Earth Loads on Steel Pipe

CHAPTER 8: Earth Loads on Steel Pipe
Source: Journal (American Water Works Association), Vol. 53, No. 8 (AUGUST 1961), pp.
1045-1080
Published by: Wiley
Stable URL: https://www.jstor.org/stable/41257176
Accessed: 12-06-2020 06:32 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Wiley is collaborating with JSTOR to digitize, preserve and extend access to Journal
(American Water Works Association)
This content downloaded from 213.55.102.49 on Fri, 12 Jun 2020 06:32:43 UTC
All use subject to https://about.jstor.org/terms
CHAPTER 8
Earth Loads on Steel Pipe

pipe, both with thin wall ning with 8.9, was derived from obser-
and with thick wall, has been vations on steel water pipe. The
used underground for nearly a century theory is basic, however, and has been
in the case of smooth-wall pipe andfound to apply equally well to corru-
half a century in the case of corrugatedgated culvert pipe. Therefore, ex-
pipe. These kinds of pipe have been amples of both types are relevant to
subjected to earth loads and live loads show validity through installation ex-
of many kinds and of high intensity perience. All observations plotted in
without any general failure by col-Fig. 8.6 pertain to steel water pipe
lapse or undesirable change in shape. installations, most of the lines being
Therefore, the load resistance of pipe Dresser coupled. Vertical deflection
in the commonly used diameters andaverages about 2 per cent, which is
wall thicknesses has been generally normal for such lines, and is confirmed
adequate and the problem of analysis by the sidewise settlement theory.
is not a vital one, except that deflec-
tion of steel water pipe must be re- 8.1. Types of Conduits
stricted to permissible limits for the
linings and coatings used and for When grouped according to degree
water tightness of the joint where of rigidity, three classes of pipe are
mechanical couplings are used. generally recognized :
Real collapse failure of steel pipe a. Rigid conduits, whose cross-sec-
does not occur under earth loads until tional shapes cannot be distorted
a condition is reached where the sufficiently to change their vertical
vertical diameter has been decreased or horizontal dimensions more than
about 20 per cent of the nominal 0.1 per cent without causing material
diameter and the horizontal diameter damage.
has been increased a similar amount. b. Semirigid conduits, whose cross-
Many thousands of corrugated steel sectional shapes can be distorted
drainage structures show that they sufficiently to change their vertical or
are virtually uncrushable under any horizontal dimensions more than 0.1
existing fill or moving-wheel load per cent but not more than 3.0 per
with or without impact when the cent, without causing material damage.
earth cover is at least equal in depth c. Flexible conduits, whose cross-
to the pipe diameter. The average sectional shapes can be distorted
vertical deflection of such structures,sufficiently to change their vertical or
even including those under railroads,horizontal dimensions more than 3.0
is about 2 or 3 per cent of the nominalper cent before causing material
diameter. damage.
The Barnard "sidewise" settlement Steel pipe, having either light or
theory, described in sections begin- heavy wall, because of its physical
1045
1046 committee report Jour. AW WA
characteristics can 8.3. Rigid Versus

always Flexible Pipe as
function
a flexible conduit, with due considera-
The differences in reaction to load
tion given for the relative rigidity of
exhibited by rigid and by flexible or
linings or coatings of water pipe.
The working data in this report semiflexible pipe affect design. The
have special application to conduits methods of design which apply to
made of steel and not to conduits one do not apply to the other.
made of other materials. Field load tests made on buried
pipe show: [1] that the load de-
8.2. Types of Loads veloped on rigid pipe is greater than
that developed on flexible pipe; and
Earth loads on pipe are of two [2] that this greater load tends to
kinds. The one more common in concentrate at the top and bottom of
Fig. 8.1. Effect of Earth Load on Rigid and Flexible Pipe
The load on rigid pipe (demonstrated at the left) is concentrated at the top and bottom
of the pipe, creating bending moments that must be resisted by the shell. With flexible
pipe {demonstrated at the right), a lesser load is developed, because it is more evenly
distributed around the circumference.
water works practice is trench load- the rigid pipe, thus creating bending
ing, the pipe being laid in an excavated moments which must be resisted by
trench and backfilled. The second the shell. The tests show that the
kind occurs when a pipe is laid lesser
on a load developed on flexible pipe :
graded or prepared ground surface[1]
and
is more evenly distributed around
a fill of earth is placed around its
andcircumference, with less concentra-
over the pipe. As the load carried
tion at any point than is true for rigid
by the pipe may differ in the two
cases, trench loading should be pipe;
dis- and [2] may be carried as a
compressive force in the shell which,
tinguished from fill loading. Further,
depending
the reaction of flexible pipe to either on stiffness, frequently is
fill or trench load is different fromresisted
the more by arch action than by
reaction of rigid pipe to these loads.
moment strength. The effect of earth
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1047
load on rigid and flexible pipe is shownFigure 8.2 shows photographs of

in Fig. 8.1. sand models made to demonstrate
Although the maximum load-carry- what takes place when normal settle-
ing capacity of flexible pipe depends ment occurs in a fill containing pipe.
to some extent on the wall thickness In these models, the horizontal lines
and its section modulus, the pipe, bywere straight before settlement oc-
deflecting, is able to make full usecurred. Their displacement gives a
of the load-carrying ability of theclue to what causes the difference in
earth surrounding it. As the pipe loading on rigid and flexible pipe.
may change shape without failure, it In the left-hand portion of Fig. 8.2,
showing rigid pipe, S represents the
transfers part of the vertical load into
a horizontal or radial thrust which is amount of settlement of fill by consoli-
resisted J)y the passive pressure of the dation. The layers above the rigid
Fig. 8.2. Effect of Settlement on Rigid and Flexible Pipe in Fill

In the photograph at the left, pertaining to rigid pipe, S represents the amount of settle-
ment of fill by consolidation. In the photograph at the right, pertaining' to flexible pipe,
D represents the deflection of the pipe as it yields under load.
earth at its sides as these move out- pipe are higher than those at the sides.
The outside layers, by sinking farther
ward. If the wall itself were rigid,
this movement could not occur with- than the layers in the prism above the
out wall fracture. It follows that the pipe, create a drag at the sides of
rigid pipe must carry the whole loadthe prism above the pipe and add to
itself or break, whereas the flexible the prism weight the amount of the
pipe divides the load with the earth frictional forces in the two vertical
enclosing it. Therein lies the in- friction planes. The load on the rigid
herent difference between rigid and pipe may therefore be greater than the
flexible behavior and the explanation weight of the earth prism directly
of why the classical bending-moment above it.
formulas apply to the analysis of rigid In the right-hand portion of Fig.
pipe but not to the analysis of flexible 8.2, showing flexible pipe, D represents
pipe. the deflection of the pipe as it yields
under load. In thisConsequently,

case, in design,
afterit should settle-
be considered that a of
ment and consolidation rigid pipe under fill, the
the
layers above thea fill
pipeor embankment
are lower may carry a than
load greater
those at the sides. The than the prism weight
weight of the
directly
earth prism resting on overthe
it by the amount of
flexible pipe
frictional force
is partially supported by drag the
between material
prism
0Il
0 1 2 .3 4 5 6 7 8 9 10 11 12 13 14 15
H_
Bd
Fig. 8.3. Computation Diagram for Loads
Cd refers to the calculation coefficient; H, to the height (ft) of fill over the top of the conduit;
and Bd to the width (ft) of the ditch at the top of the pipe. The numbers on the curves
represent: [i] minimum for granular materials without cohesion; [2~' maximum for sand
and gravel; [J] maximum for saturated topsoil; [4~] ordinary maximum for clay; and
[5] maximum for saturated clay.
at the sides of the pipe. This partial and side earth; a flexible pipe may
support is measured by the amount of carry a load which is less than the same
the upward acting frictional force prism weight by an uplift amount de-
generated in the two friction planes pendent on a similar frictional force.
and is governed, too, by the relative With rigid pipe the frictional force is
settlement of fill and deflection of additive. With flexible pipe it is sub-
pipe. tractive. The same principle applies
to flexible pipe in a trench, as ex- If, however, the conduit is rela-

plained below. tively flexible, such as a thin-walled
steel pipe, it will deflect vertically
8.4. Determination of Load and, by so doing, transmit a portion
of the load to the trench walls by
The vertical load acting on a pipe
friction and to the trench bottom by
under a fill or embankment may be
calculated in accordance with Mar-
ston's formula applying to fill or
embankment conditions (i). Alter-
natively, for flexible pipe, fill load
may be taken as the weight of the
earth prism above the pipe. The
Marston formula for embankment or
fill will not be discussed here.
When installation is made in a
trench, the load may be calculated in
accordance with Marston's formula
for ' 'ditch" conduits (1). Or, for
flexible pipe, the load again may be
taken as the weight of the earth
prism above the pipe.
The Marston ditch conduit for-
mula is :
We = CdwBd*
in which :
Wc = load on the conduit (lb/

ft of pipe)
Cd = a calculation coefficient
Fig. 8.3)
w = unit weight of ditch fill ma-
terial (lb/cu ft) Fig. 8.4. Deflection of Steel Pipe
Bd = width of ditch at top of
pipe (ft). Pipe deflection transfers a portion of earth
load to the sides and bottom of the trench.
This formula gives the total vertical В d is the trench width at the top of the pipe;
load in the ditch or trench at the level В c and Di, diameter of installed pipe;
of the top of the pipe. Marston and A, the difference between В d and Bc ; d, pipe
deflection; D2, diameter of deflected pipe;
others have shown by experiment and
and r, radius of installed pipe.
observation of actual structures that
practically all of this load is carried
by the conduit itself when the conduit settlement. The right-hand photo-
graph in Fig. 8.2 illustrates the action
is rigid, as this type is very unyielding
in relation to the earth materials be- of friction in reducing trench ]'load if
tween the sides of the structure and the trench sides are considered to be
the sides of the trench. just outside the vertical lines. In Fig.
1050 COMMITTEE report Jour. AW WA
8.4, the trench fill is shown divided

in which В c is the width of the pipe
into three prisms. It is suggested and the other symbols are as indi-
that, as a minimum loading condi- cated for Eq 8.1.
tion, if the pipe deflection d equals It should be pointed out that, for
the settlement of side fill below the steel pipe, Eq 8.1 represents the
top of the pipe, the pipe and the maximum loading conditions and
side fills may each carry the sameEq 8.2 represents the probable mini-
amount of the load per unit of width. mum loading condition which may
develop on a pipe in a trench. The
7j
actual load in a given case may lie
somewhere between these limits, de-
pending upon the soil properties and
the relative rigidity of the conduit
,-v -
compared with the side fills. For
more rigid pipe, the actual load will
approach that given by Eq 8.1, while,
for quite flexible pipe, the actual load
will approach, or equal, that given
by Eq 8.2.
I A rough estimate of the percentage
X 2 settlement to be expected in the side
fills may be obtained from Fig. 8.24,
discussed in Sec. 8.19.
8.5. Wall Thickness to Resist Earth

Load
ll
15 20 30 40 50 60 70
The bending-moment, or flexural-
Diameter- in. stress, method of analysis does not
Fig. 8.5. Relationship Between Calculated
always apply to steel pipe under earth
Height of Fill and Diameter of loads. Almost all analyses of steel
16-Gage Corrugated Pipe conduits have been directed at the
structure itself and the load it carries.
The relationship was computed by use ofThis report presents two methods of
Eq. 8.3 for a deflection of 1 per cent of pipe
design, both of which use the elastic-
diameter; a constant e = 20; lag factor
= 1.5; К = 0.10; and soil weight = 120 ring theory for analyzing the pipe
lb/cuft (5). barrel ; both are based on designing
flexible steel pipe to obtain a required
vertical deflection, and both recog-
Under these conditions, the load on nize earth assistance at the side of the
the flexible pipe may be determined pipe. The first method is by Spangler
by multiplying Marston's load expres- (2, 3), who uses a single selective
sion (Eq 8.1) by the ratio of the width constant for the modulus of passive
of the pipe to the width of the trench, resistance of the earth. The sec-
giving : ond is by Barnard (4), who analyzes
the pipe barrel both for elastic-ring
We = CdwBd*(^j strength and for confined-arch strength ,
determines the outward thrust of the works authorities do not ap-

pipe wall, and then considers the pas- ply the lag factor to the
sive deformation of the soil at the deflection caused by live
side of the pipe under this load to be load.)
a "sidewise" settlement and employs
the principles of soil mechanics It tois to be noted in Eq 8.3 that W
determine its magnitude. In the is the total load on the conduit (in
Spangler formula, it is very difficult
pounds per linear inch). Therefore,
to determine for use the proper value
to obtain design loads W for flexible
of the modulus of passive resistancepipe in trenches using Eq 8.1 and
of the enveloping soil. The Barnard
method permits use of either typical
values of engineering properties of 100 |
90
soils or any specially determined 80 •
values in a given case.
8.6. Design Factors in Spongier

:s===i::
Formula .^,1^.
The formula for the design of
flexible pipe culverts evolved by . - ..^
Spangler is:
-° tsrf&J
in which :
„_^____
• • • N. ' '
Ax = horizontal deflection of flex-
ible pipe (in.)
К = bedding constant (0.10 for . ■ • ' '
flat-bottom trench with un- . '
tamped backfill or with back- io
fill tamped to horizontal di-

ameter or spring line of pipe)
W = vertical load per unit of pipe 7
length (lb/lin in. of pipe) e

r = radius of pipe (in.)
5
E = modulus of elasticity of pipe

metal (30,000,000 for steel) 4 1
/ = moment of inertia of cross 20 30 40 50 60 70 80 90 100

Diameter- in.
section of pipe wall (in.4/lin
in. of pipe) Fig. 8.6. Relationship Between Values of
e = modulus of passive resistance Spangler e and Pipe Diameter
of enveloping earth (psi/in.) The values of e, calculated from measured
Di = deflection lag factor (1.25- test data , decreased as the pipe diameter
1.50). (Note: Some water increased (7).
1052 COMMITTEE REPORT lom. AW WA
including both live 8.6.1.

andEarth modulus
dead, e.loads
The factor
(and considering eunits of formula
in the Spangler measure-
is intended
ment) : to measure the passive resistance of
the earth at the sides of the pipe on
w-Ur-k + w*) an elastic basis. It is supposedly de-
pendent on the type of soil and the
in which : void ratio, or degree of compaction,
in that soil. Spangler reports meas-
W = total load on flexible pipe
(lb/lin in. of pipe)
We = dead load on pipe from Eq
12 1
8.1 (lb/lin ft of pipe) 10
Wl = live load on pipe as calcu-

lated by Marston's (i) meth-
ods (lb/lin ft of pipe)
Bc = pipe diameter (ft)
В d = trench width at top of pipe
i :~-Ziz:::
(ft).
The Spangler formula may satisfy

engineers who desire a theoretical | /
approach. *It must be understood,
however, that the earth assistance I 3
properties of the soil surrounding a
buried pipe are not adequately meas- 2
ured by the single factor e, the modu-

lus of passive soil resistance, incor-
porated in the formula. Because of
the inconsistent results given by the
Spangler formula, it is suggested that
its range of useful application is to ll
20 30 40 50 60 70 80 90 100
pipe 60 in. and smaller under fill e
heights of 10 ft and less.

Fig. 8.7. Relationship Between Values of
The relation between the height of Spangler e and Height of Fill
fill and the diameter of a 16-gage
corrugated pipe, computed (5) using Measured deflections showed that, for the
same pipe, the value of e increases as the
Eq 8.3, is shown in Fig. 8.5. As the fill height increases (5, 6).
diameter increases, the calculated
fill height causing 1 per cent deflec-
tion decreases to a minimum and then ured values of e in embankments (2)
increases. Experience does not con- as given in Table 8.1. Note that
firm such behavior of pipe. When tamping the side fills practically
Eq 8.3 was used to determine the doubled the value. When calculat-
thickness of a pipe 156 in. in diametering the value of e corresponding to the
to deflect 2| in. under a 20-ft backfill then available measured deflections
cover with e = 30, the figure for re- for smooth steel water pipe of many
quired wall thickness was negative. different diameters, Barnard (7) found
e to decrease as the pipe diameter in- depth and where ordinary

of common
creased (Fig. 8.6).* backfilling methods have been used.
When the measured deflections in The method of backfilling very
a carefully conducted field test (6)
flexible pipe over 4 ft in diameter may
were employed (5), to calculate thecompletely offset the designer's calcu-
lations if care is not taken. Initially
value of e using Eq 8.3, for increment
increases in fill height on the same the pipe should have its full vertical
pipe, the resulting value of e varied diameter, at least. The backfilling
from less than 30 at 1 ft of cover to should be built up evenly on each side
about 70 at 12 ft of cover (Fig. 8.7).at the same time. Backfilling from
The modulus e has been found con- one side and then the other may be
sistent with measured deflection and disastrous. Ordinary tamping up to
the horizontal diameter, plus careful
with soil properties only in degrees of
soil compaction and loading seldom backfilling above that point, will in-
reached in pipe installations (4). sure installations wherein the vertical
Therefore, e must be treated as a deflection is normal.
selective constant whose actual value
in a given case is unknown. A value TABLE 8.1
of e = 30 appears to be an acceptable
Values of Spangler e Determined by Experimen
approximation for deflection calcula-
tions for pipe 60 in. and smaller Expt.
No. Kind of Soil psi/in.
e
under cover of 10 ft or less, for either

trench or fill conditions. 1 Black silty loam (not c
2 Well graded gravel (not
Barnard (<?), using the Spangler compacted)
formula with e = 30 and load as in
3 Yellow sandy clay loam (not
Eq 8.4, developed the deflection of compacted)
60-in. and smaller for steel pipe under
4 Yellow sandy clay loam (tamped,
external load in trenches, as shown in dry) 27
Table 8.2. These data seem to check
quite well with experience in trenches
8.6.2. Deflection,
* In 1957, after a study in similitude,ing,
7a Ax is the hor
Spangler modified the denominator of the his pipe, but, wit
original formula (Eq. 8.3) by a change inexperience has sh
nomenclature which substitutes e'rz for er4, in
considered to be the vertical deflection
which e' = er. He then believed that there is
as well.
a size effect not originally recognized and that
it is more nearly correct to consider the The amount of vertical deflection
modulus of soil reaction e' to be a constant. that will cause collapse of flexible
The actual value of e' for a specific instance is
pipe is about 20 per cent of the nom-
still unknown, but a tentative value of
inal diameter. Structurally speak-
e' = 700 is suggested by Spangler for side fill
soil which has been compacted to 90 per cent ing, therefore, deflections up to 10
or more Proctor density. The corresponding per cent of diameter are amply safe.
value of e for different pipe diameters when The spun coal-tar enamel meeting
er = 700 is shown as the dotted-dashed line in
AWWA standards will safely take a
Fig. 8.6. The comparison is clear between
these theoretical values of e and values calcu- 10 per cent deflection. Corrugated
lated by Barnard from measured data on culvert installations are commonly
steel
designed
water lines, as shown by the various dots. using 5 per cent deflection.
1054 COMMITTEE REPORT Jour. AW WA
TABLE 8.2
Deflection of Steel Pipe Under External Load in Trenches*
12 3 4 5 6 7 8
d я Deflection - in,
m Wall Thickness- in.

Nominal 2% m of , „ «л e ~
Size Diam. 5-ft , Cover „ «л 10-ft e Cover ~
in. in.
Fraction Decimal Tru°k Tr¿ck Tru°k Tr¿ck

6 0.12 i 0.1250 0.01 0.02 0.02 0.02
& 0.1875 0.00 0.01 0.01 0.01
8 0.16 I 0.1250 0.03 0.07 0.06 0.07
& 0.1875 0.01 0.02 0.02 0.02
10 0.20 i 0.1250 0.07 0.17 0.15 0. 18
А 0.1875 0.03 0.06 0.05 0.07
12 0.24 i 0.1250 0.15 0.32 0.27 0.33
А 0.1875 0.06 0.12 0.10 0.12
i 0.2500 0.03 0.06 0.05 0.06
I4 0.28 i 0.1250 0.21 0.48 0.42 0.52
& 0.1875 0.09 0.21 0.19 0.24
i 0.2500 0.04 0.10 0.09 0.11
16 0.32 i 0.1250 0.30 0:60 0.56 0.68
& 0.1875 0.15 0.31 0.28 0.35
i 0.2500 0.08 0.16 0.14 0.18
18 0.36 Ä 0.1875 0.21 0.41 0.39 0.47
i 0.2500 0.12 0.22 0.21 0.26
A 0.3125 0.07 0.13 0.13 0.15
20 0.40 A 0.1875 0.28 0.53 0.49 0.59
i 0.2500 0.17 0.31 0.29 0,35
& 0.3125 0.10 0.19 0.18 0.21
24 0.48 A 0.1875 0.41 0.75 0.71 0.83
i 0.2500 0.29 0.54 0.50 0.59
I 0.3750 0.14 0.24 0.23 0.27
30 0.60 i 0.2500 0.40 0.71 0.70 0.83
& 0.3125 0.31 0.54 0.54 0.64
| 0.3750 0.23 0.40 0.40 0.48
36 0.72 i 0.2500 0.48 0.83 0.87 1.02
& 0.3125 0.41 0.71 0.75 0.87
Î 0.3750 0.34 0.58 0.62 0.72
42 0.84 i 0.2500 0.53 0.98 0.95 1.14
} 0.3750 0.44 0.80 0.78 0.92
i 0.5000 0.32 0.57 0.57 0.68
48 0.96 A 0.3125 0.52 0.87 0.94 1.09
I 0.3750 0.48 0.81 0.87 1.01
& 0.4375 0.43 0.73 0.78 0.91
i 0.5000 0.'38 0.65 0.70 0.81
60 1.20 * 0.3750 0.57 0.98 1.05 1.22
i 0.5000 0.51 0.87 0.95 1.10
* Trench width equals outside diameter of pip

backfill.
From Barnard (8)
For steel water pipe, permitting 2 Design load, W. Experience

8.6.4.
per cent deflection is good practice,indicates that the calculated deflec-
although for large-diameter tion pipe,will be more nearly accurate if W
unlined, deflection has sometimes approaches the minimum value for
amounted to as much as 10 per smaller pipe given by Eq 8.2 and the
cent on well engineered lines. maximum value for larger pipe given
Permissible deflection for cement- by Eq 8.1. In this sense, large pipe
is 48-in. and over.
lined pipe is on the order of 2 per cent.
The effect on hydraulic efficiency8.6.5. Deflection lag factor, D/. It
of a pipe deflected by as much ashas been observed that corrugated-
5 per cent is entirely negligible as pipe
its culverts continue to deflect
area is 99.75 per cent of that of slowly
a for a period of time after instal-
Fig. 8.8. Fringe Pattern Around Ring on Slip-Fit and Over-Size Holes
In a first series of tests, the hole size was a slip-fit for the ring. The fringe pattern
(at the left) indicates a relatively uniform radial stress. The stress pattern in a second
series of tests, when the diameter of the hole was -£j in. larger than that of the ring, i
shown at the right (4) .
lation. The gradual increase in de-

perfect circle and its hydraulic radius
is also 99.75 per cent. When the flection is an action somewhat similar
deflection is 2 per cent, the corre- to foundation settlement and time
sponding figure is 99.95 per cent for consolidation of soils. The increase
both area and hydraulic radius. becomes slower and slower, however,
8.6.3. Bedding constant, K. For and after a period of years virtual
field conditions usually encounteredstability is attained. Factor Dt re-
in steel pipe installations, a good lates initial deflection to final deflec-
design value of К is 0.10. The range tion. Values observed by Spangler
of К is from 0.110 for point support range from 1.38 to 1.46. A design
to 0.083 for^bedding ' the full width value suggestion is 1.25 to 1.50.
of the pipe. ' Some water works authorities do not
1056 committee REPORT Jour. AW WA
apply the lag factorzontal

to side
the thrust action resembles
deflection
caused by live load. vertical thrust action on a footing
and that the resulting soil deforma-
8.7. Barnard's Sidewise Settlement tion at the pipe side is a "sidewise"
Theory instead of a vertical settlement,
Barnard (4) determines this sidewise
A footing resting on a ground sur- settlement using empirically the prin-
face settles under a vertical load.
ciples established for estimating ver-
An initial settlement occurs immedi- tical settlement under foundations.
ately, and usually a further and slower
The distance the pipe side wall moves
settlement occurs as time passes. against the soil is termed "passive
The extent of the settlements depends deformation" of the soil. The ver-
upon the load intensity, the number tical change in diameter or deflection
and nature of applications, the size
of the footing, and the initial and / Die Pressure Die Pressure ч
final engineering properties of the
soil assuming the load. To this ex- (a) 1st Operation Forming
tent there is interaction between the
soil and every manmade structure
resting on it. i ' I
The engineer has means of design-
ing footings from knowledge of the
engineering properties of the materials Pressure ft ' JJ
in the footings. Also at his disposal
are means of estimating soil settlement vL/l Vu/
when he knows the engineering prop-
erties of the soils.
(b) 2nd Operation (c) 3rd Operation
A rigorous analysis of the problem Forming Closing
of underground conduits will probably Fig. 8.9. Operations in Forming High-
remain impractical as long as there Pressure Steel Pipe
are unmeasured variables in field
Operations shown are perf
installation, in backfilling and com-
to 36-in. size having a wal
pacting, and in precise knowledge small as 0.25 in. (4).
of soil engineering characteristics.
Under such circumstances, extreme of the pipe is two tim
accuracy is not expected. The semi- deformation. The true engineering
empirical approach is, however, justi- properties of both the steel pipe and
fied as a practical approximation. the earth envelope enter directly into
Experience shows that the vertical these calculations for deflection. The
diameter of a buried flexible steel pipe method of calculation has checked
under dead load and live load de- closely with measured deflections un-
der low, medium, and high cover.
creases and that the horizontal di-
ameter increases. The vertical loads
The calculation process is simple and
principally graphical. The method
obviously result in horizontal thrust-
has the distinct advantage of telling
ing loads against the soil at the sides
of the pipe. Such thrusts have directly
been the amount of vertical ellip-
sing stulling, or strutting necessary
measured. Postulating that this hori-
in a given instance to avoid exceeding

A 72-in. diameter, 8-gage (0.172-in.)
the final permissible maximum corrugated
pipe metal pipe under the "hot
deflection. This advantage is track"
due to at a steel mill is one outstand-
the fact that the calculated sidewiseing example. This culvert carries one
settlement is really a distance meas-of the heaviest wheel loads known -
urement, even though it is expressed 343 tons on 16 wheels - 88.5 tons of
in terms of pipe diameter. which can act on the culvert when the
wheel truck is directly over it. The
8.8. Pipe Action cover of earth, which is only 3.5 ft, im-
When there is little or no side poses an especially critical condition.
When analyzed for flexural stress by
support, a steel pipe underground the classic methods, the indicated
functions as an elastic ring loaded at
maximum fiber stress is 972,000 psi.
top and bottom. Relatively thick
Actually, the culvert is made of soft
rings carry more load with less deflec-
iron with a maximum specified tensile
tion than do thinner rings and may
strength of 44,000 psi and a yield
be said to possess greater "ring
strength of 27,500 psi. The calcu-
strength." In a given case, however,
lated fiber stress is therefore 22 times
if the ring strength is not sufficient
the ultimate and 35 times the yield
to carry the full load, the pipe deflects
strength. In spite of the figures, how-
and soil side support builds up.
ever, experience demonstrates that, as
Considering progressively pipe hav-
a culvert, this structure has a safety
ing less and less ring strength, it factor
is of four or five. The unit
seen that the resisting action in the
arch stress is approximately 6,500
different walls must change from that
psi. Careful observations showed no
of a ring loaded at top and bottom to
change during the first 9 years with a
the arch action of a totally enclosed
load passing over it several times
ring. In the latter case, the confined
daily. Now, after 29 years (1958), it
ring must possess sufficient arch
is still in service with no appreciable
strength to resist the load adequately
change in recent years.
even though its ring strength may be
Another outstanding example is. a
inadequate. Data on ring strength
60-in. diameter, 8-gage corrugated
are given in Sec. 8.12. culvert under a 77-ft railroad fill. If
the classical bending-moment formula
8.9. Arch Strength
analysis and dead-load assumptions
That a confined thin steel shell sometimes used by engineers are em-
under outside load must act in com- ployed, and if any effect of train live
load is ignored, the indicated maxi-
pression with bending resistance rela-
tively unimportant is proved by widemum fiber stress is 1,100,000 psi - 25
experience (4, 9). Two outstanding times the ultimate. Investigations,
examples of culvert performance willhowever, showed that no appreciable
serve to demonstrate the tremendous change in shape has occurred in 11
years. The unit arch stress is ap-
resistance to earth pressure possessed
by thin metal cylinders. (The fact proximately 9,700 psi.
that corrugated pipe is involved inThe validity of the statement con-
these instances does not make them cerning thin steel shell under outside
less relevant.) load can be demonstrated by labora-
1058 committee report . Jour. AW WA
tory test and by the

liveredmanufacturing
a load covering one-half the
procedure of forming steel pipe in
tube diameter.
a die. In the first series of tests, the hole
8.9.1. Proof by laboratory test. An size was a slip-fit for the ring. This
endeavor was made to demonstrate would come as near developing uni-
photoelastically that the postulationform radial pressure in the ring as
of load distribution around embedded could be expected, considering the
steel pipe used in the Barnard theoryelements and factors involved in the
was not seriously in error. A holetest. The fringe pattern (Fig. 8.8)
Jr
A A A A
AD A
43_L ;
Plane H
_1
plr 's^ i Л
I*
TIT,Point
Fadeout -^
Fig. 8.10. Soil Settlement Under Load
ГАе diagram pertains to set

for an explanation of lette
was drilled in
indicates a f-in.
a relatively uniform radial th
plate to accommodate
stress except near the area of applica- an
ring 1.25 in. in diameter, 0.034 in. tion of a concentrated load, where a
thick, and £ in. long. A slot f in. different pattern would naturally be
wide extended from the hole to the expected.
top of the plate. A steel loading In the second series of tests, the
plunger | in. wide operated in this diameter of the hole was made ^ in.
slot. The contact face of the plunger larger than that of the ring. Thus,
was ground to fit on the tube. It de- the ring was allowed to deflect about
2' per cent vertically before full proof of the tremendous arch loads
contact was established in the hori- which can be carried by smooth-wall
zontal direction. The idea here was steel pipe. The forming operations
to develop, at the extremities of the
are shown in Fig. 8.9. The pipe
is made from steel plates having a
horizontal diameter, a pattern which
would resemble the familiar onionlike length equal to the finished pipe
bulb of pressure existing under foot- section and a width somewhat greater
ings delivering load to soil. Figure than the pipe circumference. The
8.8 shows the stress pattern achieved.first operation consists of die-forming
To be noted are the beginnings of the the ends of the plate to a radius ap-
"ears," which, when full grown, are proximating that of the finished pipe
postulated as being similar to that
"bulb of pressure" under a footing
load except for having been rotated
90 deg from the vertical and converted
to a horizontal thrust.
Noteworthy are the calculated
stresses in the model ring. As an
elastic ring without side support,
placed between the heads of the test-
ing machine, it held 30 lb with a
deflection of about 2' per cent. The
maximum load in the first series of
tests was 200 lb. In the second
series the maximum was 250 lb.
The stress determined, using the
method for "arch load" - see Para-
graph 8.11 (c)- is 5,900 psi for the
200-lb load and 7,300 psi for the Fig. 8.11. Distribution of Load
250-lb load. and Thrust
Calculated on the basis of section
The diagram gives an idea of the distribu-
modulus resistance, the apparent
tion of the horizontal thrust of the pipe side
stress in the ring under the plunger against the soil. A, B, and С are refer-
load is about 3,750,000 psi for the ence points; a is one-half the bedding con-
200-lb load and 4,700,000 psi for the tact angle (4).
250-lb load.
The material in the ring had a yieldas shown in Fig. 8.9a. The second
strength of about 55,000 psi and an operation is performed in a U-shaped
ultimate of about 75,000 psi. The die, with the result shown in Fig.
ring used was not harmed. After 8.9b. In the last operation, as
being removed from the testing ma-shown in Fig. 8.9c, two semicircular
chine, it rolled on plate glass without
die sections are brought against the
showing a balance point. U-shaped pipe, forcing it into a com-
8.9.2. Proof by die-forming pipe. plete circle, with sufficient pressure
The process of manufacturing high- between the plate edges to permit
electric-resistance flash welding. The
test line pipe furnishes continuous test
final closing of the | in.

dieNaturally,
forces the pipe
weld
sections
metal from the welded joint
endure this inward
operation successfully or
toward the center of the pipe and the method of manufacture would be
outward at the top. Such operations discarded. It is necessary to recog-
are performed on 30-in. or 36-in. pipe nize the 'important part that the
having a wall thickness of as little as engineering properties of the pipe
^ о '^
Fig. 8.12. Equivalent Earth Column
The diagram shows th

See Sec. 8.13 for a
material and the die material have in Test Specimen A is tested with con-
this manufacturing operation. fining pressure equivalent to the
height of earth column Zb and axial
8.10. Barnard's General Theory pressure reaching the specimen loca-
The soil which surrounds a pipe tion, the ratio of its shortening, s¿, to
its length, /, may be considered the
in a trench or in a fill is a remolded
soil. Assume that it is required to same as the ratio of settlement S' to
determine the settlement under load the thickness of layer Zi, which it
represents. If Sa is expressed as a
P, Fig. 8.10, in a fill made of two kinds
of materials and built on a natural percentage of /, then Si is the same
ground at Plane 3. There are several percentage of L'.
ways to do this. One method is to The same reasoning applies to
make properly controlled triaxial testsSpecimens В and С and layers L4 and
of the soils involved, to determine Z,3. The sum of all the settlements
their engineering properties (11)- equals the settlement of the load P.
That method is used in this report The accuracy of the findings is in-
Case 1 Case 2 Case 3
''wv wv wv
''Щ+ "wv+ wv+

(a) No Side Support (b) Some Side Support (c) Full Side Support
Fig. 8.13. Cases of Conduits and Loadings
As explained in Sec. 8.13, wv represents the total vertical unit load o

the unit horizontal load {psi) transfer load contact pressur
for illustrative purposes becausefluenced bythe

the many uncertainties of
data on several typical remolded sampling, soils
testing, and calculations of
were available {12). Settlement may
underground load distribution.
be determined by void ratio-pressureThe foregoing calculation is an
data on soil samples or, in more estimate of initial settlement. Settle-
approximate form, by using datament due to time consolidation is addi-
developed by Proctor (13). The tive. In this report, time consolida-
latter subject is further discussed in
tion is accounted for by a "lag factor"
Sec. 8.19. and is discussed in Sec. 8.18.
In Fig. 8.10, the load distribution The left-hand portion of Fig. 8.10
in the soil due to load P is found by indicates a steel pipe as it might be
the methods of Boussinesq, and a installed in a fill such as considered
"fadeout" point is established for before. The pipe is shown as having
calculation purposes. When triaxial settled with Plane 3 and deflected
enough to occupy the the space

intensitybetween
on the vertical diameter
Planes 2 and 3 after initial settlement.
by some amount. It is affected also
by the instantaneous radius of the
For purposes of engineering design, it
is postulated that the pipe is sup- pipe in contact with the soil at a
porting the full weight of the prismgiven point, such as A , B, or C. This
of earth above it because there are is discussed in Sec. 8.15.
no "friction planes" (see Sec. S.3)The
to size of a footing is a factor in
its settlement, and therefore the
modify the load. For a flexible pipe
diameter of a pipe is a factor in its
lOOi- » '''|''|''| I I I 111
"sidewise" settlement. The greater
the size or diameter, the greater the
settlement for given load intensity in
stffiè5tEEE:!= unit per area. The settlement or
passive deformation at the pipe side
may be expressed in terms of the pipe
diameter.
The loading at the side of a pipe is
s .Kit
similar to that for a flexible strip of
infinite length uniformly loaded.
The passive deformation for a given
remolded soil with given compaction
may be determined from triaxial test
data for that soil corresponding to a
° 6 '''v«,iVi%M^- location P on the horizontal pipe
axis. Of course, the deformation also
may be determined by other methods,
if available. For a given soil, the
load to cause a given deformation
■n
becomes greater as the depth of the
soil below the surface becomes greater.
This is important in determining pipe
deflection.
10 20 30 40 50 60 80 100 150
Pipe Diameter- in.
The basic concept of the equivalent-
earth-column method of calculation
Fig. 8.14. Theoretical Height of Cover is shown in Fig. 8.12. The engineer-
"Ring Load" for Steel Pipe With
2 Per Cent Deflection
ing properties of the soil are deter-
mined for a position on the horizontal
pipe axis.
The weight of the soil was 125 lb/cuft (4).
8.11. Definition of Terms

this is the worst situation possible, so
Further discussion will be facili-
it can be used for calculation purposes.
The horizontal thrust of the pipe
tated by defining and using certain
side against the soil is assumedterms
to be
:
distributed in some such fashion as a. Ring load - the load carried by
indicated in Fig. 8.11. As the total the conduit shell as determined by
of all forces must be in equilibrium, moment-modulus calculations for ring
the force of maximum intensity on diameter changes of about 5 per cent
the horizontal diameter must exceed or less.
b. Ring aid - the horizontal thrust

Case 1. In this case (Fig. 8.13a),
toward the pipe due to active the wall thickness and diameter of the
earth
pressure ; it may be calculated pipe selected
on the to meet pressure and
other for
basis of one-third vertical pressure service requirements are such
sand and one-half for clay; it isits
that used
ring strength can carry satis-
only to reduce ring load, not arch factorily
load.all earth loads upon it with-
с Arch load - the load carried by out undue deflection. This case is
the conduit shell under external, quite frequent in water works prac-
radially acting pressure determined
by using the Barlow formula : 1>000L'''4 I 1 1 1 1 1 11
2St = pD
£00Л'УЛУ
or:
400 '''' 'X ^X V -'b

St = pR
in which :
5 = stress ■l!i«F 200V^S

(psi)
200 400 '''' V^S
'' V'' 'X О<%ОЛ^
V ^X V О<%ОЛ^ -'b' J
/ = wall thickness (in.)
p = pressure (psi) due t
cover above pipe
1ии - ^Туллудлд
D = conduit diameter (i
R = instantaneous radius of de-
S 60
formed conduit at given point ел
(in.).
The indicated stress in the pipe wall 40 ^xW
from Eq 8.5 and 8.6 is, respectively: S УлЛ
5-f 20
or:
1OI
5 = ^ 10 20 30 40 50 60 80 100 150
Pipe Diameter- in.
d. Transfer load - Fig.

the load in
8.15. Theoretical Heightexcess
of Cover
of the ring load which "Arch Load"
is for Steel Pipe With
imparted to
Full Side Support
the soil at the sides of the pipe.
e. Passive deformation -was
The stress (S) the hori-
equal to 7,500 psi (Eq.
zontal "settlement" or distance the 8.5); the weight of the soil was 125
soil is compressed by the transfer load. lb/cuft (4).
/. Contact pressure - the measure of
the force acting between metal surface tice, and no problem occurs in instal-
and soil surface in an approximately lation. Ring strength for 2 per cent
horizontal direction. deflection is given in Fig. 8.14.
Case 2. In Case 2 (Fig. 8.13b), the
8.12. Ring Strength and Arch pipe ring strength is sufficient to
Strength carry part of the earth load and live
load, but not all of it, without undue
Three different cases of pipe resist-deflection. Some side support must
ing action are recognized : be afforded by the earth.
Case 3. In Case 3 (Fig. 8.13c),

€ = strain the or
in soil (in./in.,
pipe is a completely flexible decimal of tested
thin length)
ring and by itself can carry under load of little,
very contact pres-
if any, of the live and dead loads sure intensity, Wh
without undue deflection. Full mo- Cwh = strain (per cent) correspond-
bilization of earth assistance is needed ing to given wh in Fig. 8.18-
to retain the degree of roundness re- 8.24.
quired. The enveloping forces are
essentially radially acting. The ring 8.14» Intensity oí Vertical Load
is a confined arch under compression.
It will continue to act as an arch as Total vertical load equals dead
load for height of cover, //, plus live
long as the curvature remains convex
load, impact, and so forth:
against the earth. Arch load creating
compressive stress of 7,500 psi is wv e w vd + wvt
given in Fig. 8.15. When a trench is ba

nary methods and w
8.13. Nomenclature some portion of th
of the soil in the ba
Referring to Fig. 8.12 and the dis-
the side of the pipe and tends to
cussion below, let : support it. For this case,
Wv < Wp + Wa ; Wv - Wa < Wp. . (8.10)
D = pipe diameter (in.)
If wp is not greater than wv - wai
AD = vertical deflection of pipe
Eq 8.10 should be ignored altogether.
(in., or decimal of diameter)
wv = total vertical unit load on
pipe (psi) 8 Л 5. Intensity of Contact Pressure
wVd = unit dead load on pipe (psi) Intensity of contact pressure be-
wvi = unit live load on pipe (psi) tween pipe and soil is dependent on :
F = shape factor (a selective [lj the vertical load, [2J the degree
constant) of stiffness of the pipe as defined in
Wp = unit load (psi) ring load Sec. 8.12, and [3J the instantaneous
carried by pipe when AD radius of the point of contact con-
= 0.02Z> (values from Fig. sidered - in this case, the extremity
8.14) of the horizontal axis.
wa = unit load (psi) ring aid A flexible steel ring or pipe without
afforded by active soil pres-appreciable ring strength or stiffness
sure (Case 3, Sec. 8.12) is best thought of
Wh = unit horizontal loadas(psi)
being a stress gage itself when
transfer load contactconfined
pres- by an earth envelope and
sure supporting earth loads. If the pipe
is perfectly round, there is no bending
L = effective length of equivalent
earth column (in., or in moment and the compressive stress
terms of D) in the wall is given by Eq 8.8 :
AL = passive deformation (in., or
decimal of L)
It is obvious that the same stress stress given by Eq S.St with R equal
must prevail in every portion to
ofthe
theinstantaneous radius which
pipe periphery. The vertical load occurs at the vertical axis of an as-
initiates the stress. The confiningsumed ellipse. This radius is larger
pressure in the soil must maintain than that of the original pipe and
that same stress at every point or the creates a higher stress in the pipe wall.
pipe will change shape. Therefore, Conversely, the instantaneous radius
in a truly round pipe, the confining at the extremity of the horizontal axis
pressure must act radially and be is smaller than that of the original
equal to p in Eq S.S. pipe and, as the same stress in the wall
Also, it is seen in Eq 8.8 that, for amust be maintained there to prevent
given stress and thickness, the valuepipe wall movement, the contact
of p must increase if R decreases andpressure on the horizontal axis must-
decrease if R increases. Therefore, be greater than that on the vertical
to maintain this circumferential corn- axis.
[-<
(a) ^
Fig. 8.16. Instantaneous Radiuses
Shown are
installat
pressive force equally at all points The calculation constant which ac-
after a round pipe has deflected, the counts for this condition when com-
outside radially acting confining force, bined with the figure for length of
or earth pressure, must rise at points equivalent earth column is here termed
of lesser radius.
the "shape factor." It is further dis-
In an initially round pipe of diam- cussed in Sec. 8.17.
eter D,a = b where a and b are located
The intensity of contact pressure
as shown in Fig. 8.16. As the pipe
covering Cases 1 and 2 in Sec. 8.12
deflects, the amount by which b is
can now be stated as :
lessened is added to a. The unit
vertical load on a flattened pipe wh = F(wv -wp)
creates in the wall at the vertical
diameter a tangential compression
For Case 3, wp = 0.
8.16. Length of Equivalent Earth

per cent stress has been added. On
Column this basis, the horizontal leg of the
triangle has a length of approximately
To determine the length of equiva- 6.5 times B. This distance from the
lent earth column, a "footing width"
must first be assumed. This "width"
pipe face is therefore considered the
point of practical "fadeout" /or calcu-
is taken as a chord subtended by a
lation purposes. Thus, in Fig. 8.17,
central angle of 100 deg, as shown in
В = r sin 50° = 0.766r
Fig. 8.17. Dimension В is the usual
one-half footing width used in de- Or, in terms of pipe diameter D :
termining vertical stress at selected
В = 0.383D
depths below that footing. In Fig.
8.17, Line I represents the influence The length, L, of the equivalent earth
line for calculating vertical stress column is:
TABLE 8.3 L = 6.5(0.383)D

L = 2A9D (say, 2.5D)
Values of Shape Factor, F
8.17. Passive Deformation

Values of F*
Pipe Deflection On the basis that the deformation or

% of diam. pipe Deflected Pipe Strutted
Horizontally Vertically shortening of the equivalent earth
(Fig. 8.16a) (Fig. 8.16b)
column under the fading load is equal
0.0 1.25 1.25 to that produced by the average
0.5 1.29 1.21 load over full column length :
1.0 1.33 1.18
1.5 1.37 1.14
2.0 1.40 1.11 AL = y* (2.5¿>)e
2.5 1.45 1.08
AL = 1.25whDe
3.0 1.50 1.04
3.5 1.55 1.01
When the shape f
4.0 1.60 0.98
4.5 1.65 0.95
in Sec. 8.15, is sub
5.0 1.70 0.93 AL = FwhDe
*To nearest 0.05. in which F equals

tive constant whic
ratio of instantaneous radiuses at the
under a strip when the length-to-
top and sides of the pipe.
width ratio lies between 2 and infinity,
as shown by Terzaghi (15). The The values of F for use in Eq 8.14,
stress central in each increment of В
including the constant 1.25 in Eq 8.13,
are given in Table 8.3 for round pipe
is shown in the corresponding blocks.
which has been decreased in vertical
Ten per cent of stress is yet unac-
counted for in the ninth block from diameter and also for pipe which has
the soil-pipe contact line. been increased or ellipsed in vertical
Line II represents the hypotenuse diameter for construction purposes.
of a right-angle triangle which con-For pipe normally installed round and
tains the same area (between line allowed to deflect under load, the
and horizontal axis) as that under horizontal thrust increases as the
Line I, except that the missing 10 deflection increases.
When pipe is formed, strutted, bleor

pipe tends actually to make the
stulled to increase its vertical diam-
pipe occupy a larger space in the soil
eter, and it remains that way afteras it expands horizontally and becomes
installation, the horizontal thrust de-
more nearly round. This action causes
creases as the percentage of strutting
a consolidation of the earth envelope
increases. The stress in the pipe and builds up pressure by changing
wall is less than for round pipe becausethe void ratio until the pipe becomes
the instantaneous radius of the force- fully round. A round pipe, when
deflecting vertically, actually occupies
activating top is less than that on the
horizontal axis. Therefore, a lower less and less volume as the deflection
balancing contact pressure is neededincreases and must develop side sup-
on the horizontal axis than on the port without the special aid supplied
vertical axis. This ellipsing cannot to the strutted pipe.
4^^^ Line II
i|r At
^Л
ДЛ'' и'1 glslslslailslslsls
n' / ^ i ö o d do dodo
Fig. 8.17. Determination of Length of Equivalent Earth Column
See Sec. 8.16 for discussion of
be carried too far, however, or the 8.18. Time Lag Factor

active pressure at the pipe side may
The passive deformation determined
reverse the convexity of the wall.
Equation %.% may be used to investi- using Eq 8.14 is the initial sidewise
settlement. As with vertical settle-
gate the latter condition with allow-
able stress held well below the yield ments, experience shows that the
sidewise settlements increase slowly
point. The ring strength must be with time. The increase becomes
considered.
slower and slower, however, until,
An important factor often over-
after a period of years, virtual stabil-
looked in the case of strutted pipe is
ity is attained. In a particular case
that the area of an ellipse is less than
that of a circle with the same circum- when engineering properties of a given
ference. Therefore, the vertical load soil are known, the time consolidation
causing flattening of a strutted flexi-rate can be determined. Initial and
1068 COMMITTEE report Jour. AW WA
subsequent measurements compaction control. made on

The diagram
actual pipe installations (2) show
represents the results of about 900 an
soil
increase with time of from 1.38 to 1.46 specimen tests. The actual range of
times initial deflection. A design all varieties of the remolded soils
value of 1.25 to 1.50 is suggested fortested indicates a spread of one-half
average conditions. to twice the values shown. The
actual consolidation of natural soils
8.19. Passive Deformation by Use in place in the 40-1,200-psi range of
of Diagrams indicated saturation penetration re-
sistance may be about one-third of
Figures 8.18-8.23 are included as the values shown in Fig. 8.24.
aids in design and deflection control
To use Fig. 8.18-8.24, calculate Wh
using Eq. 8.11. Determine the value
of z psi (Fig. 8.12). Enter at the
150 i
left of the diagram for a given soil

125 type, using the calculated value of Wh.
Pressure -psi y >
Trace right to the intersection of a line
(probably interpolated) representing
ï io° 47 y / / the value of z. Read the axial strain
vertically below in per cent, corre-
' /у sponding to the calculated Wh. Sub-
stitute this value for €Wh in Eq. 8.15 to
determine passive deformation at the
1 "^/^/ TÍH
side of the pipe (Fig. 8.12) :
AL = €whD
8.20. Relationship
tion to Passive Deformation
0 12 3 4 5
Axial Strain, (w
Because the vertical- per
deflection of cent
Fig. 8.18. Axial Strain in Typical Soil: the pipe and the shortening of the
Well Graded Sand earth column are now both expressed
in terms of D, it becomes advan-
Data are from Test 1-T, Ref. 12.
tageous for calculation purposes to
note the relationship. If e is ex-
calculations. The information has pressed as a percentage change in the
been selected from published data length of the earth column, then
(12, 16). Other data are available. double that percentage applies to the
Those selected are more or less typical pipe deflection. In other words, if
of remolded soils. If not otherwise a given contact pressure in pounds per
indicated in the figures, soil specimenssquare inch causes a passive deforma-
have been compacted to 98 Proctor
tion corresponding to eWh = 1 per cent,
density. the vertical deflection of the pipe will
Figure 8.24 has been included for be 2 per cent of D. If the allowable
use by those designers familiar with vertical deflection is 5 per cent, the
Proctor's methods (13, 14) of soil required passive resistance must be
dead weight and live load, exceed the

mobilized with a passive deformation
allowable
corresponding to eWh = 2.5 per cent. ring load plus ring aid range,
then from a design point of view the
8.21. Methods of Calculation earth envelope becomes a part of the
structure. Attention must therefore
To facilitate^calculation, unit pres-
be given to the bedding and backfilling
sures are expressed in pounds per
operations to the extent that sufficient
square inch (psi). The following de-
terminations are involved : strength is built into the backfill to
support the pipe properly.
Step 1 - intensity of vertical load Step 2. Ring load (wp) for 2 per
(Eq 8.9) cent deflection is given in Fig. 8.14
Step 2- ring load (Fig. 8.14) in terms of depth of cover, in feet,
Step 3 - ring aid (Paragraph 8.11b)for soil weighing 125 lb/cu ft. The
Step 4- arch load (Fig. 8.15) depth of cover may be converted to
Step 5 - contact pressure (Eq 8.11)
pounds per square inch by referring
Step 6 - passive deformation (Eqto Table 8.4. Figure 8.14 shows the
8.14 or 8.15).
Step 1. The load on the pipe, of 100 1 - i
Applied La
course, is governed by the height of Pressure -
cover, plus live load, impact, and 'g. 75

other factors. The vertical load is
calculated using Eq 8.9. Unit earth
loads in pounds per square inchf for
so
given heights of cover are shown in
Table 8.4 for soil weighing 125
lb/cu ft. The unit pressure for other
soil weights is in direct proportion to
the tabular value for a given height 0 12 3 4 5
of cover. Axial Strai
Intensity of truck live loads for the

Fig. 8.19. Axial Strain in Typical Soil:
standard highway H-20 design and Sand With Clay Binder
for the standard railroad locomotive
Cooper E-72 loadings are shown in Data are from Test 10-T, Ref. 12.
Table 8.5 for given heights of cover.
height of backfill carried without ring
Steel pipe for water service is usually
of such diameter and wall thickness aid by pipe with a given diameter and
that earth loading is not a problem. thickness. If the proposed pipe will
In the smaller sizes, the ring strength carry the entire backfill load, and any
alone is sufficient to bear the load. live load present, no further calcula-
When ring aid in the form of active tions are necessary.
earth pressure is added to the ring Figure 8.14 is based upon the stand-
strength, sizes up to about a 36-in. ard formula for an elastic ring, using
diameter in most cases need no more a constant of 0.10, which applies to
than the ordinary attention given
earth load on pipe in a flat-bottom
trench. The constant for concen-
bedding and backfilling operations.
But when the earth loads, including
trated loads top and bottom is 0.149.
Step 3. If the ring load for the

pipe almost carries the vertical loads, 1
Applie
Pressu
the engineer may elect to figure such
assistance at the side of the pipe as £ 30
will be given by some percentage of

the active pressure of the earth in the
backfill. Actually, underground pipe
has probably been getting such aid
for a long time without its being
recognized. The value of the active
pressure to be used must remain a
matter of judgment based on circum- 0 12 3 4 5
stances of the installation. Unless Axial Stra
the value is better known, the sugges-

Fig. 8.21. Axial Strain in Typical Soil:
tion given in Paragraph 8.11b may be Very Plastic Clay
followed. The foregoing statements
Data are from Test 20-T, Ref. 12.
are made on the assumption that the
trench is backfilled by ordinary meth-
ods and without tamping. wishes and if care is taken in the
Step 4. If the proposed pipe will
installation procedure to provide ade-
not carry the vertical load in ac-
quate strength in the earth envelope.
cordance with Eq 8.10, its arch load
Step 5. After the effective vertical-
should be determined from Fig. 8.15.
load intensity and the ring strength of
The loads given in Fig. 8.15 are asso-
the pipe are determined, the transfer
ciated with a safety factor of four
load from pipe to soil is determined
for steel having a yield point of 30,000
using Eq 8.11.
psi and a joint efficiency of 100 per
cent in compression. Still higher Step 6. When the intensity of the
transfer-load pressure is known, the
loads can be carried if the designer
passive deformation may be deter-
40 1
mined using Eq 8.14 or 8.15. The
value 6 or eWfi is a measure of the axial
Applied strain caused by an axial load on the L
£ 30 soil while it is at the same time being
subjected to lateral radial pressures
due to height of cover. It is at this
I 20 point that soil mechanics must enter
the problem.
.ил I I
8.22. Triaxial Shear Tests of Soils
In Fig. 8.12, an imaginary triaxial-

0 12 3 4 5
Axial Strain,
shear-test soil specimen
6 -
is assumed
per
to c
h
lie somewhere along the length of the
Fig. 8.20. Axial Strain in Typical Soil: equivalent earth column. If the posi-
Lean Clay tion С is at the contact surface of the
Data are from Test 15-T, Ref. 12. pipe, the specimen will be subjected
to the maximum axial stress existing 300 j
in the equivalent earth column, and

the confining stress on the specimen A
/ Pressure, 45 psi
will be that due to the depth z. Then
if triaxial-shear-test data, including 250
the stress-strain characteristics of
the sample corresponding to different
confining pressures due to change in
depth z, are available, it is possible to
solve Eq 8.14 or 8.15. 200
Stress-strain diagrams of certain I /

8. /
"typical" remolded soil materials S /
compacted to 98 per cent maximum Я /

density are shown in Fig. 8.18-8.21. §> 150
с /
These diagrams give the usable axial T3
3 /
/
compression in pounds per square ob
с /
/
о /
inch corresponding to given lateral
pressures in pounds per square inch, 100
with axial strain given as percentages.

The diagrams are presented as aids to
understanding rather than as solu-
tions for particular problems. Full 50 - +
200i
°0 0.5 1.0 1.5

Max
App
Deflection Л *^^~ Axial Strain, (^ -per cen

150
Fig. 8.23. Axial Strain in Crushed Slag
at Maximum Density
Data are from Ref. 16.

£ loo
.1 // / ^^~~w data on these soils and the tests

conducted by the Bureau of Reclama-
tion have been published (12). Fig-
50 If/ ures 8.22 and 8.23 show data for
graded gravel and crushed slag
for maximum compaction. Addi-
tional data have been published (16).
The deviator stress ordinates in
0 0.5 1.0 1.5
Axial Strain, Fig. 8.18-8.21
iw -perand the longitudinal
cent
stress ordinates in Fig. 8.22 and 8.23
Fig. 8.22. Axial Strain in Graded Gravel are values of Wh found by use of Eq
at Maximum Density
8.11 and associated with the proper
Data are from Ref. 16. applied lateral pressure correspond-
ing to distance z below Deflection of the

pipelinessurface
with spun coal-
(Fig. 8.12). In Fig 8.18-8.24, the tar enamel is best held to 2 per cent for
abscissas (axial strain percentage) are mechanical field joints ; however, other
values of eWh in Eq 8.15 stated in types of field joints may be deflected
percentages for associated values ofas much as 5 per cent. The deflec-
Wh and z. tion of pipe lined with spun cement-
mortar should be held to 2 per cent.
200
180
1
8.23.2. Definition of failure. At this

point, when deflection is mentioned,
160
140
120 the engineer accustomed to thinking

100-J
in terms of flexure or bending-moment
80- Indicated Saturated
_ Penetration Resistance
formulas in rigid construction is likely
to contend that permanent deflection
Before Loading- psi /
can occur only after the yield point
=- ao ЛЖ^-А-^
о / ¿s c' s y '
has been passed and that, therefore,
a pipe so stressed has failed struc-
turally and is dangerous. The sim-
plest rebuttal to this argument is to
% îs=- ÉgŽS^ - recognize that the steel in a finished
pipe has, in the manufacturing process,
been cold coiled, uncoiled, bent,
curved, or twisted a number of times
and has been stressed beyond the yield
point each time; yet, after all these
operations have been completed, the
finished steel pipe is used for all
manner of high-pressure work without
fear or hesitation.
If the engineer still is hesitant to
restress a part of the finished pipe
0 12 3 4 5. wall beyond the yield point by slightly
Mean Measured Consolidation
per cent of
deflecting it underground,
specimen
let him
depth
consider what happens to the test
Fig. 8.24. Axial Strain or Consolidation
specimen by which the pipe strength
of Soil Specimens Under Dif-
ferent Conditions is measured according to specification.
Usually it is sliced as a ring from the
The various specimens were under different
end of a finished pipe, cut at one side,
degrees of compaction subjected to pressures
uncurled from the circle into a flat
induced by applied loads or weight of
cover (4, 13). piece, and then put in a tensile-testing
machine which proceeds to show that
after once more passing the elastic
8.23« Comments on Design Factors
limit, the steel still possesses the
in Barnard Theory
specified strength. In a way, the
8.23.1. Permissible deflection. De- deflection underground is simply a
flection of unlined pipe, or of pipe finished forming operation.
lined after installation, may safely Therefore, where steel pipe such as
reach 5 per cent of nominal diameter. is here discussed is concerned, the
word "failure" must define a state of in flexible pipe have at times shown
the horizontal intensity to be about
falling short of satisfactory perform-
ance and not a state in which localized 25 per cent more than the vertical.
stresses appear to pass the yield point
The shape factor used in the formula
for transfer pressure intensity is
of the material as judged by the results
of bending-moment formula analysis. intended to account for increased load
8.23.3. Vertical loads. The results due to increase in horizontal diameter,
of many tests show that flexibleincreased
steel pressure because of shorten-
pipe frequently carries less load ing thanof vertical diameter, and in-
that indicated by the weight of creasedthe peak horizontal intensity due
to load distribution.
prism directly over the pipe. Some
measurements on flexible pipe have
In the original concept of contact
shown the full weight of the prism
pressure and "sidewise settlement,"
TABLE 8.4
Unit Earth Loads Due to Height of Cover

(Soil Weight 125 lb/cu ft)
+0 +1 +2 +3 +4 +5 +6 +7 +8 +9
Height of
Cover, H
Unit Earth Load - psi
0 0.87 1.74 2.60 3.47 4.34 5.20 6.08 6.94 7.81

10 8.68 9.55 10.4 11.3 12.2 13.0 13.9 14.8 15.6 16.5
20 17.4 18.2 19.1 20.0 20.8 21.7 22.6 23.4 24.3 25.2
30 26.0 26.9 27.8 28.6 29.5 30.4 31.2 32.1 33.0 33.9
40 34.7 35.6 36.5 37.3 38.2 39.1 39.9 40.8 41.7 42.5
50 43.4 44.3 45.1 46.0 46.9 47.7 48.6 49.5 50.3 51.2
60 52.1 52.9 53.8 54.7 55.6 56.4 57.3 58.2 59.0 59.9
70 60.8 61.6 62.5 63.4 64.2 65.1 66.0 66.8 67.7 68.6
80 69.4 70.3 71.2 72.0 72.9 73.8 74.6 75.5 76.4 77.3
90 78.1 79.0 79.9 80.7 81.6 82.5 S3.3 84.2 85.1 85.9
100 86.8 87.7 88.5 89.4 90.3 91.1 92.0 92.9 93.7 94.6
above the pipe transmitted the value of the

toshape
thefactor F was
pipe. As carrying thegiven full by weight
Barnard in anofunpublished
the prism would probably be the work as F = 1.35 for round pipe
worst case for flexible pipe, the full- whose horizontal diameter was in-
weight assumption should be used. creased 2 per cent by vertical load.
A simple empirical means of varying The original value of F = 1.35 was
the load would be to modify the esti- developed by consideration of change
mated weight of the soil prism itself. in pipe shape, load distribution at the
The height of cover over large pipe extremity of the horizontal diameter,
may be taken as the average between measured pressures at the side of in-
the horizontal diameter line and stalled pipe, and the length of the
the top. equivalent earth column from theory
8.23.4. Horizontal pressure. Meas- and as influenced by experimental ob-
urements of the horizontal pressure servations. The value F = 1.35, de-
veloped through a pipe

method entirely
likewise varies with these factors.
different from instantaneous-radius
It should be noted that, while only
considerations, is in close agreement horizontal thrusting forces are here
with F = 1.40 in Table 8.3, developed considered, the same general principles
by the latter more general method. can apply to forces radiating from
Moreover, the value of F = 1.35 was any part of the pipe periphery if the
used originally to check the measured thrusting loads are known.
deflections of many steel pipe installa- 8.23.6. Backfilling. In extensive
tions of large and small diameters, large-diameter flexible pipe installa-
thick and thin walls, under high and tions, if vertical deflection is a serious
low cover, and with and without live factor, laboratory tests on backfill
load, for the purpose of proving the material are essential, unless guessing
sidewise settlement theory sound. is preferred. Field compaction con-
trol should tie field operations to
TABLE 8.5 laboratory data unless, again, guess-
ing is preferred. , Such field control
Approximate Unit Loads Due to H-20 Truck
is now common practice in large
and Cooper E-72 Effective Live Loads high-
on Structure Under Given Heightway projects and pays off in the form
of Cover of better roads and lowered main-
tenance costs. A nominal amount of
TT . , , Load - psi y
TT Height . , of , - y backfill control can reduce pipe ma-
Cover, H
terial costs on large-diameter pipe
ft H-20 E-72* installations.
1 11 30 One approximate method of pre-
2 6 25 dicting settlement is to use the
4 3 18
Proctor (13) indicated saturated pene-
6 2 14
8 1 10
tration resistance in conjunction with
10 1 7 measuring soil dry weight, both opera-
12 1 5 tions being performed in the field.
14 4 Percentage of consolidation can then
16 4
18 3
be estimated using Fig. 8.24.
20 2
25 1 8.24. Ellipsing, Shilling, or Strutting
* Loads include 50 per cent impact. It is common practice in the high-
way and railroad field to increase the
The agreement was good between vertical diameter of round pipe during
calculated and measured deflections, installation by installing vertical col-
in the examples in Sec. 8.29, either umns inside the pipe. The amount
one, or both, values of F are used as a of ellipsing has been a matter
matter of interest. of judgment and experience. The
8.23.5. Value of e. Strain, or method herein proposed for determin-
shortening under compression, ining the passive deformation at the
soil
varies with such factors as void ratio, side of the pipe provides an immedi-
moisture content, axial load, confining ate and extremely simple method of de-
pressure, load repetition, and time. termining if ellipsing is needed and to
The "sidewise settlement" or passive what extent the pipe diameter should
deformation of the soil around the be elongated vertically. Because the
"sidewise settlement,'* or increase in

cent elongation of the vertical diam-
horizontal diameter, of the
eterpipe
would is
achieve the desired result.
similar to the deflection of a beam or From this it can be seen that percent-
bridge under load, ellipsing is theage of ellipsing distance is obtainable
equivalent of ' 'cambering." Whendirectly through the use of Fig. 8.18-
bridges are cambered, an upward ex-8.24. The pipe theoretically then
becomes round after installation.
tending "bow" is built into them, and
the amount of camber or bow is such The stulls or struts may be re-
that, when loaded, the beam becomes moved before backfill reaches full
straight or the bridge floor conformsheight if it is determined that ade-
to the design contour. In these quate side pressure has developed. If
cases, the natural deflection of the the pipe is formed and installed as an
structure under load occurs prior to ellipse with the long axis vertical, no
the time the deflection becomes ob- stulls or struts are necessary. Struts
jectionable. cannot be used with some types of
TABLE 8.6 field joints.
Backfill Classification According to Soil

8.25. Trench Conditions
Compaction Requirements
Standard
ASTM
When pipe is put in a trench in-
Percent- stead of a fill, that portion of the
age of
Compac- equivalent earth column lying be-
Class Type of Backfill tion
tween the side of the pipe and the
1 Supporting extremely heavy side of the trench undergoes passive
loads 100 + deformation which may be equal to
2 Subject to surface loads 95-97 that of the soil beyond the trench
3 Not subject to surface loads 90-93
4 Where settlement causes no
side only if the compaction between
harm 85-90
the pipe side and the side of the
trench is such that it is comparable
From Sowers {17)
with the remainder of the length of
the equivalent earth column.
The same thinking When the backfill is not placed at can
98
the passive deformation of the soil per cent of maximum density, the
at the side of the pipe. For instance, additional passive deformation in
if the axial strain in the soil reaches 2.5 excess of that shown by the soil data
per cent to develop a certain transferfigures may be estimated. Establish
load, the theoretical length of the the distance I between pipe wall and
equivalent earth column can be in- trench side. An extra amount of
creased 2.5 per cent by pulling the passive deformation will occur in this
pipe side inward by that amount. If portion of the equivalent earth col-
then, under these circumstances, the umn. This additional deformation is
pipe is stulled or strutted 5 per cent approximately 0.5, 0.7, or 1.0 times
vertically, the full transfer load will the deformation which would be cal-
be accepted by the two equivalent culated for the distance I for standard
earth columns, each with a 2.5 per ASTM D698 densities of 95, 93, and
cent passive deformation! Alterna- 90 per cent, respectively. This addi-
tively, if 2 per cent deflection were to tional-deformation figure is not a
be allowed in the pipe finally, 3 per percentage but a distance and is to
1076 COMMITTEE REPORT Jour.AWWA
be so considered in a estimating the

decrease i n vertical d iameter . Whe n
pipe deflection. internal pressure is introduced into the
For example, suppose a passive pipe, the first action is reversed and
deformation of 1 per cent occurs the at internal pressure tends to force
the side of a pipe whose wall is 12 in. the pipe to a perfect circle. The
from the trench side when the transfer internal pressure reduces the stresses
pressure is developed. If this is a caused by the soil. In a way of
48-in. pipe, the vertical deflection is speaking, the backfilling operation has
2 per cent of the diameter, or 0.96 in. prestressed the pipe against internal
The compaction is to be 90 per cent pressure. For the Case 3 (Sec. 8.12)
standard ASTM density. The addi- design condition - full arch load and
tional deflection between pipe wall no internal pressure - the pipe wall
and trench is 0.01 X 12 = 0.12 in. for will be under 7,500 psi compressive
one side. The total pipe deflection isstress if the full cover height is used
0.96 plus twice 0.12, or 1.20 in. and will be under proportionally lower
It has been found that, for standardstress for lesser heights. Whatever
ASTM D698 densities below 90 per compressive stress due to soil action
cent, the soil resistance drops off veryexists in the pipe must be overcome
rapidly. by the tensile force due to internal
pressure before the pipe wall can go
8.26. Compaction Control into tension.
For the above reason, the common
The maximum density, as defined practice of determining steel pipe
by standard ASTM D698 tests, is wall thickness on the basis of yield
that obtainable with a moderate
strength to resist internal pressure is
amount of rolling or tamping with the
sound and conservative. Theoretical
soil at optimum moisture content.
bending stresses and tensile stress
The ratio of the actual soil density tonot be totaled for steel pipe as
should
the maximum density is the percent-
must be done for rigid pipe.
age of compaction. This indicates
the degree of compaction and the 8.28. Live Load and Shallow Cover
ability of soil to carry heavy loads
without settlement. Table 8.6 shows The equation for intensity of con-
tact pressure (Eq 11 ; F = 1.40) :
classification according to compaction
requirements (17). Data on methods wh = 1.40 (wv - wp)
of trenching and backfilling for water
pipe based on a soil mechanics ap- contains a means of determining the
ring strength necessary to keep the
proach have been published (17 y 18).
transfer pressure to its reasonable
8.27. Internal Pressure limit. Where no live loads are pres-
ent, or where the depth of cover is
Buried steel pipe when empty considerable,
is ring strength as such
under compressive forces due to soil
is relatively unimportant. For live
weight and soil pressure. The out-
loads and shallow cover, however, the
side vertical forces tend to flatten the quantity (wv - wp) - that is, the differ-
pipe and create tensile and compres- ence between the vertical load on the
sion stresses in the pipe wall owing to pipe and its ring strength load - must
be such that the allowable transfer Problem 2 (testing for semirigidity

load is not exceeded. An example -Case 2, Fig. 8.13b). Will a 30-in.
will illustrate the procedure for de-pipe having a J-in. wall thickness sus-
termining the wall thickness in this tain a 5-ft cover plus an H-20 live
case. load without exceeding 2 per cent
Assume: 48-in. diameter deflection?
pipe, 2 Backfill is sandy clay.
Solution: H-20
per cent deflection, 3-ft cover,
live load, burial in sand-clay
Step 1.binder
Vertical-load intensity
as in Fig. 8.19. From Tables 8.4 and equals 4.34-psi dead load (Table 8.4)
8.5, wv = 2.6 + 4.5 = 7.1 psi. From plus 2.5 psi (Table 8.5), totaling
Fig. 8.19, the maximum Wh for 1 per6.84 psi.
cent passive deformation is taken as Step 2. From Fig. 8.14, a dead
8.0 psi. Then: load of 2.7 ft of cover will be carried
Wh = 1.40 (wv - Wp)

by ring strength. A cover of 2.7 ft
8.0 = lA0(wv - wp)
is equivalent to 2.34 psi {yq of 23.4 psi
8.0 = 1.40(7.1) - lAOwp
for 27-ft cover; Table 8.4).
8.0 - 9.94 = - lAOwp Step 3. Is this a Case 2 situation ?
wp = 1.4 psi Assume ring aid (active soil pressure)
equals ' of vertical soil pressure at
From Table 8.4, 1.4 psi corre- depth of pipe axis (6.25 ft). Pres-
sponds to a 1.6-ft cover. From Fig. sure at depth of 6.25 ft (this is height
8.14, for a 1.6-ft cover, the thickness 2, Fig. 8.12) equals 5.4 psi (interpo-
of a 48-in. pipe should be 0.375 in. lated from Table 8.4). Assumed ac-
under the live load in this soil com- tive pressure is ' of 5.4 = 1.8 psi.
pacted to 98 per cent density. If Referring to Eq 8.10:
there were no live load, Fig. 8.15 Wv - wa = 6.84 - 1.8 - 5.03
shows that this pipe could be as 5.03 > 2.34
light as 0.141 in. and still not deflect
2 per cent. As 5.03 is greater than 2.34, this
cannot be considered Case 2 and must
8.29. Examples of Calculation be calculated as Case 3 - that is,
Problem 1 (self-supporting conduit wp = 0.
- Case 1, Fig. 8.13a - where no side Step 4. From Fig. 8.15, arch load
support is assumed to be provided by is 140 ft, which is more than adequate.
the soil). What is the height of Step 5. With F = 1.35 in Eq 8.11,
cover without live load for a 30-in. wh - 1.35 (wvd + wvi)
pipe having §-in. wall thickness when = 1.35(4.34 + 2.5)
allowing a maximum deflection of 2 = 9.23 psi
per cent?
Solution. Referring to Fig. 8.14, If F = 1.40 is used instead of
the height of cover is found to be F = 1.35, wh = 9.57 psi. The differ-
21 ft. ence of 0.34 psi is too small to detect
Comment. Backfilling may be done in Fig. 8.19.
with no special attention to tamping Step 6.
because the pipe carries the full load From Fig. 8.19, it is found (by
by ring strength. interpolation between Curves 3.1 and
12.5) that for a z heightIf F =corresponding

1.40 is used instead of
to 5.4 psi (see Step 3)F =the
1.35, axial
the 3.8strain,
per cent increase in
eWfi, for Wh = 9.2 whpsi is aboutin0.8
is imperceptible Fig. 8.18.
per cent. From Eq 8.15, Step 6. Passive deformation (from
Fig. 8.18):
AL = eWhD
AL = 0.008 (30) = 0.24 in.
Я=4 Я=8 Я = 12
As this is less than the 0.01 (30)
= 0.30 in. allowable, pipe will be whpsi 4.7 9.4 14.0
satisfactory provided backfill is simi- z- psi 4.5 8.0 11.5
eWh - per cent 0.3 0.6 0.7
lar to that represented by Fig. 8.19
and is tamped to about 98 per cent Pipe deflection
of optimum density. Calculated- per cent 0.6 0.12 0.14
Calculated- m. 0.18 0.36 0.42
Comment. In the foregoing ex- Measured - in. 0.13 0.22 0.32
ample, notice that there is a leeway
of 25 per cent between the 0.24-in.
theoretical deflection and the 0.30 in.
Comment. Problem 3 is a d
allowable. Experience indicates this as far as is known from r
to be an ample lag factor, or time the Chapel Hill, N.C., pipe
factor, to account for final soil con- Considering the multiplicity
solidation and passive deformation. influencing the precise inst
Problem 3 (flexible conduit). A such a flimsy pipe, plus the
30-in. smooth -wall pipe with 0.109-in. tion of properties of sand,
wall thickness is to be installed with agreement between estimated and
tamped sand backfill. What vertical
actual figures appears quite satis-
deflection is to be expected under factory for practical purposes.
covers of 4, 8, and 12 ft, assuming that Problem 4 (large pipe, high fill).
Fig. 8.18 is typical of the backfill What deflection can be expected in
material - graded sand? an 84-in. -diameter corrugated pipe
Solution: with a wall thickness of 0.281 in.
installed with 137 ft of cover in a 100
Step 1. From Table 8.4, wvd = 3.47,
6.94, and 10.4 psi for 4-, 8-, and 12-ftper cent compacted granular fill hav-
cover, respectively. ing properties of graded gravel in Fig.
Step 2. Ring load from Fig. 8.14 8.22 ? Pipe is to be strutted 3 per cent
is less than 1 ft (ignore). vertically during installation.
Solution:
Step 3. No ring load, no ring aid
(ignore). StepL From Table 8.4, wvd= 118.9
Step 4. Arch load from Fig. 8.15, psi.
by extrapolation, is 60 ft and is Step 2. Ring load - by comparison
therefore satisfactory. of section modulus values, this pipe
Step 5. Transfer load, Wh = 1.35«w a ring load capacity about equal
has
(with F = 1.35 in Eq 8.11). Then: to 1-in. plate pipe of the same diam-
eter. Then, from Fig. 8.14, ring load
wh = 4.68 for H = 4 cover equals 8 ft, equivalent to 6.94
- 9.37 for H = 8 psi.
= 14.0 for H = 12 Step 3. Omitted, ring aid ignored.
Step 4. From Fig. 8.15, the arch

practical purposes and is on the con-
load indicated is over 200 ft. servative side.
Step 5. Because of the very high Problem 5 (large pipe, low cover).
A 96-in. -diameter water pipe carrying
load on this pipe, and because the final
transfer pressure will depend on the 150 psi pressure is to be put in a
trench under 22 ft of cover with no
final shape of the pipe, a trial calcula-
tion should be made to estimate this live load, allowing 2 per cent deflec-
final shape. The deflection shown tion. by What thickness can be used
the trial calculation will permit a with sandy-clay backfill ?
Solution. Wall thickness required
closer value of F to be selected for a
for internal pressure is ' in., allowing
subsequent calculation for transfer
load and deflection. a tensile stress of 15,000 psi (Table
Trial calculation: 6.1, Chap. 6). An old rule of thumb
for minimum wall thickness to meet
For 3 per cent strutting vertically,
F = 1.04 (Table 8.3) in Eq 8.11: the practical requirements of fabricat-
ing, transporting, and installing steel
wh = 1.04(118.9 - 6.94) pipe is that this minimum should be
wh = 116.4
not less than 0.5 per cent of the diam-
eter. For 96-in. pipe, the minimum
Step 6. Figure 8.22 does not con- thickness would be 0.48 in. As a
tain a value for z distance as great as
involved here. But if the curve show-
thickness of ' in. also satisfies the
internal-pressure requirement, further
ing laboratory maximum measuredcalculations will be based on the use of
deflection is read, eWfi = 0.6 per cent,
'-'x'. wall :
and pipe deflection is twice this, or 1.2
Step 1. From Table 8.4, wvd= 19.1.
per cent. The pipe with load on will
Step 2. From Fig. 8.14, ring load
be deflected 3 per cent minus 1.2 per
is less than 1 ft (ignore).
cent, or 1.8 per cent. By interpola-
tion in Table 8.3 between F = 1.14 for Step 3. Omitted.
Step 4. From Fig. 8.15, arch load
1.5 per cent deflection and F = 1.11
is 90 ft. As cover is only 22 ft, |-in.
for 2 per cent deflection, F = 1.122
wall thickness is adequate.
for 1.8 per cent deflection.
Final calculation:
Step 5. From Eq 8.11, transfer
load wh = 1. 40(19.1) = 26.7 psi.
wh = 1.122(118.9 - 6.94) Step 6. From Fig. 8.19, for z = 26
Wh = 136.6 ft or 22.6 psi, €wh = 1.7 per cent;
€uh = 0.8 per cent (Fig. 8.22). pipe deflection equals 2 (1.7) = 3.4
per cent.
The calculated pipe deflection is Therefore, this pipe can be held to
therefore 1.6 per cent or 1.34 in. 2 per cent deflection with fully com-
Comment. Problem 4 postulates pacted backfill of sand with clay
the conditions of the multiplate pipebinder, provided it is strutted 3.4 - 2.0
installed and observed at Cullman, = 1.4 per cent (about 1^ in.).
Ala. (19). Two years after comple-
tion, the maximum deflection meas-References
ured at the high fill was about 1.25 in.1. Marston, Anson. The Theory of Ex-
Therefore, the calculated deflection ternal Loads on Closed Conduits in the
checks the measured deflection for all Light of the Latest Experiments.
Proc. 9th Annual Meeting,Triaxial Highway

Compression Test Conditions.
Research Board (Dec. 1929). In Special Tech. Pub. 106, Am. Soc.
2. Spangler, M. G. The Structural Design for Testing Materials, Philadelphia,
of Flexible Pipe Culverts. Iowa State (1951).
College Bul., No. 153 (1941). 12. Wagner, A. A. Shear Characteristics of
3. Spangler, M. G. Protective Casings for Remolded Earth Materials. In Spe-
Pipelines. Iowa State College Eng. cial Tech. Pub. 106, Am. Soc. for
Rpts., No. 11 (1951-52). Testing Materials, Philadelphia (1951).
4. Barnard, R. E. Design and Deflection 13. Proctor, R. R. An Approximate
Control of Buried Steel Pipe Supporting Method for Predicting the Settlement
Earth Loads and Live Loads. Proc.
of Foundations and Footings. Paper
ASTM, $7:1233 (1957). 14.4, 2nd International Conference on
5. Spangler, M. G. Underground Con- Soil Mechanics & Foundation Engi-
duits - An Appraisal of Modern Re- neering, The Hague, Netherlands
search. Trans. ASCE, 113 :316 (1948). (1948).
6. Braune, Cain & Janda. Earth Pressure 14. Proctor, R. R. Design and Construc-
Experiments on Culvert Pipe. Public
tion of Rolled-Earth Dams. Eng.
Roads, 10:9 (1929).
News-Record, 111:245, 286, 348, 372
7. Barnard, R. E. Behavior of Flexible (1933).
Steel Pipe Under Embankments and
15. Terz aghi, Karl. Theoretical Soil Me-
in Trenches. Bulletin, Armco Drain-
age & Metal Products, Inc., Middle-
chanics. John Wiley & Sons, New
York (1943). p. 380.
town, Ohio (1955).
7a. Watkins, R. K. & Spangler, M. G. 16. Housel, W. S. Interpretation of Tri-
Some Characteristics of the Modulus of axial Compression Tests on Granular
Passive Resistance of Soil : A Study in Soils. In Special Tech. Pub. 106,
Similitude. Highway Research Board, Am. Soc. for Testing Materials, Phila-
Proc, 37:576 (1958). delphia (1951).
8. Barnard, R. E. Design Standards for 17. Sowers, G. F. Trench Excavation and
Steel Water Pipe. Jour. A WWA , Backfilling. Jour. A WWA, 48:854
40:24 (Jan. 1948). (Jul. 1956).
9. Report on Steel Pipelines for Under- 18. Reitz, H. M. Soil Mechanics and Back-
ground Water Service. Special In- filling Practices. Jour. A WWA,
vestigation 888, Underwriters' Labs., 48:1497 (Dec. 1956).
Inc., Chicago (1936). 19. Spangler, M. G. & Phillips, D. L.
10. Wiggin, T. H. ; Enger, M. L. ; & Schlick, Deflections of Timber-Strutted Cor-
W. J. A Proposed New Method for rugated-Metal Pipe Culverts Under
Determining Barrel Thicknesses of Earth Fills. Bui. 102, Highway Re-
Cast-iron Pipe. Jour. A WWA, 31:841 search Board; Pub. 350, National
(May 1939). Academy of Sciences-National Re-
11. Burmister, D. M. The Importance of search Council, Washington, D.C.
Natural Controlling Conditions Upon (1955).

Earth Loads on Steel Pipe

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Earth Loads on Steel Pipe

Uploaded by

Copyright:

Available Formats

CHAPTER 8: Earth Loads on Steel Pipe

Earth Loads on Steel Pipe

characteristics can 8.3. Rigid Versus

Fig. 8.1. Effect of Earth Load on Rigid and Flexible Pipe

load on rigid and flexible pipe is shownFigure 8.2 shows photographs of

Fig. 8.2. Effect of Settlement on Rigid and Flexible Pipe in Fill

under load. In thisConsequently,

Fig. 8.3. Computation Diagram for Loads

to flexible pipe in a trench, as ex- If, however, the conduit is rela-

Wc = load on the conduit (lb/

8.4, the trench fill is shown divided

8.5. Wall Thickness to Resist Earth

determines the outward thrust of the works authorities do not ap-

soils or any specially determined 80 •

values in a given case.

8.6. Design Factors in Spongier

fill tamped to horizontal di-

length (lb/lin in. of pipe) e

E = modulus of elasticity of pipe

/ = moment of inertia of cross 20 30 40 50 60 70 80 90 100

including both live 8.6.1.

8.1 (lb/lin ft of pipe) 10

Wl = live load on pipe as calcu-

The Spangler formula may satisfy

ured by the single factor e, the modu-

heights of 10 ft and less.

e to decrease as the pipe diameter in- depth and where ordinary

under cover of 10 ft or less, for either

Deflection of Steel Pipe Under External Load in Trenches*

m Wall Thickness- in.

Fraction Decimal Tru°k Tr¿ck Tru°k Tr¿ck

* Trench width equals outside diameter of pip

For steel water pipe, permitting 2 Design load, W. Experience

lation. The gradual increase in de-

apply the lag factorzontal

in a given instance to avoid exceeding

tory test and by the

Fig. 8.10. Soil Settlement Under Load

ГАе diagram pertains to set

final closing of the | in.

Fig. 8.12. Equivalent Earth Column

The diagram shows th

Case 1 Case 2 Case 3

''Щ+ "wv+ wv+

Fig. 8.13. Cases of Conduits and Loadings

As explained in Sec. 8.13, wv represents the total vertical unit load o

for illustrative purposes becausefluenced bythe

enough to occupy the the space

8.11. Definition of Terms

b. Ring aid - the horizontal thrust

400 '''' 'X ^X V -'b

5 = stress ■l!i«F 200V^S

d. Transfer load - Fig.

Case 3. In Case 3 (Fig. 8.13c),

given in Fig. 8.15. When a trench is ba

Fig. 8.16. Instantaneous Radiuses

8.16. Length of Equivalent Earth

TABLE 8.3 L = 6.5(0.383)D

8.17. Passive Deformation

Pipe Deflection On the basis that the deformation or

*To nearest 0.05. in which F equals

When pipe is formed, strutted, bleor