Steel Pipe- A Guide for Design and Installation

AWWA MANUAL M11

Third Edition

~o:k?!~@*\ American Water Works Association

Contents 7 , . . - . - : - > ; = . - .. . r < *

..... *. .

Chapter 1 1.1 I .2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10 1.11

History, Uses, and Physical Characteriodcs of SteeI Pipe. ....... 1 History, 1 Uses, 2 Physicai Characteristics, 3 Ductility and Yield S-, 3 Stress and Strain, 4 Smain in Dcsign, 7 Analysis Based on Strain, 8 Ductility in Design, 10 mects of Coid Worlnng on S- and Dudty , 10 Brite Fracrure Considemtiom in Stnicturiir Dtsign, 12 Good Practice, 15

Chapter 2 Manufacture and Testing .................................... 16 2.1 Manufacture, 16 2.2 Testing, 19

Chapter 3 3.1 3.2 3.3 3.4 3.5 3.6

Chapter 4 4.1 4.2 4.3 4.4 4.5 4.6

Hydraulics of Pipelines ...................................... 21 Formulas, 21 Caicuiations, 26 Economical Diameter of Fipe, 32 Distribution Sysams, 33 Air En-ment md Release, 33 Good Practioe, 33 Determination of Pipe Waii Thicluiess ......................... 36 I n t e d Pressure, 36 Working Tension Strm in Steel, 37 Tolerante, 38 Corrosion Aiiowance, 39 Externa1 Pressure-Uniform and Radial, 40 Minimum Wall Thicknecs, 40

Chapter 5 Water Hamrner and Pressure Surge. .......................... 51 5.1 Basic Relationships, 5 1 5.2 Checklist for Pumphg Mains, 54 . . 5.3 General Smdies for Water Hammer Control, 54 '

t . - - 5.4 Al lomce for Water Hammer, 55 . - . 5.5 Pressure Rise Calculations, 55

Chapter 6 6.1 6.2

ExternalLoad ............................................... 57 . . c . . -

- 4 . &.. . ;?y':: .;. Load Determination, 57 " - - . -

. . Deflsction Determimtion, W -. 1 ,

6.3 Buckling, 61 -. ,. 6.4 Extreme Externa1 Loading Conditions, 62 = - . > . . . . - F . , 6.5 Computer Programs, 63 A

Chapter 7 Supports for Pipe.. . . . . . . . 7. I Saddle Supports, 66 7.2 Pipe Deflection as k a m , 70 7.3 Methods of Cdculation, 70 7.4 Gradient of Supportsd Pipelines to Prevent Pocketing, 7 1 7.5 Ring-Girder ConstrucUon, 7 1 7.6 Ring-Girder Construction for Low-Pressure Pipe, 77 7.7 Installation of Ring-Girder Spans, 78

Chapter 8 8.1 8.2 8.3 8.4 8.5 8.6 8.7 8.8

Chapter 9 9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.8 9.9 9.10 9.11 9.12 9.13 9.14 9.15 9.16

PIpeJoints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Bell-and-Spigot Joint with Rubber Gasket, 86 Welded Joints, 87 Sleeve CoupLings, 88 Fimges, 89 Grooved-and- Shouldered Couplings, 89 Expmsion and Contraction-General, !M Ground Fnction and Line Temion, 91 Good Practice, 92 Fittings and Appurtenances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Designation of Fittings, 93 Bolt Hole Position, 95 Design of Wye Branches, 95 Testing of Fittings, 95 Unbalanced Thrust Forces, 95 Frictional Resistance Between Soil and Pipe, 96 Anchor Rings, 96 Nozzle Outlets, 96 Connection to Other Pipe Material, 96 Fhged Connections, 97 Valve Connections, 97 Blowoff Connections, 97 Manholes, 97 Insulating Joints, 98 Air-Release Valves and Air-and-Vacuum Valves, 98 Good Practice, 99

Chapter 10 Principies of Corrosion and Corrosion Control . . . . . . . . . . . . . . 101 10.1 GeneralTheory, 101 10.2 Intemal Corrosion of Steel Pipe, 11 1 - ,

, .

10.3 Atmospheric Corrosion, 11 1 10.4 Methods of Corrosion Control, 1 1 1 10.5 Cathodic Protection, 1 1 1

1 Protective Coatings and Linings . . . . . . . . . . . . . . . . . . . . . . . Requircments of Good Pipeline Coatings and Liaings, 115 Selection of the h p e r Coating md Lining, 1 15

1 1.3 Reaommended Cuatings and Linings, 1 17 1 1.4 h t i n g Application, 1 18 11.5 GmdPractice, 119

Chapter 12 Transportadon, Installation, and Testing. . . . . . . . . . . . . . . . . . . . 121 12.1 Transportation and Handling of Coated Steel Pipe, 121

. - . >

- -- 12.2 Trenching, 122

12.3 Installation of Pipe, 123 12.4 Anchors and Thrust Blocks, 127 12.5 Field Coeiting of Joints, 128 12.6 Pipe-Zone Bedding and Backfdl, 128 12.7 Hydrostatic Field Test, 129

Chapter 13 Supplementary Design Data and Detaib . . . . . . . . . . . . . . . . . . . . 131 13.1 Layout of Pipelines, 131 13.2 Calculation of Angie of Fabricated Pipe Bend, 132 13.3 Reiforcement of Fittings, 134 13.4 Collar Plate, 136 13.5 Wrapper-Plate Design, 138 13.6 Crotch-Plate (Wye-Branch) Design, 140 13.7 Nomograph Use in Wye-Branch Design, 141 13.8 Thmsr Restraint, 147 13.9 Anchor Rings, 151 13.10 Joint Harnesses, 151 13.1 1 S pecial and Valve Connections and Other Appurtenances, 15 1 13.12 Freezing in Pipeiines, 162 13.13 Design of Circurnferential Fillet Welds, 168 13.14 Submarine Pipelines, 170

Index, 173

This manuai was first authorized in 1943. In 1949, committee 8310D appointed one of its members, Russel E. Barnard, to act as editor in chief in charge of coiiecting and compiling the available data on steel pipe. The first draft of the report was completed by January 1957; the draft was reviewed by the committee and other authoritiec on sreel pipe. The first edition of this manual was issued in 1964 with the title Steel Pipe-Desi@ und Installation.

The 1985 edition of this rgmual was approved in June 1984. The principal changes from the 1964 edition rehted to externa1 loads on pipe, reinforcement of finings, and joint harnesses. Some of the rigorous descriptions, formulas, and calculations included in the 1964 edition were diminated where adequate references to such descriptions, formulas, and

. -

calculations were available. Some chapters of rhe 1964 edition were combined in the 1985 ,.. edition, thereby reducing the number of chapters from 19 to 13. Also, a comprehensive

< ---

index wrts added to rhe 1985 edition, and the manual's title was changed to Steel Pip-A Guide for Desi@ and Installation.

. . . . . -

This revision of the manual was approved in Jwie 1988. In addition to corrections to the ,:.- . ... . - , . - text and editing for chrity throughout, major revisions include redefinition of B' in Eq &7; . -

.- -.

. -- -

. - >&Tk - - . - redefinition of h, and Wc in Eq 6-8; revision of Sec. 13.9, including revision of Figure 13- 16

.-.

.+.- 3 - - - - - . Frank Corteilessa -.: Gary V. Heatherington

D.J. Cowling C.R. McCormick -- -

Demis Dechant Gary D. Redrnond . R. Dewey Dickson Edwin N. Seward

A.D. Harder George J. Tupac

This revision was idso approvtd by the Standards Committee on Steel Pipe and the Standards Council. The Standards Committee on Steel Pipe had the following personnel at the time of approval:

George J. Tupac, CIrwrman John H. Bambei Jr., Vice Chaiman

Dennis Dechant, Seoetary Consumer Members

J.H. Bambei Jr., Denver Water Department, Denver, Colo. G.E. 31wk Jr., GHR Engineering AswSates, Inc., Lakeview, Mass. R.S. Bryant, Department of Water and Bower, Los b e l e s , Calif. D. J. Cowling, US Bureau of Reclamation, Denver, Colo. J.L. Doane, Portland Bureau of Water Works, Portland, Ore. C.M. Frenz, M o m County Water Authority, Rochater, N.Y. Wesley Kremkau, Washington Suburban Sanitary Commission, Hyattsville, Md. Kemeth Olson, Tacoma City Water, Tacoma, Wash. E.C. Scheader, Bureau of Water Supply, New York, N.Y. G.M. Snyder, Metropolitan Water District of Southem CaIifornia, Los Angeles,

f.

General Inrerest M&s C.J. Arch, Consulting Engmeer, Covina, Calif. W.R. 3-11, Metcaif & Eddy, Inc., Arlington Heights, 111. R. Dewey D i c h n , J.M. Montgomery Consulting Engineers, Pasadem, Caiif. B.R. Eims*, Standards Engineer Liaison, AWWA, Denver, Colo. L. J. Farr, CH2M fIiU, Inc., Redding, CalZ. R.E. Gilmor, Consulting Engineer, Denver, Colo. . -. ... . G.K. Hickox, Engineering Consultant, Houston, Texas R.Y. Konyalian, Boyle Engineering Corporation, Newport Beach, Calif. -1 - G.M. Kroilik*, Naval Construction Bartalion Center, Port Hueneme, M f . G.D. Plant, Consulting Engineer, Napa, Mif. . . . . . L.T. Schaper, Black & Veatch, Kansas City, Mo. G. J. Tu- G. J. Tupac & A s e t e s , Pittsburgh, Pa. W.W. Webster, Leedshill-Herkenhoff, Inc., Albuquerque, N.M. :.K.G. Wiikes, Consulting Engineer, Escondido, Calif. --.R.E. Yomg, Robert E. Young Engineers, Sacramento, Calif. ._ .

Producer Members T.R. Brown, Rockwell Internationai, Pittsbwgh, Pa. Frank CorteUessa, Kaiser Steel Corporation, Fontana, M f . Dennis Dechant, c/o Thompson Pipe aud Steel Company, Denver, Colo.

_ A.D. Harder, Ameron Pipe Division, Portland, Ore. -George Harris, Taipecoat Company, Evanston, iil. C.R. McCormick, CRM Enterpises, Vacaviile, Calif. J.R. Pegues, American Cast Iron Pipe Company, Birmhgham, Ala. E.N. Seward, United Concrete Pipe Curporation, Baldwin Park, Calif. H.R. Stoner, Koppers Company, Newark, N. J. J.A. Wise, Canus Industria, Ltd., Vancouver, B.C.

.. . L.;:

AWWA M A N U A L un 1 History, Uses, and

5; > . ,

-. .-. . . , --

. .

-. , . ,

el pipe has been used for water lines in the United States sin- the early 1850s.' The pipe was fmr manufactured by r o h g steel sheets or plata into shape and riveting the seams.

of pipe, continued with improvements thicknesses couid be readily varied to fir the different pressure

y low tensile strength of the early steels, and the low efFiciency of . - cold-riveted seams and riveted or drive stovepipe joints, engineers initially set a safe design

S at 10 000 psi. Over the years, as riveted-pipe fabrication methods improved and r strength steeh were developed, design s m s e s progressed generally on a 4-m-1

safety factor of tensile strength, increasing from 10 000 to 12 500, m 13 750, and Finally to psi, adjusted as necessary to account for the eficiency of the riveted seam. The pipe

-= was furnished in diameters ranging from 4 in. through 144 in. md in thickness from 16 f single-, double-, triple-, md even

rying in eficiency from 45 percent to 90 percent depending on

d in 1905, had nearly supplanted riveted pipe by 1930. involved planing 30-ft long plates to a width approximately equal to half the cumference, upsetting the longitudinal edges, md rolling the plates imo 30-ft

"' long half-circle troughs. H-shaped bars of special aonfguration were applied to the mating edges of two 30-ft troughs aud clamped into position to form a full-circle pipe section.

Following the general procedure of the times, a 55 000-psi tensile-strength steel was , used. With a 440- 1 safery factor, this resulted in a 13 750-psi design stress. Lock-Bar pipe

tages over riveted p i p : it had only one or two straight seams and no round

-s. The straight seams were considered 100-percent efficient as compmed to the 4'S;percent m 70-percent eficiency g e n e d y applied to rivered -s. Manufxtwd in s h s from 20 in. through 74 in., from plate m q h g in thickness from 3/i6 h. to $2 in., Lo&-Bar played an increasingly g r a m role in the markct riritil the advent of automtic electric welding in the mid 1920s.

By the early 1930s, both riveting and --Bar methods grdudly pmsed out of the picture, and welding dominated $he field. Pipe produaed using automtic de&c Eusion welding offered thc advaniages of fewer pieces, fewer operakm, fmter production, smalIer seam protrusion, and 100-permnt oPelded+mm eflkiencp. Fa- of fusion-welded pipe foliowed somewhat the sime initial pduction sequemes as for M - B a r . Thmugh the 193th and into h e 19&, 30-ft fletes were used. By the 19509, some rms had obtained 48-ft dh, mda ii=w f o ~ ~ f t ~ ~ in p-,

h h g the developing d.ecstde of welding ip tk i9%, a new approach was mken to des@ stresses. Prior to that time, it had been c&on p d c e to work with a safety factor

- &

of Pta- 1 b& on the tensiie s t m q i d ~ As w d d pipe came into p r a d o h , the concept -

g of ~ ~ ~ - 5 0 w t of the field kmme g e m d i p accepted. - --

Spirsiily forrrred oind wclded ppewas devebped In thc early 19309 and was ~ s e d :- . . 'c --

extensivelg in diameters from 4 h. through 36 in. Welding was by ihe e l & ~ fusion L- - g-. . - methd. After Worid Wm 11, Ger~ian mcbines were impomd, andsubsequentiy domestic

ones were developed t h t could s p a y hrrn and w d h u g h dbekrs of 144 in. E:- -5; 1 3 m& 5-, 5,-

HISTORY, USES, PHYSICAL CHARACTERISTICS 3

The properties of steel rhat make it so useful are its great strength, its ability to yield or deflect under a 1 4 w u e stiii offwing fuli resismce to it, its ability to bend without breaking, and its resistan= to sho&. The water utiZity eagineer s h d d understand these properties, how they are measured, what they wiii do, and to what extent they can be relied

-

on.

- 1.4 DUCTILITY AND YIELD STaENGTH .-

Solid mmhis can be differentiated imo two dasses: d u d e and brittle. In engineering ? . r practice, these IWO classes must be mted differentlg b u s e they behave differently under 5 - load. A ductik material exhibits a ~~ pht ic deformatiun or fiow at a faairly definite 5

-

6

The top photwraph shows a section of pipe after it collapsed as a result of Ihe falure of automatic vacuum-relief val-. The r e s t w section, rounded out by water forced through under pressure, is sbown at the botiom. Figure 1-3 Sections of 94-111. Bwquet Canyon Ppeline

F

4 STBBL PIPE

stress level (yield point or yield strength} and S hows a considerable total elongation, stretch, or plastic defomtion before final breakage. With a brittle material, the plastic defomtion is not weli defined, and the ulrimate elongation before breakage is s d . Mild steel, such as

-- is used in steel water pipe, is typical of the duaile materials. (The behavior of brittle ?materials will not be examined in this manual.)

It is because of steel's ductility, its ability to yield or flex but not break, that the Bouquet h y o n pipeline shown in Figures 1-2 and 1-3 still operates satisfactorily in 1983 after 50 years of service. It is ductility that allows comparatively thin-wded steel pipe, even

;. - though decreased in vertical diameter 2-5 percent by earth pressures, to perform - satisfactorily when buried in deep trenches or under high fills, provided the true required

"- - - - strength has been incorporated in the design. It is also because of ductility that steel pipe

with theoretically high localized stresses at flanges, saddes, supports, m d joint-harness lug connections has performed satisfactorily for many years.

Designas who determine s m s using formulas based on Hooke's iaw find that the calculated results do not reflect the inregnty exhibited by the structures iiiusmted here.

reason for the discrepancy is that the conventionai formulas apply only up to a certoiin s level and not beyond. Many eminently safe stnictures and parw of structuces oontain

calculated stresses above this level. A fuli undersmding of tbe performance of sudi ures requires that the designer examine empirically the actual behavior of steel as it is

ded from zero to the breaking point. & - The physical properties of steel (yield strength ami ultimste tensile strength) used as

- the basis for design and purchase specif~cations are determined from tension tests made ona "standard specimen pulied in a tensile-testing machine. T h e strength of ductile materials, in terms of design, is defined by the yield strmgth as measured by the lower yield point, where one exists, or by the American Society for Testing and Materials (ASTM) offset yield stress, where a yield point d- not exist. For steel usually used in water pipe, the yield s t r q g h is fixed by s pecification as the stress due to a load causing a 0.5-percent extension of the puge length. The point is shown in Figure 1-4. The yield srrength of steel is considered to be the same for eithex tension or compression laads.

Ductility of steel is measured as an elongation, or stretch, under a tension load in a - testing machine. Elongation is a measurement of change in length under the load and is

, - %. .expressed as a percenuge of the original gauge length of the test specimen. . . < ,=.+-: - ,. .. - '% .- ., ... -<

-.: :, +- -. > - -

-

- , - L.-!. +.- . - . ,- E > , -

, ...

- ,;- - - =, . . - .

- 7 7 . ; .'d .

, . t: - - , , . . . , ^.---' - , - - . , .

- 1.5 STRESSANDSTMN - : - :-.., -

. .

,.= - 1 ; .:. , -

p:, , . . , > . - ' . .- ', -+

In engineering, stress is a f w e obtained by dividing a l d by an area. Strain is a length + - . - -

9 . - -

-- :--

MAGNlFlCATlON T O W MAGNiFiCATtO

ELASTIC RANGE REGION OF

PROWRT1ONAL

LOA0 ECTlONAL AEA

, =

INCREASE IN LENGTH ORiGINAL LENGTH

hape of the test piece of steel, which the test, is shawn by the bar9 drawn

HISTORY, USES, PHYSICAL CHARACTERISTICS 5

o O 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.8

TRUE STRAIN, PERCENT

Unlike canventonal stress-strain curves, both true stress and true strain have been calculated for the curves shown.

Figure 1-5 True Stress-Straln for Steel rdn Cuwe for Steel

m the l d by the actual cross-sectional ama of the bar as it decreascs in cross section under . found that the true stress never decreases. Figure 1-5 is a stress-strain

on which both true stress and true sirain have been plotted. Becsuse conventid sms-srrain di- are used commercially, only conventid diagrams are used for the remainder of this discussion.

Figure 1 4 shows various parts of a pure-tension stress-strain curve for steel such as that wed in water utiiity pipe. The change in shpe of the test piecc d d n g the test is shown by the bars drawn uader the curve. As the bar strettdies, the cross d o n decreases in area

maximum tensile smgth, at which point local reduction of ama (neckiag in) toikes

Many types of steel used in construction have stress-strain diagrams of the general :,,. form shown in Figure 1-4, whereas many other types used struchidy and for machine

pam have much higher yield and ultimate strengths, with rcduced ductiiity. Still other .,. useful eqheer ing steels are quite brittie. In general, the low-d-ty steels must be used at

rehtively low strains, even though they may have high stcength. The ssaending line in the left-hand portion of the gmph in Figure 1-4 is stmight or

nearly straight and has an easily recognized slope witb respea to the vertial axis. The break of the curve is rather sudden. For this type of curve, the point whtre the frrst

deviation from a straight h e occurs marks the proportional limit of the steel. The yield- strength is at some higher stress level. Nearly al1 engineering formulas involving stress&

a lmding such that working srsesses, as calculated, will be below &e: 7 2

STEEL PIW

Stresses md strains that f d below the proportional limit-that is, those that fa11 on the straight portion of the ascending line-are said to be in the elastic range. Steel s t n i c m s loaded to create stresses or suains within the elastic range r e m precisely to their original length when the load is removed. Exceptions may occur with certain kinds and conditions of loading not usually encountered in water utility installations. Within this range, stress increases in direct proportion to strain.

The mdulus ofelasticity, as commonly defined, is ihe slope of the ascending srraight portion of the stress-strain diagram. The modulus of elasticity of steel is about 30 000 000 psi, which means that for each incremenr of load that creates a strain or stretch of 1 pin. per inch of length, a sness of U) psi is imposed on &e steel cross section (30 000 000 x 0.000 001 = 30).

Immediately above the proprtional liihit, between it and the 0.5-percent extension- under-load yield strength of the material (Figure 1-4) lies a portion of the suess-strain diagram that may be termed the elastic-plastic range of the material. Typical stress-strain curves with this portion magnified are shown in Figure 1-6 for two grades of carbon steel used for water pipe. Electric-resistance strain gauges provide a means of studying the elastic-phtic segment of the curve. TRese and associated instruments dlow minute .. examination of the shape of the curve in a manner not possible before their development.

The elastic-plastic range is becoming increasingly important to the designer. Investigation of this range was necessary, for example, to determine and explain the

-7 . .

successful functioning of thin steel flanges on thin steel pipc4 Designs that load steel to -

'2 within the elastic-pktic range are safe only for certain types of apparanis, structures, or

.e- a.- - m parts of structures, For example, designing within this range is safe for the hinge points or

yield hinges in steel ring flanges on steel pipe, for hinge points in structures where local yielding or relaxation of stress must occur, and for bending in the w d of pipe under earth load in trenches or under high Fills. It is not safe m rely on performance within this range to

- handle principal tension stress in the walls of pipe or pressure vessels or to rely on such

- - - .

..

O 0,001 0.003 0.005 0.007 STRAIN, INJIN.

The curves show the elastic-plastic range for two grades of carbon steel.

Figure 1-6 Stress-Strain Curves for Cahon Steel

t D UI r t3

t 1 Ii 1 1 1 I I / -- I I

,PLASTIC- ]ELASTIC, + STRAIN ' STRAlN 5000 r l ~ . / ~ ~ +

Shown are the elastic and plastic portians of a stress-strain curve for a steel stressed to a given level. Figure 1-7 PIastic and Hastic Strains

HISTORY, USES, PHYSICAL CHARACTERISTICS 7

ACTUAL STRESS

- - - t . CO

- : , . - B . . (r +

- . p . + , * . -

. . . ..

. . : i- A-

Whon the total measured strain is known, the actual stress can be de'ermined by use of the s?essSJswajfi curve.

. > . * ? .,

figure 1-9 ~eterminat& of Actual Stress

. --

.?. +-

performance in other situations where the accompanying deformation is unmntroiled or m o t be tolerated. - , si,: --:-

e - ,:, - - Figure 1-7 shows the elastic and plastic portions of a stress-edn curve for a steel

stressed to a given levef. Figure 1-8 shows graphically how a completdy ficritious stress is - I determined by a formula based on Hooke's law, if the totai strain is mdtiplied by the

modulus of elasticity. The actual stress is determiued using ody the elastic strain with the mdulus of elasacity, but there is at present no way to separare theoretically the elastic and plastic strains in a structure. The ody dternative is to take the total measured strain as

:z . indicated by strain gauges and then determine the actud stress from the stress-strain curve, as shown in Figure 1-9. ,

1.6 S W N 1N DESIGN Analysis of a strucnire becames more complete when considered in terms of strain as weli as stress. For example, it has long k n known that apparent stresses calcuhted using classic formulas based on the theory of elasticity aregreatly in error at hinge-point stress levels. The magnitude of this error near the yidd-strength stress is demonstrated in the next paragraph, where the classically calculated result is compared with the measured performatlce.

By definition, the yield-strength load of a steel specimen is that load which causes a 0.5-percent extension of the gauge length. As was indicated in an earlier paragraph, in the elastic range a stress of 30 psi is imposed on the cross-sectiod sea for each microinch- per-inch increase in length under load. Because an extension of 0.5 percent corresponds w

8 STEBL PIPEvzec '

5000 pin./in., the calculated yield-strength stress is 5000 x 30 = 150 000 psi. The measuwd yield-strength stress, however, is on the order of 30 000-35 000 psi, or about one fourrh of the calculated stress.

-.- - , - Simikrly varied results between strain and stress analyses occur when the performance of steel at its yield strength is compared to the performance of its ultimate strength. There is a great difference in strain between the yield strength of low- or medium-carbon steel at 0.5-percent extension under load and the specified ultimate strength at 30-percent elongation, a difference which has a decided bearing on design safety. The specified yield strength corresponds to a strain of 5000 pin./in. To pass the specification requirement of 30-percent elongation, the strain at ultimate strength must be not l a s than 0.3 in./in., or

: =. .

300 000 pin./in. The ratio of strain at ultimate strength to strain at yield strength, therefore, . . is 300 000:5000, or 60: 1. On a stress basis, from &e stress-strain diagram, the ratio of 2 ultimate strength to yield strength is 50 m 3 0 000, or only 1.67: 1.

Actucilly, mild steels such as those used in waterworks pipe show nearly linear stress-strain diagrams up to the yield level, after which strains of 10 to 20 times the

, elastic-yieid strain occur with no increase in actual load. Tests on bolt behavior under tension substantiate this e f f e ~ t , ~ and the ability of bolts to hold securely md safely when they are drawn into the region of the yield, especially under vibration conditions, is easily

I 'explained by the strain concept but not by the stress concept. The bolts act sornewhat like extre-me17 stiff springs at the yield-strength level.

-- . ,

ALYSIS BASED O N STRAlN In some structures and in many welded assemblies, there are conditions that permit initial adjustment of strain to working load but limit the action automaticaily, eirher because of the nature of the loading or because of the mechanics of the assembly. Examples are, mspectively, pipe under earth load and steel flanges on steel pipe. In these instances, bending stresces may be in the region of yield, but deformation is limited.

In bending, there are three distinguishable phases through which a member passes when being loaded from zero to fdure. In the first phase, ail fibers undergo strain less than the proportional limit in a uniaxial suess field. In this phase, a structure will act in a completely e b t i c fashion, to which the classic laws of stress and suain are appiicable.

In the second phase, some of the fibers undergo strain greater than the proportional or elastic limit of the material in a uniaxial stress field, but a more predominant portion of the fibers undergo strain less than the proportional limit, so that the structure still acts in an 3 essentially elastic manner. The classic formulas for stress do not apply, but the strains can be

. . . , -r ' .

2 1 adequately defined in this phase.

c: -: -.

4 In the third phase, the fiber strains are predominantly greater than the elastic limit of

-j '.. . the material in a uniaxial stress field. Under these conditions, the structure as a whole no 4

longer acts in an eiastic manner. The theory and formulas applicable in this phase are being developed but have not yet reached a stage where they can be generally used. .

An experimental determination of smin charaaeristics in bending and tension was made on medium-carbon steel similar to that required by AWWA C200, Standard for Steel

-

', Water Pipe 6 Inches and ~ a r ~ e r ? Results are shown in Figure 1-10. Note that the proportional-limit strains in bending are 1.52 times those in tension for the same material.

- Moreover, the specimen in bending showed fully elastic behavior at a strain of 1750 pin./in., - . - = : which corresponds to a calculated stress of 52 500 psi (1750 x 30= 52 500) when the modulus

- of eiasticity is used. The specimens were taken from material having an actual yield of 39 000 .,- -22, ' psi. Therefore, this steel could be loaded in bending to produce strains up to 1750 pin./in.

. -; - - - and stiU possess fuii elastic behavior.

'J - - . > Steel ring flanges made of piate and fillet welded to pipe with a comparatively thin wall

- . -+- have been used successfdly for many years in water service, and this experience f o m a

HISTORY, USES. PHYSICAL CHARACTERISTICS 9

O 0.001 0.002 , - 0.003 - . 0.004 STRALN, IN./LN. -. . .

_, - -

L=! .-

~ p a ~ i ~ ~ a l nmlt (P.L.) strains in bending are 1.52 times thse in tension for the- material. - ffgme 1-10 Experlmentai Detenninatian of Strain Characterlstfcs . L . . ' _ , > .

- -

+ , , Ti. .. . . '

t- ble 1-1 M m u r n Strain in Rpe Wall Developed In PractIce

- - . -

h: M, R.E. k i g n of S e 1 Rhg F k g a for Water Works Serk+A Progress Report. Jow. A WWA, 42 10:93 1 Ig. 1950). .., . - - .

. i . .;. . m - a- - 1.. ? _. .

r- .- 9 -- &:= good background agoiinst which to make calculations. The fianges ranged from 4 in. through

1- - 96 in. in diameter. Calculations were made to determine the strain that would occur in the

..j-

pipe wall adjmnt to the h g e s . Table 1-1 shows the results. Mote from the table that, in practice, the limiting suain was always below the

s..-q.: T.-; commody recognized yield-strength strain of 5000 pin./in. but did approach it quite 4. . . dosely in at h s t oae instan=. Al1 of these flanga are suff~ciently satisfactory, however, to

2; -. . warrant their continued use by designers. The idea of desi- a structure on the basis of ultimate i d capacity from test data

-

rather than entirely on diowable stress is simply a return to m empirical point of view, a w p . point of view that early engiaeers were obliged m accept in the absence of knowledge of the .

mathemstics and statin necesssry to calculate srresses. The recent development of r 3 mathemaucal processes for stress analysis has, in some instances, overemphasized the b;

. !. impomnce of stress and underemphasized the importante of the o v e d strength of a 4z j.gr. strumre.

1.8 DUCTlLdTY IN DESIGN -.

HISTORY, USES, PHY SICAL CHARACTERISTICS 1 1

STEEL PIPE

indiated in Figure 1 - 1 1, the decrease in ductility is approximately equal to the momt of inelastic prestrain.

A stetl spcimen that has betn straintd into the 86-hardcning mge, unlmded, and ailowed to agc for several dap at m m temperahut (or for a much shorter time at a moderately elevated tcmptraturc) wili tend to follow the path hdicated in Figure 1-12 during reloading.'O This phehomenm, known as s tdn aging, has the effect of increasing yield and tensile strength while dtmasing ductility.ll

T h e effects of aold work on the strengch and dudiity of the stniaurai steels can be eliminated Iargcly by thcmd s m s relief, or ariiaealing. Such tteatment is not oilways possible; fomaattiy, it is not o f b necessary.

1.10 BRilTLE FRACTURE CONSIDERATIONS IN STRUCTURAL DESIGN

General, Considerations z . 1- . .., As temperanite decreases, an increase is gcncraiiy nokd in the yield stress, tensile strength,

. .

. -

modulw of elasticity, and fatigue smngrh of tk plate seels.* In conrrast, tbt ductility of . .

,A ,- ' C . ..

i, - . these steels, as measurcd by rcduction ia area or by elonption under load, dtmasts with

-. decreasing temperanms. Furthmnore, there is a temperarure behw which a smictural steei subjected m tensile stresses may fmcture by cleavage, with little or no piastic deformoltion, rather thain by shear, which is u s d y preceded by a considerable amount of plastic defomtion or yieiding.7

Fracture-that occurs'by cleavage at a nominal tensile stress below the yield stress is aommonly referred m as brittie fmctwe. Generaly, a brittle fracture can mxr when there is

- - a suffcicntly adverse combimtim of m i l e stress, tempemttm strriin mte, arid geometricai dkoatinuity (such as a notch) present. Other design md fabriatim factors may also have an important innuenct. Because of the interrehtion of these effccts, the e x x t combination of stress, temptratuce, notch, and other conditions that will cause brittie fracture in a given

: strumre canaot bt xleadily calculated. Consequently, designing against brittle fracture ofm consists mainty of avoiding conditions that temi to cause brittlc fracture mi selecting a steel appropriatc for tbt application. A discussion of thae facton is given in the following pamgmphs. Referaces 12,13,14, aad 15 -ver the sub@ in much more detd.

Fracture mechada offers a more direct approach for predidon of crack propagation. For rhis analysis it is assunmed tbat an internai imptrfectim iddizcd as a crack is present in the structure. By linearelastic stress d y s i s and hboratory tests on pmxackcd specimens, the applied stress that wiU cause rapid crack propagation is relatd to thc size of the imperfectim. The application of fracture mechanics has become incrcasingly useful in dweloping a fracture-control p h and establishing, on a rationai basis, the interreiated requirements of mareriai selection, design stress level, fabrication, and inspection requirements?

Conditions Causing Bnttie Fracture L . 1 : - . :.;,

. +

Piastic deformation can occur ody in tbe presence of shear srrtsses. Shear stresses are always present in a uniaxial or a biaxid s t a u of stress. However, in a triaxid seate of stress, the maximum shear stress approaches zero as the prinupd stressm approoich a common value. As a result, under equal triaxial tensile stresses, failurc occurs by cicavage mther thm

*Tbis gccrion was obtaiacd fmm rrfffenee 9 with minar editing. tShear and chvagt are usad in the mcmliwgki sense (maarisaopically) to denote dificrtnt f m

mechanisms. RefmriEt 12, as wcii as most elcmenmry twxbmh on a u r g y , d i i t b c mechpnisms.

HISTORY, USES, PHYSICAL CHARACTERISTICS 13

by shear . Consequently, trhxial tensile stresses tend to cause brittle fracture and should be avoided. As discussed in the foliowing material, a triaxial state of stress can result from a uniaxial loading when notches or geometrical discontinuities are present.

If a transversely notched bar is subjscted to a longitudinal tensile force, the stress concentration effea of the notch causes high longitudinal tensile stresses at the apex of the notch and lower longinidinal stresses in adjacent material. The lateral contraction in the width and thickness direction of the highly stressed material at the apex of the notch is restrained by the smaller lateral contraction of the lower stressed material. Thus, in addition to the longitudind tensile snesses, tensile stresses are created in the width and thickness directions, so that a triaxial state of stress is present near the apex of the notch.

The effect of a geometrid discontinuity in a strumre is generaliy similar to, dthough not necessarily as severe as, the effect of the notch in th t bar. ExampIes of geometrical discontinuities sometimes found include poor design details (sudi as abrupt changes in cross section, attachment welds on components in tension, and square-cornered cutouts) and fabrication flaws (such as weld crack, undercuts, arc strikes, and scars from chipping hammers).

Increased strain mtes tend to increase the possibility of brittle behavior. Thus, structures that are loaded at fast rata are more susceptible to brittle fracture. However, a rapid strain rate or impact load is not a required condition for a brittle fracture.

Cold work, and the strain aging that n o m l l y follows, generally increases the Wrelihood of bnttle fractures. This behavior is usuaiiy attributed to the previously mentioned reduction in ductiiity. The effect of ooId work that occurs in cold-forming operations can be minimized by selecting a generous forming radius, thus limiting rtie amount of strain. The amount of strain that can be tolerated depends on both the steel and the application. A more severe but quite localized rype of cold work is that which o c a m at the edges of punched holes or at sheared edges. This effect can be essentidy eliminated for holes by drilling instead of p u n c h g or by reaming after punching; for sheared edges, it can be elimhated by machining or grinding. Severe hammer blows may also produce enough cold work to locally reduce the toughness of the steel.

When tensile residual stresses are present, such as those resulting from welding, they add to m y applied tensile stress, resulting in the actual tensile stress in the member being greater than the applied stress. Consequently, the likelihcad of brittie fracture in a structure that contains high residual stresses may be rninimized by a postweld heat treatment. The decision to use a postweld heat treatment should be d e with assurance that the anticipated benefits are needed and will be reahzed, and that possible harmful effects can be tolerated. Many modern steels for weldd construction are designed to be used in the less costly as-welded condition when possible. The soundness and mechanical pmperties of welded joints in some steels may be adversely affected by a postweld heat treatment.

Welding may also contribute to the problem of brittle fracture by introducing notches and flaws into a structure and by causing an unfavorable change in microstrumire of the base metal, Such detrimental effects can be minimized by properly designing welds, a i n g care in selecting their location, and using gwd welding practice. The proper electrde must be selected so that the weld metal w U be as resistant to brinle fracture as the base metal.

Charpy V-Notch lmpact Test Some steels will sustain more adverse temperature, notch, and loading oonditions without fracture thm will other steels. Numerous tesrs have been developed to evaluate and assign a numerical vdue indicating the relative susceptibility of steels to brittle fracture. Each of these tests can establish with certainty only the relative susceptibility to brittle fracture mder the particular conditions in the test; however, some tests provide a meaningfd guide to &e relative performance of steels in structures subjected to severe temperature md stress

14 STEEL PIPE

conditions. The most commonly used of these rating tests, the Charpy V-notch impact test, .... , . .A: is described in this section, and the interpretation of its results is discussed briefly.

Referentes 12 and 13 give detailed discussions of many other rating tests. - . . ...

. . T h e Charpy V-notch impact test spediically evdiluates notch toughness-the resistance -. . to fracture in the presence of a notcb-and is widely used as a guide to the performance of , - steels in structures susceptible to brittle fracture. In this test, a small recranguhr bar, with a

. . . , . .-

, - * V-shaped notch of specified size at its midlength, is supported at its ends as a beam and ' , ..,

. >. . - ! ' . fractured by a blow from a swiuging pendulum. The energy required to fracture the

. ., - - -. = specimen (which can be caiculated from the height to which the pendulum raises after

breaking the specimen) or the appearance of the fracture surface is determined for a m g e of - .

- . - -.. :.- , :;. . .

- - - temperatures. The appeararice of the fracture surface is usually expressed as the percentage - .

-: - . re , -+ - - ..- * - :: . . >+ ,!: -

. , ,

of the surface that appears to have fractured by shear as indicated by a fibrous appearance. A . . S - -. .;tL-.c: shiny or crystalline appearance is ascociated with a cleavage fracture.

(such as those shown in Figure 1-13) of energy or ion of temperatwe. For the structural steels, the

ure decrease from relatively high values to rektively low sing temperature. The temperature near the lower end of the

temperahire curve, at which a selected value of energy is absorbed (often 15 ftmlb ), is temperature. The temperature at which the percentage of .

shear fracture decreases to 50 percent is often d e d the fracture-appearance transition ition temperamre. Both transiuon temperatures provide a

resistance of various steels; the lower the transition the bemr &e resismce to brittie fracture. The ductility transition temperature

transition temperature depend on many parameters (such as wmposition, rmomechanical processing) and, therefore, can vary significantiy for a

ss of steels used for spec3c applications can be determined " service performance. Fracture-mechanics methods, when used in :

study of material properties, design, fabrication, inspection, . erection, and service conditions, have proven useful. In general, where a given stel has been used successfdy for an extensive period in a given application, brinle fracture is not likely to be experienced in similar applications unless unusual temperature, notch, or stress conditions are present. Nevertheless, it is always desirable to avoid or minimize the previously cited adverse conditions that increase the susceptibility to brittle fracture.

. .

. . .

, - .

_ BO - NOTE Cunies are for mrbon steel

- .. 50 - and are tiken from re fe ren~ 13. 100 - m + .. . . 5 8o -;. - ?2 7 .> P- E ' - 3 ' ; e- -7. < m -

A. Energy Trmslth Cuwe E m - 5. Fracture Trandiion CUN 20 - #- 40-

W I 10 - m -

/- o I I -. I t 1- o 1 1 1 l l l l l

+U 4 0 -20 O 20 40 80 100 120 140 +O 4 0 -20 O 20 40 80 80 100 120 140 TEMPERATURE. 'F -.. : TEMPERATURE, 'F

... :.

Source: Btockenbrough. R.L. & Johnsion, B.G. USS Steel Design Manual. ADUSS 27-340044. US Steel Corp., Pittsburgh, Fa. (Jan. 1981). Figure 1-1 3 Transition Curves O btnined from Charpy V-Notch lmpact Tests

. .

HISTORY, USES, PHY SICAL CdARACTERISTICS 15 - - -7. .

-. ...

. . .'.-.. , -5 2.- 7.

,

. . . :

.- i.ii GOO~FRACTICE The ordinary water pipeline requirts li* stress calcularion. The commonly used internal pressures for stetl water pipe are given in Tables 4-1 and 4-2 in Chapter 4. Suggested design stresses to resist other loadings are given as guides in various chapters on the different design

When designing the details of supports, wye branches, and other specials, es pecidiy for large pipe, the engjneer wiii do weU to consider the data in Chapter 13.

The concept of designing on the basis of strain as well as stress will shed light on the behavior of steel and other mterials in many cases wkre consideration of stress done offers no reasonabk exphtion. The adon and undesirabk efftcts of stress raisers or stress concentrations-such as notches, threads, hps, and suddw chaages in cross section-will be better understd. The steps to be taken in counteracting advcrse tffects become clearer.

References 1. ELLIOT, G.A. The Use of Steel Pipe in

Water Worb. Jow. A WWA, 9:11:839 (Nov. 1922).

2. CATES, W.H. History of Steel Water Pipe, Its Fabrication and Dcsign Develop- ment, (Apr. 1971).

4. B m A R D , R.E. Design of Stacl Ring Flangcs for Water Works Service-A ProgmriRepwt.Jm. AWWA,42:10:931

&m Steel Co., B e W r n , Pa. (1946,

9. BROCKENBROUGH, R.L. & JOHNSTON, B.G. USS Steel De& M a d . ADUSS 27-3400-04. US Steel Corp., Pinsburgh, Pa. (Jan. 198 1).

10. DI^, G.E. JR. Mechanicol MetalIingy. McGraw-m1 h k Cumpany, New York (1%1).

11. CHAJES, A.; B R ~ c , S.J.; & W I ~ , G. EffeCKs of cm-Smin i t lg on Smictural Shwt Steels. Jour. of the Stnictural Div., Pm., ASCE, 89, No. ST2 (Apr. 1963).

12. PARKER, B.R. Brittle Behapnot of Engi- neering S ~ t u r e s . John Wiley ami Sons, Ntw York (1957).

13. Control of Steel Construction to Avoid Bri* Failure. WcIding Regtaxch Coun- ciI, New York (1957).

14. LIGHTNER, M.W. & VANDERBECK, R. W. Factors Involved in Brittle Fracture. Regional TecIinicai Mectiags, Americsin Iron and Steei Insritute, Washington, D.C. (1956).

15. WLPE, S.T. & B~RSOM, J.M. Fracture and Fanngare Control in Structures-Appli- catims of Fracture Mechanics. Praiucc- Hd, Inc., E n g l e w d Cliffs, N. J. (1977).

i

AWWA MANUAL 1

Manufacture and Testing

i r ' : - ;+-- ..*: - ." I.%_ '

welding md ele& fusiw welding me the most common methods used steel b m , plates, sheets, md strips into tubuloir products.

Electric resktmce welding (ERW) is done without filler materioil. The fht strip, with edges previously trimmed to provide a &m, even S& for welding, is formed progressively into a tubular shape as it mveh through a series of mils. Thc forming k done

, . cold. Welding is then e f f d by the appliation of herit and pressure. The weldjng heat for -:>. .. the h r b h ed- is generaed by resktance to &e flow of an e&c cunrnt, which can be

inuriduccd through eIectrodts or by induction. Pressure roUs force tht hcated edges togaher to effect the weld. The sq- d o n of tht pxtssure roUs forming the weld cause some of the hot weld mcmi to be exmded from tbe joint m forma bead of wtld flash botb. b i d e and outside rhe pipe. The flash is wirmally trimmed within tolerame limits while it is su hot from welding, using mechanical cutting tools contoured to the shape of the pipe (Figures 2-1 tbrough 2-5).

Electnc fusion weldMg (EFW) differs from ERW in that filler mterial is used and mechanid p-ssure is unnecessary to effect tae weld. Pipe pradud with this process can hriw straight or s p i d Mms. Straight-seam pipe is made from plate with e d p planed prirallel to each other and squrire with the ends. Curving the phte edges with crimping rolls is the first step of the forming prooess. This is foIlowed by presses that form the phte First into a U&@ trough d then into a full O-sboiped tube. The O-shped tube is then fed into a longitudid seam-weiding machint. Spiral-seam pipt is madt from coiied strip or platc by a continuous proas (Figure 26). An automatic macbinc m U s the coa, prepares the edges for wel- ami s p i d y fonns the suip into a tubular shape. As thc tubt h v e s the forming elenrtnt, the &es are j o W by fusion welding in the same submergcd-arc proatss as is generally u& in straight-sam pipe ( F i i r e 2-7). The welded rube is cut to tbt d e s i d length by an auuimatic cutoff device.

MANUFAnURING AND TESTING 17

WELDED TUBE PRESSURE-63 1

ROLL

STR l P , FROM

COlL

WELDING ELECTRODES FINAL FORMING ROLL FIRST FORMING ROLL

Source: See Figure 2-1

The current ente= the tube via slidlng contacts and flows along Vee edges to and from the weld point.

Figure 2-3 Electric Resistance Welding Using High- Frequency Welding Current

INDUCTION COlL I

POINT SEAM

h 1 A 3 WELD WELDED

Source: See Figure 2-7

Eddy current flows around the back of the tube and along the edges to and from the weld point.

Figure 2-4 flectric Resistance Welding by Inciuction Using High-Frequency Welding Current

18 STEEL PIPE

1. Edge Planing-Submerged are weld plpe begins as a flat rectangular steel plate from the plate mlii. Yhe flrst step in tmnsforrning it to p i p is planing the edges paraiiei to each ofher and square with ths ends.

2. Edge Crirnping Rolls-Here the edgesrif *e pk te are curved to Tacllitat flnat t~rrning of the ppe, F&UM die wear. and produce gresrer unllormity at the seam edgea when the plate is pressed to acyiindtical shape. The total s u m e of tha plate, both sides edge to edge, is atso inspected ultrasonlcaiiy.

] CLEAN I NO TORCH OuiDE R o L Z

4, Gkrg e--The U-

*pea'p-eenters thls ;._ p - t k :cj;:'pmkkcular diea open. " 5Tlk-w die, under

;':&&~uIFc pressure, IS - ;.- -%&yd -- &M on the U, :: ::-&Htwrrring it to a

g?iWCal shape.

5. Outside Wekiing-The O-formd plate is now fed into a longltudkiai seam wld ing machine in which me abuttlnp edges arepropedy igned. firmly p r e s d together. and welded by the submergetd arc process. Two electrodes are used, and the wJd is completed to within 3 in. of tha pipe ends.

3. U-ing Press-A sernicircular ram descends on the piate. forcing it down between rocker dies to form a U. The plate 1s slightly over-bent to allow for spring-back.

WlRE ELECTRODE FLUX /

TV CAMERA

1

6. End Wdding-Here a s in . steel plate is imched to each end of the pipe at the searn,-permlttlng me iast tew inches of the OD seam to be welded.

1 7 4 DLES GLOSE0

WATER UNDER PRESSURE 7. 8. lnside Welding-Here the welding Expsnding and Testing-TRe pipe is either mechanically or hydrostatically expanded head and a srnali TV camera are depending on t h i mil1 location. In elther case. accurate aiza and straightrtsss and improved mounted on a Iong cant i lmr transverse yleid strength are obtained by expandon. boom. As the pipe is drawn oWr the welding boom, a iV s z m at Mechanical Expander-The pipe is rnechanically Hydrostatic Expander-00th ends the oprator's control W expanded in24-in. through 27-ln. lncrernents of the pipe are sealed by mandrels. enabies him to keep fhe welding until haH Of the length is completed. The pipe Thesemicircular dios, slightly levger exaaly on the searn. Finishiw up rolis to a second expander die where the than the pipe 00, are ciosed, and the on the tab, the last few inches of remaining half of the length is expanded. The pipe ii hydraulically expanded the seam are welded. The srnall plpe length then proceeds to a hydroatatic unit against the dies. The dles are opend plates are then remowd, and the where a specified interna1 pressure is applied to and a specified interna1 preasure cornpleted weld is inspected test tbe weld for sweats or leaks. applled to t a t the wetd for sweats or inside and out. leaks.

Sourm: Carbon Steel P i p , Strudural Tubing, Line Pipe and Oil Country Tubular Goods. Steei Products Manual. American /ron end Steel lns titute. Washington, D. C. (Apr. 1982). Figure 2-5 Sequence of Operations in a Typical Double Submerged Arc Weld h e s s

MANWFACTURING AND TESTiNG 19

Figure 2-6 Sdiematk Diagrm d Process for Making Spiral-S- Plpe

G-hmailr Iiiaaram fnr Maliina Phtn Clirra

- . ..- .

.= .-

- .

. .

-- .,

. -

limitations of individual pipe manufacturers. This p m s is esmaiiy suikd for Spe in thod9 . . . , -

.

. .

-Tests of Chemicai Properties The various services to which steel pipe is put require a variety of demical compositions to

: produce the necessary characteristia. The chemical compsitions established in the

term appiied no the chemical analysis is is the d y s i s reported to the pwdmer.

-

d & e steel from the U.

- , -. . . . . - - .

' Pipe sizes mmufactud using the fusion wciding process are G t e d only by si%

,.&$-wA sted pipe standards are suited to the usud needs d water utility applications. &-er, there are other steel materials tbat may be t q d y suitable, and these can be

results are determined by testing for such elemeats as h i e been specif&d, using a ~miple obtained from the first or middle p m of the heat or blow during the

20 STEEL PIPE

It is common practie in most steel melting operations to obtain more than one la&-test ingot sample from each heat or bIow; often three or more are taken, representing the fmt, middle, and last prtions of the heat or blow. Drillings taken from the first or middle sample are used in determining the ladle analysis because experience has shown that these locations most closely represent the chemical analysis of the entire heat or blow. The additionai samples are used for a survey of uniformity and for control purposes.

Check analysis. Check andysis, as used in the steel industry, means analysis of the metal after it has been rolled or forged into semifmished or fmished forms. Such an analysis is made either to verify the average composition of the heat, to verify the composition of a lot as represented by the ladie analysis, or to determine variations in the composition of a heat or lot. Check analysis is not used, as the term might impIy, m confirm the accuracy of a previous result. Check analysis of known heats is justified only where a high degree of uniformity of composition is essentid-for example, on material that is to be heat treated. Such analysis should rarely be necessary for water pipe, except to identify or confirm the assumed adys i s of plates or pipe that have lost identity. The results of analyses representing different locations in the same piece, or taken from different pieces of a lot, may differ from each other and from the M e anaiysis owing to segregation. These permissible variations from the specified ranges or limits have been established in the applicable specification or by common practice. me variations are a natural phenomenon that must be recognized by inspectors. The methods of anaiysis commonly used are in accordance with the latest edition of ASTM ~ 7 5 1 > those approved by the National Bureau of Standards, or others of equivalent accuracy.

Tests of phydcal properties. The methods of testing the physid properties of steel pipe are established in ASTM ~ 3 7 0 2 The physical properties required are containd r in AWWA C200, Standard for Steel Water Pipe 6 Inches and ~arger? or are as otherwise - i specified by the purchaser.

. Hydrostatic test of straight pipe. Straight lengths of pressure pipe md tubing are customarily subjected to an interna1 hydrostatic pressure test. This operation is conduaed

" as a part of the regular mil1 inspection procedure to help detect defects. I t is not intended to bear a direct reiationship to bursthg pressures, working pressures, or design data, although 1 test pressures sometimes influence design pressures. AWWA C200 contains a formula for

. . determining hydrostatic test. It is customary to make hydratatic tests at the pressure required by the standard

-

during the course of manufacture of the pipe. The requirements for hydrostatic testing in ; the presence of the purchaser's inspector involve additional handling, unless tfie inspector is present during the course of manufacture. The producer, on request, customarily furnishes a certificate confirming such testing.

Tests of dimensional properties. The diameter, length, wall thickness, straight- - ness, and out-of-roundness of pipe are checked as part of the normal manufacturing procedure. Such dimensions are subject to the tolerantes prescribed in the appropriate standards or sptcifications. . .-

Referentes 1. Methods, Fractices, and Definitions for 3. Steel Warer Pipe 6 lnches md Larger.

Chemicd Andysis of Steel Products. 2 - , AWWA Standard C200-80. AWWA, ASTM Smndard A751-77. ASTM, Phil- Denver, Colo. (1980). adelphia, Pa. (1977).

2. Methods and Definitions for Mechanical - . - Testing of Sted Prcdum. ASTM Stand- -

, .

- . , .. .. . . ard A370-77. ASTM, PhiIadelphia, Pa. : .

, (1977). ' L.:

. 3% . - -

AWWA MANUAL

- -Hydraulics of Pipelines

E

>: - : ,. .: - .: . .-,~,&E-A .+ :

This chapter is p r i d y cace& wirh the flow of water in transmission conduits; it is not intended to cover flows through the complicated networks of distribution systems. &cause this manual is a pide to p d c e rather thm P textbook, historid md theoretid development of the many hydraulic flow formulas has been omitted, as has discussion of universal or ratiod formuias.

The discussions d data in this chapter are therefore restricted to the three formulas bebeved to be most commonly used in water flow d d t i o n s in the westem hemisphere. Definitions of the hydraulic md other symbols used in the following formulas are given at the end of the chapter.

The Hazen-Wllliarns Formula Probably the most popular formuia in cumnt use among waterworks engineers is the Hazen-Williams formula. This formula, fmt pubtished in 1904, is:

V = 1.318 c # ~ ~ - The head l a s hf may be caiculated from:

4.72Ql.852 L hf = ~ 1 . 8 5 ~ ~ 4 - 8 7

Tests hsve shown that the vdue of the Hazen-Wiiliams roughness coeficient C is dependen1 not oniy on the surface roughness of the pipe interior but also on the diameter of

22 STEEL PIPE

the pipe. Flow measurements show that for pipe with smooth interior linings in good condition, the average vaiw of C may be appmximated by the formula:

C = 140 + 0.17d (3-3)

However, in consideration of long-term lining deterioration, dime buildup, etc,, a Iower design value is recommended, as follows:

A graphical solution of the Hazen-Wilhns formula for C= 150 is presented in Figure 3-1 for pipe sizes 6 in. through 144 in. The multiplying factors in Table 3-1 provide a convenient means of changing the flow capacities shown in Figure 3-1 to the flows for other vaiues of C.

The Scobey Formula The Scobey formula for steel pipe, used perhaps more commonly in irrigation work than in the waterworks industry, is:

or for determining head 105s:

The recommended K, vdue for new bare steef pipt or pipe with iiaings confocming to current AWWA smdmds is 0.36. A graphid solution to the Scobey formula for K,= 0.36 is shown in Figure 3-2. Mdtiplyiag factors for odier frictim d i c i e n w are given in Tabk 3-2.

The Manning Formula

.- .

HYDRAULICS OF PIPELINES 23

0.1 0.2 0.4 0.0 1 2 4 6 10 20 40 60 100 200 4 0 0 6 0 0 1000 2000 4000800010000 1 , . , , , i HYDRAULIC GRAOIENT PER 1000 FT, FT

Flgure 3-2 Solution of Scobey Fiow Formula for K, = 0.36 (See data ln Table 3-2 for other K, values.)

Table 3-2 Muitipiying Factors for Friction C&dent Vnlue- Base Ks= 0.36+

K, vaiw 0.32 0.34 0.36 0.38 0.40 -tive discharge 1.125 1,059 1.000 0.946 0.900

*Data for use with Figure 3-2.

Figure 3-2 Solution of Scobey Flow Formula for K, = 0.36 (See data In Table 3-2 for other K, vrilues.)

T&le 3-2 Multiplying Factors b r Frictlon Coealdent Vaiue- Bace KS = 0.36*

K, value 0.32 0.34 0.36 0,W Rektive discharge 1.125 1.059 1.000 0.946

*Data for use with Figure 3-2.

n vaiue 0.009 0.010 0.011 0.012 0.013 Relative discharge 1.222 1.100 1.000 0.917 0.846

*Data for use with Figure 3-3.

26 STEHL PIPE

Computations for Fiow Through Pipe . - . . . . - .

- .

The quantity of water that will pass through any given pipe depends on the head (pressure) producing the flow, the diameter and length of the pipe, the condition of the pipe interior (smooth or rough), the number and abruptness of bends or elbows, and the presence of tees, brmches, valves, and orher accessories in the line.

- The total head, or pressure, affecting flow may be divided into four parts: velocity head - loss, entrance head loss, loss of head through friction, and minor losses due to elbows,

fittings, and valves.

Velocl ty H ead Loss ( v '/zg) Velocity head loss is defmed as the height through which a body must faH in a vacuum to acquire the velocity at which the water flows in the pipe. This loss is u s d y considered to be unrecoverable at the outlet. Numerical values are given in Table 3-4.

. .

i . . . Entrance Head Los , . , L. . . . . . . Entrance head loss is the head required to overcome the resismce at the ennance io the pipe; it is usually less than the veiwity head. When the conditions are not specified, it is ordinarily considered equal to one-half the velacity head, on the assumption of a sharp-edge entrance. Safe values for the ordinary entrance head loss may be obtained from Table 3-4 by

5 taking half the velocity head corresponding to the velocity in the pipeline. Head losses for -E

other rhan sharpsdge entrances may be found in treatises on hydraulics. . -

d.

" Loss of Head Through Friction g-:>L Friction head loss may be detemned by one of the formvlas duit have been discussed previously . (Data are given in this chapter to aid in solving the formulas.)

- . &+i: - Minor Losses Dueto Ubows, Fittings, andvalves

.. .. . . 2.. - . . . -> .

. - In long lines, minor head losses due to bends and fittings are occasionaiiy ignored. In any ,. . -

. , - given line, however, it is best to consider di losses so that no important factors will be . .

overlooked. The minor losses should always be recognized when evaluating flow tests. Total

1

Table 3-4 Theoretid Head Corresponding to Wven ~elocity -v '/tg . -

Vekooty Head Veiocity fps ft fP5

1 0.02 2 0.06 3 0.14 4 0.25 22

5 0.39 24 6 0.56 26 7 0.76 23 8 1 .O

9 1.3 32 10 1.6 34

14

Somce: hmard, R.E. Dcsign Standards for Steel Water Pipe. Joirr. A WWA, 40: 124 (Jan. 1948).

m

HYDRAULICS OF PIPELINHS 27

head loss in long lines with low velwities, the s u m of ve1wity head loss and entrance head . . - - --

*- loss, may be relatively insignifiant; in short lines with high velocities, this sum becomes very important. Ordinary tables and charts showing flow of water in pipe usually give only the fiiction head loss in straight pipe. In long lines, this is the largest loss.

In tht fid correct solution to a flow problem, the sum of ali losses must eqmi the available head, or pressure, prducing the flow. The foregoing formulas determine H or V, and the volume of fiow Q is found from:

The information contained in Tables 3-5 through 3-9 wiIl be useful when making hydraulic calculations.

Flow Through Fittings-Ecluivalent-Length Method Experiments have shown that the head lms in bends, fitrings, and valves is related to flow velocity and pipe diameter in a manner somewhat similar to that in straight pipe.

Drop per 1000 of Pipe F

LRngthofEpe 1-ft Drop

S for Sacl Water Pipe. Jm. A W WA, 40: 1 :24 (Jan. 1948).

STEEL PIPE

Tabk 3-6 Flow Equivalents

16 11 111 24.77 62 - 43 056 95.98 . 17 *-- 11 806 -26.31 M ' 44 444 " 99.08

< _ . - A

&.-F. - ', 18 12 500 27.86 66 45 833 102.17

- 19 13 194 29.41 68 47 222 105.27 E-:- ,. 3 . a 1 3 889 30.96 70 48 611 108.37 :

HYDRAULICS OF PIPELINES 29

Table 3-7 Pressure (pd) for H&s (a) -. Additional Heads

O +1 +2 +3 +4 +5 +6 +7 +8 +9 Head Bressure ft PS

O - 0.43 0.87 1.30 1.73 2.16 2.60 3.03 3.46 3.90 10 4.33 4.76 5-20 5.63 6.06 6.49 6.93 7.36 7.79 8.23 20 8.66 9.09 9.53 9.96 10.39 10.82 11.26 11.69 12.12 12.56 30 12.99 13.42 13.86 14.29 14.72 15.15 15.59 16.02 18.45 16.89 40 17.32 17.75 18.19 18.62 19.05 19.48 19.92 20.35 20.78 21.22 50 21.65 22.08 22.52 22.95 23.38 23.81 24.25 24.68 25.11 25.55 60 25.98 26.41 26.85 27.28 27.71 28.14 23.58 29.01 29.44 29.88 70 N31 30.74 31.18 31.61 32.04 32.47 32.91 33.34 33.77 34.21 80 34.64 35.07 35.51 35.94 36.37 36.80 37.24 37.67 38.10 38.54 90 38.97 39.40 39.84 40.27 40.70 41.13 41.57 42.00 42.43 42.87

4: C Source: W r d , R.E. Dcsign Standards for S ~ e l Water Pipe. Jour. A WWA, 40: 124 (Jan. 1948). .-.

,'. , ..

Tabk 3-8 Head (ft) b r Pressures (psi) Additional Htads

- 2.3 4.6 6.9 9.2 11.5 13.9 16.2 18.5 20.8 23.1 25.4 27.7 30.0 32.3 34.6 3.9 39.3 41.6 43.9 46.2 48.5 50.8 53.1 55.4 57.7 60.0 62.4 64.7 67.0 69.3 71.6 73.9 76.2 78.5 80.8 83.1 85.4 87.8 90.1 92.4 94.7 97.0 99.3 101.6 103.9 106.2 108.5 110.8 113.2

M e w r ~ Water in. k. Pi

13.6 0.49 27.2 0.98 40.8 1.47

Mercury Water in. fn. N 13 176.8 6.38 14 190.4 6.87 15 204.0 7.36 16 217.6 7.85

: Bamard, R.E. Design Standards for S t d Water Pipe. Jour. A WWA, 40:1:24 (Jan. 1948}.

30 STEEL PIPE

SHEAR GATE

Source: John F. Lenard. President. Lenard Engineerrng, Inc.

Figure 3-4 Resistance Coefficients of Valves and Fittings for Fluid Flows , . - . .

. i . . . t '

VALUES OF Vd FOR WATER AT 60"

v."- -\

108 2 3 4 5 6 0 1 0 ' 2 3 4 5 6 010' 2 3 4 5 6 810 ' 2 3 4 . 5 6 8 1 0 ' ~ - * 2 . 3 3 - 5 . 6 , B l O b REYNOLDS NUMBER (RI O.%

Soutce: Pipe Friction Manual. Hydraulic Institute, New Vork (1 954). Figure 3-5 Momly Diagram for Friction in Plpe

32 SSTI3EL PEPE

velocity, pipe diameter, and fluid viscosity. Values for f have been developed by ~ o o d ~ ~ and others. With a known f and L/D, the Darcy-Weisbach formula can be expressed as:

.Y In this equation, K is the resistance coefficient. Figure 3-4 shows values for K based on a summary of experimental data.

Examples to determine head loss HL for fittings and vdves and equivalent pipe Iengths using Figure 3-4 are as follows:

Pipe = 6 in. C = 100 Flow = 450 gpm V = 5.12 fps

CalcuIaitaon~: . , -

- 5.. , ,- -

.., ..,. - . Velwiry head: ..

.

HYDRAULICS OF PIPELINES

Aqueducts Economic studies of large aqueducts are frequently cornplicatcd by the desirability of combining different means of carrying water-for example, through open conduits, pipe, and tunnels-in the same system. Hinds4> demonstrated the use of grapbical rneans in making such studies in the design of the Colorado River Aqueduct. The methedoffinding economical slopes elaborated by Hinds had been csd previously in &e desi of rhe Owens River Aqueduct of Los ~ n ~ e l e s ~ and the Catskill Aqueduct of New York. P Penstocks An economic study to determine penstock size generally requires that the annual penstock cost plus the value of power lost in friction be minimai. The annual pwstock cost includes amortization of al1 rekited construction costs, operation and maintenance costs, and replacement reserve. A precise analyticai evduation, m b g al1 facuirs imo account, may be neither justified nor practical, since ali variables entering into the problem are subject to varying degrees of uncertainty .

Figure 3-6, which is used to determine the economic diameter for sreel penstacks and pump lines, was derived from the method presented by Voetsch m d res en.^

3.4 DISTRIBUTION SYSTEMS Methods of determining economical sizes of pipe for disrribution systems have been published.9

5 AIR ENTRNNMENT AND RELEASE Air entrained in flowing water tends to form bubbles at or near the summits in a pipcline. If not removed, such bubblcs become serious obstades to flow. The formation of a hydraulic jump in a pipe at the end of these bubbles is an importan1 reason to remove rhe air. Possible air entrainment and its removai must be considered and remedies applisd if needed. The ability of the hydraulic jump to enuain the air and to carry it w a y by the flowing water has been investigated. Quantitative data have been published10 rehting b c t e r i s t i c s of the jump to the rate of air removal. Removal of air through air valves is discussed in Chapter 9.

COOD PRACTICE Waterworks engineers should use hydraulic-friction formulas with which they are most familiar and with which they have had experience. Three of the common conventional formulas have been discussed in this chapter. In any particular case, the results c a l h t e d using the different conventional formulois can be compared. Engineers should, howwer, recogniae the increasing use of the rational or universal formulas, become familiar with them, and make check calculations using them. A practica1 d ~ c i w r value for the formulas should ix conservatively selected.

The results of flow tests will generaiiy be more usefui if related to the rational concept of : . - fluid flow. This entails more attention to relative surface roughness, water temperature,

nolds numbers, and an amlysis of test results aimed at fitting them into the frame of the ' fluid-mechanics a p p d to flow determination.

Definition of Syrnbols Hydrauiic symbols:

A = asea of pipe (sq ft) C = Hazen-Williams coefficient D = diameter of pipe (ft) d = diameter of pipe (in.)

; . f = Darcy fiiction factor

34 STEEL PIPE

a : Coat o f pipa p i r lb., i n s t i l l i d , dollors. n = B * Oiometir muitiplisr from Groph 8. b - olus d !ast powtr in dollar, pw k wh. D : Economic d i o m i t i r ir f i i t .

Ks ' n =

o : Ovarotl plant eiieraney. 9 = e) : Joint lffieieney. r : I : Lpss factor trom G w h A. p =

t, =

Wsightld ovimpi hiod imludi iq w o t i r hammar. (bid m dmign haad)

Frictim c ~ l f f i e i o n t in Seabiyh formulo (R34). Ratio oi owrwmight b w i g M o ptpi s h l l . Flow in tubie f& ver sacmd. (ai M p n h M d at turbina 1 Rotio of onnuol c o s t tobm( $me ixolanot ionl Al lowibl i t inrion, p r l Wrightid w t r i g i p l# i thicknisr( oi dnipn h iad 1 fo r to ta l Iingth. Avsraga plat l thicknns for frnpth Lo.

E X P L A N A T I O N ANO L X A Y P L E Eaomplo for plnstotk Q = le6 CFS

L ,= 2 6 L*. 1% ~f&, L*:loo'

625

r: ~ o ~ + o . ~ ~ + ~ o ~ I ~ : o . ~ T ~ s a a r = %#a 0.0119

+, L,t,*C,t,*L,t,* ... tkk K: sit%qb: ~x465iaSlor&twioxasoxaopg, ,, L , + L I * L I * . . + L n prt I+n) R27 a D . N 7 0 i 1.15

0 : 1.281 Ifron m 61; 6. 3 7 ( h BUpk C) i EeoooiiTc &O. I.ZBS ~ 3 . 7 ~ 4 7 5 I m i 4'IOm0'dip.l

n: L , + C 2 + L i * . +Ln & M E : Cdeuldad 0- d i gpdd k v * r ~ cbsi tp msrmd di- m it 8 r f w t k m this i~ampl i . Thi w b l n i shosiid k nworkrd mt i l tkir mi*

m ! o s t y " p ~ ~ ~ P t o f ~ ~ Oipriciatim ir k r t d on t h i mmiiptiwi ot m a m l ihhing f d iarninp 3* int i r is t roquired b npluce 50% o thi pipi in 45 piri . Thi onnwl 0.8 M : Cwt o t mpiniuininp inhriar and mterior surfoet wymin t nqu i r td 13 q u a i to 0.M5135 timar t h i t i r i t cort.

mi o f E r I w r i i r i o w o r w f o r ~ p ~ r , nridt w d whidt s&w ama kr i t p o n d ppi I por p a r .

Dapriciotim : Si0 Redamtion Ymuil, MI. m m r , poqi 2,4,llD. t:liihrest+[kpracioti~n+ % O B Y

Adaptd from Steel Penstocks and Tunnel Liners. Steel Plate Engineering Data Vol. 4. American lmn and Steel lnstitute cooperation with Steel Plate Fabricators Assoc., lnc., 1982. Courtesy of AISI.

Figure 3-6 Economic D i e t e r for Steel Penstocks and Pump Unes

HYDRAULICS OF PIPELINES 35

g = accelerationof gravity(32.2fps/s)hf = headloss(ft) in pipelengthL (ft)H = headloss (ft) in 1000ft of pipeKs = ScobeyconstantL = length of pipe (ft)n = ManningcoefficientQ = discharge(cfs)r = hydraulicradiusof pipe(ft)

s = T =1O~0=slopeofhydraulicgradientV = meanvelocity (fps).

Other symbols:'

b = valueof power($/hp/yr)Qa = averagedischarge(cfs)S = allowableunit stressin steel(psi)t = pipethickness(in.)a = costofsteel($/lb)i = yearlyfixedchargesonpipeline,expressedasa ratio

Ha = averageheadonpenstockincludingwaterhammer(ft).

References

Thefollowingreferencesarenotcitedin thetext.

- ALDRICH,E.H. SolutionofTransmissionProblemsof a Water System.Trans.ASCE, 103:1579(1938).

- BARNARD,R.E. Design StandardsforSteelWaterPipe.Jour. AWWA, 40:1:24(Jan.1948).

- BRADLEY,J.N. &THOMPSON,L.R. Fric-tion Factorsof LargeConduitsFlowingFull. EngineeringMonograph7, USBureauofReclamation(1951).

- CAPEN,C.H. Trendsin CoefficientsofLarge PressurePipes.Jour. A WWA,33:1:1(Jan.1941).

- CATES,W.H.DesignStandardsforLarge-DiameterSteelWaterPipe.Jour.AWWA,42:9:860(Sept.1950).

- CROSS,HARDY.Analysisof FlowinNet-worksof Conduitsof Conductors.Bull.286.Engrg.Expt.Stn.,Univ.of Illinois,Urbana,Ill. (Nov.1936).

- DAVIS, C.V., ed. Handbookof AppliedHydraulics.McGraw-HillBookCo.,NewYork(2nded.,1952).

- FARNSWORTH,GEORGE,JR. &ROSSANO,AUGUST,JR. Applicationof theHardyCross Method to DistributionSystemProblems.Jour. A WWA, 33:2:224(Feb.1941).

- HINDS,JULIAN.ComparisonofFormulasfor PipeFlow.Jour. AWWA, 38:11:1226(Nov. 1946).

- KING,H.W.HandbookofHydraulics.Mc-Graw-HillBookCo.,NewYork(4thed.,1954).

- MOODY,L.F. FrictionFactorsfor PipeFlow. Trans.ASME, 66:671(1944).

- PIGOTT,R.J.S. PressureLossesin Tub-ing, Pipe, and Fittings.Trans.ASME,72:679(1950).

- PipeFriccionManual.HydraulicInstitute,NewYork (1954).

- PipelineDesignfor Waterand Waste-water.Reportof theTaskCommitteeonEngineeringPracticein the DesignofPipelines.ASCE, NewYork(1975).

- ReportofCommitteeonPipelineFrictionCoefficientsandEffectof AgeThereon.Jour. NEWWA, 49:235(1935).

I

~

1. CROCKER,SABIN,ed.PipingHandbook.McGraw-Hill BookCo.,NewYork (4thed.,1945).

2. FlowofFluidsThroughValves,Fittings,andPipe.Tech. Paper409,CraneCo.,Chicago(1942).

3. MOODY,L.F. FrictionFactorsfor PipeFlow. AmericanSocietyof MechanicalEngineers,NewYork.

4. HINDS,JULIAN. EconomicWaterCon-duit Size. EngineeringNewsRecord,118:113(1937).

5. --- EconomicSizesofPressureCon-duits.EngineeringNewsRecord,118:443(1937).

6. BABBITT,H.E. & DOLAND,J.J. WaterSupplyEngineering.McGraw-Hill BookCo.,NewYork (1927;1955).

7. WHITE,LAZARUS.CatskillWaterSupplyofNewYork,N. Y. JohnWileyandSons,NewYork (1913).

8. VOETSCH,CHARLES& FRESEN,M.H.EconomicDiameterof SteelPenstocks.Trans.ASCE, 103:89(1938).

9. LISCHER,V.c. Determinationof Econ-omicalPipe Diametersin DistributionSystems.Jour. A WWA, 40:8:849(Aug.1948).

10. HALL, L.S.; KALINSKE, A.A.; & ROBERT-

SON,J.M. EntrainmentofAir inFlowingWater-A Symposium.Trans.ASCE,108:1393(1943).

Chapter 4

AWWA MANUAL M 1 an

Determination of Pipe Wall Thickness

- . -

, The wal thickness of steeI pipe is dected by a number of factors that will be discussed in this and succeeding chapters, including the following:

1. Interd pressure a. Maximum &sign pressure (Chapter 4) b. Surge or warer-hammer pressure (Chapter 5) .'

. .

2. E x t e d pressure ,.. , a. Trench Ioading pressure (Chapter 6) b. Fwth-fili pressure (Chapter 6 ) c. Uniform oollapse pressure, atmospheric or hydrauIic (Chapter 4) d. Vacuum underground (Chapter 6)

'." . 3. Special physical loading . . a. Pipe on saddle supports (Chapter 7)

b. Pipe on ring-girder supports (Chapter 7) 4. Fhcticoil requirements (Chapter 4) . -

The thickness selected should be that whic'h satisfies the mom swere requiremei

When desi- for internai pressure, the minimum thi- of a cyhder shouId be selected to h i t the circumferential tension stress to a c e h kvel. This stress is frequwtly termed hoop stress. The internal pressure used in design should be that to which the pipe may be subjmed during its lifetime. In a transmission pipeline, the pressure is m the dismce between the pipe centerline md the hydroiulic gmde line. If there valves, the h u m pressure on the pipe between them wiil be measured by the dis

i .'

PIPE WALL THICKNBSS 37

between the pipe centerline and the elevation of the static leve1 with the valva closed. Surge or water-hammer pressures must also be considered. These are discussed in Chapter 5. In a

... pump-discharge pipeline, rhe internai pressure is measured by the distance between the pipe and the hydradic grade line created by the pumping operation. Ressure at the outlet and the loss due to friction enter into this determination. If it is possibie to impose a pressure equal to the shutoff head of the pumps, the pressure is measured between the pipe and the shutoff grade line. Figures 4- 1 and 4-2 show typical pipeline and hydraulic grade profiles for gravity and pumped flow.

With pressure determined, the wail thickness is found using Eq 4-1:

Where: . .

+, t = minimum specifed wall thickness (in.) -:. p = pressure (psi) d = outside diameter of pipe (in.) steel cylinder (not including coatings) s = allowable stress (psi). -- . m

STATIC TEST HGL ) - - m - - - - - - - 4 STATIC TEST HGL 1 - - m - - - - - - - 4

STATIC HGL

- .&&i

figure 4-2 Relation of Various Heads or Pressures for Selection of Design Pressure (Pumped Flow)

.. .

WORKING TENSION STRESS IN STEEL Tension Stress ancl Yield Strength Modern steel technology has dowed increases in the allowable working stress for steel, with

. xhis working stress determined with relation to the steel's yield strength rather than its * ultimate strength. A design stress equal to 50 percent of the specif~d minimum yield

strength is often accepted for steel water pipe. Design criteria for penstocks h v e been - adopted by the Bureau of ~eclamation' that base design stress on 93 the minimum tensile

strength or 2/3 the minimum yield strength, whichever is least. With the use of given methods of stress d y s i s and proper quality control mesures, hese ailowable design stresses are considered conservative for the usual water-transmission pipelines. Table 4-1 illustrates grades of steel used as a basis for working pressure and the design stress as

to minimum yield point and minimum ultimate tensile strength for common steel as referenced in AWWA C200, Standard for Steel Water Pipe 6 Inches and

-

38 STEEL PIPE

Table4-t Gradesof SteelUsedinAWWAC200asBasisforWorkingPressuresinTable4-2

Table 4-2 givesthedesignerworkingpressurescorrespondingto 50percentof thespecifiedminimumyieldstrengthfor severaltypesof steelcommonlyusedin waterworkspipelines.The designeris cautionedthatthediametersandwallthicknesseslistedin thetablearefor referenceonly and do not representengineeringor manufacturinglimits.Modern steel-millcapabilitiespermitthemanufactureof almostanydiameterandwallthicknessof pipe;in practice,however,mostpipemanufacturersfabricatepipetostandarddiametersandwallthicknesses.Pipewiththickliningssuchasthecement-mortarl~ningsspecifiedin AWWA C205,StandardforCement-MortarProtectiveLiningandCoatingforSteelWaterPipe-4 In. and Larger-Shop Applied,3andAWWA C602,StandardforCement-MortarLining of WaterPipelines-4 In. (100mm)and Larger-In Place,4isusuallyfabricatedtotheindividualmanufacturer'sstandarddiameterstoaccommodatetherequiredliningthicknesses.It is, therefore,recommendedthatthepipemanufacturersbeconsultedbeforefinalselectionofdiameterandwallthicknesses. .

PressureLimitsHigh qualityin themanufactureof boththepipeandthesteelusedin itsmanufactureisrequiredbyAWWA standards.Therefore,hoopstressmaybeallowedtorise,withinlimits,above50percentofyieldfortransientloads.Whenultimatetensilestrengthisconsidered,asafetyfactorwellovertwois realized.For steelpipeproducedtomeetAWWA standards,theincreasedhoopstressshouldbelimitedto75percentofthespecifiedyieldstrength,butshouldnotexceedthemill testpressure.

4.3 CORROSIONALLOWANCE

At onetimeitwasageneralpracticetoaddafixed,rule-of-thumbthicknesstothepipewallasacorrosionallowance.This provedtobeanirrationalsolutionin thewaterworksfield,wherestandardsfor coatingandliningmaterialsandproceduresexist.It is preferabletodesignfor therequiredwall-thicknesspipeasdeterminedbytheloadsimposed,thenselect

DesignStress MinimumUltimateSpecificationsfor 50%ofYieldPoint MinimumYieldPoint TensileStrengthFabricatedPipe psi psi psi

ASTM A36 .18000 36000 58000ASTMA283GR C 15000 30000 55000

GRD 16500 33000 60000ASTM A570GR 30 15000 30000 49000

GR33 16500 33000 52000GR36 18000 36000 53000GR40 20000 40000 55000GR45 22500 45000 60000GR50 25000 50000 65000

ASTM A572GR 42 21000 42000 60000GR50 25000 50000 65000GR60 30000 60000 75000

DesignStress MinimumUltimateSpecificationsfor 50%ofYieldPoint MinimumYieldPoint TensileStrength

ManufacturedPipe psi psi psi

ASTM A53,A135,andA139 GRA 15000 30000 48000

GRB 17500 35000 60000ASTM A139 GRC 21000 42000 60000

GRD 23000 46000 60000GRE 26000 52000 66000

PIPE WALL THICKMESS 39

linings, coatings, and cathodic protection as necessary to provide the required leve1 of corrosion protection. . ..

4.4 EXTERNALFLUIDPRESSURE-UNIFORMANDRADIAL The proper wall thickness must be selected to resist external loading imposed on the pipe. Such loading may take the form of outside pressure, either atmospheric or hydrostatic, both of which are uniform and act radially as collapsing forces. Buried pipe must be designed to resist earth pressure in trench or fiil condition. These considerations are discussed in Chapter 6.

Atmosphere or Fluid Environrnents A general theory of collapse-resistance of steel pipe to uniform, radially acting forces has been d e ~ e l o ~ e d . ~ Any unreinfoxced tube longer than the nitical length can be considered a tube of infinia length, as its collapsing pressure is independent of further increase in length. The foliowing formula applies to such tubes:

- . : :: . .

. - .. _ .'*A.

_ .,I

Where: .. .-- -. - .. . - .5 : -. , - ; iL.

STEEL PIPE

Equation 4-4 is predicated on the pipe being mmmerciaiiy round, made of steel with a minimum yield of at least 27 000 psi, and having a length six diameters or more between reinforcing elementc. -

- . t . . .

. r - - - . .. . -. . -

4.5 MlNIMUM WALL THICKNESS Minimum plate or sheet thidmesses for handling are based w p o formulas adopted by rnany specifying agencies. They are:'

D t = - 288 (pipe sizes up to 54 in. ID)

.- .

t = + (pipe sizes greater than 54 in. ID) 400

In no case s W the sheli thickness be less than 14 gauge (0.ll747 in.). Ir should be mted that for pipe diameters smaller than 54 in., the use of Eq 4-5 (Pacifc

Gas and Electric formula) will result in a rhinner pipe wdl than the use of E q 4-6 (United States Bureau of Rechmation formula). For 54411. and larger pipe, the opposite is t r u ~

4.6 WOD PRACTICE ' > X ! Interna1 pressure, external pressure, special physical loading, type of Iinq and coating, and other practicd requirements govern wall thickness. Good practice with regard to internal pressure is to use a working tensile stress of 50 peraent of the yield-point stress under the influence of maximum design pressure. The stress of transitory surge pressures, together with static pressure, may be taken at 75 percent of the yield-point stress. The daigner should, however, never overlook the effect of water hammer or surge pressures in design. It is more positive and economical to select a proven coating or linhg for protedon against corrosion hazards rhan to add sacrificial wall thickness.

I r ' --a.

' 1 ' = --

C' -S

ir'. "

PIPE WALL TIfICKNHSS 41

Table 4-2 Working Pressures For Allowable Unit Stresses*

Stress pa' m WdI Wcight FipeAxis Sccaon m T'MuEB$ per Foot DJt h 4 -dUS 15000 16500 17500 18000 21000

Table 4-2 WorWng Pressures for Allowabie Unit Stresses* (continued) y . , .

42 STEEL PIPE

Stress psi wall waght Pipe& 15000 16500 17500 18000 21000 Diameterf Thicknesst per Fmt t in.' Modulus

m. in. M R a t i o (S) Worlung rssure prix 18 OD .O747 14.30 240.96 168.96 18.77 125 137 145 149 1 74

-10% 19.99 172.08 235.41 26.16 174 192 203 a)9 244 -1345 25.67 133.83 301.m 33.47 224 247 262 B9 314 .1563 29.79 115.16 348.74 38.75 261 287 304 313 365 .1793 34.13 100.39 398.53 44.28 299 329 349 359 418 .2188 41-56 82.27 483.12 53.68 365 401 425 438 51 1 .2500 47.40 72.00 549.14 61,02 417 458 486 500 583

30 OD .1M6 33.40 286.81 1 097.51 73.17 105 115 122 126 1% -1345 42.91 223.05 1401.02 93-80 135 148 157 161 188 .1%3 49.82 191.94 1 631.50 108.77 156 172 182 188 219

*Vd= have km ampted by e k a m k -puta. See agt for formulm u d . ~S~undu45in.~outsi&~sizcs,~45in.andoverartinsidediamoersk. $ M m u f a r c r s can fuma wall thiche3scs 0 t h thm shown. $WorlMg press- may be interphed or exn-aplated for other wall thic$Qesses or messe.

PIPE WALL THICKNESS 43

Table 4-2 Working Pressures for Allowable Un& Stress* (cmtinued) - - -

. - - . . . .

m WaIl Weight PipeAxis Scction DiameterT Tbicknessf m FOO~ Do/t h? M d d u s 15000 16500 17500 18000 21000

k OD .1793 57.11 167.32 1 867.28 124.49 179 197 209 215 251 .21# 69.60 137.11 2 269.64 151.31 219 24 1 255 263 306 .2500 79.44 120.00 2 585.18 172.35 250 27s 292 300 350

3 - 99.10 96.00 3Sll.28 214.09 365 /' 375 438 . --4375- 138.15 68.57 4439.73 295.98 438 48 1 5 10 525 613

.5000 157.55 60.00 5 042.20 336.15 500 550 583 600 700 32 OD .lo46 35.64 305.93 1 332.85 83.30 98 108 114 118 137

45.78 237.92 1709.04 106.81 1% 139 147 151 177 53.16 204.73 1 981.98 123.87 147 161 171 176 205 60.94 178.47 2 268.73 141.80 168 185 1% 202 235 74.28 146.25 2 758.28 172.39 205 . 226 239 246 2J37 84.78 128.00 3 142.37 196.40 234 258 273 281 328

3125 105.77 102.40 3 904.95 244.06 293 322 342 352 410 .4375 147.50 73.14 5 403.00 337.69 410 45 1 479 492 574 .5#0 168.23 64.00 6 138.62 383.66 469 516 547 563 656

37.87 325.05 1 599.62 94.10 92 102 108 111 129 48.65 252.79 2051.45 120.67 119 131 138 142 166 56.50 217.53 2 379.37 139.96 138 1 52 161 165 193 64.77 189.63 2 723.95 160.23 158 174 185 190 221 78.95 155.39 3 312.46 194.85 193 212 225 232 270 90.12 136.00 3774.37 222.02 221 243 257 265 309

.3125 112.45 108.80 4 691.95 276.00 276 303 322 33 1 386

F

sures may be interpokd or exPapolared for 0 t h wd thicknesses or strcsses.

44 STEEL PIPE

Table 4-2 WorWng Pressures for Ailowable Unlt S-* (continud)

Moment of Inertia

About Stress gs Pipe wau WeigIlf PipeAxis Scction

Diamc~rt ThicknessI rier Foot Dn/t inP Mdulus 15000 165MJ 17500 18000 21000

57 ID .2500 152.88 230.00 18421.89 640.76 132 145 154 158 184 .3125 191.31 184.40 23 103.11 801.84 164 181 192 197 230 .3750 229.82 154.00 27 814.90 963.29 197 217 230 237 276 ,4375 268.41 132.29 32 557.37 1 125.09 230 253 269 276 322 .5000 307.09 116.00 37 330.a 1 287.26 263 289 307 316 368

Values h v e been computed by ekcwinic computer. See text for formulas used. TSizes under 45 in. are outside diametcr sizes; those 45 in. and over are inside diameter sizes. SManufmurm furnish w d thiicimesses otkr than shown. SWorking pressures may be kiterpolated or extrapolated for other wall thiclumses ox smsscs.

PIPE WALL THICKNESS 45

Table 4-2 Working Prwsures for Ailowabk Unit Stressec* (continued) . ' . - - .. . . . . .

stress psi pipe Wd weight PipeAxis Section

Diamttert Thichmd ncr Foot Dn/t in? 15000 16500 17500 18000 21000

m -

46 STEEL PIPE

84 ID -3125 281.43 270.80 73551.47 1738.29 112 123 130 134 156 .3750 337.97 22.00 88 458.73 2 087.52 134 147 156 161 188 .4375 394.59 194.00 103 432.09 2 43728 156 172 182 188 219

*Val= have k n computed by ekcamic oomputer. See text for formulas u d . unda 45 in. art outside dhmm sizes; those 45 in. d ovcr m inside diameter si=.

$Manukhms furnish d r ' ' r other tfian s h . $Workiq pmsures m y be interplated or extrapokd for other wall thiclmesses or suesses.

BIPE WALL THICKNESS 47

Table 4-2 Working Pressures for Ailowable Unlt S&-* (contfnued)

stress psi RF Waii Weight Pipe A K i s Section

Dimead Thicknessf per Foot Da/t ,+ hladdUS 15000 16500 17500 18000 21000 iw. iR. (bare) h t i o ( I ) (S) Working Pnssuit psi3

STEEL PIPE i $ i J

< .

< Table 4-2 Working Pressures For Aliowable Unit Stresses* (continued) , _ . .. -, . , .. i

Moment of Irleda

About Stress pss' P i ~ e W d Weight Pipe Axis Section

Diameaq ThicknessS per Foot D d t i n ~ lModulUa 15000 16500 175M) 18000 21000 m. zk. (bare) &ti0 (1) w o k h g Pressure psi$

102 ID A875 754.08 150.36 292 350.87 5 656.12 202 222 236 243 .7500 823.14 138.00 319513.64 6174.18 221 243 257 265 - 309 .a125 892.27 127.54 346 774.99 6 92.88 239 263 279 287 -8