Professional Documents
Culture Documents
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
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
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
1046 committee report Jour. AW WA
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
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1047
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
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
1048 committee report Jour. AW WA
0Il
0 1 2 .3 4 5 6 7 8 9 10 11 12 13 14 15
H_
Bd
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
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1049
in which :
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
1050 COMMITTEE report Jour. AW WA
,-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.
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 ,
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1051
90
-° 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
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
1052 COMMITTEE REPORT lom. AW WA
i :~-Ziz:::
(ft).
20 30 40 50 60 70 80 90 100
pipe 60 in. and smaller under fill e
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1053
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
1054 COMMITTEE REPORT Jour. AW WA
TABLE 8.2
12 3 4 5 6 7 8
d я Deflection - in,
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1055
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) .
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
1056 committee REPORT Jour. AW WA
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1057
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
1058 committee report . Jour. AW WA
Jr
A A A A
AD A
43_L ;
Plane H
_1
plr 's^ i Л
I*
TIT,Point
Fadeout -^
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
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1059
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
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
1060 COMMITTEE REPORT Jour. AW WA
^ о '^
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1061
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-
''wv wv wv
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
1062 COMMITTEE REPORT Jour. AW WA
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).
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1063
2St = pD
£00Л'УЛУ
or:
in which :
(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.
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
1064 COMMITTEE REPORT Jour. AW WA
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1065
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) ^
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.
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
1066 COMMITTEE REPORT Jour. AW WA
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1067
4^^^ Line II
i|r At
^Л
ДЛ'' и'1 glslslslailslslsls
n' / ^ i ö o d do dodo
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
1068 COMMITTEE report Jour. AW WA
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
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1069
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
1070 COMMITTEE REPORT Jour. AW WA
.ил I I
8.22. Triaxial Shear Tests of Soils
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1071
200i
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
1072 COMMITTEE REPORT Jour. AW WA
180
1
140
_ Penetration Resistance
formulas in rigid construction is likely
to contend that permanent deflection
Before Loading- psi /
=- 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
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1073
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
+0 +1 +2 +3 +4 +5 +6 +7 +8 +9
Height of
Cover, H
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
1074 committee report Jour. AW WA
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1075
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
1076 COMMITTEE REPORT Jour.AWWA
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1077
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
1078 committee report Jour. AW WA
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
Aug. 1961 STEEL PIPE DESIGN AND INSTALLATION 1079
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
1080 COMMITTEE REPORT Jour. AW WA
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