l2l 2A : arc, in radians, of unstiffened shell in plane saddle
degrees or zero degrees at the shell acting in the longitu- effective against bending
dinal direction. This only applies to unstiffened shells.
The vessel must meet the allowable with or without ores-
sure. At Mid-span
l- *'-
":-t*-[..(, I *ll,AL
+
"41ll o: longitudinal bending stress at midspan
3L
I +
IJ
The longitudinal bending stress at midspan has the same
(4-3)
(4-r)
n \
--il \
.-11--T-
- I
N l-/
|| zT\-[ll
ll/r\rr.,
tt/ | \lll
ffi-["
'Y V-t-+
Figure 4-3. Bending moment diagram for a horizontal vessel developed by Zick
l2l.
168 Mechanical Design of Process Systems
;&p
z
z
6 o.os-
=
=
o
z
z
UJ
)
t-
zgJ
,'u o'o2-
=
E
1.O
RATIO A/r
The Engineering Mechanics of Pressure Vessels 171
L Unstiffened shell with saddles awav from head (A If A/r>1, then K6: O.42Z2e-a ot710 (4-11)
>R) where d : deg
If A/r < 0.5,then K6 : K3/4
''v- - 11-ctr \ L+H a /L-H -2A'l Otherwise, use Figure 4-3.
/
:in," I
'(\7r-@+slnqcosd/ ,o-r,
1= r,# (; . rq],
g & B in degrees (4-6)
r! :
,j.
tangential shear stress, located at an angle of B/20
This stress occurs only when t}re shell is stiffened by the . - r(#)'+ 2cos2B
head and when the head is located less than one shell ra- (4-12)
dius from the saddle. The rnaximum shear is located at ^/^\:i
srnpcosp, .lslnpl
an angle of (l9l2O)P as shown. a \B I
o6 3 0.8o"1
-
3Q I sin']" I (4- 13)
"s 3'-1u - 6e,1 Lr .o, + sin " cos
"l
Circumferential Stress at Horn of Saddle
08 S oall
-,'QIu
2(r - = CA)'?
p.s) (t - CAXb + 1.56(r(t - Ca;101
'o'' r
02 +op =
(metal or concrete) 02
ff
Tangential Shear Stress
Location ot Saddle Supports
03' q4 < 0.8 oall
For thin wall vessels with large diameters, it is desir- o5' 06 < 0.8
able to locate the saddles close to the head, where A = Circumferential Stress at Horn Saddle
ID/4, using the stiffness of the head. Although arbitrary o7 < 1.5 o"1
on what a thin shell is, and Zick [2] does not define the
term, a shell is generally regarded as ',thin" when D/t Circumferential Stress at Bottom of Shetl
:
> 100, where D shell diameter and t shell thick- : oe q 0.5 (or1-i") * Compressive Yield
ness. For shells where D/t < 100 and the distance from
the head tangent to head tangent is rather large (approxi-
mately L/r > 10), the saddles are best spaced when the
longitudinal bending stress at rhe saddle, or, equals the
longitudinal bending stress at midspan. o2. Undei no cir- Zick Stlffenang Rings
cumstances should the distance from the saddle center
line to the head tangent, A, exceed 0.25L. When the Zick stresses in a vessel become excessive
and the location of the saddles no longer is a factor be-
A listing of allowable stress criteria is siven in Thble
cause the stresses are below the allowable stress, then
4-3. Each of the previously menlioned stress values
two options are available-increase the vessel wall thick-
should be evaluated with this table and the appropriate
ir code.
ness or add stiffening rings. Almost always it is more de-
sirable to add stiffening rings because it is cheaper to add
a few rings than go to a larger size shell thickness, par-
Wear Plate Deslgn ticularly with expensive alloys. Also, if the vessel is sub-
jected to external pressure , the Zick rings can act as ex-
One of the first things to consider when designing a ternal pressure stiffening rings as well as Zick rings.
horizontal vessel is the need for wear plates. Too often Referring to Figure 4-3, if two Zick stiffening rings
these plates are "auromatically" included with no are located on each side of the saddle, then
lhought given to their necessity in each application. Wear
plates involve material and labor expense and are a waste Ln,n : l.Jb Vfl, It
if not needed.
Wear plates are not required if two criteria are met: Lr* : r, ft
The circumferential stress at the horn of the saddle must
be less than 1.5 times the allowable stress, and the ring The stress in the ring is
compression stress in the shell over the saddle musr bi
less than one half the minimum yield strength. These cri- _- _ -KuQ KuQr ,
teria can be written as follows: nA- n7,
o1 1 7.5 o^x Where = l-/c for ring in the plane of the saddle,
Z
oe ( 0.5 o, 6n tn.'
Table 4-4 shows minimum allowable shell thicknesses z : I,-,/d at saddle horn at tip or flange of
required for horizontal vessels without wear plates. The stiffener ring, in.3
I*- : moment of inertia of stiffening ring about
values are based on using a fluid 1.75 times the weight "
axis x-x, in.a (includes wear plate
of watet and the metal has a minimum yield of 30,000 thickness if one is used)
psi and an allowable stress of 17,500 psi. For vessels in : cross-sectional area of stiffening ring, in.2
seismic regions wear plates should always be used to n : number of stiffener rings per saddle
minimize stress concentrations at saddle plate-shell junc- r : mean vessel radius, in.
ture. K6 : previously defined
The Engineering Mechanics of Pressure Vessels 175
Table 4-4
Minimum Shell Thickness Required lor Horizontal vessels Without Wear Plates
lD (in.)
78 84 90 96 102 108 114 120 132 144 156
8
r/s in.
r/r in.
l0 I
t2
t4
l6
_r- I
5/r in.
lllro in.
Not€s
1. The above table is based on the following:
a. vessel is tully loaded with a fluid of specific gmvity of 1.5.
'72.
b. The ratio of the shell outside radius, R., to shell thickness, t, is R-lt >
c. vessel weight is computed with not€ (a.) and hemispherical heads.
d. Vessel material has the following properties:
d,i" y,.rd = 30,000 psi and o.rr* = 17,500 psi
2- In seismic zones 3 and 4 wear Dlates should be used.
174 Mechanical Design of process Systems
In compression, oro is negative, strengthened with stiffener or web plates. but often too
many are used. which increases laboi and material costs.
oleAB(0.5o., In the past, saddle plates have been purposely over-de-
signed to guard against uncertainty. This is no longer re-
In tension, o,6 is positive,
quired, since literature on flat plate theory has increased
o'e * oo ( o.1 [tension]
with mounting experimental data. One such organization
that has engaged in extensive research is U.S. Steel [4].
where B = ASME compressive stress (see ASME Section Figure 4-4 shows a typical saddle configuration for a
8 Division l) horizontal vessel. Section A-A shows that only an effec-
o", : compressive yield stress (see ASME Section g tive portion of the member will resist compression.
Division 2) shear. and bending loads because when rhe member is
op : internal pressure stress (includes wear plate loaded, the outside fibers ofthe web plates and the center
thickness if one is used) of the saddle plate -shown by rhe sh;ded areas in Figure
4-4-go into the plastic range. The rest of the plate area
In defining the parameter K7, it must be noted that the is still in the elasric range because of residual stresses
Zick stiffening rings can fit on either the inside or out- that were created by non-uniform heating during rolling
side of the vessel shell. Many clients object ro the rings or welding. Presently, this "effective" area can be deter-
being external to the vessel surface because of aesthetici. mined only by experiment. Equations 4-9,4-10, and 4-
However, after insulation is applied, the rings are no 14 are used in saddle design as follows:
longer visible. We will consider rhe rings in both ways.
The constant K7 is defined as follows: b":KL (4-15 )
For a ring in the plane of the saddle- where b" = effective width, in.
K; = plate buckling coefficienr for either
+ 1.0 compression, shear, bending, or a
0.340, 0 : 120" combination of these loads (see Figures 4-5,
Kr: 0.303, 0 : 150' 4-6, 4-7, 4-8,4-9, and 4-10).
0.250,0 : 180. i : c, b, s, or a combination of these characters,
where K. : plate buckling coefficient for compression,
dimensionless
For rings adjacent to saddle- Kb : plate buckling coefficient for bending,
For internal rings, dimensionless
Ks : plate buckling coefficient for sheaq
\: -1.0 dimensionless
(0.27 | , 0 : 120.
Kz: 10.2t9, 0 :150' o
{0. r+0, o = 180' 4(t - CA) tb" + 1.56(r(t - CA) fI
-.lt*'gl.,rrL
L(r - LA)'
< 8R (4-10)
STEEL SADDLE PLATE DESIGN
d"
t/ll
lffl
b
- b"-:l
fI t---------1 I
\r .,-.-lN I
'.-lF 1l"'l*" -
|
'-ff I
*lJL.._"1
-.T----
sections A-A and B-8,
shaded areas are in the
plastic range.
elevation view
Figures 4-5 to 4-10 are courtesy of United States Steel Corporation. --'t------------- - --'l-----------_-
USSC makes no warranties, express or implied, and no warranty as
to the merchantability, fitness fot any particular purpose, or accuracy I --r
--i cAS€ 3
r- --
F ---l CISE 4
F-
I
of the information contained in any material reproduced herein from
l _.1-...---------.1= 5
its Steel Design Manual. In the event of any liability arising out ofthe
publication of such material herein, consequential damages arc ex- -l casE 5 F
cluded. ri\
loaoEo FtXE0
E
\\' ---- EDGES
LOADED EDGES
z.
E \ SIMPLY SUPPORTEO
j i\ I
\\.
ta.'a
aaTro oF
EENDING STBESS
-TOU\IFOR\,I MJNII\,4UI\,IBUCTI.NGCOLfFICITNI.'I,
COVPBLSSION
ST-8ESS, UNLOAOED EDGES UNLOADED
LOAD ING Sll\,4PLY SUPPonTED EDGES FtxED
\l
T-- .]
t7/
tl -t!ft.
{PU8E BENOING)
3H.
jr:=-2l3r,r 5.00
,.1 \
Fry]=
!l r, = r/3f I E/ 0.50
\t_-_____tr/
Ir r-r____-_r_: I,
F= f: = f, t= rp,,..
. -... "^iiil.*
--.....,--jtoN) 4.0
.VALUES Figure 4-6. Buckling coefficients for flat
GIVEN AAE BASED ON PLATES HAVING LOADED EDGES
S{I\4PLY SUPPORTED AND ARE CONSERVATIVE FOR PLATES plates under compression and bending. (Cour_
HAVING LOADED EDGES FIXED. tesy of U.S. Steel [4].)
n
i LONC EDGES FIXED,SHONT
ED6ES SIMPLY STIPPOfi TEO
Figure 4-8. Buckling. coefficients for stiffened plates under 0.6 0.8 1.0
t
uiform compression (one longitudinal stiffener at mid-point). 2.O
14 I
13
12
tl
i
z I
tr
U
o
o
z
=
f
34
2.8
30
4 6 810 12 14 16 18
3.0
NONDIMENSIONAL PABAMETER, O
2A I
26
j 24
22
: 20
o
(, 18
=
t6
-
12
2.8
3.0
F
z I
q
n
o
NONDII\,4ENSIONAL PARA]\IETER, d
shere di : saddle plate length normal to vertical axis of thickness of saddle plate, in.
stiffener (web) plate, shown in Figure 4-4. effective width of saddle plate that is
modulus of elasticity, psi perpendicular to the web plate, in.
Poisson's ratio
effective saddle width, in. with
saddle plate thickness, in.
d.=d,(0.25+0.91\)
t R/3
Qt" lo'.,n.
,: I\24.600 A-) 7l
Lrl
where b : Lz
Q: load on saddle, lbs
,'rete foundation or steel structure with anchor bolts the internal stress in the circumferentiai direction is com-
embedded in the concrete or steel. Normally, a vertical bined with the bending and tensile (or compressive)
\'essel must be at least thirty feet tall to be classified as a stresses. Writing this expression we obtain,
"tower." This height is used because thirty feet is the old
"= -(.*J'H-(x)
first wind-zone demarcation in code use. However, (4-27)
smaller vertical vessels are governed by the same design
criteria, but are not usually referred to as towers.
The various phenomena that affect towers in normal where Z: section modulus of the shell cross section, in.l
operation make their design complex and worthy of ex- A: area of the shell cross section, in.2
perienced engineering personnel. Therefore, towers
should never be taken lightly by any design office, be- Substituting Equations 4-24, 4-25, and 4-26 into Equa-
cause a failure could result in massive loss of material
tior 4-27 produces
and possibly lives.
, : * (ryf - (":zlur..
_ , h- iLr,
_ @-zB)
MINIMUM SHELL THICKNESS REOUIRED
FOR COIIBINED LOADS ipo\ /
"io^lra
to \
"J
- - \4,/= -t DitrDr D,:t/
\nrtO"
High-speed electronic computers now provide de-
tailed, exact solutions to complex mathematical prob-
Iems, and so have replaced the "strong arm" approxima- ,( 2w \ (4-29)
tions of yesterday. An example is solving the equations = \",tnt * Ot1
of the moments of inertia and section modulus. Before or
the advance of computers, the following expressions
were used to quicken computations on a slide rule or a /po\ / to o"t'.1 \
small electronic calculator: ' - \+r/- \norD,-r D,rrD"r - D,2r/
I = nR3t; exact: I: #,o"" D'a) (4-24) ,( 2w \ (4-30)
= \""fr5 * t1
.I Referring again to Figure 4-13, we summarize the fol-
Z : rRztt exact = z = ,a (gd ar'"'D | 4-15 | lowing:
A= exact: A:
l,o" - 1. For the tension or windward side,
?Irt; D,) (.4-26)
Using R as the mean radius minimizes the error and ' /po\ / ro o"v \
using R as the outside radius results in considerable er- \ +t / \rrr D" - D,tt D"2 - D,2t /
ror. Solving for the thickness or stress with the exact for-
mulations involves iterative analysis, which is a key at-
- lzwl+ DJ/ (4-31)
\"(D"
tribute of today's computers. or
The minimum shell thickness required for internal or
external pressure alone is often not sufficient for addi . - /.o\ I 16 D.M \
tional stresses induced by bending moments and weight ' \+, /- \ro{D" --J,DJ + D-l-,/
loads. Bending stress is a result of static wind, dynamic
wind gusts, vibration or seismic response spectra. In de-
sign the engineer takes the largest bending moments in-
- /zw\
\"r(DJ-D)/
t4-1)\
duced by one of the following: wind, vibration or seis- 2. For the compression or leeward side,
mlc.
Referring to Figure 4-13, we analyze the stress ele- /po\ i ro o.r,a )
ment depicted. The maximum stress resulting from inter-
\4tl \Tt(D" + D,XD.r I D,r)/
nal pressure occurs along the x-axis, i.e., the hoop stress
is twice the longitudinal stress. Wind, vibration, and/or - lzwi /4-11)
seismic forces cause the shell to bend about the z-axis, so \"(D" + D)/
182 Mechanical Design of Process Sysrems
l--lo l- clF
o
u, 3lt ll, =ll
stN ll'
U' ld"d
l^'l l= x 'E'
o ' ,. f---l /
1t--l /
-3
- -\'l II< l<-l --r'l tr l<1_ -
ttGt
q)
o
l I
^i^
L-
s-^ > E-
tt,
lo o-l I ci o i.;
Nls
" lci r o tk
th
at
\ | lr
--41ll < lr>- -
o"
*
irlol<r-
rl
/
I
I
q)
o
tt f x
to 3 ielo ^r
EO
C'
Rl+--i- ",*
,ll ;l*lll
^"-'l= i' Ii.- - l''
' oo- +
bp - !l < l<L ' {i ll I'
-rdo[<r-
!'<f flti lt ti
r.u 3
o)
-t'- --l-
o. i1
" Rlr x !:J t€
I
a,l
E"
O,o
3
5x
^, i rine*
o
rl)
o i -t-
u
r r N
E(/,
il= "-1
- ' f---------1r
)
Fl_"_12 ^lil+
-\1il <lK- .Jt ql I/
- +;l
o
lr it I:+ ..-t-''_ Ir
o-
t-l
I
Y\
x ol
o- td . pl*
o
q)
a9
or l:
!: (1, IL
U'|,,l
The Engineering Mechanics of Pressure Vessels 183
OT
the cost-plus contractors seek to standardize designs and
use lower pressure vessel code allowables. As with wear
'- /ro\
._ / ro o.rra \ plates on horizontal vessels, most lump-sum contractors
\4"/- \""(DJDJ(DJ +-3/ would elect to omit them whenever possible to save ma-
- lr*\ (4-34)
terial and labor costs. This phiiosophy is becoming in-
creasingly popular with recent economic upheavals and
\"r(D" + DJ/
increasing international competition.
3. For vacuum vessels the maximum stress occurs on Types of skirt supports are shown in Figure 4-14. Fig-
the compressive side, such that Equations 4-29 and 4- ure 4-14b shows the most common and desirable skirt,
30 become since the shear is eliminated by the type of attachment.
This type is used primarily on short vertical vessels.
The skirt is designed to resist loadings caused by bend-
ing and the tower weight. Writing the expression that de-
- l2wl
scribes this we obtain
(4-35)
\"(D. + DJ/
and
l6MD" , 2W
(4-38)
SUPPORT SKIRT DESIGN no(D" + 5tD"'+ D-5 "'@. + D)
The design ofvessel skirts is one area in which design- Once again, Equations 4-37 and 4-38 must be solved by
ers disagree philosophically. Lump-sum contractors seek iteration. Normally, these equations do not govern the
to use higher allowables and thus less material, whereas skirt thickness, as the reaction of external bolting and
t,/
l\l l
Jt tE, i
f*- "---|
r?
lzA,l
1I
IT
t[
liill Iil
J IIL
F-r:-'-_i r-*-l
k----il--+l
NOTES:
all dimenslons in inches BTHK to be evaluated by
eouations 4-57 or 4-60 all welds to be size "t"
BOLT
SIZEABCBH GH CHLLJMNP
'I
31lz 'l4a 14q 1112 2 4e 31lt 6 9e 51lz 5 31la 3 31lz 74s3
'l tl' 3{+ 148 17k '1518 21k 9e 33/e 6 Ye 53lq 51lc 31lz 3 3{+ 731rc3
'l1la 4221s1c2112 4e 33lq 6 4t 6 5112 33t+ 3qo 4 1'tlq th 3
1q8 41lq 21la 2118 148 2sla llz 41le 7 4q 61lq 53lq 4 33lq 4318 7112 1lq 3
1112 4tlz 21lc 21la 2 2sl+ 112 41lt 8 3lq 6112 6 41lq 41la 4112 81|q4
11la 5 2112 21lz 2th 31lq 5lB 5 9 1 7 6tlz 4eh 4alc 43lq 9 5h6 5
2 51lz 24c 23lc 21h 31lz 3lc 51lz 10 1 7112 7 Stlq 51lz 6 '12 {e 6
2tl+ Sslc 3 24c 23lt 34a 3l+ 53lc 12 1 Telc 7112 53lq 5718 6112 13 {e 7
21h 61lc31l+334118 4e 64a 13 11/e 81lc 74c 64a 6 7 144rc8
23lq 61lz 3rlz 3 Srla 4112 1 71lz '14 11lq 8112 I 61+ 6{e 71lz 16 1lz 9
7 3glc 31lc 31lz 4gla 11k 8 16'l1la98r/+777slq 18 llz 10
Figure 4-15. Typical designs and dimensions of chair and base plates
186 Mechanical Design of Process Systems
ANCHOB BOLT
MAXIMUM TENSILE
FOUNDATATION UPLIFT FORCE- q
MAXIMUM
JI E.
COMPBESSIVE <l
FORCE = nFc I
I
zt<
I
COMPRESSIVE +-
FOBCE
TENSILE FORCE DISTRIBUTION
DISTRIBUTION CENTROID
CENTROID
prevents the strength of the bolting in the concrete from factor. The modulus of elasticity of steel is approxi-
becoming fully developed. It is advisable to set the bolts mately 30 x 106 psi and that of concrete approximately
at least 18 inches apart. To accommodate this minimum 2.O to 4.O x 106 psi. Defining the ratio of the two as n.
spacing a wider base ring with gusset plates can be used we wnte
or the skirt can be tapered with a conical skirt. As shown
in Figure 4-14, with a tapered skirt the apex angle should F
(4-43)
not exceed 15 ". F
€s
ring plate used to distribute these load-s around the cir-
cumference of the bolt circle. Anchor bolts normally and e, = e. because of the base plate-concrete bond we have
vary in diameter from one to three inches-bolts smaller os(induced) = noc(induced) (4-44)
than one inch are more likely to strip or shear off; bolts
larger than three inches require large wrenches and cre- Listed in ?ble 4-7 are values of the moduli ratio n and
ate excessive problems for construction personnel. For the various concrete mixes from Brownell and Young
these reasons it is desirable to attemDt to adhere to the [3]. Figure 4-16 shows a detail ofthe compressive force
one to three inch size range. of the concrete, F", multiplied by the value of n shown
In the case of a concrete foundation, the relative opposite the maximum tensile stress, Fr of the base plate
strength of the concrete to steel becomes a significant steel.
The Engineering Mechanics of Pressure Vessels 187
Table 4-6
Bolis, Threaded Parts, and Rivets
Tension [51
Allowable loads in kips
Bolts and Rivets
Tension on gross (nominal) area
Nominal Diameter. d. in.
ASTM Fi 3la 4s 'l1la 11la 13ls 11lz
Designation Ksi Area (Based on Nominal Oiameter), in.'?
0.3068 0.4418 0.6013 0.7854 0.9940 1.227 1.485 1.767
A307 bolts 20.0 6.1 8.8 \2.O 15.'7 19.9 29.7 35.3
A325 bolts 44.0 13.5 19.4 26.5 34.6 43.'7 54.0 65.3 77 .7
A490 bolts 54.0 16.6 23.9 32.5 42.4 53.',1 66.3 80.2 95.4
.\502-l rivets 23.0 '7.1 to.2 13.8 18. I 22.9 ?8.2 34.2 40.6
A502-2,3 rivets 29.0 8.9 12.8 r7 .4 22.8 28.8 35.6 43.1 51.2
The above table lists ASTM specified materials that are generally intended for use as structural fasleners.
For dynamic and fatigue loading, only A325 or A490 high-strength bolts should be specified. See AISC Specification. Appendix B. Sect. 83.
For allowable combined shear and tension loads. see AISC SDecification Secl. l 6.3.
Table 4-7
Design of Supports lor Vertical Vessels [31
Values of Constants q,
C", Z, andJasa
Function of k Average Values ot Properties ot Three
C" Cr ZJ Concrete Mixes
0.050 0.600 3.008 0.490 0.760 Water Content oi n
0. 100 0.852 2.881 0.480 o.'766 U.S. Gallons 28-day Ultimate 30 x 106 Allowable
0. 150 1.049 2.772 0.469 o.7'7 | Sack
per 94Jb Compressive Compressive
0.200 1.218 2.66r 0.459 0.776 ot Cement Streigth, psi Ec
Strength, psi
0.250 1.370 2.551 o.418 0.7'79 7tlz 2000 l5 800
0.300 1.510 a Ana 0.438 0.781 63/c 2500 t2 1000
0.350 1.&0 2.333 o.427 0.783 6 3000 10 1200
0.400 1.765 2.224 0.416 0.784 5 3750 8 1400
0.450 1.884 2.t t3 0.404 0.785
0.500 2.000 2.000 0.393 0.786
0.550 2.113 1.884 0.381 0.785
0.600 2.224 r.765 0.369 o.784
'188 Mechanical Design of Process Systems
ork = I
(4-4s) oq.*, : (sFC)t*X**] (4-s4)
tro"
D" - (skirt OD)
using L = , ln. (4-s5)
Equation 4-45 is solved by iteration using the following
steps: Thke values for C", C,, Z, and j in Thble 4-7 for a
given value of Z. Normally, k = 0.333, C" : 1.588, we solve for the base plate thickness, BTHK,
C. = 2.376, Z = 0.431, and j : 0.782 to start the pro-
cess. Then the following equations are solved: BrHK = L (4-s6)
/^^\
M _ (W,r(z)l!!l
[tf,]"'
" lt)l :
rt--/\ where ou1 allowable working stress for base plate metal,
(4-46)
psi
.. lBcl
rJrl;l
\'.1
(Ah)N
'' r(BC)
(4-47) By using Equation 4-56 one assumes no gusset plates on
the base plate-skirt connection. To reduce the required
F base plate thickness in Equation 4-56 the additional
Ir = / -\ (4-48) strength of gusset plates can be used, because with the
rt,r l!91c, gusset plate stiffening the base plate at the skirtjuncture.
\2/ the base plate between the gusset plates can be consid-
fc: fi +wE ered as a rectangular, uniformly loaded plate with two
edges simply supported (at gusset plates), a third edged
BPW : base plate width, in. fixed (skirt side), and the fourth edge free. The deflec-
tions and bending moments are tabulated by Timoshenko
Bpw:(D.)-(Di) (4-4e) [6] and are shown in Thble 4-8. The process of using gus-
2
set plates to stiffen the base plate is begun by making the
number of gusset plates equal to the number of anchor
tz = BPW - t; (4-s0)
bolts. Doing this we write
J|L = --
f,
(4-s2)
RAT:!SG
/^l
(h - (Enxt ))
(]Jrc.r M- : 1E,1o"1.""(SG)'? (4-58)
where SFC : compressive stress on concrete at the bolt
circle and M,' : lE,1o"1."-(L)2 (4-59 |
The Engineering Mechanics of Pressure Vessels 189
Where o"1*n,; is determined by Equation 4-55, using the The minimum skirt-to-base plate weld size is determined
greater of M, and M, we have by
and
This iteration can be repeated as many times as desired to
reduce the base plate thickness. In normal practice, it is € :
w = 2F* weld size (4-65)
unusual to use more gusset plates than anchor bolts.
The bearing pressure on the base plate must be :
where o"1 smaller of the allowable stress values for the
checked to prevent exceeding the allowable compressive
base plate and skirt metals
stress of the concrete. Computing the uplift force on
M: moment at base plate induced by wind or
each anchor bolt we have, seismic forces, in.-lb
D.r : outside diameter of skirt, in.
--' 96MD" WF ,, (4-61)
N1D"z a P:; N'''
ANCHOR BOLT TORQUE
N" WF 12MD"
.
o^ -_ ----:: + -----: +
A" A. 21,
DSI (4-62) There have been many recipes proposed for the com-
putation of bolt torque over the years. The mystery of
bolt loads is unveiled by such authorities as Bickford [7]
o" ( 1,200 psi for-.weight and wind
where, M : ftJbs
and Faires [8]. Their extensive research into bolt loading
produced the following recommended formulation:
r[(D")'?
T: CDFi, in.lb (4-66)
where A" = - (D)'?]
4 where C: 0.20 bare steel
C: 0.15 for lubricated bolt
t- r[(D")a - (DJa] D= nominal bolt diameter, in.
64 F1 : anchor bolt uplift force, lbs
190 Mechanrcal Design ol Procg55 g151snlt
2.3<+<2.6
groutl-i
-f-- Boltom of sleeve
concrete or top of concrete
L = 17Du
Figure 4-17. "J" and "L" type anchor bolts are used for small vessel..
Referring to Figure 4-13, we see that the stress ele- such that the leeward side governs, then
:lent in the shell is affected by the combined loads more
::r the longitudinal direction than the circumferential di- loo-o*-o*l < Bfactor
:ection. However, for the longitudinal axis the internal
rr external pressure stress is governed by the relation If Equations 4-70 and 4-71 hold,
WIND DESIGN SPEEDS ing hit by one is extremely small; however, nuclear
sites are designed to withstand tornado winds.
The procedures for determining wind design speeds 2. Using site and structure factors calculate the design
for structures, towers, and stacks varv from Counirv to wind speed. The factors on rhe ANSI l98Z tesr
country, depending on how well records have been kept. used are as follows:
The wind velocity is a function of the temperature gradi- (a) Importance Coefficient, I, a hospital or nu-
ent and terrain roughness. The first representation of a clear plant would be designed moie conserva-
mean wind velocity profile in horizontaily homogeneous tively than a barn on a farm.
terrain is the power law, first proposed in 1916. This law (b) Variation of wind speed with building height
states and surrounding terrain.
(c) Gust response factor.
,t: r.\r:)" (4-74)
(d) Velocity pressure coefficient, K2.
3. Test a model of the tower and its surroundings in a
wind runnel. Even though rhe 1972 ANSI stindard
where Vo : mean wind speed at a reference height Z does not mention this, the 1982 version sDecifies
Z. = reference height (normally 33 fr orl0 m)
certain requirements lor wind tunnels.
Q: a constant dependent upon roughness of
terrain
Z= height above ground These three options are new to both the ANSI-ASg. I
standard and to the three building codes-the Uniform.
Other proposals have been made to determine wind Basic. and the Standard. The larrei three codes do not au-
speeds. Simiu [9] states that the logarithmic law is a su- tomatically adopt newly revised ANSI standards, thus
penor representation of strong wind profiles in the lower making for inconsistency in wind code provisions in the
atmosphere. What is relevant to the reader is to be famil- United States.
iar with whatever standard is used. The discussions and The basic wind pressure in the ANSI-A5S .l-19j2 rs
examples presented in this text are slanted toward those
standards in the United States. However, the technioues q3a= pv2l2 : (0.5X0.00238)(5,280/3,600fV30
and base principles of engineering are applicable to all : 0.00256 v3o @-'75,
countnes.
In the United States there are four basic codes sovern- where q.s : basic wind pressure at 30 ft, above grade
ing wind- ANSI A58. | 1982. the Uniform, thi Basic, lb/ft,
V36 : basic wind speed, mph
and the Standard Building Codes [10]. The ANSI-
A58.1- 1982 differs from the ANSI-A58. I - 1972 in that
The effective velocity pressures of winds for buildings
three optional methods of determining wind design loads
and structures, qF, is
on a structure are given [11]. These options are as fol-
lows:
9r : KzGrQ:o (4-76 t
l. Choose a design wind speed (50-year mean recur- where K2 = velocity pressure coefficient that depends
rence interval) off the U.S. map provided on the upon the type of exposure and height Z above
document. The national map is a graphic display of the ground
isopleths (lines of equal wind speed) of the maxi- Ge : dynamic gust response factor
mum values of the mean speed for which records
have been kept. i.e.. basic wind speeds rhat can be In the 1982 ANSI-A58.1 Code the effective velocin
expected to occur within a particular period. This pressure for wind is partially a combination ol Equarion.
"particular period" is called the return period. The 4-7 5 and 4-76,
problem with a nalional map consisting bf isopleths
is that localized wind speeds can vary as much as
qz : 0.00256 KzGV)2 \417)
30 mph over the speed shown on the isopleth (par-
V= basic wind speed, mph
ticularly in mountainous regions).
I: importance factor I :
Hurricanes are fairly well accounted for on these A value of V can be approximated for the United States
maps. Tornadoes are considered to be nonexistent, from the isopleths shown in Figure 4-19.
because it is not economically feasible to design an One of the major differences between the ANSI
entire building for tornado wind speeds. The rea- A58.1-1972 and 1982 is how the velocitv Dressure coef-
son for this is that the probability of a structure be- ficient, K2, is determined. In the 1972 Cod'e the value ot
J
o;
-;
\-\ .E e. b
oot;9
\\ ;] b9;.0
*-t!cb
\ ; !b69
\\ 3. iaEg
o-i=H9
a'*-w*
ird\* ii:# .s i
\px /\,r\ *i *tid
"E
i4' 9\i ^;it
!ii.s
-
\F 9!.o
) '-+!-€ : R: P
2 : \8. E i a; :
f : ]{ ;
I" .: IEE:
i-'q
s -i Hf;n*
:,'' s 18 ;:
;
g;ni:
I-* l(U*963 :
$ " !;i;
o-!o
8
let o -:
6 6 r. E'=-
i" f i:.E
4:' ig ?i
>iif o
z
-6 r No
6 O
o
5 J
z
.9
TL
.l
!L
:
rl' I jll\
194 Vechanical Design of Process Syslem,
Table 4-9
Velocity Pressure Exposure Coefficient, Kz [1 1l
Height above
Ground Level, Z
K7 is a linear function of the height Z from heights of where a = velocity of sound = 0 in our case, because winci
thirty to nine hundred feet. This results in a triangular speeds are extremely low compared to sonic speeds
wind distribution on the tower. In the 1982 Code the
value of K2 is a parabolic function (can be approximated This equation shows that there is a relationship dic-
with a step function) for wind loading depicted in Table tated by the dimensions of the parameters involved. Ap-
4-9 and for dynamic gust response, K7 is governed by plying dimensional analysis makes the equation
lhe power law, Equation 4-74.
,r*
lz\2'
\r,)
forz > ls feet -t;
-tpvt,
P(Y2!2 ,}:.
Kz= @-78) where each of the two components is a dimensionless pa-
rameter. The equation can be solved for the first dimen-
forZ < 15 feet sionless combination by
Equation 4-79 (see any basic fluid mechanics text) and ob- Table 4-10
tain Exposure Category Constanls [111
Exposure Category 4 Do
3.0 1500 0.025
B 4.5 1200 0.010
2 c 7.0 900 0.00s
D 10.0 700 0.003
Cp pYz12
orF: (4-80)
F: q2GCpAg (4-81)
tO. 2 4 8105 2 4 ato6 2 4 8tO7
whele qz : wind pressure at height Z, EgrJation 4-77, Ae
_ lb/fc $rface
(d)
G:
sanded
gust response factor for main wind-force
srrface
resisting systems of flexible structures ---Smooth
Cp = force coefficient -
Ar = cross-sectional area of tower and other
attachments, ft2 Figure 4-20. The curves depict the influence ofthe Reynolds
number, corner radius, and surface roughness on the drag co-
The gust response factor, G, when multiplied by the efficient, square to circular cylinders; r is the corner radius
mean wind load, produces an equivalent static wind load and K is the sand grain size [9].
196 Mechanical Design of Process Systems
112 p!2
Figure 4-21. A sequence ofpressure fields forming around a cylinder at Nq6 = I 12,000 for approximately one third of one cycle
of vortex shedding (Flow-Induced Vibration by R. Blevins. @1977 by Van Nostrand Reinhold Company, Inc. Reprinted by
oermission.)
lp ?tr.,/s \r2
_11 =
2.35(C,- )0 5
-
tJ: L,.o.l +t +
11
| (Zl301rt"
e-82)
\p 1 + 0.002ci
S: structure size factor (Figure 4-23)
where p: structural damping coefficienr (percentage of average horizontal dimension of the building
critical damping). For normal working stress or structure in a direction normal to the wind.
conditions, 0.01 < P < 0.02 for towers. ft (see Example 4-2)
The Engineering Mechanics of Pressure Vessels r97
Iv
MEAN VELOCITY: V ^,--l
l'--4-l
|-J
OUnOt'O".?
3
5
15 )165 tt
Figure 4-22. Diagram of relationship between gust duration and gust diameter.
platform
DE = effective diameter
of area resisting
wind
insulation OD
d = plattorm angle
Figure 4-258. Wind area and force calculations for conical sections.
tions, and the third case of the end couple represents the
case of couples produced at the shell section junctures by
the translation and rotation of the upper sections. This
WIND-INDUCED VIBRATIONS ON TALL
combined loading is shown in Figure 4-26.
Adding the three cases we obtain the following: TOWERS
Table 4-12
Cantilever Beam Formulas
Formula
Uniform 1 w--.'
dITTtrM ^ w!2
: wf'
Load
T- 6EI 8EI
End
Load
1i : Q/'
2El
= Qi,
3EI
2
End ,-\
4
:vd
Couple
T-) EI
=
lul{
2ET
^ wl2 , w/,
, Mo{
-2EI-Er' l iw{ -, wr +M)
6EI EI\6
A:0t, '
I\- olr
\2"1 "tw.t
I'+ w/
I'+M| \
A=
Er, l\6 2 |
tall stacks by Baker at the turn ofthe century. Since then, What is clear in wind tunnel tests and field observa-
many advances have been made in the field of aerody- tions is that at low Reynolds numbers the tower is dy-
namics allowing designers to adequately design tall namically stable, vulnerable only to forced vibrations
structures. This chapter discusses tall process towers and and at higher Reynolds numbers a possibility of self-ex-
Chapter 5 discusses tall stacks. The differences between cited vibration will be present. From many field obser-
the two will become more clear in the following discus- vations it can be concluded that the first peak vibration
sion. Staley and Graven u5l summarized the state ofthe amplitudes occur at the critical wind velocity Vr, which
art of wind vibrations. Their studies indicate that even corresponds to a Strouhal number of 0.2 with the forced
though vortex excitation of higher modes has been ob- vibration as the basic source of excitation. Thus. it is sis-
tained in wind tunnel tests, existing free-standing stacks nificant that the peak amplitudes of vibration determined
have always been observed to vibrate during vortex exci- by forced vibration theory are in very good agreemenr
tation at a frequency and with a mode shape associated with field observations. This will be seen later in this
with the fundamental mode. Furthermore. the shaDe of chapter in Example 4-4.
the dynamic lorce amplitude or existence of nearly con- Even though the Rayleigh method is the industrialll
stant frequency over the height of the stack (or "lock- accepted method for the present, there are other methods
in") implies that dynamic response will almost entirely used to describe the vibration phenomena of tall process
be induced by the first mode. Staley and Graven con- towers and stacks. One such method was devised by N.
cluded that all higher modes should be neglected in the O. Myklestad, a great pioneer in the theory of vibrations.
dynamic analysis and that the frequency and associated The Myklestad method used in cantilever beams is es-
critical wind velocity ofthe fundamental mode should be sentially a Holzer procedure applied to the beam prob-
considered. For this reason the Rayleigh method is the lem. Its strong point is utilizing field and point transfer
industrially accepted method because it is used to deter- matrices to obtain relations that govern the flexural mo-
mine an approximate value for the lowest natural fre- tion and vibrations of lumped-mass massless elastic
quency of a conservative system based on an assumed beam systems. This method is used in such applications
confisuration of the first mode. as aircraft wings where the structural component is sub-
The Engineering Mechanics of Pressure Vessels 2O1
jected to high Reynolds numbers. Since we have already magnification factor we must consider some basic prin-
delineated the difference between cylinders subjected to ciples.
high and low Reynolds numbers and the fact that modes Consider Figne 4-27 in which a system with a single
higher than the fundamental mode can be neglected, the degree of freedom is subjected to viscous damping and
Myklestad method has lost favor to the Rayleigh method. an externally imposed harmonic force. The spring is
We are primarily interested in forced vibration peak am- denoted by stiffness k, the friction coefficient by c, mass
plitudes of relatively low natural frequencies. Although by m, displacement by x, impressed force as F sin cJt so,
the Myklestad analysis is excellent for relatively clean we have
aerodynamic surfaces such as wings and missiles, its
practical use in process towers with attached ladders, -X+.x +ki: Fsin<rt (4-89)
platforms, and piping is questionable. Even for stacks.
From the theory of differential equations we know that
low Reynolds numbers allow for the fundamental mode
the solution of Equation 4-89 is the superposition of the
to dictate. general or complementary solution of the homogeneous
Before the Rayleigh method is applied to our analysis,
Equation 4-89 and the particular solution of the same re-
let us summarize some basic precepts. Equation 4-80
lation. Writing this in equation form we have
calculated the pressure force exerted on a cylinder by a
static wind. When dynamic effects settle in maximum ac- X=X"*Xp
tual amplitudes, these amplitudes often exceed those un-
der static conditions. The net result is to multiply Equa- where X" is the complementary function and Xo is the
tion 4-80 bv a masnification factor. To understand the particular solution. This classical differential equation is
T"
*, = ]+ r|",1 lr-eol
I*
i
-7- -,
I
,-il]ur= r6"
_L-
-r ,r
,=
--l[: r-i- 14.=s\"
-r
I
[. 6!
4_.=o"lr,-.+r,.1
l)'. I_-r 6.
4_; qlL,.+r,-,+r, .l .L A
olFrt
== IA
4-.=qlq.,+ r'.+t'-..r.-"1 L*
Figure 4-26. Schematic diagram of wind loadings and deflections of a tower.
202 Mechanical Design of Process Systems
we have
x., (.4-9t)
"T -,,f-1_12r*
The maximum actual amplitude X of forced vibration
f fStru"t = forcing function is obtained by multiplying the static deflection X,, b1
fraction X/X,,.
The fraction or ratio X/X* is called the dynamic mag-
nification factor, D. These formulations indicate that the
nondimensional amplitude X/X,, and the phase angle, 0.
damper-represents tower's
are functions of the frequency ratio r and the damping
stiffness
factor f and are plotted in Figure 4-28. These curves in-
dicate that the damping factor has a large influence on
the amplitude and phase angle in the frequency region
Figure 4-27. The vibration of a tower is modeled as a sinsle near resonance. From Equation 4-91 we see that at reso-
degree of freedom. which i5 exposed to an exrernally impos=ed nance the dynamic magnification factor, D, is inversell
harmonic force and subjected to viscous damping. proportional to the damping ratio, or
I
n-'
K
M
Letting X:
.(T -l t+ (2rt Freouencv r.tio. = (;/o
Figure 4-28. The dynamic magnification factor versus the
frequency ratio for various amounts of damping. (From Slruc-
and tan 0 tural Dynamics by M. Paz. @ 1980 by Van Nostrand Reinhold
I -r Company, Inc. Reprinted by permission.)
The Engineering Mechanics of Pressure Vessels 203
The damping ratio, €, is not known and extremely dif- The force coefficient can be readily obtained from Fig-
iicult to measure at best. A practical method for experi- ure 4-29. Equation 4-92 yields the maximum transverse
mentally determining the damping coefficient of a sys- force per unit area of the projected surface of a cylinder
tem is to initiate free vibration, and measure through at resonance.
decreasing amplitudes of oscillatory motion, as shown in Equation 4-93 may be rewritten with the velocity in
Figtre 4-29. This decrease or decay is termed the loga- miles oer hour as
rithmic decrement, 6, and is defined as the natural loga-
rithm of the ratio of any two successive peak amplitudes, F= 0.00086(CrD)(H)V1'?, for air at 50'F (4-94a)
The evaluation of damping from the logarithmic dec- These equations apply when the top third of the tower
rement is given analytically by is the controlling length. Often, the top fourth of the
stack may be best to use as the controlling length. An ex-
X(t) : Ce-fdr cos(@Dt - cr) ample ofthis would be a section on top ofthe tower that
is one fourth the total tower height and is significantly
It can be shown [17] that the dynamic magnification greater in diameter than the section below (see Example
factor, D, and the logarithmic decrement, A, are related 4-4). Thus, for the top foufih of the tower Equation 4-93
using the previous expression as becomes
^T (4-92)
F= 0.00065(CrDXd)(H)Vr'?, for air at 50'F (4-94c)
and
Most research data available for practical use are pre- (4-94d)
0.07728pCeDVr'?(dH)
sented in terms of the logarithmic decrement, 6. Table 4-
13 provides values of 6 versus D for various structures.
where d = outside diameter of either upper r/: or r/+ of
These values are acceptable for use in actual design of tower, ft
process towers and stacks. H= total height of tower, ft
Applying the dynamic magnification factor to Equa- Vr : first critical wind velocitY,
tion 4-80 we have Vr : 3.40d/T, mph
T: first period of vibration, Hz
CeDpV2fz
(4-93)
p: density of air at any specified temPerature,
lb/ft
F
z
=
9
h
Table 4-13
Conservative Values for Logarithmic Decrement
and Dynamic Magnilication Factor for Tall
Process Towers
Logarithmic Dynamic Magnitication
Decrement Factor
6 D
Low damping: rocky-stiff
soil, low-stressed pile 0.052
support, or structural
frame
Average damping: moderately
stiff soil, normal spread 0.080
For tall, slender towers of constant diameter, the first the static equilibrium point. For the potential energy o;
period of vibration is given by the expression the system, the reverse is true. Thus,
T: (l/0.5l)(WHa/gEI)o 5
(4-95) (K.E.)-,, = (PE.).,- = total energy of the system
where g= 32.2 ftlsec The resulting equation will readily yield the natural fre-
H: total height of tower, ft quency of the system.
To estimate the period of vibration using the Rayleigh
The Rayleigh method applies only ro undamped sys- method the tower is considered as a series of lumpec
tems, but is found to be sufficientlv accurate for comDut- masses. These lumped masses are determined by consid
ing the fundamental frequency of process rowers. e;en ering the weights of
though towers have varying shell thicknesses down the
Iength that result in unevenly distributed mass and stiff- 1. Shell and heads
ness. The Rayleigh method is basically the conservation 2. Trays and internals
of energy, i.e., the total kinetic energy of the system is 3. Manways and nozzles
zero at the maximum disDlacement but is a maximum at 4. Insulation and fire proofing
The Engineering Mechanics of Pressure Vessels 205
''-
, Kil 7.58r. vE
u. = w, lI1 +n. )l
* *,lvr-*,\
-\ 2 6oDt
(4-100)
\2 I
Figure 4-30. A tower modeled as a sectionless beam with distributed lumped masses.
Mechanical Design of Process Systems
Table 4-14A
Vibration Deflections Based on the Coniugate Beam Method
,f
+ t'+ +*t*t +
{
+
I
+
I
+
l)
t-, I r_, L, l- 15 ' -1.
|
w.w"vqq
Xs, = o,
P
Mr r, t4r M, S, +A, = A, ia, + e,\ Pt*l.z: Itt
E'It ZE)l) \?/
Lr=Pr
M2 = W1L1 , RI+R' rr 19 /v. M,\ 52*41 :,A2
/a, + a.\ Pz * ir: : r:
-' 2 EzI: \E,I, Err,/ t?,
2
xL2: P2
XL2: 52
M1 : *L) I:
W1(L1
+ w2L2
-
-2
R, +R. M: irur,- M, \ 53+A.4 = Aj
14i14)
Ptl p+: pt
E:I: \E.I- E,I,/ \21
2
xL3 = Pr
XL3: S1
xLn: S"
lttrttl
w" w'r w.{
V Y "Y V V
Ds, = o,
Mi li M/Eili = Mrdx/Eili Pi p = M;d2xlE;11
Mr :0 Ir Mr Mr - ^ sr +Ar:Ar /e' +
ErIt 2El1 l?l
xLq:P1
xL2 = 52
: 2
W1(L1 + L2 + Lr) + W2L2 xL4 = P4
xL4: 54
lrl
\2 2l E:I: s5 Dsi
2
+w,l&+n.+!l xL5: P5
'\2 2l xL5 = 55
+ &l
-\21
+ w.lR.
I,
lvl" = lvl^ r+L", r,sr --, -, E t/\
M" /tut"-, tnl"
W,-l _jI L" = P^
Ik \EJ" 2
2
xL":S" S"=A"
M.+ r = M"+L" @ : ebrupt section break
k : n + (number of abrupt section breaks)
<Fw, y = (pi)(12) ft Infigureabove,k=n*1
t:l
Mechanical Design of Process Systems
Table 4-14C
Centroids ol Shell Volumes
- b(4a - 3t)
':'(T i
Conical Section
k-t--l
The vortex shedding frequency is given by process columns, because these vessels usually hare
many external, attached appurtenances. What is more
f,"D 0.2v
(4-101) commonly done with towers is to stiffen up shell sectiont
to offset ovaling resonance. See Chapter 5 for more in-
formation on ovaling.
where v : 45 mph or 66 fps
where s = Strouhal number : 0.2 for this application 1. If the critical wind velocity, V1, exceeds 60 mpf.
then a vibration analysis is not required. Very feu
To counter ovaling vibration, ovaling rings or helical cases of severe vortex excitation have been ot -
strakes are added. These normallv are not oractical for served for wind velocities in this ranee.
The Engineering Mechanics of Pressure Vessels 209
q- 4.
H/d ) 15 process columns I
+- lows:
+ LD,
\ ZW
vibration analysis must
be performed
i
w o15 (4- 105)
:
rather than brittle. The result is that while more risorous I : moment of inertia, ft
analyses are very helpful in determining design ciiteria. g= 32.2 ftlsec2
practical design procedures are simplifications of the
complex dynamic phenomenon used as'.quasi" static When Equation 4-108 is applied to sreel wirh a value of
criteria applied with elastic srress limits. :
E 30 x 106 psi we have
The Uniform Building Code 1982 [10] requires thar all
freestanding structures in seismic zones to be desisned
/ \, t::--
1= 17.65e v t0-") l[)'^1tz*o'' (4- l09
and constructed to with5tand a total lateral force t-base
shear) given by
\D",i !
r
r
and for E : 29 x
V: ZIKCSW (4- 106)
106 psi,
,: ln)' \,F-4D-
This is done in the Uniform Building Code by using the
ratio of the fundamental elastic period of vibration of the
tower, T, to the characteristic site period, T,.
(4 Il2r
Formulations used to determine the fundamental natu-
ral period ofvibration for seismic response vary as to the fEo, +
\100/ a.y
3
E
xllo
ollR
;llo
o||oll0
;''ll,ro
Ell'
"ll:
!l]f;
tl
oLJ
nt .€
ol 65
ol !o
Ll 6
6l
NI -R5C
R
ol
-
6l
212 Mechanical Design of Process Systems
Ao, A.y : differentials in the values of a and .y, VERTICAL DISTRIBUTION OF SHEAR
from the top to the bottom of each FORCES
section of uniform weight, diameter,
and thickness. 6 is determined from
For towers having an overall height-to-base-width ra-
each concenttated mass. Values of a. 6.
and "r are shown in Table 4-15. tio greater than 3.0, a portion of the total earthquake
force. V. shall be applied ro rhe top of the tower aciord-
In applying Equation 4-ll2 the following factors ing to the following relationships:
should be considered:
For,
n Ifa tower's lower section is several times wider in di-
h
ameter and shorter than the upper sections, then the
tower's period can be more accurately determined by ;<3.0,F,:0
computing the upper section's period, assuming that
h
the tower is fixed as to translational and rotational dis- 3.0<:
l)
< 6.12. F. = 0.07 TV (4-ll4)
placement. If a tower's shell diameter or thickness is
significantly larger than that of the supporring skirt,
the period calculated by Equation 4-112 may be h
6.12, F, : 0.lsv
overly conservative for earthquake design and a more ;>
accurate method may be desirable.
D For conical tower sections the Mitchell eouation can- where F, : total force applied at top of structure
not be used because of lack of data for the
'coefficients h: overall height of tower, ft
a, B, and 7. The Rayleigh equation (Equation 4-97) is D= diameter of tower, ft
more comprehensive and ubiquitous in application.
V= total base shear from Equation 4-106
The characteristic site period, T,, falls into the following where Lr,-, : length of section below shear force, ft
tlme: Ci - L,lZ for a cylinder
/.\f 2 ^zrrr' + rr'l
^21
0.5 < T. < 2.5 sec
c - lil lrl-+ foracone tsee Figure 4-33r
\+/Lri+rlr,+r;I
When T. is not properly established, S is taken as 1.5,
except when T exceeds 2.5 seconds, S can be determined For an illustration of seismic analysis, see Example
by assuming a value of 2.5 seconds for Ts. 4-3.
The Engineering Mechanics of Pressure Vessels 213
Table 4-15
Coefficients for Determining Period of Vibration of Free-Standing Cylindrical Shells
Having Varying Cross Sections and Mass Distribution'
nx h"
H H
I .00 2.103 8.347 l .000000 0.50 0. 1094 0.9863 0.95573
0.99 2.02r 8.12l 1.000000 0.49 0.0998 0.9210 0.95143
0.98 1.941 7.898 1.000000 0.48 0.0909 0.8584 0.94683
0.97 1.863 7 .678 l .000000 0.4'7 0.0826 0.7987 0.94r 89
0.96 1.787 '7
.461 1.000000 0.46 0.0'749 o.74r8 0.93661
0.95 1 .'7 t4 7 .248 0.999999 0.45 0.0678 0.6876 0.9309'7
0.94 1.642 '7
.O3'7 0.999998 0.44 0.0612 0.6361 o.92495
0.93 1.513 6.830 0.999997 0.43 0.0551 0.5872 0.91854
0.92 1.506 6.626 0.999994 0.42 0.0494 0.5409 o.911'73
0.91 1.440 6.425 o.999989 0.41 0.0442 0.4971 0.90448
0.90 1.377 6.227 0.999982 0.40 0.0395 0.455'7 0.89679
0.89 1.316 6.O32 0.9999't I 0.39 0.0351 0.416'7 0.888&
0.88 1.256 5.840 0.999956 0.3 8 0.0311 0.3801 0.88001
0.87 1 .199 5.652 o.999934 o.37 o.o2'7 5 0.3456 0.87088
0.86 1.143 5.467 0.999905 0.36 o.0242 0.3134 0.86123
0.85 1.090 5.285 0.999867 0.35 0.0212 0.2833 0.85105
0.84 1.038 5. 106 0.999817 o.34 0.0185 0.2552 0.84032
0.83 0.988 4.930 0.999154 0.33 0.0161 0.2291 0.82901
0.82 0.939 4.758 o.999614 0.32 0.0140 0.2050 0.81710
0.81 0.892 4.589 0.9995'76 0.31 0.0120 0.1826 0.804s9
0.80 0.847 4.424 0.999455 0.30 0.010293 0.16200 0.79t4
0.'79 0.804 4.261 o.999309 0.29 0.008769 0.14308 0.7716
0.78 0.762 4.1o2 0.999t33 0.28 0.00-t426 o.12516 0.7632
0.'77 0.722 3.946 0.998923 0.27 0.006249 0.10997 0.7480
0.76 0.683 3.794 0.998676 0.26 0.005222 0.09564 o.'7321
0.75 0.646 3.645 0.998385 o.25 0.oo4332 o.0826'7 0.7 r 55
0.74 0.610 3.499 0.998047 0.24 0.003564 0.07101 0.6981
o.73 0.576 3.356 0.997656 o.23 0.002907 0.06056 0.6800
0.72 0.543 3.217 0.997205 0.22 0.002349 0.05126 0.6610
0.7 | 0.512 3.081 0.996689 o.21 0.001878 0.04303 0.@13
0.70 0.481 2.949 0.996101 0.20 0.001485 0.03579 o.6207
0.69 0.453 2.820 o.995434 0. l9 0.001159 0.02948 0.5992
0.68 o.425 2.694 0.99468 r 0. 18 0.000893 0.02400 o.57 69
0.6'7 0.399 2.57 | 0.993834 0.17 0.000677 0.01931 0.5536
0.66 0.374 0.992885 0. l6 0.000504 0.01531 0.5295
0.65 o.3497 2.3365 0.99183 0. 15 0.000368 0.01196 0.5044
0.64 0.3269 2.2240 0.99065 0. 14 0.000263 0.00917 0.4783
0.63 0.3052 2.1r48 0.98934 0. 13 0.000183 0.00689 o.4512
0,62 o.2846 2.OO89 0.98789 0.12 0.000124 0.00506 0.4231
0.61 o.2650 1.9062 0.98630 0.11 0.000081 0.00361 0.3940
0,60 o.2464 1.8068 0.98455 0.10 0.000051 o.00249 0.3639
0.59 o.2288 1.7107 o.98262 0.09 0.000030 0.00165 0.3327
0.58 o.2122 1.61'7'7 0.980s2 0.08 0.000017 0.00104 0.3003
0. 1965 1.52'79 0.97823 0.07 0.000009 0.00062 0.2669
0.56 0. l8l6 1.4413 0.97 573 0.06 0.0000M 0.00034 0.2323
0.55 0.1676 1 .3579 0.97301 0.05 0.000002 0.00016 0. 1966
0.54 1.1545 I .217 5 0.97W7 0.04 0.000001 0.00007 o.159'1
0.53 0.1421 1.2002 0.96688 0.03 0.000000 0.00002 0.1216
0.52 0.1305 1.1259 0.96344 0.02 0.000000 0.00000 0.0823
0.51 0.1196 1.0547 0.959'73 0.01 0.000000 0.00000 0.0418
0. 0. 0. 0.
r- ,, E wA. . ,ruiE
t {#l
pB
'vi,cher rormura:
ti--S;;-
214 Mechanical Design of Process Systems
r"o= OO SMALL END that the cone-to-cylinder stresses computed by the equiv-
r-= OD LARGE END alent circle method are very close in magnitude to those
computed by more exact methods. Because of its close
approximate answers and simplicity, the equivalent cir-
cle method is normally the method used for treating con-
TRUNCATED CONE ical sections in towers. The method will only be outlined
here, as others l22l have already derived it.
Figure 4-33 shows how the sections of a truncated
cone and a conical head are approximated by an equiva-
lent circle, which is used to compute the section modulus
and moment of inertia. These formulations are used in
tower design and are demonstrated in the examples that
follow.
Conical shells used in tower sections have a half anex
,."=[*&l anglecv ( 30degrees. Whenh. ( 0.10H, rhecon..an
be approximated by considering two cylinders shown
with dotted lines (Figure 4-34). In pracrice, stiffening
rings must be used when required by the vessel code.
CONICAL HEAD
From Equation 4-1 at the saddle, ring compressive stress in shell over saddle
., = 10*13910)
[t# (, ,,1] [H#]
_ 7,828.981
(0.941)[0.375 + 1.56[(2 1X0.941)]0rl
or = 50.501 psi
.I I - cos(u4) I
lr - +
oo : -!D : iilllt'.t : 1,67i.50
:
I.990 sin (l t4) cos (l14,
I Ill;,
I ltt
I rtt
r rrtl
ltttr
r
I
SHELL AND
HEAT EXCHANGER
l
TTJBE I
l:'
_!_rt-!-!
'.. -7 -------+
"oo t$
I
'roP aEo 1.,
-'..
,tt
__ _.j--L___ -------lr:l-
___i-L__
____--L__
csruliiinny
,"r"*"" r*-------lli-I: |
-T I
BOTIOMBEO
-ti
.__i1___
-Tl
"-*.-,J:l
l*IU
F.-i-----+r l"l
N.RMAL LrourD .."r. I
ri
-f
--l FI
I
+-
Figure 4-36. Tower analyses ensembles.
The Engineering Mechanics of Pressure Vessels 217
9z
: 0.00256 Kz (IV)'z; V = 100 mph Effective Cross-sectional Area (Figure 4-37)
Bottom portion,
K,:2.58f
- l1\''' = o.rot
\900/
D. : 47.O il:'. + (12.'75 + r2) in.
ForZ)15ft, + (12.75 + in. + in. :
11) 12 107.50 in.
/ 7 \0286
K,'. : 2.58 l:l Gust Response Factor
\900/
I /szzor\
From Thbles and 5 of A58.1-1982, c = 1""'-- I { l27.ool
\r04.2921
I= 1.0
o,
'" = 0.00256(2.58)1" \900/
I (100F From A.58.1 Table A9, s: 1.00
I7
: 66.048
\0.286
(eoo-1 . z > ts rt
- 0.5 Ih (10.5X0.981)1104.2921
r:_=#=lu.t4J
f
sv (1.00X100)
6" THtcK
u'r","*-rr*I.$ INSULATION
'*"u'o''on
ffJ
Y
Figure 4-37. Effective cross sectional area.
214 Mechanical Design of Process Systems
1o4,292ft _
ir rs.:os\
+- _>..
9: \ t'? /: o.oe5:c
I 18.305
: 9.859
h 1o4.292
#----->r
ffi: ooo"
I- 0.0055, y= 6.145
*--->..
o =
oo.*,(fr-o|"'",0r,,.
s : l.l
For a tower with many attachments and connecting pip-
ing,
9:9zG
\+-{)-->
f00t '|
Figure 4-38 shows the wind pressure distribution q plot- l5.oort
ted along the tower length.
+;
Figure 4-38. Wind force distribution (q) along tower is para
bolic above 15 feet. Section wind force distributions are com-
Centroid of a Spandrel Segment bined into a force vector located at the centroid of the wini
section.
The centroid ofa parabolic segment is shown in Figure
4-39. Applying the general equation, Equation 4-117, to
our case we obtain
. t^ \
i _ tn + Ql llaqn I ngn.tl ,.t .t .r (4-117)
2(n + 2(}\ \dqn + nqn r/
From Figure 4-39 we obtain a general expression for the Using this equation we compute the wind force distribu-
composlte area, tion. From Figure 4-20, Cr = 0.6. Solving for sectior
properties we have the following:
t7 -7 .\
Af - ::l-----:l--I {dqn + nqn r). lb/ft (4-118) I7 \2t1
(n+q)
'" :66.048
aa 1.-l
\900/
Solving for the wind resultant force acting through each
centroid we obtain, Section A
F: AO"CI Z: 104.292 tt
- The Engineering Mechanics of Pressure Vessels 219
)|
) l^l
NINI
il:
$l- il-
o
-l;l I
st?l
.l+l IN
il- Nlcl
,_l-l
Nt
ql
1l
!L!
tN+
-,I
I
--t^ ^lN
^,1 rf.f ^ |
:l r',.'NS
@ YI 1F
qo N
Ntd N.
=x":
I
r< I Lr-'
N I.t re Nir++
t:-o
6
T-l
'-l
g^1 Jii
Q,ni+++
F
I
"l
ol I
i- Ll
u,- !L >
.el hl rN tL>
(JO (l
gilrl
E
(!
:l' l
:-^ ldl
<t)
o* l.
-___t91
tl : ^l q)
N
,Yl.l
6 tt
€ - t: !
rr llld
il-
F[?
'iYrl
Y
isl
dlB N' ll 'i! | -;-
tql
o.l' nu' xl
<ft
tl
d;d N
Rt
(G -----i:--
q --J9
'l
.9
i l'l
=N Nl.
o- tN lt
rl-
o \t-l
IN
IN
:* ll
()+
'N
220 Mechanical Design of Process Systems
--l 28o
/rr}4 rq?\0 Section D
9z 66.048 l:: : : 35.659
\vwi
q":48.852
q" (l.726X3s.659) : 6r.547
( / qR \0 286
e"-r : (r.726)ffiul- (66.048):44.573
9n- t 726t l:;l 166.048) = 60.461
Section B Section E
q" = 60.461
/ qn \o 286
(66.048):43.097
o. , = 1l.726] l^il (66.048) = s9.007
\vw/
+ (2X43.097)l
^_
(8.00x7)(60.461) + (2X59.007)l
: 4.009 ft
: 1.881 ft
Section C Section F
q" = 59 007
z: zl (2X7X59.007) + (2)(48.8s2)
,0, ,0,
eI(2) (7 )(43.o97) + (2X39. 943)
32 [ 1 + (7)(s9.007) (2X48.8s2) ] nl
32 | ('7 )|43.097) + 12)(39.943)]n
oo,
= 22.128 ft
The Engineering Mechanics of Pressure Vessels
Section G Section J
q" = 39.943
35.347
28o
/ro oo\0
C" | = (1.726) l-:::::l (66.048) : 38.378 : 35.347
\vlru/ 9n_r
(3.00)ft?X39.943) + (2X38.378)l At
:
q(2" - Z" ,) = (3s.347X15.0)
^ - 530.205 lb/ft
: 118.786 lb/ft ll :
2
z.so r,
F= ArD"Cr
Section H
- 2s6
/r?
9"_r : (1.726)
on\o
l:::-:::; (66.048) = 36.63s FB = (481.103) {'t]^ool to.ut = 3.055.004 lb
\7wl \ rl /
+
^_
(3.00)t(7x38.378) (2X36.635)l 0l
Fc : (2.468.&0) {l?lq.
^ (0.6) = t5,675.864 lb
\tzl
: ll).972lblft
FD = (6r0.73e) {toLtol ,0.u, = 3,2E2.i22 rb
\12 I
-i : e l2IlI38.3Z8rj i?I36.61s)l ,r nn,
321 (7x38.378) + (2)(36.635t'- "-' 501
Fr- = n65.919r l'07 ro.ur = Eel.8r5 rb
: 1.507 ft
\12l
FF = es6.ii3) (ue- ) ,0 u, : r ,5e5. r55 rb
\ 12 /
Section I
F6 = (u8.786) {'ol=to) to.u,t = 638.475 lb
q":36.635
\ Lz I
/, < 286 r," = (n3.e7z) (!Zi2t\ ,o u, : 612.6oo rb
q"_r = (1.726) lj':Y:l^^\o
t66.048) = 35.347 \ /'
\vUU/
(2.00)[(7)(36.63s) + Ft = (72.6si) I'ol=tol to.u, = 3e0.746 lb
^_
(2X35.347)]
r2 \ I
Solving for section moments we use the following ex- 1 "OD 16 BWG TUBE
presslon:
n- I
+ F,z. (4-8s)
Mc : 39,560.960 + (5504.421X43.5)
+ (15,675.864)(22.128) = 625,g7g.rt tt-tt
Figure 4-40. The tube bundle is modeled in banks ofconcen-
MD : 625,878.792 + (2r,180.285)(12.75) tric circles used to approximate the section moment of inertia.
+ (3,282.722)(6.430) : 917,035.328 ft-lb The tube bundle enhances the section stiffness.
ME : 9r7,035.328 + (24,463.007)(3.75)
+ (891.815)(1.881) : 1,010,449.109 ft-lb
Mn : 1,010,,149.109 + (25,354.822)(7 .00)
+ (1,s9s.155)(3.s2s) : I,193,555.784 ft-lb
Mc : 1,193,555.784 + (26,949.9'77)(3.OO)
+ (638.475X1.506) : |,275,367 .258 tt-tb For l-in. OD i6 BWG tube,
Mu: 1,275,367.258 + (27,5 88.4s2)(3.00) I : 0.0210 in.a : metal cross section
+ (612.600Xr.491) : 1,359,046.001 ft-lb
A: 0.191 in.2 : metal area
Mr = 1,359,046.001 +
(28,201.052X2.00)
+ (390.746X1.003) :
1,415,840.023 ft-lb K: number of tubes per circle
: n : number of circles
Mr 1,415,840.023 + (28,591.798)(15.0)
+ (2,849 .8s2)(',7 .s0) : 1,866,090.883 ft-lb Thus, from the parallel-axis theorem the composite mo-
ment of inertia is
Section Moments of Inertia
I: DKIG + AL)
Section a-rl+-in. t
Values of I are tabulated in Table 4-16.
,." " ffs+.sol, _ ile\"]
64 1\t2 I
: 0.756 rta For enclosing shell, :/a-in. t,
\12/J
I " - (54.00)41 - 23,676.070 in.'
Section b-3/a in. t 64'1s+.zs)o
)-l
Lt' = rr4.i35.44r in.a + 23.676.070 in.'
The shell and tube heat exchanger section moment of
inertia is approximated by a set of concentric circles of : 138,411.511 in.a
tubes. The concentric circle pattern approximates that of
the exchanger tube sheet. Using the parallel-axis theo- or for the total cross-sectional area
rem, we arrive at the section moment of inertia. Refer-
ring to Figure 4-40 we analyze the exchanger as follows: Ib : 6.675 fC
The Engineering Mechanics of pressure Vessels
-
Table 4-16
Values of I for Tube Bundle
16.364.299 t6,364.299
14.127 .503 30,491.802
12.104.531
10.284.531 s2,880.927
8.574.543 6r,45s.470
7. r38.100 68,593.s70
5.871.903 74,465.473
;
17.00
;i#.uZr
2 89.000 0.191 0.021 86 r.765.174 .7s,230.641
Section g-r3/r6-in. t
Section d-5/s-in. t
: 0.917 fta
: 1.208 fta
lso Deflections
! _ tb.292tt ltz.++s.+ngxo.zoztl
o Ia + "' (i.157 x lo") [- I
ydla h.
: 2.416 x 10-s ft
$ lo a"\
.-
(2.787
(8.oot
10'9
[(e,oss.oo+)(a.oo)
[ 8
rho=26.1925 in : 3.088 x l0 5 ft
tto:29.1109 in
51.o in
1o=
. (43.50)'? I (15,675.864X43.5)
Figure 4-41. The equivalent radii for the skirt sections. ' (6.402 10') [
(s,s04.421)(43.5) soo)l
+'32-) + Q2lOo
: 0.055 ft
64 L\ 12 / \ 12 /l
, (24,463.OO7)(3.75)
-J-rl (917,03s.328t
Section j
= 0.002
/sr.oo + :s.rrr\
'' \ 2 cos 6.934' /
(7.00), [(r,sqs.rssxz.oo)
t-
L' j'" l/so.uor\t
ll"" '" 1 '1hs.ozo\*l
"" "1 | = 7.846 fta (4. 635,10)L 8
64 L\12 12
i \ iI
(2s ,354.822)(7 .0O) , (1,0r0,449.109t
-2)
Values of each section's wind force, shear, bending
moment, and moment of inertia are summarized in Thble
4-17 for the entire tower. : 0.006 ft
1l
Table 4-17
Summation of Section Wind Force, Shear, Bending Moment and
Moment of lnertia
Section Fr (lb) Qi (lb) Mi (ft{b) li (fta) El (lb-ft'?)
a 2.449.4t7 2.449.417 7,118.113 0.756 3.157 x 10'q
(6.2e2)(8.00) [ (3,05s.004)(8.00)
(rJ8i ,, 109 t 6
3'l 2 I
,
t-
(27,s88.4s2X3.0) , (r,27s,367.25U]|
-)l (5,504.421)(43.s)
+ 39,560.960]
= 0.001 ft = 0.027 ft
: 2.626 x t0 a ft = 0.148 ft
I
" (15.00), I (2.849.852)( 15.00) ^ r70.542rr3.75,1{8q1.815)r3.?51
orn:-l
(J.276 l0'9 L 8 't- r+.::ot tolL .
(28,59r.798)(r5.00) , 1,415,840.023)l , _ q tr,u:S.:28
-21 (
_ r 24 .46 3 .001 )\ 3 .15
2
: 0.006 ft : 0.060 fr
226 Mechanical Design ol Process Systems
/po\ / ron^na \
\4tl - \?r((D" - D,XDa + Di,
(8r.2s2)(3.00)[(638.4 75X3.00)
^1
Dt
(5!45 x rO l- lzw\
6 - \"<n. * o, 1
[rrsO.Orrs+.orl I r6(54.50,{l2r{7.trA.rr:rl
^ _ 184.2s2t(3.00r fr6l2. 60x3.00)
-r8-(6J02^toq)L- 6
t {4x015) I = [.{015x t08J0)6J86.2t]
+
(27,588.452)(3.0)
2 ' ,-'-,ra'''" ,aol
ao., -'"1 - [tr(0lsxl08
2(1.8t4) I
d
: 0.052 ft o= 8,100 ! 161.009 - 42.57
: L(4)(0.37s) I
0.067 ft
, [ 16(54.75)(39.s60.960)(12) ]
":
rt \-,1 * \-,r = [r{ 0J7s x 10850X5-9{3J63)l
L/" L/ -1
o:t-l krso.ox+z.orl
o = -11,900.01 psi compression
[ (4)(0.62s) I
_ (1s0.00x42.00) _
*l
-t
(L 6) (43.25) (9 r7,03 s .328\ (12)
r(0.688)(8s.375)(3,&s.39r)
op
2(O.75)
4,200.00 psi
- t-l
I zt+:,+rs.oor I Section g-t:71u-in. g
lr(0.688t(85.375r1
! o:1-l krso.otr+z.orl
2,520.00 11,320.361 - 470.584
l(4)(0.813) I
13,369.777 psi tension < 18.800 psi : o,
. 6(43. 625) (1,27
116(43.625)(
1. |,27 s,3
s.367
67 .258) (r2)
.258)(t2
- 9,290.944 psi compression =[
"(0-813X85525X3'667140
(150'0x42'o) Izt+s.ozl.+ssr I
2(O.62s)
= 5.o4o.oo osi - t"(0-8t3l8s.6rtl
228 Mechanical Design of process Systems
q= 1 ,937 .2'l ! 13,319.972 - 439 .64 D": OO of base plate : 106.75 + 2(2.375)
: 111.50 in.
o: 14,817.602 psi tension < 18,800 psi : da
D; : ID of base plate = 111.50 - 2(6.j 5)
q: -ll,822.342 psi compression : 98.00 in.
All section stresses are less than the allowable stress of -rn2 - nll =
18,800 psi. Thus, the tower thicknesses are acceptable. A. '-o------1- 2221 .302 in '
4
fr4l
" "",'- "'
I. = -/na : 3.059.323.380
64
_ 8MD,, W.
First we determine the size and number of anchor bolts ' N(Di + Dl) N
(4-.+ l
reouired.
_ (8)(12X1,866,090.883X1 1 1.s0) 63,815.727
(4-42) (.12)(22,036.250) n
^,: [+-*]/*".
Using an A- 193-87 high-strength bolt with an allowa-
= 70,219.061 Ib
I
(4_1j
= 1.615 in.'? < 1.680 in.2
lo.
. r( 106.75 )
20,000 = allowable working stress for steel, psi
lJOrr sDaclng
t2 I ,000 psi
2.939.611
weld size w (2)(9290.0s)
= 0.158
t
BTHK = t4.75t l -I v?6q
fr
" '""|
lqrlo
=
r
1.613 in. Use at least a 3/re-in. weld on each side of skirt.
L 20.000 I Anchor bolt torque is determined by Equation 4-66.
-
For lubricated bolts with Fel-Pro C-5A,
Make base t 15/8 in. thick
C: 0.15
Solving for the compression ring thickness using Equa-
tion 4-63 we have T: (0.15)(1.62s)(64,605.803) : 15,747.664 in.-tb
or
[ (b4,605.803x2.37s) I ^-^-
[4{20.000)14.75 - r 1.25I 1.625'1].j T: 1,313 ftlb with torque wrench
@-64) t, x
q
"00X
t'z /
1.500 ':a'4? 'z0"000)l
t = Q.672 in. < r3lroin. skirt thickness at chair
Fi = 64,605.803 Skirt thickness meets chair ring reaction
cnterla.
Grr : 9.00 in.; G* = 4.25 in. A sketch of the skirt and chair design is shown in Figure 4-42.
230 Mechanical Design of Process Systems
f 11td BoLls
^-$.-87
Section e
ul"r"'", - o.rr)
t : s.857 in. = 0.488 ft -
'-2-
24oo
= l2.ooft
8oo:4.ooft Section m
i," =
'"2
_ 7.00
Ld = 1.3'75 + 4.00 : 5.375 ft 2
Section n
v,=17s=0.875ft
"2
Li : 4.00 + 0.875 : ft - 3.00
4.875 '"2
Section i : +
L13 3.50 1.50 = 5.00 ft
'2 :350:
Section o
t, l.75oft
: 6.247 rn.
Section p
t = 0.521 ft
"2
3.75
= l.X/l tf L,. : (3.00 1.546) + 9.420 : 10.874 ft
Lro : 17.00 - 9.420 : 7.580 ft
Mechanical Design of Process Systems
l
H $
.'
R
s;^x^x^^x^^^;^;xx q P
9
H
9
E .-
3
N
9
€
|
! -
H: i :.
q ;iI qS
p .. 9
Vortex.lnduced Vibration
* : 6r?ffir,rral : 4.523 x ro-5
Referring to Figure 4-43 we have the following:
M.:0
687 ,472
: 1.038 x 10 4
(4.32 x 10)(1.533)
Mb: (0.423)(2.95 8) : 1.251 *o-t
M" : 1.251 + (1.814)(6.50) : 13.042 kip-ft
Ma : 13.042 + (15.201X5.00) : 89.047 kp-ft
8s3,091
M" : 89.047 + (16.192X13.00) : 299.543kip-ft ,4 ?) lft"t, r Sll' - 1.288 ' 10-4
"
Mr : 299.538 + (29.004)(13.375) = 687.472 ktp-ft
Mc : 687.46'7 + (30.813Xs.375) : 853.091 kip-ft t't _ 1.O24.t26
r '
Mh : 853.086 + (35.084X4.875) = 1,024.126 kip-ft
=
A Jx) = 1.546 to
Mi - |,024.r2r + (36.016)(2.625) = 1,118.668 kip-ft
Ivl = 1,118.663 + (37.s32)(2.229) : r,202.322 ktp-ft l'118 668 :
M,. = ,2o2 .322 + (37 .913)(6 .646) : | ,454 .292 ktp-tt !" 1.689 r lo '
1
Mt : |,454.292 + (43.171)(800) = 1,799.660 kip-ft
'- (4.32 x l0e)(1.533)
M. : 1,799.660 + (4s.028)(5.375) = 2,M1.685 kip-ft
M" : 2,041.685 + (47.662)(s.o0) = 2,279.995 Y,tp-ft <ro -
| .202.322
.lo"ltt.gtot
M. : 2,2'79.995 + (50.684)(3,046) : 2,434.379 kip-ft 14.32
MF = 2,434.379 + (s 1.937X10. 874) : 2,999. r42 kip-ft
Ms^, = 2,999.142 + (63.816)('7.58) : 3,482.867 kip-ft I L<L )A) _14?|v|^4
' (4.32 r t0")(0.917)
M' (30
- = E.I,
{, , x tO")tb/in.,(t44tin.b/ft,
r,, - f'7q9'660 4.tt2 t ro "
: (4.32 x 10,X1 .013) -
4.320 x lOe \blf€
!,,. 204l'685
1,251 : 3.830 x 10-7 ' -4.258xlo1
(4.32 x l0)(0.756) \4.32 ^ l0')(1.1l0)
13,042
r 1_70 00<
169 x l0 4
(4.32 x 10)(6.675)
= 4.523 x l0-1 <r4 - (4J2 /' 10\1108) - -' '
&= = r.345 x 1o 5 ,
r,< - 2,434,379
er+;ffi1j3,, (4.32 v l0)( 1.s32)
--r.u78xl0a
The Engineering Mechanics of Pressure Vessels
+
r. . (4.121482'q67
(,1.369 3.678)( r0-4)
l.l5o , Io a
Sr+ (3.046) = 1.226 x l0 r
/ 10)(7.01l)
/. + l,l,\ , (3.678 x 10 1+ 9.902 x l0 ) (10.874)
^ = 18,*t
J, l-l Li Sr: =
2
\rl
(3.83 x l0 )(2.958) : 2.538 x 10-'
5.665 x l0 '
2
Sro : (9.902 x l0 5 + 1.150 x 10 r)
(7.58)
t(3.83 x l0 ) + (4.523 x 10 ?)l
S:= (6.s0)
: 8.111 x 10 1
2.?15 x 10 6
3.476 x 10 5
: 8.111 x l0
Ar a
5"') -
,-1
l0 N+: 1.726 x l0-22 + 3.476 x l0-56 :: 1.730 x l0-'
b.2sr -
+ ) /\
s8 : (1.546 L689)(10 e.625) : 4.246 x r0-l p'\21
= lA'+ A'.
lr-,
s, : (1 689 + L457Xl0r)
e.2zg) = 3.506 x r0 I r,. : (ry, : 3.s74
" ro ,
)(7.58)
8.111 x 10 1
+ 3.349 x 10
+ 10-) (6.646) P,. :
Sro =
(1.457 3.671)( : 1.704 x 10 r ( t ( 10.874)
.262 x l0-2
Sr 9!\?9(8.00) : 3.113 x ro l
r
t,,\.349 x10r + 4.575 x l0 1 (3.046)
s,, =
(4 112 + t258xl0 ) (s.37s) : z.z4g x 1.o l
r , I
: 1.207 x l0 r
234 Mechanical Design of Process Systems
= 2.283 x l0 |
_n
P":( |.726 x. t0 2
+ 1.730 x 10
( 13.00) Section weights and displacements for computing the
, ) tower's period of vibration are listed in Table 4-18.
The first period of vibration, T, is determined as fol-
: 2.246 x l0-1 lows:
x ,\ l0-
,=r"\E
1.730 10-, + 1.730
P,:(
t (s.00)
(4-98)
: 8.650 > l0-2 leDwv
--__-4
- The Engineering Mechanics of Pressure Vessels
Table 4-18
Tower Vibration Def lections
Dellection
Section y (in.) w (lb) wy (in.lb) Wy' (lb-in.'?)
14.683 423 6.2 r0.909 91.194.771
14.064 1,391 t9 ,563 .024 5,134.3'70
2'7
_ (34.7 64)l(7)(6r. +
s47) (2)(54.8o8)l
:2,087.559 . Finally, if the Reynolds number is greater than approx-
(2+7) imately 350,000, a vibration analysis is not required, be-
cause in these regions the vortices break-up. In our case.
/r rr nn\
F= t2087.559r l'-;;""1 (0.6) = l? )55 000 rh-
\ rz I Nn" _ DcVp
or
where D. : effective wind diameter at top r/: or r/+ o:
F*,0 : 13,256 lbr > 1,608.56 lbr : F" tower : 127.0 in.
critical wind velocity = 15.292 ft/sec
Since F*;"6 ) F,iu.",ion, the wind stresses are greater 0.071 lb./fc
lhan those at resonance vibrarion, so no further vi6ration 1.285 x l0 5lb-/ft-sec
analysis is required.
If the vibration amplitude force had been greater than /rzzoo\ "f fi tts.292t /n\ rh
the maximum wind force, further investigation would lj:i--:,
\ Lz " I l:' lrO.Orr,)
\sec/
l
have been required. Dynamic stresses ar the-crirical wind
N"" : 11-
: 894,215
velocity can be approximated by taking the ratio of the (1.285 x 1o-1 ;lb.
It-sec
vibration force amplitude, F", to the maximum wind
force amplitude, F1,, and multiplying this ratio by the
bending stress term in Equation 4-29. The pressure Since N*" : 894,215 > 350,000, a vibration analvsis is
stress, which is a primary stress, and the weight load not required since we are outside rhe range of vortex for-
stress in tension are unaffected. Shear is not considered matron.
in the equation because it is almost always negligible. Vortex formation has been observed at NRe > 3.5 x
Defining the ratio of vibration force amplitude to the 106, but wind velocities encountered would not cause
maximum wind force amplitude as R. Equirion 4-29 be- Reynolds numbers that high.
comes
The term G7 is given in Table 8 of ANSI-A5S.1-1982 where G21u"ry = average value of G7 across the height of the
rhat is determined by the following expressions: tower.
Now, for many, if not almost all cases, For zone 4,2 : t,I = 1, K : 2.0, W : 15,571 lb
Since the tower is not of uniform thickness, equation
c>G, 4-108 cannot be used. Either the Rayleigh equation
(Equation 4-97) or a modified form of the Rayleigh
This is certainly true as one moves up the tower in com- Equation, the Mitchell Equation (Equation 4-112), can
puting Gz. It can be safely said that only be used. For illustration purposes the Mitchell
Equation will be applied and then compared to the more
G ) Gau,er accurate Rayleigh method.
238 Mechanical Design of Process Systems
Using values in Table 4-15 we determine the values to Solving for V we have
be used in Equation 4-112. Connecting piping exerts a
concentrated load o12.7 kips at the support point mid- v: (1)(1X2.0X0.078X1.434X1s,571) : 3,484.0 lb
way in Secrion @-@. using the values in Table 4-15 we
construct Table 4-19, Using Equation 4-114 to find F,, we have the following:
111 \2
F- : (V - Ft) -YYhY : 13.484 - 522.60)
w)hv
: 0.673 sec/cycle 515,380
u00/ D*'*"
i=
I
Using the more accurate Rayleigh method, Equation
4-97 , the value of T is : 0.0057 wr,hy
T : 0.734 sec/cycle
in which the Mitchell Equation is in 8.3 % error (which is To solve this equation we must set up the table shown
quite normal). For application ofthe Rayleigh Equation in Thble 4-20. After determining the values for W, h",
see Examples 4-2 and,4-4. W)h). F,- and V, we solve for the seismic moments using
Now, we must solve for the bending moments induced Eouation 4-116:
by the seismic forces. First we find the base shear using
Equation 4-106. To accomplish this we have the follow- M- Vxi Ly_r + Fx Ci Dt
rng: M, - (0.30)(5.083) + (0.49)Q.s42) : 2.770 2.'770
M3 : (0.49)(7.50) + (1.19X3.75) : 8.138 10.908
1""'r l:125 Ma : (1.68)(5.417) + (0.18X2.708) = 9.s88 20.496
(r.i" - l.U
= 1.t25 < t.5... K - 2.0 M5 : (1.86X2.00) + (0.09x1.00) = 3.810 24.306
M6 = (1.95X8.00) + (0.28)(4.00) = 16.72o 4r.026
M7 : (2.23X4.00) + (0.16)(2.00) : 9.240 50.266
Flexibility facror = C: -]- = O.OZA Ms : (2.39X10.00) + (0.32)(s.00) : 25.sm 75.766
15(l1tr': M, : (2.71X10.00) + (0.22X5.00) 28.200 : 103.966
M16 : (2.93)(9.083) + (0.23)(4.542) 27.658 : 13r.624
The characteristic site period, T,, is determined by a Mrr : (3.16)(8.00) + (0.10X4.00) 25.680: rs7.304
soils consultant to fall within the following range: M12: (3.26)(2.917) + (0.01)(1.458) 9.524 : 166.828
0.5<T.<0.55
To be safe, we will use the lower value of 0.5. Now,
The wind moment for an 80-mph wind was calculated
T O ?14 ro be 106,716 ft-lb. Since 166,828 ft-lb > 106,716 ft-lb
_l s - ;;u.)u = 1.468 > 1.0 in which Equation 4-tl3b seismic phenomena govern.
The skirt and base plate analysis is identical for seis-
applies. Thus, we have mic and wind analyses. Just as in Example 4-2, the seis-
mic forces and moments are used instead of the wind
forces and moments. In the case of this tower a thicker
: + 0.6 El El'
\TJ-
S 1.2 0.. base plate was welded on, the number of gusset plates
\TJ
were doubled, and anchor bolts of a high strength alloy
S: 1.2 +0.6(1.468) - 0.3(1.468f were used to meet the seismic criteria. In an earthquake
zone other than zero, a comparison of seismic to wind
s: 1.434 should always be made.
t
-
The Engineering Mechanics of Pressure Vessels
Table 4-19
Numeric Integraiion ol Period
ot Vibralion, T sec/cycle
whv Aa P WAa +
E l9l'to"
kipsfft H
WAB/H
u0i
0.380 0.0412
0.000
A:0.,140 B:0.261
240 Mechanical Design of Process Systems
Table 4-20
Wind Load Distribution
x"'o-ri. Lzt ztttp w\-r
kips !r w,h, Fx v, *,1-u *,oTi
72 ffi
+
\: 0.49 2.770 2.770
3.289 63.2 207 .86 l. l9
+ .': 1.68 8.138 10.908
0.542 56.7 30.73 0. 18
-\3
1.86 9.588 20.496
+ ->s,* i 0.284 53.0 15.05 0.09
+ +2s.k-
1.0 48.0 48.00 0.28
1.95 3.810 24.306
+ 1.0' -l
0.645 42.0 27 .09 0.16
2.23 16.720 41.026
+ $
1.109 1.5 l .66 0.01
3.26 25.680
9.524
157.304
166.828
15.571 5 15.38 3.270 3.27
+
@
-+
+ o
5l
The Engineering Mechanics of Pressure Vessels 241
A phone call from a plant manager reveals that an ex- D" : [24.50 in. + 2(4.5) in.] + [3.50 in, + 2(4) in.]
isting tower needs to be analyzed for wind vibrations.
The tower was designed, built and installed overseas and D. : 33.50 in. + 11.50 in.
is vibrating so badly all the natives drove off the plant
site in fear of the tower falling over. D" = 45.00 in. = 3.75 ft
The tower with the appropriate wind load distribution
Zone 7-Section I
is shown in Figure 4-44. The tower is divided into wind
zones at 30 ft,40 ft, and 75 ft and according to shell di-
ameter and thickless. The variation of wind zones based D. : 136.625 in. + 2(4.5) in.l + [6.625 in. + 2(5) in.]
on the shell diameter and thickoess is necessary since the + [3.50 in. + 2(4.5) in.] + [6.625 in. + 2(5) in.]
tower's section moment of inertia will vary.
To begin the analysis we start with defining the effec-
D" = 46.625 in. + 16.625 in. + 12.50 in. + 16.625 in.
tive diameter of each section as illustrated in Figure 4- D. = 104.875 in. : 8.740 ft
45. Thus we have the following:
Transition Piece-Section 2
De:[32in.+8in.] + 14.5 in. I 2(2) in.l 48.5 in.
: 4.042 ft
Referring to Figure 4-45,
Zone 3-Section 5
5" : [25.25 in. + 2(2.5) in.] + [4.5 in. + 2(2.5) in.] ,"r:( 18.375 + 12.375
+ [3.5 in. + 2(3.5) in.] 2 cos 26.565'
D" = 30.25 in. + 9.50 in. * 10.50 in. req 17. 190 + D.q = 34.380 in
Zone 4-Section 4
Iz : 5791.250 in.a : 0.279 fta
D" : 25.25 in. + 2(2.5) in. = 30.25 in.
\ : 1, - : r,400.ri2 in.a
D" :
Kz4.sq4 e4.00)11
2.521 ft
9" = 125.25 in. + 2(4.5) in.l + [4.5 in. + 2(2.5) in.] + \ = L64Kz4.i5)o - (24.00)11 = 2,133.181 in.a
[3.5 in. + 2(3.5) in.]
l\'
* "; %: *,*
".". ' '"*
T(
Sections I and 2
y : f,11zt.z+ty4 - (27.72141
= 0.565 fll
+ (268.547x1.5
- 8) - (J4s.41) l:l Mr = 59,362.095 + 206,693.985 + 4,450 + 62,621.t41
\21 + 27,478.86s + 9,287.16r + 10,161.920
+ 1,006.128 + 6,555.227 + 3,016.168
+ (3e.328) lll
\zl Ma = 390,632.690 ftib
M: : 39,586.031 + 139,216.401 + 4,450 Sections 8 and 9
+ 26,163.00 + 9,380.176
+ 2,551.197 + 1,397.&4 + 19.664 :
Me Q 88.425)(75.292 + 2.9 17) + Q,690.r72)(.1 6.833
M5 = 222,7&.113 ft-Ib + ).:917) + 4,450 + (1,453.50)(43.083 + 2.9t7)
+ (721.5s2x38.083 + 2.9t7) + Q68.547)(34.583
Section 5 and 6
+ 2.9t7) + (349.41r)(29.083 + 2.917)
+ (39.328)(25.583 + 2.917)
+ (522.662)(r2.sQ + 2.917)
M6 : (788.425)(50.209 + 8.0) + (2,690.172)(5r.75
+ 8.0) + 4,450 + (1,453.50)(18.0 + 8.0) -, 17s4.042\A + 2.gt.l,) + ,Zt+.OOr(2.717\
+ (72r.5s2)(13 + 8) + (268.547X9.5 + 8) \z I
+ (349.411)(4 + 8) + (39.328X0.5 + 8)
Ms : 61,661.931 + 214,541.217 + 4,450 + 66,861.00
/^\ + 29,583.632 + 10,070.513 + 11,181.152
+ rs22.662t lll + 1,120.848 + 8,079.832 + 5,215.709
\21 + 401.003
Mo : 45,893.431 + 160,737.'177 + 4,450
+ 37,79r.00 + t5,152.592 + 4,699.573 Ms = 413,166.837 ft-lb
+ 4,192.932 + 334.288 + 2,090.648
M6 = 2'7Q,892.241 ft-lb
(99o.o99xro)
*,
ff _ _.-l:
106,090.721 _ 001658 fr r, 'T ".. ,., . "l = 0.t3055 rt
622.662t8t+ 222.764.1131 .
(sf ks,szz.rso)(s)
-'6 -
(56.50X9.083) ft6.833.596)(9.081)
(4J?6 x t06x0x9)t 3 14.176 x t0\0260)t 2
(388,738)(8) t+'t.OOt.SZ]
+
8 *' 2 l-= 0.00989 ri +
(572 730x9 083)
+ 2io.8s2.z41f = 0.143 ft
6l
(8f ko,:ro.s:+Xs)
Af?6 x ro)(oJ??)-t 3 (71.583X8) ft2,+oo.rz r'11r;
(4.176 x 10)(0.565) Lz
+ t522.66UG)
*' Zzz.lo+.tnl
8 2 = 0.0r I 15 ri
)- +(7s4'y2)(8) + n2,s4i.481: 0-09071 rt
6l
-'* - 17 )
3.58r(2.g t67 [rS. rOO.:OStrZ.l rOrr
(4.176 x 10)(0.260) 1+, tuo x to"1o,+ogt 2
+
(572.730)(9.033)
+270,8s2.24rf : 0.012 ft * 121!!!)t2!tfr)+
6) J90.632.691 = 0.05r74 ft
82j
(8), 6 : total deflection at top of vessel
[tr.+oo. rzox
t' - ,aJ6 v 1g\05sr[ a r
: \-
.l 2y, + \- LtA, I
: 0.743 ft
(7s4.042)(8) 332,s47.481 :
+
821 + 0.00507 ft
: 8.910 in. at top for static gust wind
(2.gt667tt [rt. roo.:os x z.r reor I
18 - t+.tzo x ronxo,+oott :
Referring to Figure 4-47 , we determine
t274.u2\2.st667, * 39!.632.69] Mr :0
- 821 = 0.00201 ft
-, --- 6
{388.738x8)
+, l+/,uo
, ,,
".1.s2l = 0.09560 fr T:M/I
I
(48.s0x8) ft6.310.934)(8) Mt 32'786 :
' (4.176 x t0')(0.177)[ 2
T,
'' = tz - rv.5t2.54
0.279
246 Mechanical Design of Process Systems
T^ :M.: 75,t75 _
I,105,514.71
I3 0.068
:Mo: 17 4,987 _
T4 1,698,902.91
r4 0.103
M'= 265,104 _
= I5 0.139
|,907 ,223.02
:Mu= 361,288 _
T6 2 ,041 ,r7 5 .14
I6 o.t77
T- :Mr: 487598 _
3,7 50,7 53.85
-q-
11 0.130
---->4.71k
:Mt= 633,032 _
T8 | ,r20 ,4r0 .62
Is 0.565
'748,922
:M,: _ 1,872,305.00
-------e Ie 0.400
782,@
Tro =M'o=
Ie n /An - 1,956,660.00
s. = M dx/I
--r.2.71k +
Sro
_ (1,956,660.00 1,872,305.00)
(1.458)
------€)
----------->2 .48x :2,791,315.49
(1,872,305.00 + t,r20,410.62)
(5.458)
2.01k 2
------e
-> 8,167 ,120.93
-> :20,022,921.55
-------€
_ (.3,75O,7s3.8s + 2,041,175.r4)
(8.862)
S7
344k
-----e
--> : ,5 664 017
(2,O41,175.14
15
+
-----l3.54K Sr: 2
1,90',1 ,223.O2) (8.00)
(117,512.54\
Po: (56,645,395.32 + 72,438,9U .96)
Sz: 6.961\
2
(8.00)
a
Q9,002.40 516,337 ,533.2
Pz=
(112,620,414.7 + 113,029,417.1)
56,&5,395.32 + 15,793,592.&
2
(6.961)
72,438,987.96
785,374,239.6
72,438,987 .96 + 16,227 ,566.69
88,666,554.65
88,666,554.65 + 18,579,266.73
: 2,034,868.99
rc7,245,82t.4 ttto
lL2,620,414.7 + 409,002.40
ps : 39,557,941.95 + 172,393,524.9
rt3,o29,4r7.r = 211,951,466.9
pt : 211,951,466.9 + 388,274,143.9
: @0,225,61O.8
o=(*, )u t4 =
:
ffi0,225,610.8 + 516,337,533.2
|,116,563,144
4,105,659
-..T 3.40 d
,7 53
: 0.950 ft : 11.405 in.
(43' x tOt
q6
3,1,39,457 ,578
:
L= 16";-=re.24tt
(4.32 x l0e)
0.727 ft = 8.721 in.
":z1
|,1t6,563,r44 :
From Equation 4-101, at resonance
Table 4-21
Values for Determining Tower's Period of Vibration
Detlectlon
1_!:-! w
13.59 4.7 r0 64.008.90 869.880.95
1.289.33 14.711.26
8.72 710 23.63t .20 064.06
5.r2 2,480 .60 65.01 l.7 t
3. l0 2,010 6.231 .00 19.316.10
1.67 2,230 3.724.10 6.219.25
0.59 3.444 2,029.60 1.197 .46
0.11 1 54n 389.40
0.01 1.9i0 19.10 0.19
Dtr : 114,02s.23 Dwy, = r,182.443.81
First Period of Vibration, T
ILwy' i 11. t82.441.8r)
r = z,r \/etrwv = zr 1/(386.4X114,020.23, = 1.03 sec/cycle of= O.9j Hz
tou J t
The Engineering Mechanics of Pressure Vessels
I50 It TAN/TAN
T
lrot
I
l-
I
i\
i-8
A, f.i.,---
" i"['i"'
wi=3 46 tl
tttl
ABCD
Combining with fluid weight the total weight, v: 2(270,966.86) : 54r,933.73 in.3
The total liquid volume above Section B{ is vru : 484,029.86 in.3 + 3228.72 in.3 : 487,258.58 in.3
Vr : 10,611,534.46 in.3 + 541,933.73 in.3 The total metal weight above Section B-C is
= 11,153,468.19 in.3
-^^-..^ lb
Metal Volume Above Section B-C
Ws = (487.258.58)in.i (0.2833)
ft3
- r(7?.s\2[rt * - tr#J :
The total liquid volume above Section B-C is
u B7,oe7.7e in.3
w,- _ (11.151,198.I?)in.r
|
(62.4) E (1.4)ftj
The inside volume was determined from calculating the rn.3
,723
liquid volume as being 135,483.43 in.3 Thus the metal : 563,869.78 lb
volume for a single head is
The total weight above Section B-C is
Yu = 137,097.43 in.3 - 135,483.43 in.3 = 1,614'35 in.3
Wr = 137,894.18 lb + 563,869.78 lb : 701,763.96 lb
I h = 0'50 in'
*f0a6.'12"\ - 0.551I
(12x3.46X0.50)
V= 20,205,196.86 in.3
(12)(3.46X0.s) + 2(0.sxls - 1)l
The metal volume is
= 0.59 in.
Vr,a =
:
20,205,196.86 in.3 - 19,963,181.93 in.3
AIso,
242,O14.93 in.3
where \ = (l)[9"
\w,/ \o*i
1{OTATION
^
= (u*;)(':'-'*f': oo' dimension from saddle centerline to tansent of
head (Figure 4-2) ft, in.
w" : (3.46)(12)(0.25 + 0.91 (0.41)'?) : 16.74 1n. effective area of concrete, ft2
b plate width (Equation 4-15) in.
10.392( l2) BP bearing pressure, psi
n= = 7.449 -use E stiffeners BPW base plate thickness, in.
16.74
C constant for bolt torque (Equation 4-66), di-
l5l'ooo mensionless; friction coefficient (Equation 4-
-8=
o^, : 43.875.0 osi 89) dimensionless; structure period response
factor (Equation 4-106) dimensionless
FB = (8)[7.5 + 2(0.597X0.s)](43,875.0) CA: corrosion allowance, in.
: 2,842,W7.0 lb c.: critical damping factor (Equation 4-90), di-
mensionless
Since 2,&42,047 lb > > 351,000 lb, eight stiffeners are cs= compressive strength of concrete (Thble 4-7),
sufficient. psi
The Engineering Mechanics of Pressure Vessels 253
D= diameter (Equation 4-27), in.; dynamic magni qF : velocity pressure of wind on structures (Equa-
fication factor (Equation 4-9 1), dimensionless tion q-i6), rcJf(
Dr : effective wind diameter (Figure 4-22), in. g:o : basic wind pressure at 30 ft, lbrift'?
D. : outside diametet in. S : Strouhal number used (Equation 4-102), di-
D, = inside diameter. in. mensionless; structure size factor (Equation 4-
E: welding joint efficiency (Table 4-2), dimen- 82)
sionless: modulus of elasticity. psi T= bolt torque as defined (Equation 4-66), in.-lb
F= wind force (Equation 4-94) Tr : exposure facior for wind (Thble 4-11), dimen-
F; : bold uplift force (Equation 4-39), lbr sionless
f. : natural frequency of a ring (Equation 4-100), t6q : compression plate thickness (Equation 4-63),
Hz in.
f, : vortex shedding frequency Equation 4-101, Hz t8 : gusset plate thickness, in.
Gr : dynamic gust response factor, dimensionless tr, : head thickness (Equation 4-7), in.
Gg = gusset plate height (Equations 4-39 and 4-63), ( : shell thickness (Equation 4-1), in.
ln. vo : theoretical ovaling velocity (Equation 4-102),
G* : gusset plate width (Equation 4-63), in. mph or ft/sec
H: depth of vessel head (Figure 4-2), in. vr : first critical wind velocity (Equation 4-94),
I: moment of inertia (Equation 4-24), in1 ;occu- mph
pancy importance factor (Equation 4-106), di- v30 = basic wind speed at thirty feet used as design
mensionless wind speed (Equation 4-75), mph
I" = moment of inertia of effective area of con- W= vessel weight (Equation 4-40), lbr
crete, in1 x.t : static deflection of a spring acted upon by a
K: coefficient of buckling for shear (Equation 4- force (Equation 4-90). in.
15 and Figure 4-3), dimensionless xO = displacement as a function of time (Equation 4-
k: dimensionless parameter for concrete (Thble 4- 90), in.
7) y= total lateral displacement of tower (Equation
K' = plate buckling coefficient (Equation 4-15), di- 4-88, Figure 4-21), in., ft
mensionless Z: elevation or height above a reference point,
Kz : velocity pressure coefficient (Thble.4-9 and such as the ground (Equation 4-74), ft
Equation 4-78) Z= reference height in which basic wind speed is
L: length of a horizontal vessel from seam to considered (30 ft or 10 m), ft
seam (Figure 4-2), ft, in.
L" = effective column length (Equation 4-19), in.
M: bending moment, in.lb, ft-lb Greek Symbols
m: bolt spacing (Equation 4-39), in.
Mc : compressive bending moment in the shell of a a : ir - (tr 1180)(012 + B/20) (Equation 4-6), de-
horizontal vessel (Figure 4-2), tt-lb grees
Mr : tensional bending moment in the shell of a hor- B= (180 -012), degrees
izontal vessel (Figure 4-2), ft-lb A= t
(?./180x5di 12 30), degrees
N: number of anchor bolts (Equation 440), di- 6; = lateral translational deflection oftower, (Equa-
mensionless tion 4-88 and Figure 4-26), in.
: d : angle of contact of saddle with shell (Figure 4-
P buckling load for compressive loading (Equa- 1), degrees, radians; rotational displacement
tion 4-18), lb6; probability of exceeding wind of tower (Figure 4-26), degrees
design speed during n years (Thble 4-11) and
Appendix A), dimensionless
\ = (t/bxE/ocil used in Equation 4-18, dimension-
less
Pu : annual probability of wind speed exceeding a p : radius of gyration : (I/Af .5
given magnitude-see (Appendix A), dimen-
6 = general term for stress, psi
sionless o" : allowable stress values (Table 4-3) psi
R: mean radius of shell (Figure 4-2), ft, in. d. : allowable stress induced on concrete (Equation
Ri : inside vessel radius (Equation 4-13), in. 4-40), psi; general tern for compressive stress
& : outside vessel radius (Equation 4-73), in. (Equation 4-16), psi
r: inside radius of vessel (Figure 4-2), ft ogp = critical stress in a flat plate defined in Equa-
Q: reaction at saddle (wl2), lbl tion 4-15, psi
Mechanical Design of Process Systems
: elastic buckling stress (Equation 4-16), psi; 10. International Conference of Building Officials, Unl-
28-day ultimate compressive strength of con- form Building Code, Whittier, California, 1982.
crete (Thble 4-7), psi 11. American National Standards Institute, Inc., "ANSI
oE : stress due to weight, lbr A58.1-Minimum Design Loads for Buildings and
oP : pressure stress induced by either internal or Other Structures- 1982," New York.
external pressure, psi; longitudinal stress in 12. Kuethe, A. M. and Schetzer, J. D., Foundations of
Equation 4-67 , psi Aerodynamics, John Wiley and Sons, New York,
o. : tensile stress in steel, psi 1959.
ow : stress induced by wind or earthquake response 13. Blevins, R. D., Flow-Induced Vibration, Van Nos-
spectra, psi trand Rheinhold Co., New York, 1977.
oy : minimum yield stress for a ductile material, 14. Macdonald, A. J., Wind Inading on Buildings, Ap-
psi plied Science Publishers, Ltd., London, England,
z= Poisson ratio for a given material, dimension- 1980.
less 15. Staley, C. M. and Graven, G. G., The Static and
d: concrete bearing parameter (Equation 4-20), Dynamic Wind Design of Steel Stacks, ASME 72-
dimensionless Pet-30, New York.
16. Vierck, R. K., Vibration Analysis, Harper and Row,
New York, 1979.
' 17 . Paz, M., Structural Dynamics, Van Nostrand
REFERENCES Rheinhold Co. New York, 1980.
18. Australian Standard 1170, Part 2-1983 SAA Load-
l. ASME Boiler and Pressure Vessel Code, Section ing Code, Part 2-Wind Forces, p. 55.
VIII Division I , American Society of Mechanical 19. Timoshenko, S., Young, D. H., Weaver, W., Vibra-
Engineers, New York. tion Problems In Engineering, John Wiley and Sons,
2. Zick, L. P., "Stresses in Large Horizontal Cylindri- New York, 1974.
cal Pressure Vessels on Two Saddle Supports," 20. Higdon, A., Olsen, E. H., Stiles, W B., Weese, J.
Welding Research Journal Stpplement, 1971. A., and Riley, W. F., Mechanics of Materials, John
3. Brownell, L. E. and Young, E. H., Process Equip- Wiley and Sons, New York, 1976.
ment Design, John Wiley and Sons, New York, 21. Mitchell, Warren W., "Determination of the Period
1959. of Vibration of MultiDiameter Columns by the
4. U.S. Steel, Steel Design Manual, U.S. Steel, Pitts- Method Used on Rayleigh's Principle," an unpub-
burgh, Pennsylvania, 1981. lished work prepared for the Engineering Depart-
5. American Institute of Steel Construction, Manual of ment of the Standard Oil Company of California.
Steel Construction, Eighth Edition, AISC, Chicago, San Francisco, California, 1962.
Illinois. 1980. 22. Bedna\ H. H., Pressure Vessel Design Handbook,
6. Timoshenko,5., Theory of Plates and Shells, Mc- Van Nostrand Rheinhold Co.. New York. 1981.
Graw-Hill Book Co., New York, 1959. 23. Boardman, H. C.. "Stresses at Junction ofCone and
7. Bickford, J. H., An Introduction to the Design and Cylinder in Thnks With Cone Bottoms or Ends,"
Behavior of Bohed Joints, Marcel Dekker, Inc., Pressure Vessel and Piping Design, coTlected, pa-
New York, 1981. pers, ASME, N.Y., 1960.
8. Faires, Y. M., Design of Machine Elements, The 24. Weaver, William, Jr., "Wind-Induced Vibrations in
Macmillan Co.. New York. 1962. Antenna Members," American Society of Civil En-
9. Simiu, E. and Scanlan, R. H., Wind Effects on gineers, Paper No.3336, Yol. 127, Part 1, N.Y..
Stuctures, John Wiley and Sons, New York, 1978. N.Y., 1962.
Appendix A
v"' : RiL
|to' _ ,inol - panial volume shown (A-l)
2 \180' rl In snaoeo regron
L: length of cylinder
I--x^L?-q
Examplg lFigure A.tl ' ait --
For a cylinder with 144-in. ID find the partial volume
of a fluid head of 60 in., if L : 100 ft:
v,' :
(721!zoo) der.
r80 - sin (160.81")l
2 ["tlggrsri [ I
Yp : 7,707 ,650.2 in.3 : 33,366.5 gal
PARTIAL VOLUIIE OF A
HEIIISPHERICAL HEAD r(3sft3(50) - l00l : 64,140.85 in.3 : 277.7 ga|
,,rP _- -------------
rry':(3Ri -D)
(A-2)
J
Example
V. = partial volume shown in shaded region
For horizontal volume in Figure A-2b find partial vol-
Example ume for a head with Rr :
50 in. and y 35 in.' :
For vertical volume in Figure A-2a find partial volume 277 .7
for a head with Ri :
50 in. and y = 35 in.:
= 138.85 gal
255
256 Mechanical Design of Process Systems
^
K, =
ll4 - 2(1.0)
)O.U ln.
z
-=-:
a: 159.43" : 2.78
I ln,
lv
tl I
t?
PARTIAL VOLUMES OF SPHERICALLY
DISHED HEADS --
--.-{-} -
Horlzontal Head
The partial volume of a horizontal head (Figure A-3) is
v="lJGt:lT-{p-v-F ryl (A-l) Figure A-3. Partial volume of spherically dished horizontal
neaqs.
Vertical Head
The partial volume of a vertical head (Figure A-4) is
-.
v: ?rv(3x2
--:--:----------
+ v2)
(A-4)
6
or
,, rry2(3o - v)
(A-5) Figure A-4. Partial volume of spherically dished vertical
3 heads.
n
Appendix A: Pressure ry'essel Formulations
Yi = 6.786"
Figure A-5.
(91.21)(562 - 6.7862)
u = (I93)'-(Rl
- n1i (4-6)
6RI
.
v^= . 2r(tDD)R''?
' ' rRl _______:
Ilv uj I
(A-8)
-
2 l' 3(IDD)'?j Figure A.7. Partial volume of vertical elliptical head.
258 Mechanical Design of Process Systems
,, - 2rQ6.50)(53.U2 - -z-'lteo-
zrt53.0) [
H=IDO-KR
D
horizontal knuckle region
PARTIAL VOLUilES OF
TORISPHERICAL HEADS
oH ?rY(3x'z + Y'z)
' 6'"(r^2 + 4ru2
v": '" + 12)
" + 6
(A-11)
wherey:IDD-KR
Volume of Knuck-Cylinder Region @ (Figure A-12) Figure A-11. Sketch for example partial volume calculation
of horizontal torispherical head.
V1 : V6+ V6
Horlzontal Head Example The head is vertical and the liquid level is 18-in. Deter-
mine the partial volume.
A 102-in. @ OD flanged and dished (torispherical) From the vessel head manufacturer's catalog we deter-
head made to ASME specifications (KR ) 0.60p and mine the following:
KR > 3th, tr, = head thickness) is spun from l-in. plate.
The head is horizontal and the liquid level is 35-in. deter-
mine the partial volume. p : 132 in., KR = 3 in.,IDD = 20.283 in.
From the vessel head manufacturer's catalog and Fig-
llR trl 5l
ure A-12 we determine the following: R, = '-- - 2.=-" -', = 67.50 in.;
p= 96 in., KR : 6.125 in.. IDD : 17.562 in.
x = 67.50 - (31 - H2lo5 : 66.446 in.
too
R, = : 50in., L = 96.0 - 17.562 = 78.4J8 in.
=z For kluckle-cylinder region,
From Equation A-14 we have
r,, = Rr :61.50;ri
= Ri - KR : 67.50 - 3.00 : 64.50
3
67.50 + 64.50
f
rm=-=ob.ul
(78.438)(50' - 15') , /.) < r1, 14(6.125)
T JT = + 15.0)l :2.283 in.
h 120.283 - (3.0
+ (5o.oo - 6.12s) + (50.00 - 6.125f1
" -l-'' l(67.50), +
vv = -()
)9,11
) 4(66.0)2 + (64.5011
Vr = 14.091.,14 in.r = 147.59 ga. o
* z(17.283)[3(64.500)'? + (17.283)'?]
Vertical Head Example 6
Yv = 31,247.726 in.r + 115,645.832 in.3
A 138-in. d OD F&D (flanged
and dished) head nor
made to ASME specifications is spun from 1llz-in. plate. Vv = 146,893.558 in.r : 635.903 gal
Appendix A: Pressure Vessel Formulations
Cylindrical Shell
Longitudinal Joint
i= PR D_ oEt
oE + 0.4P '-R-O3t
Circumferential Joint
PB" 2oEl
'- ^
2'E + 1AP Ro - 1.4t
t= PDo
^ 2oEl
2oE + 1-BP D. - 1.8r
t= PRo o_ 2oEl
2dE + O-8P
qEt
| _ 0.885P1 ^
'-;E+o-sP 0.885L - 0.8t
Conical Section
r= PDo
- 2 cos o(oE ^Y=- 2SEl cos a
+ 0.4P) Do - 0.8t cos c
Mechanical Design of Process Systems
Cylindrical Shell
Longiludinal Joint
PRi
t=
oE - 0.6P I'ti + u.bt
Circumferential Joint
PRi
t= ' -F;- o.4t
2oE + O.4P
1-\
i-
2:1 Ellipsoidal Head
^
Or
2oEl
+ 0.2t
lt-----Ti -'------t l'-
^ 2oEl
R + 0.2t
Conical Section
Values ot Factor M
UT 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 3.25 3.50 4.00 4.50 5.00 5.50 6.00 6.50
M 1 .00 1.03 1.06 1 .08 1.10 1.13 1.15 1.17 1 .18 1.20 1.22 1.25 1.2a 1.31 1.34 't.36 1.39
10.5 11.0 1.5 '12.0 13.0 14.0 16.0 164s
Ur 7.OO 7.50 8.00 8.50 9.00 9.50 10.0 1
M 1.41 1.44 1.46 1.48 1.50 1.52 1.54 t.co '1.58 1.60 1.62 1.65 1.69 1.72 1.75 1.77
. The maximum allowed ratio: L-t= D. When Ur > 16?3 (non-ASME Code construction), the values of M may be calculated by
i
I
M= 1
/ fL\
oit.!;/
I
xrl -@
Appendix B
A standard is a collection of current practices, past ex- One of the most widely accepted international standards
periences, and research knowledge. Standards that are is the Australian Standard 1170. Part 2-1983. SAA
developed by consensus groups (e.g., ASTM, ANSI), Loading Code Part 2-Wind Forces.
trade associations (e.g., AISC, ACI), or government The Australian Standard I 170 is more applicable to the
groups (e.g., HUD, CPSC) carry more authority than process industries because in it are shape factors for
other standards because they reflect wider ranges of ma- geometries that are more common in that industry, e.g.,
terials. circular shapes. However, before applying the shape fac-
The ANSI A58.1-1982 is a collection of information tors of the Australian standard to the ANSI or any other
that is considered to be the state-of-the-art in the design national standard, one must be very careful to correctly
of buildings and other structures. Local and regional convert the factors. This is because the codes have dif-
building codes adopt portions of the ANSI standard for ferent basis upon which these factors are deiermined,
their own use. These local and regional codes are devel- and a direct application of other parameters is not possi
oped to represent the needs and interests of their respec- ble. This is discussed later after we discuss the basis for
tive areas and are written in legal language to be incor- the various standards.
porated into state and local laws. Because these building
codes are regional or local in scope, they often do not
include everything in the ANSI standard, which is na- CRITERIA FOR DETERMINING WIND
tional in perspective. For this reason, one must be cer- SPEED
tain that a local code written for one area is applicable to
the site being considered. Wind is caused by differential heating of air masses by
The ANSI standard does not have as much authority as the sun. These masses of air at approximately one mile
the ASME vessel codes, and, unfortunately, does not above the ground circulate air around their centers of
have a referral committee or group to officially interpret pressure. At this altitude, the velocity and direction of
the document. Therefore, one must make decisions the wind is almost entirely determined by macro-scale
based on past experience and accepted methods of de- forces caused by large scale weather systems. Below this
sign. The ANSI standard (Paragraph 6.6, p. 16) states gradient height, the wind is modified by surface rough-
that in determining the value for the gust response factor ness, which reduces its velocity and changes its direction
a rational analysis can be used. A note below the-para- and turbulence. A secondary criterion, except for ex-
graph states that one such procedure for determining the treme wind conditions, is the temperature gradient,
gust response factor is in the standard's appendix. The which affects the vertical mobility of turbulent eddies
note at the top ofthe appendix (p. 52) states clearly that it and therefore influences the surface velocity and the gra-
is not a part of the ANSI 458.1 mirninum design stan- dient height. Therefore, the exact nature of the surface
dard. What all this implies is that one may follow the wind at any point depends, first, on the general weather
guide of the ANSI standard's appendix or use another ra- situation, which determines the gradient wind and the
tional analysis, which includes another wind standard. temperature gradient, and, second, on the surrounding
Thus, one caz use another standard for design purposes. topography and ground roughness which, together with
265
266 Mechanical Design of Process Systems
Figure B-2, Ratio of probable maximum wind speed averaged over t seconds to hourly mean speed.
Thus, the 1.18 factor would have to be used in the 2-sec Category A-A very restricted category in which the
gust code if that code were to be substituted for Appen- wind speed is drastically reduced. Most petrochemical
dix A of ANSI A58.1-1982. and power facilities do not fall within this category.
Similarly, the Canadian code we must convert to ob- The wind force is reduced because the structure is con-
tain an equivalent fastest mile wind speed from the mean sidered to be among many tall structures. One exam-
hourly. The mean hourly implies that the wind moves an ple would be a ten-story building in downtown Man-
average of 100 mph across the anemometer in a period of hattan, New York, where the taller buildings would
3,600 sec. Reading Figure B-2 we have V'/Vru* = 1.6. block the stronger air currents.
Thus Category B-A classification that encompasses some tall
structures, but not enough to block the majority of
wind gusts. An example of this category would be a
lj: ozor tower in the midst of a large petrochemical facility
where there were other towers that would block some
which yields an equivalent velocity of 76.9 mph. With of the wind force. A forest surrounding a tower is an-
the Canadian code one must use 0.769 in use of shape other example.
constants and the various other parameters when using Category C-The most common classification for petro-
with ANSI A58.1. A comparison of the major wind chemical applications. This category is open terraln
codes is given in Thbles B-Z, B-3, B-4, and B-5. where the tower would receive full impact from the
wind with minimum ground resistance to the wind. An
example of this category would be an open field or an
alrDort.
Category D-A classification for wind moving over wa-
ter. A beachhead, in which there is flat beach up to a
ANS| A58.r-1982 WIND CATEGORIES row of buildings would be in Category D. Miami
beach, from the ocean front up to the facade of hotels,
In the ANSI A58.1-1982 there are four wind catego- is a good example. Behind the hotel fronts would be
ries-A, B, C, and D. The categories are described as Category C. Another example of this classification
follows: would be a tall vertical vessel on an offshore structure.
Mechanical Design of Process Systems
Table B-2
Malor U.S. and Foreign Building Codes and Standards Used in Wind Design
Code or Standard Edition Organization Address
Australian Standard I 170, 1983 Standards Association Standards House
Part 2-Wind Forces of Australia 80 Arthur Street/North Sydney,
N.S.W. Australia
British Code of Basic 1972 British Standards British Standards Institution
Data for Design of Buildings Institution 2 Park Street
(cP3) London, WIA 285, England
Wind Loading Handbook 1974 Building Research Building Research Station
(commentary on CP3) Establishment Garston, Watford, WD2 7JR, England
National Building Code 1980 National Research National Research Council of
ofcanada (NRCC No. 17303) Council of Canada Canada
The Supplement to the 1980 National Research Ottawa, Ontario K1A OR6
National Building Code of Council of Canada Canada
Canada (NRCC 17724)
ANSI A58.1,1982 1982 American National 1430 Broadway
Standards Institute New York, New York 10018
Uniform Building Code 1982 International Conference 5360 South Workman Mill Road
of Building Officials Whittier, California 90601
Standard Building Code 1982 Southern Building Code 900 Montclair Road
with Congress International Birmingham, Alabama 35213
1983 rev.
Basic Building Code 1984 Building Officials and 17926 South Halsted Street
Code Administrators Homewood, Illinois 60430
International, Inc.
Table B-3
Reference Wind Speed
Australian British Canadian United States
Feference 1983)
(SAA, (BSl, 1982) (NRCC, 1980) (ANS|, 19s2)
Averaging time 2-3 second 2-second Mean hourly Fastest mile
gust speed gust speed
Equivalent reference 118.4 I18.4 76.9 100
wind speed to fastest
mile 100 mph
'l'*"1iil
Table B-4
Parameters Used in the Maior National Standards
Australian British Canadian United Siates
Parametel ,1 1982)
Wind Speed
Terrain roughness 4 4 3 4
l,ocal terrain Yes Yes None None
Height variation Yes Yes Yes Yes
Ref. speed 2-sec gusts 2-sec gusts Mean hourly Fastest mile
Wind Pressure
Pressure coefficients Tbbles in Thbles, includes Figures and Thbles, figures
appendix includes figures tables in and notes
figures commentaries
Gusts
Magnitude Gust speed Gust speed Gust effect factor Gust response factor
Spatial correlation Reduction for None Gust effect factor Area averaging
large area
Gust frequency Dynamic consideration Dynamic Dynamic Dynamic consideration
for h/b > 5 consideration consideration for h/b > 5
not included for h/b > 4
in. or for
h> 400 ft
Analysis procedure This standard is consid- Overall a very good An excellent wind Although the appendix
ered by many the best code, its weakest part standard. The is technically not con-
for use in the process is the lack of dynamic analysis procedure sidered a part of the
industries. Figures and consideration . is straight-forward standard, it contains
tables are easy to read. and the docu- figures difficult to read,
The standard actually ments-code and namely Figure 6. For
provides the user with supplement con- many structures the
equations to cutves. tain tables and fig- data extend beyond the
The analysis procedure ures easy to read. limits of the curves in
is straight-forward. Figures 6 and 7. In the
method in the appendix,
one must assume an ini-
tial natural frequency,
resulting in an iterative
process. This method is
extremely difficult in
designing petrochemical
towers without the use
of a computer.
270 Mechanical Desisn of Process Svstems
Table B-5
Limitalions of Codes and Standards
Code or Standard Statement of Limitation Location
Australian Standard "Minimum Design Loads Title
I 170, Part 2- 1983 on Structures"
National Building "...EssentiallyaSet Guide to the
Code of Canada of Minimum Regulations . . ." Use of the Code
(NRCC, 1980)
British CP3 ". . . Does Not Apply to Section 1 (Scope)
Buildings. . . That Areof
Unusual Shape or Location
For Which Special Investigations
May Be Necessary . . ."
United States "Minimum Design Loads . . ." Title
ANSI A58.I "Specific Guidelines Are Given Paragraph 6. I
For. . . Wind Tunnel Investigations
. .. For Buildings. .. Having
IrregularShapes..."
Uniform Building "The purpose . . . is to provide Section 102
Code minimumstandards..."
Basic Building "The Basic Minimum Wind Speeds Section 912.1
Code (BOCA, 1984) Are Shown in Figure 912.1 . . ."
Standard Building "The Purpose of This Code is to Preface
Code, 1982 (SBCCI, Provide Minimum Requirements . . ." Article 1205.2(a)
t982) "The Building Official May Require
Evidence to Support the Design
Pressures Used in the Design
of Structures Not Included in
This Section."
Appendix C
Properties of Pipe
272 Mechanical Design ol Process Systems
+
PROPERTIES OF PIPE
Th6 tollowilg tormulqs dre used in lh€ computorior of th6 volues i Tlr€ lsEilic ste6ls rnay be sbout 5% les!, dDd the crEte.itic stdin_
,bo\|'n in lhe toble: legs sleels qbout 27o greate! thon the values shown in tbiE tqble
t weighl which ore bcsed ort weigbts lor cdrboIt steel.
ot pipe per toor (pouDds) 10.6802(D-0
weight ol lPcler p€r foot (pou!ds) 0.340sd, * achedule numben
=
squqre leet ou&id€ ludoco per loot : 0.2518D
squorc leet inside surlace F€r loot 0.26r0d Stordord weight pipe qnd schedule rlo qte the scrae in oll sizss
= lhrough lo-isdr; lrom lz-ilch thtough 24-irch, stqndard {eight pip6
inside drea (squdre hches) 0.785d,
dred o{ rnetcd (squore irches) 0,785(Dr-d?) has d croll thicloess oI %-ircb.
moment ol ir6rtia (inches.) 0.049r(D.-d) Extro BtroDg woight pipe ond schedulegO dla the sdEe in oll
sires
A^E o' lhrough 8-i[ch, llom 8-inch thlough 24-trch. ertrd strong weight
saction moduluB (inches3)
= 0.0982(D.-ci.) pipe bos a wall thicloess oI ){-incb.
D
lodius oI gyrqtion (i!ches) Double €nr(l 3troDg weight pip6 bss no c-orrespodding schedule
= 0.25t/D,'D,+--
auEbe!,
An = oted of Eetql (square i4ches)
d = inside didoeter (incb€s) o: ANSI836.10 st€el pipe schedule Dub.b€rs
D = outside diqrn€ter (incheB)
R, b: ANSI 836.10 steel pip€ DoDilrol woll thichress dosiglqtio!
= radius ol gFcrion {iiche3)
t = pip€ woU thickness (inches) c: ANSI 836.19 6tdiDle3s steel pipe schedule uuEbols
nordnol
piF .ize achedule wcll i!3ide
inside metdl
Bq li sq tt Feiqht moD€ttl aoction rardiur
thick. dioEr- outaid6 inlide stoight
ou|lide o[ wcter OI Eodu- gyrc-
surtdce, aurldce, F!Il, psr It. inertic, lu& lior|,
b in. in"
rq. rn sq. in
Per lt po! tl Ib* iE
in
I0s 0.01s 0.307 0.0740 0.0548 0.r06 0.0804 0.186 0.0321 0.0m88 0.00437 0.127t
% 40 srd 40s 0.068 0.269 0.0568 0.0720 0.106 0.0705 0.215 0.0216 0.00108 0.0052s 0.1215
0.405
80 xs 80s 0,0s5 0.2I5 0.036{ 0.092s 0.106 0.0563 0.315 0.0157 0.00t22 0.00600 0.llt6
l0s 0.06s 0.410 0.1320 0.0970 0.141 0.1073 0.330 o,0512 0.00279 0.01032 0.1694
% 40 srd 40s 0.088 0.364 0.1041 0.1250 0.111 0.0955 0.425 0.0451 0.00331 0.01230 0.1528
0.540
80 xs. 80s 0.119 0.302 0.0716 0.1574 0.111 0.0794 0J35 0.0310 0.003?8 0.0139s 0.1547
ss 0.065 0.7I0 0.396 0.1582 0-220 0.1859 0.538 0.I716 0.01197 0.0285 0.2750
% l0s 0.065 0.545 0.2333 0.12{6 0.t77 o.t427 0.423 0.1011 0.00585 0.01737 0.2169
0.675 40 ;;; 40s 0.0st 0.493 0.1910 0.16t0 0.r77 0.1295 0.568 0.0827 0.00730 0.02160 0.2090
80 xs 80s 0.126 0-423 0.1405 0.2173 0,t77 0.1106 0.739 0.0609 0.00862 0.02s54 0.199r
0.065 0.710 0.3959 0.1583 0.220 0.1853 0.538 0.171 0.0120 0.0285 0.27S0
r0s 0.083 0.674 0.357 0,1974 o.220 0.1765 0.571 0.1547 0.0I431 0.m4I 0-2892
% 40 ;; 40s 0.10s 0.822 0.304 0.2503 0.220 0.1628 0,851 0.1316 0.0r710 0.0407 0.2613
0.840 80 xs 80s 0.147 0.sd6 0.2340 0.320 0.220 0.1433 r.0€8 0.1013 0.02010 0.0178 0.2505
160 0.187 0.466 0.1706 0.383 0.220 0.t220 1.304 0.0740 0-022t3 0.0s27 0.2402
n(s 0.294 0,252 0.0499 0.504 0.220 0.0660 1.714 0.0216 0.02125 0.0s77 0.2rs2
0.06s 0.920 0.655 0.20u o.275 0.2409 0,684 o.2a82 0.02451 0.046? 0.3{9
l0s 0.083 0.884 0.614 0-2321 0-275 0.2314 0.857 0.2661 0.02970 0.0s66 0.343
10 ;;; 10s 0.113 0-821 0.s33 0.333 o.275 0.2t57 l.r3l 0.2301 0.03?0 0.0706 0.334
i.050 s0 xs 80s 0.154 0-7 42 0-132 0.435 0.275 0.1913 1.414 0.r875 0.0448 0.08s3 0.321
160 0.218 0.614 0.2961 0.570 0.275 0.1607 1.937 0.1284 0.os27 0.1004 0.304
xxs 0.308 0,434 0.1d79 0.718 0,215 0.1137 2.441 0.0641 0.0s79 0.1104 0.28{0
0.065 1.185 1.t03 0.3{4 0.310 0.868 0.478 0.0500 0,0760 0.443
10s 0.109 1.097 0.915 0,413 0.344 0.2872 1.401 0.409 0.0757 0.ll5r 0.428
I 40 40s 0.133 I.049 0.86{ 0.494 0.344 o,2746 1.679 o,37 4 0.087{ 0.1329 o.121
J.3t5 80 80s 0.179 0.s57 0.t19 0.639 0.3{4 0.2520 2.t72 0.31t 0.1056 0.1606 0.107
t60 0.250 0.815 0.522 0.836 0.344 0.213{ 2.444 o.2261 0.t2s2 0.1900 0.387
xxs 0.358 0.599 0.28r8 1.076 o.314 0.r570 3.659 0.1221 0.140s 0.2t37 0.36t
0.06s 1.530 1.839 0.326 0.434 0.{01 1.107 0.797 0.1038 0.1250 0.55{
r0s 0.109 t.142 1.633 u.531 0.434 0.378 1.805 0.7tl 0.r60s 0.1934 0.5s0
r% 40 l'; 40s 0.140 1.380 1.496 0.669 0.434 0.361 2.273 0.618 0.1s48 0.2316 0.540
I.66'0 80 80s 0.r91 l27A 1.283 0.88r 0.{34 0.335 2.991 o.24t8 0.2913 0.s24
160 0.250 1.160 r,057 1.I07 0.43{ 0.304 3.765 0.{58 0.2839 0.342 0.506
*ts 0.382 0.896 0.631 1,534 0.434 0.2345 5.2t1 o.2r32 0.341 0.411 0.472
rt4 0.065 t.770 2.161 0-375 0-497 0.463 1.274 1.067 0.ts80 0.1663 0.649
1.900 r0s 0.109 1,682 2.222 0.613 0.497 0.440 2.08s 0.962 0.2469 0.2599 0.634
*Couftesr of ITT Gtinncll.
F
Appendix C: Properties of Pipe
{0 40s 0.145 1.6r0 2,036 0.7ss 0.{97 0.421 2.718 0.882 0.310 0.326 0.623
80 xs 80s 0.200 1.500 t-761 r.068 0.497 0.393 3.63r 0.765 0.39r 0.{12 0.605
rh 160 0.28r 1.338 1.406 1.429 0.{97 0.350 4.859 0.608 0.483 0.508 0.581
J.900 xxs 0.400 1.100 0.s50 1.885 0.49? 0.288 6.408 0-112 0.568 0.598 0.s19
0,525 0.8s0 0.567 2-287 0.197 o.223 7.710 0.218 0.6140 0,6470 0.5200
0.650 0.600 0.283 2.551 0.497 0.157 8.678 0.123 0.6340 0.6670 0.4980
0.065 2.215 3.96 o.472 o.622 0.588 1.604 1.7I6 0.3t5 0.2652 0.817
l0s 0.109 2-ts7 3.65 0.116 o.822 0.565 2.638 1.582 0.499 0.120 0,802
;; ;; 40s 0.154 2,087 3:36 r.075 o,622 0.541 3.553 1.455 0.666 0.s61 0.r81
2.3r5 80 xs 80s 0.218 1.939 2.953 t.417 0.822 0.508 5.O22 1.280 0.868 0.73t 0.756
160 0.343 1.689 2.240 2.190 o.822 o.442 7.141 0.971 I.163 0.979 0-729
xx!; 0,436 r.503 t.774 2.656 o,622 0.393 9.029 0.769 1.3I2 1.104 0.703
0.552 1.251 t-229 3.I99 0.622 0.328 10.882 0.533 1.442 r.2140 0.6t10
''''. ...' 0.587 1.001 0.187 3.641 o.622 0-262 12.385 0.341 1.5130 1.2740 0.64d0
0.083 2.709 0.128 0.75s 0.709 2.175 2.49S 0.710 0.4s4 0.988
l0s 0.120 1.039 0.753 0.690 3,531 0.988 0.687 0.975
;; :.. 40s 0,203 2.469 4,75 t.701 0.753 0.646 5.793 2.076 1.530 L064 0.9d,
2% 80 xs 80s 0.276 2.323 4.24 2.251 0.753 0.60s 7.661 1.837 1.925 1.339 0.924
2.875 ta: 0,375 2.t25 3.55 2.915 0.753 I0.01 1.535 2.353 1.637 0.894
)0(s 0.552 1.771 2.464 4.03 0.753 0.451 13.70 1.087 2.872 1.s98 0.84{
0.6?5 t.525 1.825 4.663 0.753 0.399 15.860 0.792 3.0890 2.1490 0.8140
0.800 t.275 t,276 5.2t2 0.753 0.334 t7.729 0.554 3.2250 2.2430 0.7860
0.083 3.334 4.73 0.891 0.916 0.873 3.03 3.78 1.301 o,144 1.208
l0s 0.120 3.260 8.35 t.274 0.916 0.8s3 4.33 3.6r ta22 1.011 1.195
;i d; 10s 0.216 3.068 7.39 2.228 0.916 0.803 7.58 3.20 3.02 t.124 1.164
3 80 xs 80s 0.300 2.900 3.02 0.916 0.759 10.25 2,864 3.90 2-228 r.136
3.500 160 0.437 2.62A 5.42 4.21 0.916 0.687 14.32 2-348 5.03 2.876 1.091
xxs 0.600 2.300 4.15 3.17 0.916 0.602 18.58 1.80t 5.99 3,43 1.o17
o;125 2.050 3.299 0.916 0.537 2t.487 1.431 6.5010 3.7150 1.0140
-' 0.850 1.800 2.543 7.073 0.916 0.471 24.Os? t.103 6.8530 3.9160 0.9840
5S 0.083 3.834 r.021 t-017 1.00{ 3.47 5.0r 1.960 0.980 1.385
3y2 r0s 0.120 3.t60 ll.l0 1,463 t-o47 0.98{ 4.57 4.81 2.756 1.378 L.312
*Un 40 ;; 40s o.226 3.548 9.89 2.680 t-041 0.929 9.ll 4.28 4.79 2.351 t.337
80 xs 80s 0.318 3.364 8.89 3.68 t.047 0.881 12.51 3.85 6.28 3.t4 1.307
xt(s 0.636 2.728 5.845 6.721 1.047 0.716 22.450 2.530 9.8480 4.92d0 1.2100
0.083 4.334 14.75 1.152 l.l78 1.135 6.{0 2.8u 1.249 1.562
tGs 0.120 {.260 14.2S 1.178 l.ll5 6.17 3.96 1.162 1.549
0.188 4.t24 13.35? 2.547 1.178 1.082 8.560 5.800 5.8500 2.600.0 1.5250
;; s;; 40s 0237 4.826 t2-73 3.-17 1.178 1.054 10.?9 7.23 3.21 1.510
80 XS 80s 0.337 3.826 I1,50 {.{l 1.178 r.002 l{.98 4.98 9.61 t.477
4.5N t20 0.437 10.33 t.178 0.949 18.96 4.48 5.18 1.445
0.500 3.s00 9.62r 6.283 1.178 0.916 2r.360 4.160 12.71t0 s.6760 l-1250
t60 0.531 3.438 9.24 6.62 1.178 0.900 22.51 4.O2 13.21 5.90 1.416
)c; 0.674 3.152 7.80 Lr0 1.178 0.825 27.51 r5.29 6.79 t.371
0.800 2.900 6.602 9.294 t.178 0.75S 31.613 2.864 16.6610 7.1050 1,3380
0.925 2.650 5.513 t0.384 ].l78 0.694 35,318 2.391 17.7130 7.8720 1.3060
5S 0.109 5,345 22.11 1.868 1.t56 1.399 9,73 6.95 2.498 r.929
10s 0.134 5.295 22,02 2.245 l.{s6 1.386 ?,77 9.53 8.43 3.03 t.920
;; -:. 4os 0.258 5.047 20.0r 4.30 1.455 1.321 t1.82 I5.17 5.{5 1.878
80 xs 80s 0.375 4.813 18.t9 6.lI 1.156 t.260 20-74 t.89 20.68 1.43 1.839
t20 0.500 4.563 16,35 7.95 1.456 1.195 27,01 7.09 25.74 9.25 1.799
r60 0.625 4.313 14.61 9.70 1.456 Ll29 32.98 30.0 10.80 !.760
,ots 0.7s0 4.063 t2s7 t1.34 1.456 1.064 38.55 s-82 !2.10
0.875 3.813 I1.413 12.880 1.4s6 0.998 {3.8I0 4.9S1 36.6450 13.1750 1.6860
- 1.000 3.563 l{.328 1.456 0.933 t'|.'134 1.232 39.11l0 14.0610 t.6s20
274 Mechanical Design of Process Systems
5S 0.134 to.482 86.3 4.52 2.815 2.744 r5. ts 37.4 63.7 11.8S 3.75
I0s 0.16s 10.420 85.3 5.49 2.815 2.724 18.70 36.9 I4.30 3.74
0.219 10.312 83.52 '1.24 2.815 2.10 24.63 36.2 100.46 I8.69 3.72
20 0.250 10.250 82.5 8.26 2.815 2.683 28.04 35.8 Ir3.7 2l.r6 3.71
30 0.307 10.136 80.7 I0.07 2.815 2.654 34.24 3S.0 137.5 3.69
40 std 4;; 0.36S 10.020 78.9 ll.sl 2.8r5 2.623 40.48 34.I 160.8 29.90
l0 60 xs 80s 0.500 9.750 7 4.7 16.10 2.815 2.553 54.74 32.3 2t2.0 39.4
10.750 80 0.593 s.564 7L8 I8.92 2.815 2.504 64.33 3l.l 244.9 45.6 3.60
I00 0.718 9.314 22.63 2.815 2.438 76.93 29.5 288.2 53.2 3.56
t20 0.843 9.064 64.5 26.24 2.815 2.373 89.20 2e.0 324 60.3 3.52
0.875 9.000 63.62 27.!4 2.815 2.36 92.28 27.6 333.46 82.O4 3.50
t40 1.000 8.750 60.1 30.6 2.815 2.191 .04.13 26.1 68.4
160 l.I2s 8.500 56.7 34.0 2.815 2.225 u5.65 24.6 399 ?4.3 3.43
L2s0 8.250 53.45 37.31 2.815 2.18 126.82 23.2 424.t7 79.65 3.39
1.500 7.75D 47.15 43.57 2.8I5 2.03 t48.I9 20.5 478.59 89.04
0.I56 12.438 t2t.4 6.17 3.34 20.99 s2.7 t22.2 19.20 4.45
10s 0.180 I2.390 120.6 7.t I 3.34 3.24 24.20 52-2 I40.5 22.93 t.44
;i 0.250 t2.2s0 u7.9 9.84 3.34 3.21 3s.38 r91.9 30.r 4.42
30 0.330 t2.090 114.8 12.88 3.34 3.t1 43.7'l {9.7 248.5 39.0 4.3S
;,; 4;; 0.375 12.000 I l3.l 14.s8 3.34 3.14 49.56 49.0 219.3 43.8 4.38
40 0.406 I1.938 III.9 15.74 3.34 3.13 53.S3 48.S 300 47.1 4.37
80s 0.500 I1.750 I00.4 19.24 3.34 3.08 65.42 47.0 362 4.33
t2 ;; 0.562 11.626 106.2 2r.52 3.34 3.04 73.16 46.0 401 62.8 4.3r
)2.750 80 0.687 I1.376 101.6 26-O4 3.34 2.978 88.51 44.0 475 74.5 4.21
0.7s0 I1.250 99.40 28.27 3.34 2.94 96.2 43.r 510.7 80.1 4.25
100 0.843 11.064 96.1 31.5 3.34 2.897 07.20 41.6 562 88.r 4.22
0.875 11.000 95.00 32.64 3.34 2.AA 10.3 4t.l s78.S 90.7
120 t.000 10.750 90.8 36.9 3.34 2.414 25.49 3S.3 642 100.7 1-t7
140 1.125 10.500 86.6 4I.l 3.34 2.749 39.68 37.5 701 109.9 4.13
L250 I0.250 82.50 45.16 3.34 2.68 53.6 35.8 75s.S 4.09
r.312 10.t26 80.5 47.1 3.34 2.651 60.27 34.9 781 r22.6 1.01
Appendix C: Properties of Pipe 275
noEit'al
pipo rirc .chedule wqll ilride sq tt rq It
lreight
weisht aeclion !adiu!
idside |rretol iagide ol modu- qryr6-
outside |tumb€r' tbicL- diqra- outgide
lurlcc€, auddc6, !'er IL pe! ll, inerlid, lu& Uon
didr!€ter
itr. i!- aq, in aq. in per lt perlt tbt lb in. inJ ia.
i|r b
0,156 13.6S8 t47.20 3.67 3.S8 23.O {.90
t0s 0.r88 13.624 145.80 8.16 3.57 27.7 63.1 194.6 2't.8 4.88
0.210 13.580 144.80 9.10 3.67 3.55 30.9 62.8 216.2 30.9 1.57
0.219 13.562 111.50 9.48 3.55 32.2 225-l 32.2 4.57
l0 0.250 13.S00 143.1 10.80 3.67 3.53 36.71 62.1 255.4 36.5 4.86
0.281 13.438 141.80 l2.ll 3.S2 4t.2 285.2 40;I 4.85
20 0.312 140.5 t3.42 3.67 3.50 45.68 60.9 314 4{.9 4.91
0.344 13.3I2 t39.20 14.16 3.87 3.48 50.2 50.3 344.3 t9.2 4.83
l4 ;; 0.375 13.250 137.9 16.05 3.67 3.41 54.57 59.7 s3.3 4.82
14.@o 40 0,{37 13.r25 r35.3 18.62 3.67 3.44 63.37 s8.7 429 61.2 4.80
0.469 13,062 134.00 19.94 3.67 3.42 67.8 s8.0 156.8 1.79
0.s00 13.000 t32,7 2t.21 3.67 3.40 72.09 57.5 484 69.1 4-74
;; 0.593 12,814 r29.0 24,98 3.67 3.35 84.91 80.3 4.74
0.625 12.750 t27.1 26.25 3.34 8S.28 55.3 589 84.1 4;13
80 0.750 12.500 t22.7 3,61 3.27 108,13 687 98.2 4.69
t00 0.937 12.t28 38.5 3.67 3.17 130.73 50.0 825 117.8 4.63
I20 1.093 ll.8r4 109.6 44.3 3.57 3.09 150.67 47.5 930 I32.8 4.58
140 1.250 rr.500 103.s 50.1 3.67 3.01 t70.22 45.0 tt27 l{6.8 4,53
160 1.406 11.188 98.3 3.67 2.929 189.12 42.6 l0l7 159.6 4,48
notlrindl
pip6 .iz€ schedule woll idrids 6q It sq lt rroight Ino|'lent aection tcdiut
thick- di(rm- idaide met(ll weight
outside ir16ide ol nodu- gytq-
sutldce, 6urlcce,
po!It,
sq in. aq in tbf
pe! ll, in€rti(r, lus, tior
rn. b i!L per lt perlt lb ia.. in.
r20 1.500 17.000 227.0 47.2 5.24 4.45 296.37 98.3 3760 376 6.S6
20
I40 1.750 16.500 213.8 100.3 s-24 1-32 341.10 92.6 4220 422 6,48
20.000
160 1.968 I6.064 202-',l Iu.5 5,24 4.21 37S.01 87.S 4590 459 6.41
0.188 2L.824 367.3 12.88 5.76 44 t59.1 766 69.7 1,71
t0s 0.218 2I.564 365.2 t4.92 5.76 5.65 158.2 88S 80.4 7.10
i; 0.250 21.500 363.1 17.t8 5.?6 5.63 157.4 l0l0 91.8 7.63
20 io 0.375 2r.250 354.7 25.48 5.76 5.56 87 I53.7 1490 135.4
30 xs 0.500 21.000 346.4 33.77 5.76 5_50 l15 150.2 1953 t77.5
0.625 20.?s0 338.2 4t.97 5.76 s.43 143 146.6 2400 2t4.2 7.56
22 0.7s0 20.500 330.1 50.07 5.37 170 l{3.1 2429 zs1 -2 ?.52
22.000 ;; 0.875 20.250 322.1 58.07 5.76 5.30 197 r39.6 3245 295.0 7.17
80 1.125 19.750 306.4 73.78 5.76 5.17 25r 132.8 4029 366.3 7,33
100 I.375 19.2s0 231.0 89.09 5.76 5.04 303 t26.2 475S 432.6 7.31
120 I.625 18.750 276.1 104.02 5.76 4.91 354 u9.6 493.S 7.23
I40 1.87s I8.250 261.6 1t8.55 5.76 4.78 403 113.3 6054 550.3 7.15
r60 2.t25 17,750 247.4 I32.68 5.75 4.65 451 t07.2 602,1 1.07
l0 0.250 23.500 434 18.65 6.28 63.41 188.0 l3l6 109.6 8.40
20 srd 0.375 23.250 425 27.83 6.28 6.09 94.62 183.8 1943 161.9 8.35
0.500 23.000 415 6.28 6.02 125.49 180.1 2550 2r2.5 8.31
30 0.562 22.876 4ll 41.4 6.28 5.99 140.80 178.1 2840 237.0 8.29
0.62s 22.750 406 5.28 5.96 t56.03 t78.2 3140 281.4 4.21
io 0.687 22.628 402 s0.3 6.28 5.92 vt.r? t74.3 3420 245.2 8.2S
0.750 22.500 398 54.8 6.28 5.89 186.2{ t72.4 37I0 309 8-22
:: 0.218 23.564 436.1 16.29 6.28 s5 r88.9 lt52 96.0 8.41
24.000 0.875 22.250 388.6 63.54 6.28 5.83 216 42S6 3s4.7 8.18
60 0.968 22.064 382 70.0 6.28 5.78 238.11 165.8 4650 388 Lt5
80 I.218 21.564 365 47.2 6.28 296.36 158.3 5670 473 8.07
100 1.531 20.938 344 r08.I 6.28 5.48 367.40 1{9,3 6850 57I 7.96
t20 l.8I? 20.316 328 126.3 6.28 5.33 429.39 141.4 7830 7.87
t40 2.062 19.876 3t0 r42.1 8.28 5.20 483.13 134.5 8630 719 7.r9
160 2.343 19.314 159.4 6.28 5.06 5{1.94 t27.O 9460 788 7.70
0.2s0 25.500 s10.7 I9.S5 6.81 6.68 221.1 I646 126.6 9.10
10 0.3I2 25.376 505.8 25.18 6.81 6.64 2t9.2 20?6 r59.7 9,08
srd 0.37s 25.250 500.7 30.I9 6.81 6.61 I03 217.1 2418 I90.6 9.06
20 xs 0.500 25.000 490.9 40.06 6.54 r36 2t2.8 3259 250.7 9.02
28
0.625 24.750 481.1 49.S2 6.8I 6.48 l6s 208.6 40I3 308.7 8.98
26.000
0.750 24.500 471.4 59.49 6.8t 6.4I 202 204.4 4744 364.9 8.93
0.875 24.250 461.9 69.07 6.81 6.35 235 200.2 54S8 4I9.9 8.89
L000 24.000 452.4 78.54 6.81 6.28 267 I96.1 6149 4?3.0 8.85
l l25 23.750 443.0 8?.91 6.81 6.22 299 192,1 6813 524,1 8.80
0.250 29.500 683.4 23.37 7.85 7.72 79 296.3 258S 172.3 10.52
30
l0 l0s 0.3I2 25.316 617.8 29.19 7.85 7.69 99 293.7 3201 213.4 10.50
30.000
srd 0.375 29.250 672.0 34.90 7.85 9 251.2 3823 2S4.8 t0.48
20 xs 0.500 29.000 650.5 46.34 7.85 7.59 r58 286.2 335.5 10.43
30 0.625 z8.'ts0 649.2 57.68 7.8s 7.53 281.3 62I3 4t4.2 10.39
Appendix C: Properties of Pipe 277
nomincl
schedule
woll inside sq It sq It weight rddius
pipe si:e inside rrlelal weight
lhick- didtn- oulside inside ol modu- gYra-
outside per Il,
neat, sultcce. per ft inertid, Iug, tion,
diamelet,
ilr. sq. in, Bq. in. per ft per It lbf tb in.' in.3 in.
b
40 0.750 28.s00 637.9 68.92 7.85 7.44 234 276.6 137 r 49t.4 10.34
30 0.875 28.2s0 620.? 80.06 7.85 7.3S 272 27 t.B 8494 566.2 10.30
30.000 1.000 28.000 6I5.7 9t.l l 7.33 3t0 267.O 9591 63S.4 10.25
1.t25 27 .',t 50 604.7 t02.05 7.85 347 262.2 10653 't t0.2 r0.22
0.250 31.500 779.2 24.9s 8.38 8.25 85 337.8 3l4 t 196.3 tt.22
l0 o.312 3r.376 7'13,2 31.02 8.38 8.21 106 335.2 3891 243.2 u.20
rio 0.375 3t.250 766.9 31,2s 8.38 8.18 t2'l 332.5 4656 291.0 I l.l8
20 XS 0.s00 31.000 754;1 49.48 8.38 Lll 168 327.2 6l{0 383.8 I l.l4
32 30 0.62s 30.750 742.5 61.59 8.38 8.0s 209 321.9 7578 473.6 I LoS
32.000 10 0.688 30.624 736.6 8.38 8.02 230 319.0 8298 518.6 I1.07
0.750 30.s00 730.5 73.63 8.38 7.98 250 3t6.7 8990 561.9 11.05
0.875 30.2s0 718.3 85.52 8.38 7.92 291 3l1.6 t8372 648.2 I l.0I
1.000 30.000 706.8 97.38 8.38 7.85 331 306.4 n680 730.0 10.95
LI25 29.750 694.7 109.0 8.38 7.ts 371 301.3 1302s 814.0 10.92
0.2s0 33.500 881.2 26.50 8.S0 4.77 90 382.0 371s 22t.9 11.93
10 0.312 33.376 s74.9 32.99 8.90 8.74 tt2 379.3 4680 275.3 I l.9l
srd 0.375 33.2s0 867.8 39.61 8.S0 8.70 135 3',18,2 s597 329.2 II.89
20 XS 0.500 33.000 855.3 52.62 LS0 8.64 l?s 370.8 7385 434.4 I1.85
34 30 0.625 32.750 841.9 65.53 8.90 223 365.0 9124 s36.7 I1.80
34.000 40 0.688 32.624 835.9 72.O0 8.90 8.54 245 9992 587.8 I1.78
0.7s0 32.500 82S.3 78.34 8.90 8.51 359.5 10829 637.0 r 1.76
0.875 32.2s0 8r6.4 91.0t 8.S0 8.44 310 354.1 l2s0l 735.4 tt.12
1.000 32.000 804.2 103.67 8.S0 8.38 353 348.6 141t4 830.2
I.I25 31.7s0 79r.3 116.13 8.90 8.31 395 343.2 15719 924.1 I I.63
0.250 35.500 98S.7 28.1r 9.42 9.29 96 429.1 4491 249.S 12.64
l0 0.312 35.376 s82.9 34.S5 9-42 s.26 ll9 426.1 309.1 t2.62
;,; 0.37S 35.250 975.8 42.01 9-42 9.23 143 423.1 6684 370.2 12.59
2D xs 0.500 35.000 962.1 9.42 9.16 I90 4t1.1 8785 48S.I 12.55
36.000
30 0.625 34.750 948,3 69.50 9-42 s.l0 236 4I l.l t0a72 604.0 t2.51
40 0.750 34.500 934.7 83.01 9.42 9.03 282 405.3 I2898 716.5 12.16
0.875 34.250 920.5 96.50 9.42 8.97 324 399.4 I4903 82',1.9 12.42
1.000 34.000 907.9 109.96 9.42 8.90 374 393.6 I685I s36.2 12.38
Ll25 33.750 894.2 123.I9 9.42 8.89 419 387.9 r8763 1042.4 t2.34
0.250 41.500 1352.6 32.82 10.99 I0.86 l12 586.4 7t26 339.3 I4.73
srd 0.375 41.250 r336.3 49.08 10.s9 10.80 167 s79.3 to627 506.1 14.71
20 xs 0.500 41.000 1320.2 65.I8 10.99 I0.73 222 s72.3 14037 668.4 14.67
42 30 0.6?5 40.7s0 1304.1 81.28 t0.99 10.67 565.{ 17373 827.3 14.62
42.000 40 0.750 40.500 1288.2 97.23 r0.99 10.60 330 558.4 20589 985.2 14.59
I.000 40.000 I256.6 r28.81 10.99 10.47 438 544.8 27080 1289.5 14.s0
1.250 33.500 r22S-3 I60.03 10.s9 10.3{ 544 53t.2 33233 1582.5 t4.41
1.500 39.000 1194.5 I90.S5 10.99 t0.21 649 5I7.S 3918I 1865.7 t4.33
278 Mechanical Desien of Process Svstems
To determine the seight per foot of any piping Erample. For 4" pipe rvith 4" nominal thickness
insulation, use the pipe size and nominal insulation insulation, F : .77. It the insulation density is
thickness to find the insulation l'eight factor F in the 12 pounds per cubic foot, then the insulation rveight
chart shorvn belorv. Then multiply F by the density b .77 x 12 : 9.24lb/tt.
of the insulation in pounds per cubic foot.
20 .70 .96 1.23 1.50 1.79 2.09 2.40 2.73 3.06 3.40
24 .83 1.13 t.44 2.10 2.44 2.80 3.16 3.54 3.92
Root Area of .068 ,126 .202 .302 .419 .693 .889 1.293 1.7 44 2.300 3.023 3.?19 4.619 5.621 6.724 7.918
Thread, sq, in.
Max. Safe Load,
lbs. at Rod 610 1130 1810 21L0 3??0 4960 6230 8000 11630 15700 20700 21200 33500 41580 50580 60480 ?1280
Temp. of 650"F
Appendix C: Properties of Pipe 279
Au-r'
z /\
i.
?
w
{l\
z E-I
4/ a^
B
t_J-----,
\]J
Temperature Range 'F
tr{agnesia
z Calcium
F
Combina-
tion
z
Fiber-
Sodium
type is ueight in
Boldface
ffi pounds. Lightface t]'pe benerth
weight is veight factor for
z
,t & insulation.
Instrlation thicknesses and
weights are based on average
conditions and do not constitute
Njs a recommendation for specific
thicknesses oI materials. Insula-
tion Neights are based on 85/6
{|s.:ssr magnesia and hvdrous calcium
silic&te et 1l lbs,i cubic foot. The
listed thicknesses and Neights of
7 T}
'-11 combination covering are the
sums of ihe inner layer of dia-
F 4l tomacecus earth at 2l lbs/cubic
foot and the outer layer at
N
/9N
11 lbs,/cubicfoot.
Insulotion rveights include al-
z.( lorvcnces for wire, cemerrt, can-
vas, bands and paint, but not
special
- surface finishes.
To find the weight of covering
on flanges, valves or fittings,
multiply the \veight frctor by the
uoight.pcr foot of covering nsed
@tr\
on slrarght prpe.
Vf,tve \veights 3re rpproxi-
mate. When possible, obtain
qJ Neights from the nranufacturer.
Cast iron valve $eights are for
z
F f'^
t+,!
z HJ
3 -4L.
E:::t
F
ttl
n_Lt
{- i--r
\LJ
Tenrpcraturc Range "F
Nom. Thick.,In.
Ma,gnesia
! Calcium
! Silicate
o
z
uon
Fiber-
Sodium
I effi
z Insulation thicknesses and
weights are based on average
conditions and do not constitute
a recommendation for specinc
fs-is$ thicknessesof ma,terials- Insula-
tion weights are based on 85%
! T:lii--qF magnesia and hydrous calcium
silicate &t 11 lbs/cubic foot. The
.-al listed ihicknesses and i{eights of
z /A combination covering are the
6ums of the inner layer ol dia-
4 tomaceous earth at 21 lbs/cubic
foot and ihe outer laycr at
11 lbs/cubic foot.
,N
7 /> Insulstion weights include al-
Iowances for wire, cement, csn-
vas, bands end peint, but not
speeial
- surface finishes,
To find the weight of covering
on flanges, valves or fittings,
1.<3 multiply the weightfactor by the
werghl per loot ol coverrng used
@l[' )
on strargnt prpe.
Valve weights are approxi-
mate. lVhen possible, obtain
$'eights from the manufacturer.
Cast iron velve weiqhts arc for
fl.2
IJJ L.R. 90' Elbow
.E 1.1 1.{ 1.8
n
z^"{"u
S.R. 90' Elbow
.6
.3
.8 1
E
E
{_O L.R. 45' Elbow .2 .2
r-i\
(, F!-+ Tee .6 .6
3.1
.6
3.7
.6
Temperature Range 'tr' t00-1c9 200-299 300-309 400-499 500-599 600-699 700-799 300-s99 c00-3c9 1000-1009 1100-1200
!r
u'N
u,r'
z
Ih
d-J.-t
-r--r-\
z
/>
fin
{_L_!
Temperature Range oF
Magnesia
z Calcium
F
5 Combina-
tron
z
Fiber-
Sodium
't
@
rfl
on-slr&lghl prpe.
valve wergnF are approxl-
mate. When possible, obtain
weights from th; msnuiacturer.
[],._/ Cast i.on valve weights are for
flanged end valves: sGel weiehts
+<t for ielding end vaives.
All nsnsed fittios. flanced
valve and flange weigF* inclide
FsO the prcportional weight of bolts
or 6tuds to make up all joints,
' 16 lb cu. It. density.
fr
A
u-r'
7
F w
{T\
z E'
-l -/.>\
F--1
' /-A
q-!_,
\]J
Temperature Range "F
Magnesia
z Calcium
o
Combina-
f tron
z
Fiber-
Sodium
'l
@
flr)
on straight pipe.
Valve 1Aeights are approxi-
mate. Whe[ possible, obtain
weights fron the manufscturer.
Cast iron velve weiqhts sre for
flanged end valves; sGeI *eights
+<i for welding end valves.
+ 16
t4
lb cu. lt. density.
All flanged fitting, flanged
valve ond fiange iveights include
the proportional \ieight of bolts
o! studs to make up &ll joints,
284 Mechanical Design of Procesr Systems
|4 {I/
zr\{it
E
S.R.90'Elbow
.3
.5
.3
4.4
.3
5.4
.3
: {1\
r.'.'g Tce
7.4 12.2 14.8
3 .8 .8 .8 .8
rl F4q l9
Lsteral 1.8
3.7 4.7
(-r__) Rcducer .3 .3 .3 .z
\JJ cup .5
.1.8
.5 .5
3.7
.5
'li nrpcrrlur. ncngc'F 100,14r 200-:0c 300-3c9 100-lm 500-599 600-699 700-7s9 800-80s 900-g?9 1000-1099 1100-r200
Magnesia
\orn. Thllk., In. I 1 1k 2 2 2% 3 3 3 3% 3%
? Calcir.rm LLs Ft t-25 2.08 3.01 3.01 4.07 5.24 s.24 5.24
Y Silicete
\ont.'.t'hick., IIL 2\ 3 3 3 3% 3%
rffi
O r-Fn
Screwed
Slip-On
ot 9 17 9
1.5
tl
t7 20
1.5
20
1.5
6l
6l
102
ll3
$eieht is weight fachor for
insul&tion.
Insulation thicknesses and
? s{ ils Welding Neck 1.5
19
1.5 1.5 1.5
38
1.5 1.5 1.5 weighis are based on ave.age
average
.onditions and do not constitute
nstttule
9 19 l9 36 60 99 a recommendetion for specific
sPecific
N-i.s l,ap Joint 1.5 1.5 1.5 1.5 thicknesses of materials. Insula-
Insule-
l0 l9 l0 20 24 24 38 6l r05 tjon Neights are based on>n 85%
857a
qF{i.llqn Blind 1.5 1.5 1.5 macnesia and hvdrous calcium
26 46 32 53 67 98 r50
"ili;r. 't l1 lhs/crrhin foot
rot. The
listed thicknesses and $eights
ights of
o
2 l^a S.R- 90' Elbow 3.9 4 3.9 4.1 4.3 4.6 rre th
cornbination covering are the
sums of the inncr layer of dit
dia-
E /'11
L.R. 90' Elbow
30
4.3
50
4.3
40
4.3
63
4.3
tomrceous earth at 21 lbs/cubi
)s/cubic
tvcr at
Ioot and the outer laJcr a
B,N
u /9N 45" Elbow
41
3.6
2E 46 60
3.8
93
3.9
135
4
11 lbs,lcubic foot.
Insulation rveights includel rL-
al
for \rire, cenrent, crn
< .: lorvences
ves,.blnds and Irrrirrtr buL not
no
ti] 39 67 E1 102 l5l 23E
E BJ Tee 5.9 6 5.9 6 6.2 6.9 strccrcl
- surlace nnlsnes.
ll---J To hnd the Neight of covefnlg
coverinl
t55 260 410 on flanges, valves or fitting:
fittings,
{-<t Flanged Bonnet
Gate
66
7
70
I
125
1.8 5 5.5 multiply the Neight factor
welghl per loot ol coverlng
b\ the
)r bv th
rng usei
usect
70
t55
1.8
r20 t50
495
440
on straight pipe.
Valve weights are approxr-
spproxi
mate. When possible, obtain
obtair
acturef.
weights from the ma,nufacturer.
j r\J Check 7.2 +.4 4.8 4.9 5.n Ls are for
Cist ilon valve weiehts fo
flansed end valves; steet. weights
weight
+<t Pressure Seal
Bonnet- Cate
208
3
235
for weldine end valves.
All flrnged 6tting,
rc
flanged
flange.
Pressure Seal r35 180 valve and llanse rleiqhtss include
includ
3 of bolts
the proportionlel \r eight.of bolt
Bonnet-Globc or studs to mirl(e uP all Il joints.
Jorntt
* 16 tb cu, ft. densitY,
-
4.4
.6
El#
: {l\ Tee
9.9
.9
t2.6
.9
20
.9
/.-N
Irt Reduce!
3.1
.a .3
6.9
.3
2.t 2.a
cuP .6 .6 .6
Temperature Range 'F r00-199 200-209 300-399 400-499 500-599 600-699 700-799
800-899 900-999 1000-1099 1100-1200
agnesta
Nom. Thick., In. 1 1% 2 2% 2ti 3 3 3% 314 3%
: srlLcate Lbsi/Ft r.E3 1.83 3.71 4.EE 4.88 6.39 6.39 7.80 7.80 7.80
)mbina- Nom. Thick., In.
2% 3 3 316 3% 3%
6
z
tit Lbs/Ft 6.49 E.7l 8.7 | r0.6 r0.6 10-6
Nom. Thick., In. r% 1X
Fiber- 1 1 2 2 3 3 3% 3%
Sodium Lbs/Ft z.4l 2-41 3.65 5.07 5.07 E.66 8.66 r0.62 10.62
Pressure Rating Cast Iron Ste"l-_-
psr t25 250 150 300 400 Boldface _tvpe is *eighi in
,ffi Screwed
Slip-On
or 13 2l 13 21
600 900 r600 2500
Pounds. Llghtf:rce
lvelglrt.
msul& on.
is
t)pe benecth
\aeight fcctor for
l€ Pressure Serl
Bonnet-Globe
* 16 lb cu. ft. density.
3
,\ll fluhged fit tins, fllrnsc,l
vxlve xnd flxnge rrcigl'rs inclu,le
thc proporlional weight of bolts
or studs to mcke up:rll joints.
286 Mechanical Design of Process Systems
\\ attr-l-bs/I t
f'2
!x
tr2
o
z {,\
F t-i .t
z
{i\
HI
e-
\IJ
'l'cmtx,miurr lLrngc'Ir
Ilagnesia
z Celcium
o
F
I Coml)inl- Nom.'l'hick., In.
iion
z
Fiber-
Sodium
@ ,lr1
on straight, pipe.
Valve wcights arc approrii-
mcte. When possiblc, obtrin
lveights from thc manuf&cturer.
Cast iron valve lvcights &rc for
(-!j
z |
F
w
fl-\
15.6 r7 .7
z
4'e.
B
,-'1-l
c_i_)
a-1--r
t{
To find the rveight of covering
on llanges, volvcs or fittings,
multitt]'thc wcight f$ctor by thc
\reight pcr foot of covoring used
@ on straight pipe.
Vdve rveights arc
mate. When possible, obtain
opproxi-
J
0 Flanged Bonnet
Check
weights from the manuflcturer.
Cast iton valve rveights are for
flonged end valves; steel rleights
++3 for welding end valves-
* 16
rc lt cu. ft. density.
All flangetl Iitting, flrnged
vslvc and flange weights include
the proportional weight of bolts
or studs to rnake up all joints.
288 Mechanical Design of Process Systems
gJ-f
z
{n
a-1J
z
{1\
E:cl
E_=_=r
'
!._!____,
3 @ on straight pipe.
Valve $eights are {rppror -
mate. When possible, obtain
ir) weights from the manufecturer.
Cast iron valve weights are for
flenged end valves; steel weights
+<i for selding end valves.
All ffanged titting, flanged
A
e,
T,Jr'
z
F
F uJ
z
{T\
r';J
,
lA
/t\
rFr
\iJ
Temperature Range 'F 1r00-1200
Magnesia
z Calcium
F
Combina-
2 tron
Fiber-
Sodium
All flanged
vrlves;stecl \\'eights
for \\elding enLl valves.
fittine, flanged
vslve and flongc Neights jnclude
FsO the proportionlrl lveight of bolts
* 16 lb cu. ft, density, or studs to make up all joints.
290 Mechanical Design of Process Systcms
IA
(,
z //\
k w
{i\
E.-I
.l 4'd',
!-l_,
t,t!
N'Iagnesia
z Calcium
'l
@ on strsight pipc.
|alve \rcishts ore luorori-
matc. \\'hcn- possil)le,
-;l)trirr
\r0iqhls from thc n)rnufscturcr.
ll.J Crrst iron rrlvc $cights rfe for
nuj
4 S.R. 90" Elbow
80
3
104
2
7fh
F:- L.R. 45" Elbow
60
1.3
7a IEl
1.3
r32 167 360
7
- J,l
s
180 273
) Lateral 5.4 5.4
' 33 44 94
Reducer .7
30 38 t9
Crp
Temperature Range 'tr' r00-199 200-299 300-399 400-499 500-599 600-699 700-7s9 800-899 900-999 1000-1099
1100-1200
Nom. Tbick.,In. 1% 1% 2 2tz 3 3 3% 4 4% 4%
{-iryTiu
Y Silicate Lbs/Ft 6.04 6.04 E.13 10,5 12,7 15.1 t7.9 17.9 20.4 20.4
s\"ssF
96 177 11E 209 261 341 | 475 928 1775 tion Neighis are l,rsrd on E5%
1.5 | 1.5 mrgncsiu and hrdrous calcium
o ,-{t silicxte at lL lbs/'cubic loot. The
2Ld
,. Al
S.R. 90' Elbow
265 453
5.2
345
5
509 669 E15 1474
6.2
listed thickncsscs and Ncishts of
combinrtion covering aie the
sums of thc inner loier of dirr-
| lA L.R. 90' Elbow
375
6.2 6.2
485
6.2
624
6.2
l59E
6.2
tomlceous ctrth at 21 lbs/cubic
EA
z&4 45" Elbow
235
4.3
3E3
4.3 4.3
414
4.3
469
4.5
?0sI rr24
4.7 | 4.8
foot and thc outer la\'er at
11 lbsTcubic foot.
Insuhtion rvcights includc al-
,to\l'anccs lor $lrc, ccncnt, can-
dflq Tee
403 684 513 754
7.8
943
8.3
136 t
8.7 9.3
v:rs, hanrls lrnd Drint, but not
spccial surface firishes.
To lin,l bhc ur'rglrt of coveling
Flanqed BonDet on flxrgcs, vrlvcs or fittirrae.
1.{3 687
7.a
l29a
4
1015
5
1420 2155
7
2770 4650
8 mult;l'h tl,c \reiglrt i,, tul l,\.the
\\(iAht l)cr foot ol coverirrg uscd
*@ Flaneed Bonnet
Globi or Angle
80E
9.4
r200
9.5.
7r0
5
1410 on strrlalrt DrDc.
Vrtr e rrcights
m:*c. \1'herr possil'le,
rtc errnlori-
obt:rin
j rqJ
Flanged BoDnet
Check
674
9.4
I160
9.5
560 720 1410
7.2
2600
8
3370
8
r'cights from the m:rnuilrcturcr.
Crlst iron vtlye wciqlrts &rc lor
{={t Pressure Seal
Bonnet-Gate
1975
5.5
2560
6
45t5
7
flangctl end v0lves: stccl \eichts
Jor rrcltiine cnd vclves. -
.\lt flerrgcd fitting, fl:rnged
ts0 Pressure Seal
Bonnet-GIobe
vnlvc rn4 lixfigc \rcigirts include
thc proportionrl Ncight
or studs to make
lcight of bolts
mrkc up rll joints.
rrn all ininlq
' 16 lb cu. ft. den6rty,
292 Mechanical Design of Process Systems
{?
z
|. fh
z
{t}
EJJ
t -=;t
/-\
\t/Temnr.r:1turc Rrngc 'F
Nlaguesia
2 Calcium
F
Conlbina-
Nom. Thick.,In.
tlon
z
Fibe!-
Sodium
tlpe is rrcight in
Boldlacc
ffi pouncls. Lightface tl'pe bencalh
$eight is \eight f.rctor tor
insulation.
z Insulation lhicknesscs and
6{rls $eights arc besed on
conditions and do not constitute
average
@
r)
on straight, pipe.
Valve rveights are approri-
mete. When possible, obtain
scigbts from tha manufrcturer.
CasL ilon valve Neights are for
flanged end vrlvcs; steel *eights
+<t tor rYelding end valvcs.
G
t4
n
L!_r'
z
F f>\
L4J
z
{l\
e4'4
B
r't\
!+i
f--.+--l
Sodium
E
t
@
fi1
werght.pcr foot uf covcriDg uscd
on strxrqht DlDe,
Valve rri-iel,ts al,rrroxi-
^re 0l,tcin
m.Ltc. \\'hcn I'ossil)le,
weights from th_e manui:rciuror-
Cast iroD vrlve $cishts cre for
+<t flangcd cud valvesi stiel $eiehts
{.!-r'
z
F
F
f>\
a-+-!
z
{T\
I-5:I
B
-t\"
&\
\JJ
'fonrl)erllturc lhrlac'Ir
Magnesia
2 Calcium
o
F
Combin.r- .\oro. Thi, k.,ln.
f tron
z
Fiber-
Sodium
z
stfN$ Fcights arc bascd on average
conditions and do not constitute
Nls c lecommcnd:ltion for
thickncsscs of mstcriols- Insuls-
tion $eights &re b:rsed on 85%
specific
@ on.sirlrigLt pipe,
v srvc \{crgn[s crc apl)roxl-
mate. \Vhen possiblc, obtain
IU \\cights from the manuf&cturer.
Cast ilon velve \Yeights &re for
Pipe-Lbs/ I t
$ atcr-l,bs,,lft
to
z f\
w
{l\
z L=I
F4'1
f-l
LJ-!
Magnesis,
z Calcium
Combina-
!ton
z
Fiber-
Sodium
'
rc
16 Ib cu. ft. deDsity.
-\ll fitngcrl Iitting, ilrrngcd
vlllve &n(l llllngc ryci'alrts il)(ludc
thc prol)ortionrl \ycight of l)olts
or studs to mrke up ull joir)ts.
296 Mechanical Design of Process Systems
\\'rll Dcsigrr,rtiou
f,.d
1,!J
{G
z t\
t-
w
{i}
1_'*,.1
z
14'1
/i\
-t
Lr----t
i--t
\*t"J
'I-cnDer:lturc llcngc "F
Boldfrre is weight in
ffi lleight.
,t\pe tJpe benexth
pounos, Lrgnlttce
is $eight factor for
Insulation thicknesses and
z
j
+r[1$ \reights are based on average
conditions &nd do not constituie
N+S a recommendction for specific
thicknesscs of msterials. Insule-
tion $cights are b.rsed on 857,
l:N magncsia and hydrous calcium
silicste rt 1l lbs/cubic foot. The
listed thicknesses and \eiqhts of
combination covering are the
sums of the inner laver of dia-
/14 tomaceous certh at 2l lbs/cubic
foot and the outer laver at
,N ll lbs/cubic foot.
z
/> Insulation *eights include al-
,loNMces 1ot wlre, cement, can-
l,, .{ vas, bands and paint, but not
D---S speciel surface finishes.
ffi
To find the geieht of coverins
on Banges, vrlvis or 6ttinss]
muJtipll the rreight factor by the
\aeiglrt.lrrr foot ol covering used
@
fi1
on sLrsrght DlDe.
\'Rlvt $ciehts rre annroxi-
mxtc. \\'hen- possiblc,
-obtain
\'eights from thi manuflcturer.
3 Cast iron valve \icights :rre for
frengcd end v:rlvcs, steeJ *eights
J-<t lor \reldrng end vslves.
fif
u-r'
7 h
IL4J
F
{l}
E=:l
F
-4\"
,TI
ri\
u/
r-r--r
A tion
3 ;r:r:::::
Fiber-
Sodium
z
t4 nation covering are the sums
of the inner later of diatoma-
3 ceous ealth it 21 lbs/cubic
foot and the outer laver at
,N i1 lbs/cubic foot.
z aglg/> Insulation weiphts include
allowances for w_ire, cement,
canvas, bands and paint. but
B,s
Ht not special surface finishes.
To find the weiqht of cover-
ing on flanges, valves or fit-
F<] tings, multiply the weight fac-
tor by the weight per foot of
@
lll')
covering used on straiqht Dipe.
Valve weishts are aoorbii-
mate. When- possible. bbtain
weights from manufacturer-
Cast iron valve weights are
for flanged end valves: steel
+<f weights forweldinqendvalyes.
All flanged fittlng, flanged
valve and flange welghts in-
FqJ clude.the prolo.rtiohal weight
oI oorls or studs to make uD
* 16 lb cu. ft. densit\-. all joints.
298 Mechanical Design of Process Systems
ff
&?
f^
F w
z
{i\
E::I
F -4\.
t-t-!
f---Fr
\iJ
Temperature Range "F
trIxgnesia
Cclcium
Combina-
!ron
Fiber
Sodium
#
z of the inner lafer of diatoma-
F ceous earth at 21 lbs/cubic
F foot and the outer laver at
11 lbs/cubic foot,
,N Insulation weiehts include
/9N allowances for wlre, cement,
z canvas, bands and paint.
-6nishes.but
D',
F-Jl'il
hot special surface
To find the weisht of covei-
ing on flanges, v-alves or fit-
tings, multiply the weight fac-
l"<3 tor by the weight per foot of
@
covering used.on straight pipe.
v arve welghts are approxt-
mat€. When possible, obtain
weights from manufacturer,
ll
m Cast iron valve weishts are
for flanged ehd valves; steel
+<i weightsforweldingend valves.
All flansed fittins. flansed
l 16
rcli cu. ft. density.
valve and-flanse wiiehts "in-
clude.the propo-!tional- wei ght
ol Dolts or studs to make ur)
all joints,
if,
45
u-r'
z ! ii
z
{i\
lj:I
i .4\"
\tJ
-!----l
Fiber-
Sodirrm
{!-r'
z
F
I i)
{l\
L-Li
2
b E:-:t
f,t\
ri\
\tJ
Temperature Range .F
Magnesia
Calcium
Z Siliccte
l UOmOrna-
5 tion
Fib€r-
Sodium
d
z of the inner lay€r of diatoma-
F ceous earth at 21 lbs/cubic
tr foot and the oute! layer at
11 lbs/cubic foot.
z
A
9.4
Insulation weights include
allowances for wire, cement,
canvas, bands and paint, but
D' .f not special surface finishes,
B_{i To find the weight of cover-
ing on flanges, valves or fit-
Fdl tings, multiply the weight fac-
tor by the weighi per foot of
covering used on straight pipe.
@ Valve weiEhts are approxi-
mate. When possible, obtain
weights from manufacturer.
D Cast iron valve weights are
for flanEed end valves; steel
+<i weights forweldingendvalves,
All flanged -fi tting, flanged
'
rc
16 lt cu. ft. density.
varve ano nange werEhts ln-
clude the proDortional weisht
of bolts o; stjuds to make-up
all joints.
!r
A
TJ-/
z
F
tr
b
{T\
z
F
//\"
E_=_=iI
"t\
\IJ
Temperature Range "F
Magnesia
Calcium
Fiber-
Sodium
r
rc
16 lb cu. ft. deDsity.
valve and flange weights in-
clude,the proportional weight
oI Dolrs or stucts to make uD
all joints.
302 Mechanical Design of Process Systems
Water-Lbs/Ft
{t/
/\
F
I tt
tr
{}
dJ
; 44"
L_r-!
\tJ
Temperature Range "F
Ilrgnesir
Crlcilrm.
z
F
J Com
!ton
z
Fiber-
Sodirm
feet of water kgs/sq meter 304.8 g.ains {troy) grains (avdp) 1.0
feet of water pounds/sq ft 62.43 grains (troy) grams 0.06480
feet of water poLrnds/sq in. 0.4335 grains (troy) ounces {avdp) 2.0833 x 10-1
leet/nin cms/sec 0.5080 grains (troy) pennyweight (troy) 0.04167
feet/min feet/sec 0.01667 grains/l.J.S. gal parts/million 17.118
teet/min kms/hr 0.01829 Srains/U.S. gal pounds/million gal 142.86
feet/ rn in meters/min 0.3048 graans/ lmp. gal parts/million 14.286
feet/ rnin miles/hr 0.01136 Srarns dynes 980.7
feet/sec cms/sec 30.48 grams grains 15.43
feet/sec kms/hr 1.097 grams joules/cm 9.807 x 10-!
teet/sec knot5 0.5921 Srams joules/rneter (newtons) 9.807 x 10-r
feet/sec meters/min 14.29 Srams kilograms 0.001
feet/sec miles/hr 0.6818 Srarns milligrams 1,000.
leet/sec miles/min 0.01136 grams ounces (avdp) o.03527
leet/sec/sec cms/sec/sec 30.48 grams ouhces (troy) 0.03215
feet/sec/sec kms/hrlsec 1.097 Srams poundals 0.07093
feet/sec/sec meters/sec/sec 0.3048 Srams pounds 2.205 x 10-r
feet/sec/sec miles/hrlsec 0.6818 grarns/cm pounds/inch 5.600 x 10-l
feet/ 100 feet per cent graoe 1.0 grams/cu cm pounds/cu ft 62.43
Foot - candle Lumen/sq. meter 10.764 grams/cu cm pounds/cu in 0.03613
Btu r.286 x 10 l grams/cu cm pounds/mil-foot 3.405 x 10-'
foofpounds ergs 1.356 x 107 grams/ liter graans/gal 58.417
foot-pounds gram-calories 0.3238 Srams/ liter pounds/ gal 8.345
np-nrs 5.050 x l0 ' grams/ liter pounds/cu ft 0.062427
foofpounds joules grarns/liter parts/mallaon 1,000.0
foot.pounds Kg-ca{ones 3.24 x 10-. grams/sq cm pounds/sq tt 2.0481
foot-pounds kg-meters 0.1383 gram'calofle5 8tu 3.9683 x 10-!
foo!pounds kilovr'att-hrs 3.766 x 10 , Sram-catofles ergS 4.1868 x l0'
foot'pounds/ min Btu/min 1.286 x 10 ! gram-calories foot-pounds 3.0880
foot-pounds/ min foot-pounds/sec 0.01667 gram-catofle5 horsepower-hrs 1.5596 x 10 .
toot-pounds/min 3.030 x 10 -' gram-calones kilowatt-hrs 1.1630 x l0-.
toofpounds/min kg-calories/ min 3,24 x 10 . gram-calo es watt-hrs 1.1630 x l0-3
foot-pounds/man kilowatts 2.260 x lO- 5 gmm-calories/sec Btu/hr t4.2a6
foot-pounds/sec 8tu/hr gram-centimeters Btu 9.297 x 10 .
foot.pounds/sec Btu/min o.o7717 gram-centimeters ergs 980.7
foot-pounds/sec horsepower 1.818 x 10-' gram-centimeters joules 9.807 x l0-!
{oot-pounds/sec kg'calories/man 0.01945 gram-centimeters kg-cal 2.343 x 10-r
foot-pounds/sec kilowatts r.356 x 10-' gram-centimeters kg-meters l0 -5
Furlongs miles (U.S.) 0.125
furlongs rods 40.0
turlongs feet 660.0
H
Hand Cm. 10.16
necrares 2.47 |
hectares sq feet 1,076 x 10'
gallons cu cms 3,785.0 hectograms grams 100.0
Sallons cu feet 0.1337 hectoliters liters r00.0
Salrons cu inches 231.0 hectometers meters 100.0
gallons cu meters 3.785 x 10-' hectowatts watts 100.0
Sallons cu yards 4.951 x 10-1 hennes mrllihenries 1,000.0
gallons liters 3.785 Hogsheads {British) cubic ft. 10.114
gallons (liq. Br. Imp.) gallons (U.S. !iq.) 1.20095 Hogsheads (U.S.) cubic ft. 8.42184
gallons (U.S.) gallons (lmp.) 0.83267 Hogsheads (U.S.) Sallons (U,S.)
gallons of water pounds oJ water 8.3453 horsepower Bt!/min 4?.44
gallons /m in cu ft/sec 2.228 x lO- I
gallons/min ho15epower foot'lbs/ min 33,000.
liters/sec 0.06308 horsepower foot-lbs/sec 550.0
gallons / m in cu ftlhr 8.0208 horsepower (metric) 0.9863
horsepower
gausses lines/sq in. 6.452 (542.5 ft lb/sec) (550 ft lb/sec)
gausses weoers/sq cm 10 | horsepower horsepower (metric) 1.014
gausses webers/sq in. 6.452 x 10-l (550 ft lb/sec) (542.5 ft lb/sec)
gausses webers/sq meter 10-. horsepower kg-calories/min 10.68
gilberts ampere-turns 0.7958 horsepower kilowatts 0.7 457
gilberts/cm amp-turns/cm 0.7958 horsepower watts 7 45.7
gilberts/cm amp{urns/jn 2.021 horsepower (boiler) Bt!/hr 33.479
gilberts/cm amp{urns/meter 79.58 horsepower {boiler) kilowatts 9.803
Gills (Britash) cubrc cm. t4?.07 norsepower-nrs Btu 2,547.
gills liters 0.1183 norsepower-nts ergs 2.6845 x 10r'
gills pints (liq.) 0.25 norsepower-nrs loot-lbs 1.98 x 1Cl'
Grade Radian .0t571 horsepower-hrs gram-calories 641,190.
G€ins drams (avoirdupois) 0.035s7143 norsepower-hrs joules 2,684 x 10.
1|
Synchronous Speeds
Frsoucncy x 120
Synd'ronout Spced : N;;|T;G;-
FNEOUENCY TREOUENCY
31 2 .8 176.5 71 97 .3 8l,l
35 200 166.7 76 91.7 78.9
Temperature Conversion
NOIA Thc c.ntlr .olsm'| of .'rmbcrt in boldl.ce r.ter3 to the tempe.ot'rre i. degree3, €irher Cenriorodc or fohre.heir, which ii i! dcti..d ro convcrt into th.
othe. .col.. lf .o.v.dine from Fohrenh.it ro Ce.ligrode degreei, lhe equivolen. tempe.oture will be found in rhe tefr coiumn, whit€ if Gonve.ri^s trom d.er.er
Cenligrodc to desrc$ Fohrenhi.t, th. on.wer wi be found in thc cotumn on rhe righr.
Ke/'q
Hs Abr. Hg Ab3. PSIA
E
o\
tr I I I
8
b
8
3 a5
e
3
q 5
E
R
o - d
b
:o
ao I 8
;E 6 6
I
EI A b -o
>! d
8 R
6
5€ E
I(l)
9p b
EO
o, -:
a:
8
q R
6 8 t5
6
Eg
8 P 3
ou b
iR
I
6 5
oer b
t'g
8
z
F
ao
o I
6 I
P
tr
0 F
F
C)
tt
zt! te ol
i€ .:ra
G
sj
8p Et
r= ):
:is6 3 oq
J9
€+ E -o
(9 5E Ee
.o
lo u3snnN NfAl
I L
Index
315
316 Mechanical Design of Process Systems
l:r.= r
gust response, dynamic, 194 Newtonian fluids, 21, 30,32 41. 1,19-l{l"t. l-!:.
gust response factor, 195, 196,217,236-231 145,147
gust size, 196 non-Newtonian fluids. See Hydraulics.
isopleths, 192- 193 Non-Newtonian fluids.
Kutta-Joukowski theorem, 195 Strouhal coefficient vs., 85
loading analysis, quasi-static, 196 vortex shedding, for, 83-85
logarithmic law, 192
parabolic area, centroid of, 219 Saddle plate design, 174- 179
parabolic function, 194 application of , 249 -252
peak values, types of, 196 boundary conditions for, 178
power law, 192 buckling coefficients for, 175- 178
probability of exceeding. 196 effective area, 174, 178
response spectra, 198 effective width, 173, 178, 179
return period, 192 horizontal react\on, 179, 252
similarity parameters, 195 stiffener plates, 174, 119
structure size factor, 196, 197 stress, criterion for residual, 178
surface roughness, 195 stress, (in-) elastic buckling, 179
tower U.S. Steel design method, 174-179
cross-sectional areaof, 198 wear plate requirements, 215
fluid force exerted on, 194-195 web plates, 174
gust velocity vs. structural response, 197 Seismic design
natural frequency of, 197 baseplate design, 238
wind area section properties, 219 coefficients, Mitchell, 210, 213
wind force distribution, 218 coefficients, structure tYPe, 210
wind distribution criteria, quasi-static, 210
parabolic, 194, 218-219 compared to wind, 238
triangular, 194 Mitchell equation , 210, 212
wind load compared to Rayleigh equation, 231-238
applications of, 215-231, 241-245 moments, equation for, 212
equivalent static, 195 occupancy importance factor, 210
mean, 195 period, characteristic site, 238
weld size, skirt-to-base plate, 189 period, vibration
welding, joint efficiencies for, 161-165,172 numeric integration of, 238 239
Zick analysis, 166, 215 tower,210,212
bending moment diagram, 167 Rayleigh equation, 212
compressive B-factor, 174 compared to Mitchell equation, 231-238
constant, circumferential bending moment, 170 seismic zone factor/map, 210, 2ll
head used as stiffener, l7l shear forces
saddle support location, 172 earthquake force, total, 212
shear stress in head/shell, 171 lateral force, equation for, 212
. shell vertical distribution of, 212
stiffened by head, 171 site structure interaction factor, 210, 212
unstiffened, saddles away from head, 171 equation for, 212
stiffening rings, 172, 174 skirt design, 238
stress, allowable compressive, 166 structural period response factor, 210
stress, circumferential compressive, 171 Uniform Building Code, 209-210
stress, location of, 168- 169 Skirts, 185
tangential shear, 167- 171 controlling criteria for, 184
wear plates, 171- 172 cost-plus contractor, 183
design of, 183
in, 132-135
Residual systems, heat transfer lump-sum contractor, 183
in piping, 154- 155 stress equation, 183
Reynolds number, 195, 2OO, 2Ol, 236 supports, 185
drag coefficient vs., 203 thickness, 183- 184
322 Mechanical Design of process Systems
lri:r
Zick analysis, 166, 215 unstiffened, saddles away from head, 171
bending moment diagram, 167 stiffening ings, 172, 174
compressive B-factot 174 stress, allowable compressive, 166
constant, circumferential bending moment, 170 stress, circumferential compressive, 171
head used as stiffener, l7l stress, location of, 168- 169
saddle support location, 172 tangential shear, 167- 171
shear stress in head/shell, 171 wear plates, l7l-172
shell
stiffened bv head. l7l