Stp163-Eb 31066

Copyright by ASTM Int'l (all rights reserved); Sun Jan 11 13:11:58 EST 2015
Downloaded/printed by
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further reproductions au
SYMPOSIUM ON
PERMEABILITY OF SOILS
PRESENTED AT THE
FIFTY-SEVENTH ANNUAL MEETING
AMERICAN SOCIETY FOR TESTING MATERIALS
Chicago, 111., June 15, 1954
Reg. U. S. Pat. Off.
ASTM Special Technical Publication No. 763
Published by the
1916 Race St., Philadelphia 3, Pa.
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further re
COPYRIGHT, 1955
BY THE
Copyright by ASTM Int'l (all rights reserved);

Printed Sun Jan 11 13:11:58
in Baltimore, Md. EST 2015
Downloaded/printed by April, 1955
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further reproductions auth
FOREWORD
The papers and discussions in this Symposium on Permeability of Soils

were presented at the Eleventh and Seventeenth Sessions of the Fifty-
seventh Annual Meeting of the Society in Chicago, 111., on June 15, 1954.
The Symposium was under the sponsorship of Subcommittee R-4 on Physical
Properties, under the chairmanship of Mr. Edward S. Barber, Civil En-
gineer, Arlington, Va., of ASTM Committee D-18 on Soils for Engineer-
ing Purposes. Mr. A. W. Johnson, Soils and Foundation Engineer, High-
way Research Council, Washington, D. C. acted as Chairman for the
Eleventh Session; while Mr. Harold Allen, Chief, Nonbitummous Section,
Bureau of Public Roads, U. S. Dept. of Commerce, Washington, D. C. pre-
sided over the Seventeenth Session.
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further reprod
NOTE.—The Society is not responsible, as a body, for the statements
and opinions advanced in this publication.
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No furt
CONTENTS
PAGE
Introduction—Edward S. Barber 1
Principles of Permeability Testing of Soils—Donald M. Burmister 3
Discussion 21
Water Movement Through Porous Hydrophilic Systems Under Capillary, Electric
and Thermal Potentials—Hans F. Winterkorn 27
Discussion 36
A.Low-Head Permeameter for Testing Granular Materials—E. G. Yemington 37
Permeability Test for Sands—T. Y. Chu, D. T. Davidson, and A. E. Wickstrom 43
The Permeability of Compacted Fine-Grained Soils—T. W. Lambe 56
The Permeability and Settlement of Laboratory Specimens of Sand and Sand-Gravel
Mixtures—Chester W. Jones 68
Discussion 79
Measurement of the Hydraulic Conductivity of Soil In Place—Don Kirkham 80
Measurement of Permeabilities in Ground-Water Investigations—W. O. Smith and
R. W. Stallman 98
Discussion 115
Determination of Permeability of Granular Soil by Air Subjected to a Decreasing
Pressure Differential—Arthur S. Weaver 123
Selected References on Permeability—A. I. Johnson 131
Copyright by ASTM
S R KUMAR
This page intentionally left blank
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further reproduc
SYMPOSIUM ON PERMEABILITY OF SOILS
INTRODUCTION
BY EDWARD S. BARBER1
In principle, determination of the permeability tests are compared, de-

permeability of soils is quite simple. scribed, and evaluated by formulas.
However, due to natural variations of Correlations are presented between per-
material in place, it is often difficult to meability and density and gradation of
relate tests on small samples to larger granular materials. A new sampler and
masses. In sampling soils it is hard to a device for testing under small gradients
prevent disturbing the moisture or are described. The importance of relating
density or particularly the structure of tests to field conditions is stressed.
the soil. Changes in the air or chemical or The test value is expressed as length
organic content of the permeating fluid divided by time in a variety of units, but
can cause large differences. Migration it is generally called coefficient of permea-
of particles may occur both in the field bility, although hydraulic conductivity
and laboratory. While some variables is suggested as being more consistent
can be arbitrarily controlled or elimi- with other fields such as electrical and
nated in the laboratory, it is often neces- thermal conductivity. While the variety
sary to consider them in field applica- of field situations seems to preclude a
tions. single standard test method, it should
The Symposium includes papers dis- be possible to increase the consistency
cussing the importance, evaluation, and
of results by recommendation of pre-
control of most of these factors. Field
1
ferred practices.
Civil Engineer, Arlington, Va.; chairman Previous work is covered in a list of
of Subcommittee R-4 of Committee D-18 on
Soils for Engineering Purposes. selected references.
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) 1 pursuant to License Agreement. No further reproductions authorized.
This page intentionally left blank
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further reproduc
PRINCIPLES OF PERMEABILITY TESTING OF SOILS
BY DONALD M. BusMisxER1
The permeability of soils is a most im- CONTROLLING ENVIRONMENTAL AND

portant physical property since some of IMPOSED CONDITIONS
the major problems of soil and founda-
tion engineering have to do with the A basic fact in soil and foundation en-
recognition, evaluation, and proper gineering is the inherently variable and
handling of drainage problems encoun- complex character and behavior of soils
tered in the design and construction of and the dominating influences of en-
structures. These problems include drain- vironmental and imposed conditions
age of highways and airports, seepage upon the responses of soils. Soil engineers
through earth dams, uplift pressures should realize that they are actually
beneath concrete dams and structures dealing with a very unconventional and
below ground water level, unwatering of in many respects a very unusual kind of
excavated sites to permit construction engineering material. Hence in contrast
in the "dry," seepage pressures causing to the essential uniformity of the com-
earth slides and failures of retaining mon structural materials, the predict-
walls, etc. In all of these, the permeability of their behavior within the range
ability characteristics of soils have a of common working stresses, and the
controlling influence on the effective marked constancy of their properties for
strength properties of the soils and on all common conditions of usages unaf-
their responses under stress, and hence fected by external conditions, soils should
on stability conditions. Drainable soils not be expected to follow such simple
will act essentially as "open systems" conceptions and patterns of behavior.
with free drainage and fully effective These facts may be summarized in two
basic concepts (l) :2
shearing strength, whereas soils of low
permeability may act as "closed sys-
tems" under rapid application of stress, 1. The character and responses of soils in
any particular situation not only are prede-
with the development of pore pressures termined by and are a part of the environ-
and reduction in shearing strength. mental conditions prevailing in that situ-
The determination of the permeability ation, but they are always markedly
of soils is therefore a most important conditioned and modified in direct response
aspect of soil testing. The purpose of this to inevitable changes in those prevailing
paper is to formulate into a more com- conditions by the new controlling conditions
plete form certain attitudes, concepts, imposed by the structure itself.
and principles of a fundamental and 2. In each situation, as a particularized
comprehensive approach in permeability case, the character and potential behavior
testing of soils and to increase the ade- of the soils (soil tests) must be considered
directly in relation to the specific conditions
quacy, reliability, and practical value of
permeability data. 8 The boldface numbers in parentheses refer
1
Professor
Copyrightof Civil Int'l
by ASTM Engineering, Columbia
(all rights reserved); to 13:11:58
Sun Jan 11 the listEST
of references
2015 appended to this paper,
University, New York, by
Downloaded/printed N. Y. see p. 19.
3
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further reproductions authorized.
4 SYMPOSIUM ON PERMEABILITY OF SOILS
that control in the environment: whether and varied character of environmental

different conditions are inherent in different conditions to be expected in natural
locations in the natural situation, or whether situations and of the changes in condi-
the imposed conditions are different. tions imposed by construction of struc-
The drainage characteristics of residual tures, which may govern drainage condi-
soil deposits are determined by the char- tions, a most important and fundamental
acter and structure of the soils in the step in soil investigations is an accurate
different horizons of the weathering soil and complete visualization and appraisal
profile. These soil profiles have regular of each situation with regard to the
patterns and sequences of horizons with nature, relative dominance, and favor-
their character and thickness varying able or adverse aspects of the controlling
with the type and extent of weathering. conditions. Such an appraisal requires
The character and structure of the more detailed and adequate exploratory
different soil layers forming the soil pro- information on subsurface soil conditions,
files of sedimentary deposits are deter- and more adequate and reliable soil test
mined by the constantly varying charac- data than commonly provided and con-
ter of geological processes of deposition sidered necessary.
operating to form soil deposits. CONTROLLED TEST METHODS AS APPLIED
But of equal or even greater im- TO PERMEABILITY TESTING
portance in drainage problems is the
dominating influences of the changed At present little is really known about
conditions imposed by the construction the fundamental properties of soils and
of structures; these may be favorable, of their responses in the original and
neutral, or actually detrimental in char- changed environment. The difficulty is
acter with regard to the behavior and created by the inherently complex and
responses of soils. The type and charac- variable character and responses of soils
ter of construction may materially and by the fact of removal of soil samples
change the pattern of subsurface drain- from their natural environment. Parallel-
age conditions. The lack of or the type ing the two concepts stated previously
of drainage system, its method of in- regarding the conditioning of the charac-
stallation, and its effectiveness may play ter and behavior of soils by their natural
governing roles. The time-delay aspects and changed environment, there is a
of drainage phenomena may be of major third basic concept regarding the condi-
importance in construction. These con- tioning influences imposed during soil
ditions should be recognized, investi- testing (3, p. 12):
gated, evaluated, and taken into account
3. During the observations and measure-
hi design and construction. It is there- ments of properties and responses of soils hi
fore important to give consideration to soil tests, the observed properties and re-
(1) the fundamental behavior charac- sponses themselves are conditioned, modi-
teristics of soils, in which permeability fied, and changed by the very procedures
plays a dominating r61e, and (2) certain and test conditions used to an important but
controlling conditions which govern unknown degree, unknown because there are
drainage conditions. These conditions no absolute response bases for references.
Thus the observed and measured properties
and behavior characteristics are outlined and responses are not necessarily those hav-
and discussed in detail in a previous ing any significant and direct relations to
paper (l, pp. 251-252). the actual properties and responses of the
In view of the large
number, range, soils under the controlling conditions pre-
BURMISTER ON PERMEABILITY TESTING OF SOILS 5
vailing or imposed in the natural situation. fied average test conditions, their single
Therefore potential behavior only can be fixed routine procedures for each method,
learned from soil tests. and their emphasis on ease of applica-
Thus there are major difficulties in- tion and on reproducibility of test data
volved even in the so-called simpler soil by different individuals.
tests. In view of these three concepts, it The application of controlled test
should be realized that soil is a material methods involves essentially introducing
which does not readily permit transla- representative Test Conditions, which
tion of responses made under one set of will properly condition each soil specimen
relatively arbitrary and even artificial by restoring, as nearly as possible, the
conditions in a standard soil test into original conditions that control in the
reliable predictions of responses under natural environment, and by establish-
quite a different set of actual conditions ing thereafter in sequence the controlling
in the field. Furthermore, the field condi- conditions to be imposed by the con-
tions are not necessarily constant but struction of structures, in accordance
may vary markedly from place to place with the following three basic principles
for the same type of soil and with the of controlled test methods:
seasons. The real soil testing problem is 1. To appraise and to evaluate as
not to formulate a so-called average or completely as possible the real nature
norm of test conditions, but rather and degree of control of the environ-
reasonably to bracket the actual possi- mental conditions in the natural situa-
bilities in each situation. Therefore, as a tion and of the new conditions imposed
basic and realistic approach, the proper- in sequence by the construction of struc-
ties and responses of soils should be de- tures.
termined by soil tests performed under 2. To translate this information into
conditions essentially equivalent to those appropriate test conditions which are
to be expected under actual field condi- representative of and valid for a particu-
tions, both as to probable character and lar situation, and which can be adjusted
bracketing range of conditions. This is and brought into significant agreement
the essential nature and purpose of the with and definitely made to fit the prob-
controlled test methods in previous able actual environmental and imposed
works by the author (2, pp. 83-89; 3, pp. conditions, as the responses of the soils
11-14), which state broad flexible princi- become evident and are disclosed during
ples of soil testing and describe general the test.
procedures, but which permit a wide ,3. To apply these test conditions by
latitude for making adjustments in the techniques and control in sequence dur-
soil that are subject to control at the ing the conduct of a soil test, first, to
discretion of the soil engineer, without determine accurately and completely the
restricting methods to specific fixed pro- representative character and responses
cedures and types of apparatus. This ap- definitely impressed upon the soils by the
proach represents a conception of a new natural environmental conditions; and,
kind of standard soil test as a valid basis second, to determine the responses of the
for soil engineering investigations. Its soils that would have direct, reliable, and
principal objectives are reliably to repro- valid applications in predicting the field
duce responses representative of the responses and performances of the soils
natural situation. This is in marked con- under actual conditions imposed by the
trast to the usual conceptions of standard construction of structures.
soil test methods withbytheir oversimpli-
Downloaded/printed Soil testing is such an essential part of
soil investigations for engineering works environment, of the structure itself, and
that every soil test should be treated es- of soil test conditions upon soil responses
sentially as a work of discovery under should establish the fact that such sim-
careful control. In view of the importance plicity and essential identity of action
and real character of soil testing, it and responses can seldom, if ever, be
should be considered design in the true expected in the case of soils, whether in
sense of the word, because it involves the natural situation or in a standard soil
judgment in the practical applications of test made in accordance with usual con-
these three principles of soil testing, and ceptions. In this era of fully demon-
because adequacy and reliability of the strated value and use of basic scientific
results of soil tests are not just simple developments, scientific caution, common
matters of routine applications of stand- sense, and creative engineering imagina-
ard test methods. Controlled test meth- tion should reject the idea of simple
ods are therefore considered fundamental answers to admittedly complex and
and realistic in soil and foundation en- difficult questions and problems. In soil
gineering because they particularize each engineering adequacy should not be
situation (1) to disclose and to evaluate permitted to become a fixed idea, but
the known and unknown conditions that rather it should be constantly and con-
control; (2) to provide specific answers sistently revised upward to keep pace
that are representative of and directly with increases hi knowledge and experi-
applicable to each situation; and (3) to ence and to stimulate further advance.
obtain the highest degree of agreement
between the predictions of behavior and Permeability Flow:
responses of soils and the actual ob- In order properly to conduct perme-
served soil phenomena. ability tests, and to interpret and to use
There may be some difficulty in under- test data, consideration should be given
standing the essential need for this to the nature of hydraulic phenomena.
apparently radical departure from the This paper is concerned with perme-
usual conceptions of standard soil test- ability flow, which takes place primarily
ing. Many engineers engaged in soil work through saturated soils under gravita-
have had a structural background. As a tional forces or under a pumping head
consequence they may quite naturally with the water everywhere in the region
be inclined to believe that analogous of permeability flow under a positive
simple conceptions and treatments of hydrostatic pressure. In certain im-
soil phenomena and soil testing fit the portant cases such flow can take place
facts and are acceptable as a valid ap- through partially saturated soils. Some-
proach. To a very large degree, attitudes times under high-vacuum well-point
and conceptions predetermine judgments pumping in the immediate proximity of
and practice. The almost universal ac- the well-points, there may be a region
ceptance and use at present of standard where flow occurs under a negative pres-
soil tests are really predicated on the sure or suction, but with the voids of the
premise and fallacy that there is a soil flowing full of water. Capillary flow
simplicity and essential identity of of water, in contrast, takes place under
action and responses of soils as norms in capillary forces primarily through a
a standard soil test and under actual field continuous interconnected system of thin
conditions. A consideration of the fore- capillary moisture films, principally at
going three basic concepts concerning the the grain contacts, with the water
important conditioning influences of the everywhere in a state of capillary ten-
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No furthe
sion, the distinguishing characteristic of long enough, the entrapped air will be
capillary flow. gradually dissolved in the water. Only
There are two important categories of the steady state of flow with the voids
flow, the one of principal importance be- full of water can be analyzed with any
ing flow of water below the permanent degree of exactness by present concepts
ground-water level with the voids full of and working hypotheses.
water (100 per cent saturated). The other The principal realms of flow of water
now becoming of importance in soil en- through soils are laminar flow or stream
gineering is the flow of water below a line flow, and turbulent flow. In laminar
temporary elevated ground-water level flow, viscous forces shape the character
or free water surface with different de- of flow with velocity proportional to the
grees of saturation and air-clogging of hydraulic gradient. Turbulence is initi-
the soil voids. Important examples of this ated in soils at considerably lower
latter category are: (a) flow of water velocities than usually recognized in the
;TABLE i—REALM OF VALIDITY FOR DARCY FLOW OF WATER IN GRANULAR
SOILS.
Sieve. 3 in. 1 in. I in. No. 10 No. 30 No. 60 No. 200
Dio , mm 76.2 25.4 9.52 2.0 0.59 0.25 0.074 0.02
GRAVEL SAND SILT

coarse medium fine coarse medium fine coarse fine
Realm of flow of water Practically always tur- Darcy laminar flow only Always laminar flow for
bulent flow. for H/L less than the range of H/L
about 0.2 to 0.3 for found in nature.
the loose state and
0.3 to 0.5 for the
dense state.
through river banks or levees caused by form of eddies and vortices in the larger
rising flood stages of a river; (6) flow of void spaces, due to expansion, contrac-
water through earth dams caused by tion, and change of direction effects.
rising water level hi a reservoir; (c) flow This turbulence results in increased re-
of water through gravel drains or base sistance to flow and larger energy losses.
courses beneath pavements during peri- Based upon these considerations, only
ods of large infiltration of rainfall; and one type of flow, designated the Darcy
(d) the rate of infiltration of ram water type of flow or Darcy flow, is stable in
downward into soil. In these cases per- character. It is described and strictly
meability flow applies some distance limited by four basic conditions: (a) the/
back from the advancing front of capil- laminar realm, (ft) the steady state, (c)l
lary flow, where pressure in the water flow with the soil voids 100 per cent\
has become positive hydrostatic. saturated (no compressible air present),
In the initial stages of these categories, and (d) flow with the continuity condi-
the flow is in a transient state and is tions and basic equation of flow satisfied
time-dependent. After a period of tune, in a soil mass in which no volume changes
depending on conditions, the transient occur (consolidation) during or as a re-
state of flow approaches the steady state sult of flow. Permeability experiments
of flow with the final establishment of an (5) have established the realm of validity
equilibrium
Copyright byfree
ASTMwater surface.
Int'l (all rights IfSun
reserved); flow
Jan 11 of EST
in13:11:58 the2015
Darcy flow (Table I).
the partially saturated
Downloaded/printed by condition persists Darcy, as a result of his basic work on
8 SYMPOSIUM ON PERMEABILITY OF SOULS
the mechanics of flow of water through predominance of the coarse, medium or

porous media in 1856, first stated the fine fractions, as defined in Table I; and
basic law of flow, which is absolutely gen- (c) the predominating grain shapes and
eral in its application within its realm of surface characteristics. Since every recog-
validity: nizable and identifiable soil characteristic
is certain to play an important and even
dominating role in soil behavior and
responses, as a basic concept it is im-
The velocity, v, is an average over-all portant to identify accurately and
velocity computed on the basis of the completely the soils under consideration.
quantity of flow, Q, in a time, /, and of The finest soil component and fraction,
the entire gross cross-sectional area of in general, appear to dominate perme-
the soil column, A. The hydraulic ability phenomena, because they tend to
gradient, H/L, is expressed as the ratio determine the sizes of the soil channels
of the head of water, H, causing flow by their void clogging effects. Further-
to the length of the soil column, L, in more, it is important to give each soil a
which the head is lost. The Darcy co- precise, significant, and distinguishing
efficient of permeability, K0, is there- soil name (6, pp. 7-24) that will convey
fore an over-all value, which provides an accurate information on those aspects of
adequate, reliable, and stable basis for soil character in permeability investiga-
reference and for comparison of different tions for present purposes as well as for
conditions of flow. All other conditions future comparisons and correlations of
of flow and their corresponding coeffi- permeability phenomena.
cients of permeabilities, designated by The character of the soil structure in
appropriate subscripts, are from the very natural deposits and compacted fills is
nature of the phenomena inherently first of all identified by its degree of com-
unstable in character. pactness on a relative density basis (5,
pp. 1249-1255). As a fundamental and
Dominating Influences of Soil Material practically useful concept in soil mechan-
and Soil Structure: ics, relative density describes the signifi-
The identification of the soil material cant state of compactness of the grain
(6, pp. 9-16) and the soil structure (5, structure of granular soils. Furthermore,
pp. 1249-1255) are presented elsewhere. relative density relationships provide sig-
This discussion is limited to a considera- nificant unifying bases for interpretations,
tion of the influences of character of gran- evaluations, and practical applications
ular soils that are relatively incompres- (5, pp. 1255-1268). Such relationships
sible under flow of water, as required give a clearer insight into and a better
under validity condition (</), explained understanding of soil behavior and re-
above, for the Darcy flow. For all finer sponses, as controlled by individual char-
grained soils, there is an important time acteristics of soils and by the conditions
delay in reaching a stable, consolidated inherent in natural situations. In addition,
soil structure. This makes the testing where no rigorous mathematical treat-
problem for such soils more difficult. The ment is possible because of the nature of
identifiable characteristics of the granular soil, these relationships permit a more
soil material that govern permeability complete statement to be made of the
flow are: (a) its composition and propor- physical laws governing all granular soil
tions of the gravel,
Copyright sand,Int'l
by ASTM and(allsilt com-
rights
phenomena by means of comprehensive
reserved); Sun Jan 11 13:11:58 EST 2015
ponents; (6) the gradation
Downloaded/printed by of these com- graphical presentations with relative
ponents SRexpressed
KUMARin terms of the relative
KONATHALA (Indian Instdensity
of Tech as the common
Chennai) pursuant unifying
to Licensebasic
Agreement. No
argument.. Relative density must be The relative density diagram given in
defined on the basis of the maximum Fig. 1 can be used conveniently for de-
and minimum densities (5, pp. 1250- termining relative densities and unit
1255), as definite physical limits for each weights. Experiments and correlations
granular soil and as the basic references (5, pp. 1250,1251) have shown that it is
for the 100 per cent andO per cent relative only in exceptional and unusual cases
that natural deposits of granular soils
have been found in a more dense state
than the laboratory maximum density
reference, or in a more loose state than
the laboratory minimum density refer-
ence. The bulked densities of moist soils
artificially placed in embankments would
then have a minus relative density by Eq
2. The maximum density obtained for
granular soils by vibration methods near
resonant frequencies represents practi-
cally the maximum attainable by any
method. This maximum density, there-
fore, is a stable reference for 100 per cent
relative density.
Studies have shown (s, pp. 1252-1255)
that the different geological processes
tend to form soil deposits in relative den-
sity states characteristic of the soil ma-
terial deposited. There is a significant and
fundamental relative density concept,
FIG. 1.—Relative Density Diagram.
namely, the maximum and minimum
density states, because they do represent
densities, .respectively, in the denning practical physical limits for each soil,
equation: also fix the limits of soil behavior and
responses. All natural soil phenomena for
practical engineering purposes will fall
Relative Density: between these limits.
The character of the soil structure in
natural deposits is determined by the
nature of the geological processes of soil
formation and by the character of the
soils deposited. Thin layering and the
degree of anisotropy of the layers govern
where: the permeability properties of the layers
e = the voids ratio of the soil, hi detail, particularly when visible.
W = the corresponding unit dry weight,
L -= minimum (nonbulked) density ref- PERMEABILITY - RELATIVE DENSITY
erence, RELATIONS
D = the maximum density reference, In order to disclose the nature and de-
and gree of control of the soil material and the
N = the density
Copyright by of the Int'l
ASTM soil (all
in the
rightsnat-
soil structure upon permeability phenom-
ural or compactedbystate.
Downloaded/printed
ena, investigations and research were
carried out in 1943 and 1948 (5, 7). Test relative densities and at an intermediate
conditions were carefully formulated for relative density in the region of 70 per cent,
the research on permeability to satisfy using appropriate techniques to obtain
the three principles of soil testing and the reproducibility and uniformity of conditions
four basic conditions of validity of the (8, pp. 111-113).
Darcy type of flow, explained earlier in 3. To produce at each relative density an
FIG. 2.—Distinguishing Characteristics of Grain Size Curves.

Fineness, range of grain sizes, and type.
this paper. These essential test conditions isotropic, homogeneous soil mass with prac-
are listed as follows: tically no segregation effects or nonvisible
anisotropy due to placing and compacting
1. To bracket as to character and range successive layers, and in order to provide the
of soil material from "coarse SAND" to basic stable permeability references, just
"coarse SILT," synthetic granular soils were enough moisture was mixed into a test speci-
made up to produce by regular steps definite men (0.5 to 1.0 per cent by weight) so that
grain size distributions commonly encoun- the soil would not flow freely from a funnel
tered in practice, as shown in Fig. 2. with a $-in. spout for spreading thin layers.
2. To bracket
Copyright by the full Int'l
ASTM range(allofrights struc- Then
soil reserved); by11additional
Sun Jan mixing
13:11:58 EST 2015and testing with
tests were made atby0 and 100 per cent the funnel, the soil was partially dried and
ture, Downloaded/printed
BlTRMISTER ON PERMEABILITY TESTING OF SOILS 11
brought to the correct moisture state, so that mentioned above obtain for the coarser soils
it would just flow freely from the funnel to and the higher values for the finer soils.
form successive $-in. layers for the loose 8. To insure a steady state of flow for
state. The specimen was covered to prevent each new head and plotting point in accord-
further loss of moisture while compacting ance with validity condition (6), the quan-
each layer in the higher relative density tity of flow was measured only after a stable
states. head condition in the manometers was at-
tained.
NOTE.—This isotropic condition can be
checked at the completion of the permeability NOTE.—In making the setup for the test,
test by evacuating the water out of tie test speci- great care was taken to insure that the head
men. Any light and dark alternating streaks manometers, tubes, and connections were free
are evidence of segregation of fines and of aniso- of air and were operating satisfactorily.
tropy, the dark streaks being the finer segre-
gated fractions, which have a larger water-hold- It should be evident from these test
ing capacity. True natural anisotropy cannot be conditions that the common falling head
successfully duplicated in the laboratory.
type of permeability test is inherently
4. To hold the initial relative density state unsuited for permeability testing of
without volume change during saturation granular soils, because under a high
of the specimen and during the permeability starting head test condition No. 7 is not
test to satisfy condition (d) for the validity satisfied. Even in the constant head type
of the Darcy type of flow, a light spring of test in the laminar region of flow, it is
pressure of about 2 psi over the area of the not generally possible to go back and
specimen was permanently applied through
pick up a consistent point under a lower
a suitable screen device to the top of the soil
specimen prior to measuring the initial height head. This is evidence that there has
of specimen and attaching the cap to the been some disturbance effects to the
permeability device. soil structure, even when test conditions
5. To saturate the specimen completely No. 4 is reasonably satisfied.
in accordance with validity condition (c), the Typical permeability test curves ob-
permeability device was evacuated under full tained by constant head testing are
attainable vacuum (28 in. of mercury or bet- shown in Fig. 3; these define perme-
ter, if possible) for 10 to 15 min to remove ability flow conditions between the
the air. This evacuation was followed by limiting maximum and minimum density
slow saturation of the specimen from the
states. The region of the Darcy flow is
bottom upward under this vacuum.
6. To insure against air-clogging during clearly defined in Fig. 3, in which the
the test, de-aired water was used, which was coefficient of permeability, K0 , for each
obtained from a special filter tank. relative density state is a stable con-
7. To insure laminar flow conditions in stant determined from the linear portion
accordance with validity condition (a) and of the curves for low values of H/L
Table I, a sufficient number of points were below the critical value defined by the
obtained under constant head testing by curve separating the laminar flow region
varying the head in small steps below a from the turbulent flow conditions. The
gradient H/L of about 0.2 to 0.3 for the research of Kane (5, 7) has established
loose state and about 0.3 to 0.5 for the dense the fact that for the range of these soils
state (J-cm increments) in order to define
the Q/At versus H/L permeability curve in the region of validity for laminar flow
the Darcy region of laminar flow. Thereafter conditions is limited to hydraulic gradi-
the head was increased in larger steps to de- ents below 0.2 to 0.3 for the loose state
fine the permeability curve in the region of and below 0.3 to 0.5 for the dense state,
turbulent flow. The lower values of H/L the lower values being for the coarser
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No f
FIG. 3.—Determination
Copyright ofrights
by ASTM Int'l (all Coefficient of Permeability
reserved); Bracketing
Sun Jan 11 13:11:58 ESTthe
2015 FIG. 4.—Determination of Coefficient of Permeability Bracketing
Relative Density Limits.
Downloaded/printed by the Relative Density Limits Under Air-Clogging Condition After Re-
saturation.
soils. These values are much lower than 9. The specimens were drained by evacua-
commonly realized. tion from the bottom for 15 min.
It is also apparent in the region of 10. The specimens were then resaturated
turbulent flow in Fig. 3, where the per- from the top down under aerated normal
meability curves depart from the linear atmospheric conditions at low heads.
relations of laminar flow, that a coeffi- Under these specific conditions, the
cient of permeability is definitely not a permeabilities hi the region of laminar
constant but is hydraulic gradient-de-; flow were found to be from 25 to 50 per
pendent. Therefore in analyses of natural cent of those for the completely sat-
situations, where the natural hydraulic urated state of the Darcy type of flow.
gradient in any localized region exceeds They tended toward the 25 per cent
the above critical values, the only prac- value for the finer grained soils, as shown
tical solution is obtained by the direct in Fig. 4, using as references the Darcy
use of the experimental permeability flow permeability curves for the maxi-
curves in the region of turbulent flow. mum and minimum densities. Investiga-
Such regions can be defined by Flow Net tions showed that the permeability char-
analyses. The direction and magnitude acteristics under conditions of aeration
of the errors involved by the use of a were markedly dependent upon the
constant Darcy coefficient of perme- history and sequences of events in ob-
ability in the turbulent region is de- taining partial saturation of the soil,
pendent upon which quantity is fixed namely, (1) the initial moisture state
and controls in the situation. For exam- between the limits of air dry and almost
ple, for the case a-b in Fig. 3, if H/L complete saturation; (2) the rate of
at a is fixed, then the quantity of flow saturation; (3) the direction of satura-
obtained at b' under turbulent condition flow as affecting displacement of air
;ions would be greatly overestimated. and whether capillary "pull" is with or
On the other hand, for case c-d, if the against gravity; (4) the soil material and
quantity desired at c is fixed as from a the relative density state, as affecting
pumped well, then the required H/L at the sizes of the void spaces; and (5)
d' to produce this quantity under probably other unsuspected aspects of
turbulent conditions would be greatly partial saturation.
underestimated, due to large energy The degree of saturation now in-
losses. These facts would have important creases under increasing heads due to
implications in estimating seepage quan- the compressibility of the air in the void
tities and forces hi pumping and stability spaces. It is evident that the perme-
problems if regions of turbulent flow are ability testing of soils under such condi-
present. tions is complex and difficult. In order to
In order to investigate the nature of obtain comparable, representative, and
the influences and the kind and degree useful test results, the program of testing
of departure from the Darcy type of would have to be well formulated with
flow under conditions of partial satura- each series of tests made under a specific
tion with air-clogging of the voids, the set of test conditions, which bracket cer-
following two additional test conditions tain limits only of the phenomena. Many
were applied at the completion of each practical problems in soil engineering
regular permeability test to simulate which involve seepage flow under such
conditions of infiltration of rain water conditions, as noted previously, are
into moist becoming
soil: by ASTM Int'l (all rights reserved);
Copyright of importance.
Sun Jan In order
11 13:11:58 EST 2015 ade-
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. N
14 SYMPOSIUM CN PERMEABILITY OF SOILS
quately to analyze such problems, a great in each natural situation that are repre-
deal more will have to be learned about sentative and directly applicable with
the real nature of such phenomena, and regard to the probable character and
the range of conditions that control in range of conditions that may control.
FIG. 5.—Permeability - Relative Density Relationships.

Dm = Hazen's effective size.
C'r = range of grain size defined by mean slope.
Type = letter designation of symmetric or asymmetric shape of grain size curves.
NOTE.—See Fig. 2.
natural situations. Because the perme- The results of such investigations prob-
ability characteristics under such condi- ably could be best established and
tions are not constant but are essentially presented in terms of the kind, degree,
time-, gradient-, and per cent satura- and probable range of departures from
tion - dependent, test conditions would the Darcy coefficients of permeability as
have to be formulated carefully to a stable basis for reference for practical
bracket specific limits and possibilities purposes in such soil engineering work.
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further reproductio
Basic Patterns of Permeapility Relations: range of relative density from the loose
On the basis of two series of investiga- to the intermediate to the dense state.
tions in 1943 and 1948 for the Darcy type The Darcy coefficients of permeability
of flow, two basic patterns of perme- are plotted vertically on a logarithmic
ability relations were established (Figs. scale to cover the full range of values
5 and 6). These two basic patterns of against relative density on an arithmetic
permeability - relative density relations scale to form a consistent pattern of
were first presented in 1948 as one of five curves with fineness of the soil, as indi-
examples to illustrate the importance cated by the values of Dw noted in the
FIG. 6.—Relations Between Permeability and Hazen's Effective Size, Z?ie.

Coefficient of permeability reduced to basis of 40 per cent relative density by Fig. 5.
and practical uses of relative density in right margin of Fig. 5. The basic pattern
soil mechanics, and were discussed very of permeability-/?™ relations for a con-
briefly from that viewpoint (5, pp. 1263- stant relative density of 40 per cent was
1265). They are discussed here from the obtained by interpolation from Fig. 5 and
point of view of permeability phenomena. is given in Fig. 6. The logarithm of
The basic pattern of permeability - rela- permeability is plotted vertically against
tive density relations is given in Fig. 5, the logarithm of DIO horizontally to de-
the heavy-line curves bracketing the com- fine the reference lines of soil character
mon range of soil material from "coarse and the general drainage characteristics
SAND" to "coarse SILT" from the re- for ratings of soils.
search of Kane in 1948 (7), and the light- In order properly to interpret the
line curves covering a wide range of permeability patterns of Figs. 5 and 6
composite gravel-sand-silt soils from the with regard to the controlling influences
research of 1943.
Copyright by Both
ASTMseries covered
Int'l (all rights the of soil
reserved); character,
Sun Jan three
11 13:11:58 ESTdistinguishing
2015
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No fu
and significant size characteristics of enclosed between the grain size curve
granular soils are defined on the basis of and the mean slope equal and balanced
grain size distribution curves, namely, independently for the upper and lower
fineness, range of grain sizes or mean branches and tails of the gram size curve,
slope of curve, and type of grading or as illustrated in Fig. 2 for four different
characteristic shape of grain size curve. types of grain size distributions. Since
These three size characteristics of soils the vertical intercept is always 100 per
are necessary and sufficient to define cent, the range of grain sizes, C r , may be
grain size distributions of granular soils, conveniently and significantly defined as
and they are entirely independent of the number of "coarse," "medium," and
each other. Studies (6, pp. 18-20) have "fine" fractions in Fig. 2 intercepted on
established the significant fact that grain the horizontal scale between the 100 and
size distributions of soils are not hap- 0 per cent terminal points of the mean
hazard chance phenomena but are de- slope. The mean slope and d have a
termined by and are characteristic of direct correlation for Type-S grain size
each different geological process of soil curves with the "standard deviation"
formation, and that these three charac- used in statistical analyses.
teristics significantly reflect these facts. The type or shape of a grain size curve
Hazen's effective size, D\0, has been is an index of the symmetry or of the
widely used as an index of general fine- kind and degree of asymmetry or "skew"
ness of soils, particularly in permeability in the distribution of grain sizes, which
phenomena, because of the control of this are characteristic of certain geological
fine fraction in its clogging effects in the processes of soil formation. The almost
void of the soil. Actually, however, cor- symmetrical Type-S grain size curve is
relations show that Z>5o (50 per cent characteristic of the distributions so
size) would be more significant in Fig. 6, commonly found in sands and coarse
resulting in a narrower reference band. silts which have been formed by the as-
An index of the range or spread of grain sorting action of sedimentation in flow-
sizes should be representative of the en- ing water or quiet water, by wave action,
tire grain size curves for all types of soil and by wind action. When the gravel
gradations, not only of the bulk of the content of sands exceeds about 10 per
soil material but also of the coarse and cent, the grain size distributions are
fine "tails" of the curve with* regard to markedly asymmetric with a predom-
their relative importance and control of inating "tail" of gravel. The types of
behavior. Hazen's uniformity coefficient common and significant grain size curves
was found to be too restricted and un- are given characteristic letter designa-
representative in character and to be an tions, which are sufficient for most
unsatisfactory basis. The effective range practical purposes of analyses in soil
or spread of grain sizes, designated C r , investigations, as shown in Fig. 2 (b) (5,
however, may be defined on a satisfactory pp. 1266-1267).
basis of the "mean slope" (5, pp. 1266- Certain important and significant facts
1267) of the grain size curve, in accord- regarding the controlling influences of the
ance with common engineering principles. character of the soil material and of soil
The mean slope is readily determined structure are disclosed by the perme-
graphically by using a transparent scale ability - relative density and the perme-
and making the plus and minus areas ability-Die patterns of Figs. 5 and 6 for
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further rep
the Darcy type of flow. First, there is a soils possess the flattest permeability-
generally consistent pattern of decreas- relative density curves, that is, the least
ing permeability with fineness and de- change in permeability between the loose
creasing values of Dw evident in Fig. 5. and dense states, and the curves are al-
This pattern is more clearly defined hi most linear in character. With increase
the permeability-Z>io relations in Fig. 6, in the range of grain sizes to a value of
where permeability values are reduced Cr or 1.7 (spread of four sieve sizes), there
(interpolated from Fig. 5) to a significant is a noticeable increase in the change in
and common 40 per cent relative density permeability with increase in relative
basis, in order to obtain an essential and density from the loose state toward the
consistent basis for comparison and inter- dense state, particularly above 70 per
pretation and for rating soils with regard cent relative density. Furthermore, it is
Jto their significant drainage characteris- clearly evident that the respective heavy-
tics. Otherwise there can be no proper line permeability - relative density curves
basis for comparison. The 40 per cent for Cr of 0.9 and 1.7 are essentially
relative density, which is the dividing parallel to each other for the full range
value between the loose and medium of £>io or fineness of these soils. This indi-
compact states, as noted in Figs. 1 and cates that the change in permeability
5, was chosen as a significant common with increase in relative density is gov-
basis because so many granular soil de- erned principally by the range'of grain
posits possess natural relative densities sizes, as the only variable quantity be-
between about 30 and 50 per cent, the tween the two sets of curves. With larger
lower value being more representative of values of C r , the influences of the range
the finer granular soils and the higher of grain sizes on the light-line perme-
value of the more gravelly soils. ability curves become more pronounced
Second, there is a consistent pattern in over the full range of relative densities,
Fig. 5 of decrease in permeability with particularly for values greater than 70
increase in relative density. As a basic per cent, where there is a marked and
fact of permeability phenomena, it is characteristic curvature downward to-
evident from the pattern of Fig. 5 that ward the 100 per cent relative density.
relative density can provide a unified and This is due to the more pronounced
comprehensive basis for comparing and clogging effects of the finer soil grains
evaluating permeabilities of different in the void spaces of the soil toward the
soils deposited or placed in different de- maximum density state, because with
grees of compactness. increasing range of grain sizes the maxi-
Third—and of equal importance and mum density also increases markedly
significance—this permeability - relative (5, pp. 1266-1267). The general steepen-
density pattern discloses the controlling ing of the permeability curves toward the
influences of the range of grain sizes, C r , minimum density is due to the more
upon the change of permeability with marked particle separating effects of the
increase in relative density, a fact that finer gram sizes on the gram structure in
is made clearly evident by the use of the the loose state with wider ranges of grain
more representative and significant sizes and Cr greater than 4.
definition of this size characteristic. For In the permeability-Z>io relations of
the narrowest range of grain sizes with Fig. 6, the influences of the range of
Cr equal to 0.9 (spread of two sieves sizes grain sizes are also evident in defining
only on the scale at the top of Fig. 2), the two reference bands for Cr of 0,9 and 1.7,
respectively. For the 40 per cent relative ability phenomena, and the character of
density basis of Fig. 6, which is at the the practical problems involved with
upper limit of the loose state, the refer- regard to their relative dominance and
ence band for the wider range of grain control. Such test conditions, including
sizes with CT of 1.7 lies above that of the Nos. 4 to 8 given previously, will yield
narrower range of grain sizes with Cr of permeability data having maximum
0.9. This increase in permeability is due reliability and usefulness.
to the greater particle separating effects Where relatively few undisturbed
of the finer sizes on the grain structure samples can be secured and tested, the
in the loose state as the range of grain test results can be used to test the gen-
sizes increases for soils having the same eral validity of Figs. 5 and 6. Then the
value of Dio. The influences for larger permeability information obtained from
values of CT could not be evaluated from undisturbed sample tests can be reliably
the light-line curves of the 1943 data, supplemented by making two tests on
because there was no regular pattern by each of a series of soils bracketing the
steps in the grain size distribution of range of soil character in the situation at
these natural gravelly soils. For the relative densities of 0 and 70 per cent
widest range of grain sizes with CT of under proper test conditions, such as
5.2, the points fall below the uppermost Nos. 2 to 8 listed above, in order to de-
reference band. Thus, not only is the fine a series of permeability - relative
permeability pattern with Dw made density curves such as given in Fig. 5.
more clear in Fig. 6 by using the com- By entering these permeability - relative
mon 40 per cent relative density basis, density curves with the bracketing ranges
but the important controlling influences of field rektive densities established for
of the range of grain sizes of the soil ma- these soils, reasonably reliable estimates,
terial upon permeability - relative den- bracketing the possibilities in a given
sity -Dw relations are significantly re- situation, can be obtained for the prob-
vealed in Figs. 5 and 6 for practical able range of permeabilities for design
purposes. purposes. The scope and reliability of
Figs. 5 and 6 can thus be extended and
PRACTICAL ASPECTS OF PERMEABILITY broadened. However, the only way to
INVESTIGATIONS determine reliably the influences of the
In investigating drainage, seepage, and degree of anisotropy of soils is to make
stability problems, it is essential to ob- permeability tests on large undisturbed
tain adequate and reliable permeability samples in the vertical and horizontal
data that are representative of and will directions in order to establish repre-
bracket the range of soil character, the sentative values of KY and Kh .
range of field relative densities, and the In certain cases it may be desirable to
range and character of the conditions obtain preliminary estimates of perme-
that control. For important projects it abilities of a large number of soils from
is advisable to obtain and to test as large grain size distribution curves, using the
a number of undisturbed samples of size characteristics Dw and C r . By esti-
granular soils as possible for their perme- mating a range of permeabilities bracket-
ability properties. Test conditions can be ing certain narrow ranges of similar soils
set up for each situation by a careful and from Fig. 6 at 40 per cent relative
complete visualization and appraisal of density, this range of K^ can be inserted
the soil conditions,
Copyright by ASTMboth environmental
Int'l (all rights reserved); SuninJanFig. 5. Permeability
11 13:11:58 EST 2015 - rektive density
and imposed, the nature
Downloaded/printed by of the perme- curves, which conform in general slope to
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further reproductions
the pattern curves of approximately the limiting values and combinations are the
same range of grain sizes, CT, can then more unfavorable and most likely to
be interpolated for this range of KM . govern in the particular situation. This
Estimates of the probable range of per- is not generalization as commonly used
meabilities can then be obtained for any in present practices. The major problems
bracketing range of field relative densi- in soil engineering are to remove the
ties for the given soils. The range of ignorance factors from investigations by
relative densities of granular soils may more adequate and reliable soil test data
be estimated from an interpretation of and thereby to reduce the spread in
the records of the driving resistances of working values, but insuring a real and
the sampler in blows per foot, preferably adequate but not excessive margin of
in blows per 6 in. (5, pp. 1257-1259, safety in design and construction of
Fig. 4, and Eq 6). structures.
Thus these permeability - relative den- By a reasoned and consistent applica-
sity - Dio patterns provide a powerful tion of controlled test methods and by a
and useful tool for supplementing and careful and complete appraisal and
estimating permeability information in evaluation of each situation, the favor-
soil investigations. Due to ignorance fac- able aspects can be recognized and full
tors—such as lack of fully adequate and advantage can be taken of them. The
reliable information and lack of full possibilities of improving conditions with
comprehension of the real soil phenom- regard to any adverse aspects can be fully
ena—a considerable spread in working explored and planned for in order to
values may be necessary in order to avoid construction difficulties and haz-
bracket the probable limits of soil charac- ards. In any case they can be recognized
ter, of behavior and responses, and of and be taken into account fully in the
controlling conditions in a particular planning and design by fixing in advance
situation. These working limits should be the safe limits and time sequences for de-
definitely and carefully established by a sign and construction of structures. This
reasoned consideration of known condi- is learning how to work with nature by
tions that control and of all reasonable fitting foundation and earthwork design
possibilities that may be inherent in the and construction m.thods to actual
situation. The influences of this spread in conditions, in order to achieve (1) closer
working values on the adequacy and agreement between predictions of soil
reliability of the outcome and practical behavior and the actual observed phe-
applications of an investigation should nomena, and (2) higher standards of
be carefully and completely assayed and excellence, greater economy, and more
evaluated, particularly as to which enduring structures.
REFERENCES
(1) Donald M. Burmister, "The Importance of Testing," Symposium on Consolidation
Natural Controlling Conditions upon Tri- Testing of Soils, Am. Soc. Testing Mats.,
axial Compression Test Conditions," Tri- p. 83 (1952). (Issued as separate publica-
axial Testing of Soils and Bituminous tion ASTM STP. No. 126.)
Mixtures, Am. Soc. Testing Mats., p. 248 (3) Donald M. Burmister, "The Place of the
Direct Shear Test in Soil Mechanics,"
(1951). (Issued as separate publication Symposium on Direct Shear Testing of
ASTM STP NO. 106.) Soils, Am. Soc. Testing Mats., p. 3 (1953).
(2) Donald M. Burmister,
Copyright by ASTM "The Application
Int'l (all of
rights reserved); (Issued
Sun Jan as
11separate
13:11:58publication
EST 2015ASTM STP
Controlled Test Methods
Downloaded/printed by in Consolidation No. 131.)
20 SYMPOSIUM ON PERMEABILITY or SOILS
(4) Donald M. Bunnister, "Soil Mechanics," (Issued as separate publication ASTM STP
Vol. I, Columbia University, New York, NO. us.)
N.Y.U952). (7) H. Kane, "Investigation of the Permeability
(5) Donald M. Burmister, "The Importance Characteristics of Sands," Thesis No. 558
and Practical Use of Relative Density in for Degree of Master of Science, Depart-
Soil Mechanics," Proceedings, Am. Soc. ment of Civil Engineering, Columbia Uni-
Testing Mats., Vol. 48, p. 1249 (1948). versity, New York, N. Y., June, 1948.
(6) Donald M. Burmister, "Identification and (Not published.)
Classification of Soils—An Appraisal and (8) "Suggested Method of Test for Maximum
Statement of Principles," Symposium on and Minimum Densities of Granular Soils,"
the Identification and Classification of Procedures for Soil Testing, Am. Soc. Test-
Soils, Am. Soc. Testing Mats., p. 3 (1951). ing Mats., July, 1950, p. 111.
DISCUSSION
MR. E. D'APPOLONIA.1—The reliability consideration: the first deals with the

of the results related to relative density reliability of test results based on relative
depends on two factors: (a) the experi- density as a common unifying argument,
mental errors made in the determination and the second deals with the practical
of the natural, minimum, and maximum problems of adequately rating large
dry unit weights, and (6) the procedure numbers of soils with regard to their
used to determine these unit weights. potential permeabilities.
The first factor can be minimized and The remarks of D'Appolonia with re-
kept reasonably constant by the experi- gard to the first aspect are well taken.
menter. However, the procedures used to There should be a recognized method for
determine minimum and maximum unit determining the maximum and minimum
weights vary with each experimenter, and densities as the 100 and 0 per cent lab-
widely different results for the same soil oratory references for relative density.
are obtained. It has been shown2 that A method of test was suggested by the
differences in unit weights of 2 per cent writer in "Procedures for Testing Soils,"
will mean a 10 per cent difference in rela- (8)3. This method has been used and
tive density. proved in testing granular soils in direct
The work of any one investigator may shear, triaxial compression, and perme-
be consistent. But similar work, con- ability tests over the past 15 yr as a
ducted by another person using different satisfactory basis which yields results
testing procedures, would not give the reproducible within ±1.0 per cent.
correlations between permeability and Due to the nature of the defining relative
relative density discussed in this paper. density equation, Eq 2, with differences
Before the results of research of this hi both numerator and denominator, er-
nature can be utilized and properly com- rors in the determinations of any of the
pared with other similar work, it will be three quantities are considerably magni-
necessary to standardize the testing fied. From experience in the use of rela-
procedures for the determination of tive density, it is believed that errors in
minimum and maximum densities. the determination of the three quantities
MR. DONALD M. BURMISTER (author's within 1.0 per cent will not be reflected
closure).—There are two important in errors in relative density out of pro-
aspects of permeability testing of granu- portion to its great practical usefulness
lar soils that should be given further in soil mechanics (2).
1
The maximum and rninimum density
Associate Professor in Civil Engineering, limits, however, are more than just re-
Carnegie Institute of Technology, Pittsburgh,
Pa. producible values; they have a definite
s
Elio D'Appolonia, "Loose Sands—Their physical significance, namely, they are
Compaction by Vibroflotation," Symposium on
3
Dynamic Testing of Soils, Am. Soc. Testing The boldface numbers in parentheses refer
Mats., Copyright
p. 138 (1954). (Issued
by ASTM as (all
Int'l separate
rights publica- to the
reserved); Sun list13:11:58
Jan 11 of references appended to the paper
EST 2015
tion ASTM STP No. 166.)by
Downloaded/printed and continued at the end of this discussion.
S R KUMAR KONATHALA (Indian Inst of21 Tech Chennai) pursuant to License Agreement. No further reprodu
characteristic limiting values in which ^ to f Ib per cu ft higher. For practical

granular soils can exist in natural de- purposes, the suggested method yields a
posits. As such, they fix the limits of stable and representative maximum
natural granular soil behavior and are density reference, to which may be added
used as basic references for shearing, about 2 Ib per cu ft for the probable true
consolidation, compaction, as well as for value. At least the suggested method is
permeability phenomena for granular much simpler and is more readily per-
soils. Seldom in 20 yr of experience have formed.
natural granular soil deposits been found In a research project undertaken by
in a more loose state than the minimum Mr. Cartwright (10), an attempt was
density reference, excepting possibly made to duplicate natural minimum
loess deposits, which are formed under densities attained by sedimentation
rather peculiar geological conditions; processes in quiet water conditions. This
also, seldom have granular soils in nat- was thought to be the more probable
ural deposits been found in a more dense minimum density condition in natural
state than the maximum density refer- soil deposits. Flowing water, wave ac-
ence, except under very unusual geo- tion, and wind action all tend naturally
logical conditions. These characteristic to produce higher natural densities than
density limits are not haphazard phe- the 0 per cent relative density reference.
nomena but are definitely determined by A special set of test conditions was set
grain size distribution and grain shape up for this research to produce a very
characteristics of granular soils (2, pp. uniform loose layer of granular soil with
18-19). The natural relative densities of no layering or normal segregation effects
granular soils are determined by geo- of the fine and coarse fractions by the
logical processes of soil deposit formation sedimentation process. This was accom-
and by the accompanying grain size plished by allowing a uniform, continu-
distribution and grain shape characteris- ously fed cloud of sand to settle by
tics, as the inherent characteristics of the sedimentation to the bottom of a special
granular soil material. cylinder of lucite that could be separated
Research by Mr. Jolls (9) has de- at a distance of 2 in. from its bottom.
veloped a number of significant facts After measurement of the average thick-
concerning the maximum density ref- ness of the sand layer deposited, the soil
erence. First, the maximum density ob- was evacuated to reveal the degree of
tained by a variable speed vibrator near layering present.
resonant frequencies was about 2 Ib per It was found, as evidenced by light
cu ft higher than that obtained by the and dark streaks, that only a thin layer
"vibro-tool" method used in the sug- of a few millimeters in thickness at the
gested procedure (8). This value repre- top and bottom showed segregation and
sents about the maximum attainable by layering effects. In six of eight sands
any means and therefore is a suitable bracketing the grain size limits of com-
maximum density reference. Second, by mon sands, a higher loose density was
varying the thickness of the vibrated obtained than in the suggested method
successive layers, it was shown that for by spreading dry sand in f-in. layers by
practical purposes, the specified "|-in." a large spout funnel. In the two coarser
thick layers achieved consistent maxi- soils of the eight soils tested, the density
mum densities and that for J-in. thick was found to be lower than the minimum
layers,Copyright
the maximum
by ASTM density wasrights
Int'l (all only reserved);
densitySun
reference.
Jan 11 These
13:11:58density values
EST 2015
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement.
DISCUSSION ON PERMEABILITY TESTING OF SOILS 23
TABLE II—TENTATIVE CRITERIA FOR RATING SOILS WITH REGARD TO DRAINAGE, CAPILLARITY,
AND FROST-HEAVING CHARACTERISTICS. CRITERIA FOR SOILS IN A LOOSE TO MEDIUM
COMPACT STATE.
"some fine
"little Silt" Silt" "some Clayey
Fineness Identification "trace fine "trace Silt" (coarse and "little Clayey Sill"
Sand" fine) Silt" Clay-Soils
Fissured dominating
Clay-Soils
Approx. Effective Size, Z>io mm. 0.4 0.2 0.2 0.074 0.074 0.02 0.02 0.01 0.01
Drainage free drainage drainable by drainable drains poor to
under gravity good to slowly impervious
gravity good fair fair to
excellent poor
Approx. range 0.2 0.04 0.006 0.0004 0.0001
k, cm per sec 0.5 0.10 0.02 0.001 0.0002
DR—40 PER CENT TkttATA Wnllc
Well poin ts successful
Capillarity negligible slight moderate moderate to high
high
Approx. rise, Hc, ft:
DR—0 per cent 0.5 1.5 7.0 15.0
DR—40 per cent 0.5 1.0 3.0 10.0 25.0
Frost-Heaving Susceptibility non-frost slight moderate to objectionable objectionable
heaving objection- to moder-
able ate
Ground water within 6 ft. or He/2 *-Doubtful-» ••—Drainage—»
and/or
Protective
Installa-
tions
required
Station Depth Identification:
13 1-6 yb mfS, 1-fS, t-mfc
14 3-9 b cmS, t-fc
15 2-8 yb mfS, t++-cf&, t-fo
16 16-22 yb mfS, t +-cfs, l + -mfc
17 15-21 ybmfS, s -cfs
18 4-10 ybmfS, 1-cfs SYMBOL FOR if OF IDEN-
19 2-8 bmfS TIFICATION NAME (6)
20 6-12 bmfS, t:cfs Major Compcments (first):
21 8-12 b mfS, t + -cfs G, S, S
22 8-12 b mfS, t-mfc Minor Compo nents: G, s, s
23 8-12 b cmS. t~-cfs Fractions of Components:
24 8-12 b cmS. t~-cfs coarse to m edium, cm
25 1-6 b mfS coarse to fiiic, cf
medium to ine, mf
fine, f
Proportion Te•rms:
and a- 35--50 per cent
some s- 20--35 per cent
little 1- 10--20 per cent
trace t- 1--10 per cent
P.R. Classification A-3
Casagrande Classification SP + and — ne,irer upper or
lower lim it.
Corp of Engineers Uniform Color :bro wn,b;yellow-
Classification —Sand brown, yl >; gray, g.
Potential Drainage, Capillarity, and Frost-Heaving Ratings of Subgrade Soils to Depths of 3 to 5 ft below proposed
Subgrade Elevation.
Controlling Conditions in Natural Situations:
1. Identification and character of soils, geological origin and processes of formation.
2. Climatic conditions—seasonal precipitation, temperature and evaporation conditions, normal depth of freezing,
.average number of days below freezing temperature consecutively.
3. Soil Profile—type and pedological character, horizons, discontinuities, depth to rock. Stratified deposits—se-
quence of layering and character of soils, thickness of layers, uniformity or variability in lateral extent, lenticular
strata, depth to rock.
4. Surface and subsurface drainage conditions, ground slopes, seepage zones, ground water level and probable sea-
sonal variations at critical times.
5. Granular soils—natural compactness and coherence due to silt films at the grain contacts. Clay-soils—natural
consistency, structure and state of aggregation, fissured character and fragmentary structure as controlling drainage
conditions.
6. Natural moisture content, degree of saturation of natural and compacted soils after normal capillary saturation
under conditions imposed by structure at critical times.
7. Relative permeabilities of different horizons or strata in natural deposits, or soils compacted in thin layers in
subgradesCopyright by ASTM Int'l (all rights reserved); Sun Jan 11 13:11:58 EST 2015
and embankments.
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further reproductions author
were from 1 to 2 Ib per cu ft higher or information needed for making practical

lower than the minimum density refer- and effective applications. This is in
ence. This research method was a difficult contrast to simplified "all-purpose"
and time consuming one. Therefore, as a classifications with their broad generali-
practical method, which has physical zations, confused mixing of factual and
significance, it is believed that the sug- interpretative information, and their
gested method of spreading the dry soil, fixed and inflexible conceptions of ade-
or using test condition No. 4 listed for the quacy. The shortcomings and limitations
permeability test to avoid or to elim- of classification have been discussed
inate segregation effects, would yield elsewhere (6, pp. 4-8).
sufficiently representative and signifi- A rating of granular soils with regard
cant minimum density references for to permeability, capillarity, and frost-
different granular soils for relative heaving susceptibility is given in Table
density determinations. II to illustrate the potential value, possi-
The second aspect of practical im- bilities, and practical usefulness of rat-
portance and use in soil engineering is ings. A rating to be of real practical value
that of adequately rating granular soils must be based on specific criteria de-
with regard to their permeabilities in veloped by experience or through re-
natural deposits and in compacted em- search and investigation to define the
bankments. This is a very practical prob- limits of the ratings. A rating is essen-
lem, because, in general, only a few tially interpretative information. In the
permeability tests can usually be made very nature of things, the criteria and
on a given project within, time and the limits of the ratings must be brought
economic limits, but information on the up to date at frequent intervals as
potential permeabilities of all soil sam- knowledge of soil phenomena increases.
ples obtained would be essential in order Furthermore the adequacy of ratings
to obtain clear and definite conceptions should not be permitted to become an
and information on the range and kind inflexible and fixed idea.
of permeability problems to be encoun- The primary basis for a rating is an
tered. The idea and conception of rating adequate identification of soils that pro-
soils have come more prominently into vides factual information on the inherent
use in the past few years. A rating of characteristics of the soil material and of
soils with regard to the fundamental the soil structure that govern soil be-
behavior characteristics of soils is more havior. Criteria derived from the identi-
specific, significant, and practically use- fications of soils, either the identified
ful than is attainable by the broad gen- finest fraction of a component from visual
eralizations of classification systems so examinations and identifications of the
commonly used at present. The basic soils in the field or laboratory, or the
question is more than that of the degree value of the effective size, Dio, from
of generalization permissible in soil work grain size analyses, then permit the rat-
and of how simply and easily applica- . ing of soils with regard, for example, to
tions can be made, but rather is one of drainage characteristics and permeabil-
attitudes and conceptions regarding the ity through the permeability-/?™ rela-
place and meaning in soil engineering of tions of Fig. 6 of my paper on a common
the two basic analyses of identification 40 per cent relative density basis. Such
of soils, as factual information of first a specific rating provides a systematic
importance, and of rating of soils, as basis for making effective and definite in-
necessary and important
Downloaded/printed by interpretative terpretations for practical purposes. Still
DISCUSSION ON PERMEABILITY TESTING OF SOILS 25
more specific information can be obtained upon the range of permeabilities for most
directly from Fig. 6 if it is required for successful action, as established by ex-
drainage investigations. perience. In the case of capillarity, the
It should be noted in Table II that approximate effective heights of capil-
capillarity or potential capillary rise de- lary rise are given for the common range
pends essentially upon the same criteria of relative densities. Such useful informa-
as does permeability, but varies as to tion can be amplified and extended with
magnitude in exactly the reverse order. increase in understanding of soil phe-
Frost-heaving susceptibility depends nomena. These are distinct advantages of
upon both drainage and capillarity char- the rating conception.
FIG. 7.—Possible Coarse and Fine Grain Size Curve Limits of Classifications for Purposes of
Rating of Soils with Regard to Drainage Characteristics.
acteristics of soils with the region of In order to further illustrate the prac-
principal susceptibility and maximum tical value and usefulness of ratings of
objectionable action in the silt sizes, soils with regard to their drainage and
where permeability is not too low nor associated characteristics, ratings have
capillarity too high. Ratings of soils been made in Table II of subgrade soils
should never be made apart from the in a cut section of a highway. The rela-
criteria of the ratings, but should be tively narrow spread in the limits of the
made a part of the rating table, as is done rating for each soil is to be especially
in Table II; otherwise, the ratings lose noted, being determined either by the
their real significance and usefulness. possible spread in DW or by the identifi-
The ratings also contain additional prac- cation of the finest soil fraction (coarse,
tical and useful information regarding medium, or fine) of the finest component
the limits
Copyrightofby applicability of certain
ASTM Int'l (all rights reserved); Sun(sand
Jan 11or13:11:58
silt). EST
A major
2015 problem in soil
methods of draining by
Downloaded/printed soils, which depend engineering is consistently to reduce the
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further reproduction
spread in the working values of soil in the doubtful regions and certainly in
properties and ratings due to ignorance the regions of poorer drainage and ob-
factors by more adequate identifications, jectionable frost-heaving characteristics.
testing, and analyses of soils. The com- Thus potential drainage and frost heav-
parative spread in the D\0 values inherent ing become actual problems if the normal
in classification methods is illustrated in depth of freezing exceeds 1| ft, or the
Fig. 7, where the possible grain size ground water is within about 6 ft of the
curve limits are given for the class of subgrade level or about one half the
soils hi which most of the soils rated in height of capillary rise of the soils, which-
Table II fall. The degree of generaliza- ever factor controls. The rating provides
tion of classifications now becomes evi- a tangible and practical basis for judging
dent. In contrast, the direct use of grain each situation, as to its kind and range
size curve by sieve analyses or the ac- of drainage and frost-heaving problems,
curate and complete identification of and as towihe types of installations and
soils greatly reduces the spread hi the protective measures that are likely to be
comparative ratings in Table II. It most effective.
should be evident that more adequate The ratings also provide a basis for
and specific information is obtainable comparison of expected performance of
by the identification-rating method of any section of highway, for example,
analysis than by classification methods. with the actual observed performance
Furthermore an appraisal of the con- in a condition survey after a year or two
trolling conditions in the situation, such of operation hi service. Thus it would be
as listed in Table II, will determine what possible to build up an authoritative
significance should be attached to these body of valuable information. The rat-
ratings of potential behavior in a partic- ings, interpretations, and appraisals of
ular situation. Estimates can then be probable actual performance, and the
made of the probable actual susceptibil-
judgments regarding the requirements
ity and behavior of soils, as a basis for
for drainage and protective installations
determining what treatment is necessary
to improve the qualities and responses of could be checked, modified, and extended
the soils, or what protective measures, to cover actual conditions observed in
such as installations of drains, base the field under different climatic and soil
courses, or frost-heaving protection, may conditions. Thus ratings could be made a
be required. Depending upon the prox- powerful and practical tool for the soil
imity of ground water to the subgrade engineer in making the most effective use
level during the season of highest ground- of knowledge and experience and for im-
water level, drainage installations and proving conditions and practices in soil
protective measures should be installed engineering.
REFERENCES
(9) F. K. Jolls, "The Vibratory Characteristics (10) P. Cartwright, "Minimum Density Studies
of the Maximum Density of Sands," Mas- of Granular Soils," Master of Science
ter of Science Thesis No. 658, Dept. of Thesis No. 657, Dept. of Civil Engineering,
Civil Engineering, Columbia University, Columbia University, New York, N. Y.
New York, N. Y. May, 1953. May, 1953.
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No
WATER MOVEMENT THROUGH POROUS HYDROPHILIC SYSTEMS
UNDER CAPILLARY, ELECTRICAL, AND THERMAL POTENTIALS
BY HANS F. WiNTERKORN1
SYNOPSIS
The importance of the interaction between solid internal surface and pore
water for water transmission under hydraulic gradients and electrical and
thermal potentials is pointed out. This interaction results in the establishment
of a restrained water phase possessing characteristic mechanical, thermal, and
electrical properties. If only a hydraulic gradient is established, the restraint
can be expressed as a volume factor which, however, is also a function of the
hydraulic gradient, especially at high values of the latter. The characteristic
thermal and electrical properties of the interphase are the necessary conditions
for thermo- and electro-osmotic flow. The basic theories of these phenomena
are presented hi a simple manner. The considerations presented and the equa-
tions derived hold only for such soils or similar systems that do not possess a
significant gas phase.
The movement of liquids through WATER TRANSMISSION UNDER

porous solid systems depends on the HYDROSTATIC PRESSURES
proportion and geometrical characteris- Systems possessing pores of sufficient
tics ortEe~^^Jspace~tHeÊyi£al prop- size to allow the moving fluid to act in
erties of tHe Iiquior7and the înteraction accordance with its statistical or mass
between liguHTand solid internajjsurface, properties and endowed with geometrical
as well as on thTen^^potentml inThe properties that result in laminar flow are
direction 61 flow, the cross-section con- usually treated as analogous to systems
sidered, an<Ttne time_aJloare3T The first of capillaries of uniform diameter. The
three factorsT are oTlntrmsic physical im- latter simplified systems must, of course,
portance because they represent proper- possess the total porosity and the same
ties inherent in the system. The last two liquid transmission properties as the pro-
can be varied at will and are, therefore, totypes. In establishing the equivalent
not characteristic. The energy potential capillary system, it is customary to use
factor may be varied at will but also can the Poiseuille-Hagenbach equation:
and often does have a definite physical
importance, especially when the in-
tensity of the liquid-solid interaction
varies with distance from the pore walls. where:
1
Department
Copyright byofASTM
Civil Engineering,
Int'l (all rights Princeton
V = volume of liquid in cubic
University, Princeton, N.
Downloaded/printed by J. centimeters,
27 Chennai) pursuant to License Agreement. No further reproduc
S R KUMAR KONATHALA (Indian Inst of Tech
ij = coefficient of viscosity in equation, not more and not less. This

poises, assumption is that the solid surface holds
r = inside radius of the capillary the first contacting layer of liquid
tube in centimeters, molecules so strongly that they are fixed
pi — pz = pressure head over the dis- and immovable. If the fixation extends
tance /, over a large number of molecular layers,
/ = length of tube, and the Poiseuille equation and Eq 4 derived
t = time in seconds. from it are no longer strictly valid; nor
If both sides of the equation are multi- are they strictly valid in the case in
plied by « c , the number of capillaries per which the first contacting layer of liquid
square centimeter of cross-section normal molecules are not fixed and are, there-
to the direction of flow, and divided by fore, subject to slippage. In most actual
the time /, setting at the same time the soils, the fixing effect extends to a very
considerable distance; on the other hand,
gradient equal to unity, then nonfixing and slippage is observed in
the following equation is obtained for the certain resin stabilized soils. This paper
water transmission per unit gradient per will be limited to porous systems in
second per square centimeter: which there exists a considerable interac-
tion between the solid and liquid phases,
chief representatives of which are sys-
tems composed of clay soils and water.
The author had the privilege on two
If the porosity of the system, n, is intro- occasions within the last two years to
duced and knowing that with uniform discuss the unique properties of the wa-
and constant cross-section of the capil- ter substance that enable it to interact
lary tubes the product rV X n0 = n, and strongly and hi various ways with solid
n mineral surfaces capable of ionic ex-
also that rz = , Eq 2 can be written
WCT ^ change (l, 2).2 As a matter of fact, most
as follows: physical tests on cohesive soils, including
those for hygroscopicity, consistency,
consolidation, and shear characteristics,
probe into the condition of the water,
Since w2 and nc are constants for a spe- and actually are tests for the interaction
between solid surface and water mole-
cific system,
J
can be introduced cules. With respect to this interaction,
the available experimental evidence
and an equation for the transmission renders it very probable that:
constant, k, thus be obtained: 1. Below the hygroscopic moisture
equivalent, the water substance is dis-
solved as the monomer (H20) in the
solid mineral surfaces.
In this equation one factor is repre- 2. Between the hygroscopic moisture
sentative of the properties of the solid to approximately the plastic limit, the
and the other of those of the liquid. This water substance possesses properties
equation, however, can hold only for similar to those of a melt.
systems in which the solid-liquid inter- 2
actionCopyright
is exactly The boldface numbers in parentheses refer
by as assumed
ASTM Int'l (allin the reserved);
rights deri- toSunthe Janlist
11of
13:11:58 ESTappended
references 2015 to this paper,
vationDownloaded/printed
of the Poiseuille-Hagenbach
by see p. 35.
WlNTERKORN ON WATER MOVEMENT THROUGH SOILS 29
3. Above the plastic limit, the water Putnam soils are plotted in Fig. 1 against
phase behaves more and more as an the square of the porosity. This figure
ionic solution but has peculiar properties shows that essentially a straight-line re-
because of the fixation of the negative lationship exists between the perme-
charges on and in the solid surface (2). abilities, at the higher porosity and lower
Depending upon the particular system pressure values, and the square of the
and the intended purpose, the study of porosity. Continuation of the straight3
the water-solid interaction may be fo- line through the abscissa leads to an
cused on one, several, or all of its im- intercept which physically has the mean-
mediate consequences. These are: ing of the square of that portion of the
1. Water fixation reduces the volume
available for viscous flow.
2. The highly electrical character of
the interaction phase indicates use of
electrical potentials as tools for studying
the condition of water in such systems
and as practical tools for drainage (1).
3. The fixation of the water molecules
decreases their capacity of utilizing
kinetic energy. As a consequence, part
of the heat content possessed by free wa-
ter must be given off on occasion of the
fixation of the water molecules and be-
comes evident as heat of wetting. This
and the negative temperature coefficient
of the heat of wetting indicate use of
thermal potentials as scientific and en-
gineering tools.
4. The interaction between thermal
and electrical molecular phenomena
heralds the existence and potential use-
fulness of thermoelectric effects (2). FIG. 1.—The Coefficient of Permeability, k,
For normal permeability studies in as a Function of the Porosity, n, of a Series of
Remolded Homoionic Putnam Soils.
saturated flow, the problem of liquid- liquid volume that has been fixed by the
solid interaction can be scientifically re- wall influence. However, since the degree
duced to a consideration of the effect of fixation is an inverse power function
of the fixed water volume. If the number of the distance from the wall, the volume
of equivalent capillaries per cross-section of immovable "water" decreases with
could be kept constant while changing increasing hydrostatic pressure. The data
the porosity, n, then the transmission
coefficient, k, should be directly pro- mains * If the thickness of the fixed water layer re-
constant while the porosity and pore sizes
portional to the square of the porosity. change, then the theoretical line obtained by Eq.
This condition is more or less fulfilled if 1 will not be entirely straight. There exist, how-
the permeability is calculated from con- ever, good theoretical reasons backed by experi-
mental evidence which indicate that the thick-
solidation data as long as the applied ness of the fixed water layer is a function of the
pressures are too low to cause consider- curvature of the solid surface to which it is at-
tached. See paper by author on "Studies on the
able plastic flow
Copyright by of the Int'l
ASTM soil(all
mass itself.
rights Surface
reserved); Behavior
Sun Jan of Bentonites
11 13:11:58 and Clays," Soil
EST 2015
Data Downloaded/printed
obtained in this byway by Winter- Science, Vol. 41, No. 1 (1936). Undoubtedly, in
the actual phenomenon, there are still other modi-
korn Sand Moorman
R KUMAR (3) on homoionic
KONATHALA (Indian Inst offying
Techfactors
Chennai) pursuant to License Agreement.
involved. No furthe
TABLE I.—INTENSITY OF WATER Table I are functions mainly of the type
FIXATION FORCES, KG PER
SQ CM (5). and proportion of clay minerals and
First molecular layer 25 000 organic matter, and of the type and num-
Hygroscopic water 50 ber of exchangeable cations (4). This is
Permanent wilting point 12.5
Wilting point (dead water) 6.25 illustrated in Table II. The minimal
Vacuum moisture equivalent; 0.55 water capacity of a soil, which is denned
TABLE II.—HYGROSCOPICITY, Ws , OF VARIOUS SOILS AND SILICATES.
HTGBOSCOPICITY OF DIFFERENT Soit TYPES, G H2O PER 100 o SOIL (6)
Soil Grain Size,

mm ws Soil WH
Quartz 2-1 0.055 Fine quartz sand 0.03

Sand 1-0.5 0.057 Sandy soil (Kummo) 1.06
0.5-0.25 0.085 Loamy soil 1.40
0.25-0.17 0.101 Sandy loam 2.09
0.17-0.11 0.131 Silty loam 3.00
0.11-0.07 0.168 Clay loam 6.54
0.07-0.01 0.203 Low moor 18.42
Heavy clay (Java) 23.81
HYGROSCOPICITY OF DIFFERENT ALTJMINO SILICATES (7)
Mineral WH
H-Montmorillonite ... 32,7

H-Putnam clay '" 25:5
H-Davidson clay 21.4
H-Permutite . 28.2
plotted hi Fig. 1 were obtained by using as the amount of water in grams held
in each case pressures from 1 to 8 atmos, against gravity by the molecular (elec-
the lower ones for the high porosity trical field) forces of the particles of
values and the higher ones for the low 100-g soil, can be calculated hi first
porosity values. Since plastic flow at the approximation from the hygroscopicity:
high pressures undoubtedly had some
effect on the number of equivalent capil- This minimum water capacity is hide-
laries hi these systems, the curved parts pendent of the particle arrangement and
or lower ends of the &-lines are influenced of the menisci that are functions of
by the plastic properties of both the particle shape and arrangement. Actual
entire system and the adsorbed water. minimal water capacities of soils with
Including the volume effect, Eq 3 can different clay contents are shown in
be written for the higher porosities and Table III.
lower pressure gradients as: In accordance with the experimental
and theoretical evidence presented, the
transmission of water through saturated
hydrophilic systems can be expressed by
The intensity of the fixation forces for an equation of the type of Eq 5. It must
different water conditions are indicated be understood, of course, that both the
hi Table I. Thebyamounts
Copyright ASTM Int'lof(all
water
rightsheld at
reserved); liquid
Sun Jan viscosity,
11 13:11:58 17,
EST and
2015the fixed volume
the different stagesby characterized hi
Downloaded/printed function, C2 , decrease with increasing
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further reprodu
WiNTERKORN ON WATER MOVEMENT THROUGH SOILS 31
temperature. The constant, Ci, contains h = height of capillary rise,

among other items the number of equiva- 7 = density of water, and
lent capillaries per cross-section, nc. For g = gravity constant.
most actual cases, nc remains practically Making the usual assumption of a zero
constant. The constants of a particular wetting angle, the following is obtained:
system are best obtained from experi-
mental data plotted as in Fig. 1.
WATER MOVEMENT UNDER CAPILLARY
SUCTION POTENTIALS Now let us consider the problem from
The volume of the fixed water plays an energy point of view. Assume that the
a r61e not only when water is moved as internal surface of the capillary has ad-
TABLE III.—HYGROSCOPICITY AND MINIMAL WATER CAPACITY OF
DIFFERENT SOILS.
Size Composition, per cent
Soil Type Provenience Sand Minimal

Silt Clay wz Water
Capac-
Coarse Fine ity
Sand Greenlo, Holland 69.6 28.0 1.0 1.4 0.8 7.3

Moraine sand Barlo, Holland 66.8 24.1 6.4 2.7 1.6 15.6
Loamy sand Bahri, Egypt 57.3 25.7 8.8 9.0 2.2 16.3
Loam Hedel, Holland 6.1 51.2 21.4 21.3 4.7 30.6
Loam Onderdjik, Holland 4.6 29.8 27.9 37.7 7.2 59.3
Clay Kafr Dimetun, Egypt 1.6 6.2 20.0 72.2 14.4 66.1
a result of differences in hydraulic pres- sorbed its minimal water capacity and
sure but also when movement is the re- possesses essentially the surface energy
sult of capillary and physicochemical suc- (T-E X area) of an equal area of free
tion forces. This became evident in water. The water of the reservoir is re-
previous studies of the problem of water strained from moving immediately on
attack on dry cohesive soil systems (8) insertion of the capillary; it is released
and of water accumulation underneath only after a zero wetting angle has formed
pavements (9). These problems, there- at the inserted end of the capillary. Then
fore ne d not be treated here. However, the restraint is removed and the water
one simple aspect of capillary rise appears permitted to rise in the capillary. After
worthy of mention. The simplest treat- the rise has terminated, the meniscus
ment of capillary rise is based on the with the zero wetting angle has been
force equilibrium equation applied to the translocated by the height h. In addition,
condition at the end of the rise, namely: an amount of water r2irh of unit weight
7 has been moved against the gravity g
cos a 2-icrT-g = rzirhyg
over an average distance of the free
where:
a = angle of wetting between liquid surface energy 2rirhT^ has been used up:
and wall, in the case of water and Loss of free energy:
soil usually assumed as zero,
r = radius of capillary,
Gain of potential energy:
TF = surface tension of water,
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No fur
Since TE is numerically equal to TV fore hi the total heat capacity between

and since the force equation holds true, portions of the soil system that differ in
only one half of the free energy involved temperature. If for a certain volume ele-
in the process has been utilized. This ment at the higher temperature T"2, the
equipartition of energy appears to be of total heat content:
fundamental importance and to hold
true also in cases where the surface
energy lost is, as in initially dry soils,
considerably greater than that of an and at the lower temperature T\, the
equal surface of water. total heat content:
EFFECT OF TEMPERATURE POTENTIALS
At the beginning of this paper, it was
pointed out that moist soil systems hen a certain amount of heat t/"2 —
possess mechanical, thermal, and elec- Ui = Q would become available by
trical properties. However, by keeping transferring the volume element from
the temperature of the system constant, Tz to T\, neglecting work involved in
it is possible to treat moisture transmis- volume change. This heat is akin to a
sion under hydrostatic pressure poten- heat of fusion. This Q has an entropy of
tials as a problem solely of fluid mechan-
ics, reducing the interaction between ¥- at the higher temperature and one of
lz
solid surface and water molecules to a
correction factor with respect to the — at the lower temperature; according
liquid volume that takes part in the to the second law of thermodynamics,
transmission. An analogous approach is the entropy of a system tends to de-
possible in the case of thermal potentials crease. Two physical possibilities for such
applied to mineral-water systems, having decrease are available since the system
no significant gas phase, hi which the can transmit both heat and water. Con-
water is present hi such proportions that sidering the water transmission as the
even the molecules most distant from the predominant one for the moisture condi-
solid surface are still under a certain, tions envisaged, it can be stated in ac-
though relatively low, "fixation." It has cordance with the second law of thermo-
been pointed out that such a restraint dynamics that a maximum amount of
decreases the kinetic heat energy that free energy, TFmax, is available for the
can be stored by the water molecules and transfer of liquid from the location hav-
results in heat of wetting and hydration ing the temperature T2 to that having
when liquid water molecules are placed the temperature T\ :
in contact with dry, solid soil constitu-
ents.
In soils this heat of hydration has a
negative temperature coefficient. If, or:
therefore, a soil without significant air
voids but possessing a uniform moisture
content, in the above specified state of
restraint (best in the vicinity of the
plastic limit), is subjected to a tempera- If Q represents the latent heat involved
ture gradient,
Copyright by the latter
ASTM will
Int'l (all rightsestablish inJan
reserved); Sun the11change
13:11:58of
ESTrestraint
2015 of 1 g of water
a difference by heat and there- incident to a temperature change of d/,
hi the specific
Downloaded/printed
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further reproduct
WlNTERKORN ON WATER MOVEMENT THROUGH SOILS 33
then Wiaea. is the free energy available for water. Only subsequent increase in tem-
moving 1 g of water from a location at perature at the low-temperature points
(T + dT) deg to one at T deg. The may change some of the 'shifted water
quotient of a free energy and the volume into free water. This, however, can be-
with which it is associated represents a come an important feature in soils of low
pressure or a suction. Using the metric permeability under hydrostatic pres-
system and taking advantage of the fact sures (9). A special case of thermo-osmotic
that in this system the weight of a unit water transmission occurs in the forma-
volume (1 cu cm) of water equals unity tion of ice lenses in soils. This case has
(1 g), the following is obtained for the been treated in detail in a recent dis-
maximum pressure or suction: cussion (10).
WATER TRANSMISSION UNDER
ELECTRICAL POTENTIALS
The highly electrical character of the
This suction value can be inserted into mineral-water interaction phase renders
any Darcy type formula: soil-water susceptible to movement if an
electrical potential is applied. The general
and practical aspects of this phenomenon
have been surveyed recently by Casa-
where: grande (11). The physicochemical factors
v = volume of liquid transmitted in that play a dominant rdle, especially in
time t from location with tempera- the case of moisture contents falling
ture (T + dT) to one with tem- within the plastic range (water under
perature T, restraint), have been treated theoreti-
kD = Darcy coefficient of permeability cally and experimentally by the author
in cgs system, (1, 2, 12). Because of this sufficient and
dl = distance in which temperature falls recent coverage, the subject is not fur-
by dr, and ther discussed here. However, it should
/ = time in seconds. be pointed out that all experimental and
t\T theoretical evidence available proves the
If a temperature gradient, — = 1, is general correctness of the picture on
Q£
employed, and both sides of Eq 6 are soil-mineral-water relationships that
divided by /, a thermo-osmotic transmis- has been developed over the last 25 years
sion coefficient is obtained: by the soil physicists and colloid chem-
ists. Thus, in systems of low moisture
content in which all water is under con-
straint, well defined minimum voltages
must be applied before water can be
If k0 is known, then from an experi- moved out of the system. These thresh-
mental determination of kT , the value Q old potentials correspond to "yield"
may be determined. It should be em- pressures in plastic systems (4).
phasized that in the analyzed process
only a shift or rearrangement of water in ELECTRICAL CONSEQUENCES DERIVING
the system is being dealt with, and this FROM APPLICATION OF A THERMAL PO-
comes to an end when the entropy of the TENTIAL TO MOIST HYDROPHILIC
entire system has reached its minimum SYSTEMS
value Downloaded/printed
for the prevailing by
conditions. The In an electrical system, such as a
water Sinvolved
R KUMARis notKONATHALA
free but restrained moist
(Indian Inst of clay
Tech soil, in which
Chennai) the toelectrical
pursuant License Agreemen
1
charges are asymmetrically distributed cases the effects of both the electrica
between the predominantly negatively and thermal potentials must be consid-
charged internal solid /surface and the ered. The simplest general equation is:
predominantly positively charged water
phases, any interference can be expected
to have electrical consequences. The
latter may, on their part, set up mechan- where:
ical forces. Hence, it was certain that v = volume of water moved through
application of a thermal potential on a unit cross-section,
moist soil would result in the establish- ko = Darcy transmission coefficient,
ment of an electrical potential (2, 12).
This phenomenon appears to be of i = hydraulic gradient
definite importance in normal and ab- kf, = electro-osmotic transmission co-
normal plant and animal physiology and efficient according to Winterkorn
in thermal methods for curing of dis- (12),
ease; it is only of secondary importance dE = increment in electrical potential,
for the present Symposium. More in- d/ = increment in distance normal to
formation on this subject is found in the cross-section,
reference 2. dT = increment in temperature,
SIMULTANEOUS ACTION OF kf = thermo-osmotic transmission co-
SEVERAL POTENTIALS efficient, and
/ = time.
It is obvious in the light of the evidence Although Eq 8 is correct in form and
presented that a general equation for sufficiently accurate for practical pur-
water transmission through porous soils poses, the purist, who understands the
must have mechanical, thermodynamic, derivation of the thermal and electrical
and electrical factors. In first approxi- members, is disturbed by the fact that
mation, and this seems to be the best in this derivation implied use has been
that one can do at the present tune, a made of the Poiseuille equation, whereas
simple superposition of the effects of the actual flow picture is different from
these factors can be employed. Instead that postulated by-Poiseuille. The en-
of doing this in one equation, it is gen- gineer, however, need not be concerned,
erally preferable to evaluate separately since k6 and &T as well as &D are and
the effect of each factor. In the absence should be obtained experimentally; and
of a thermal potential and of an applied the scientists should know that here, as
electrical potential when flow occurs as in many other cases, the final form of the
the result of hydrostatic pressure differ- equation is more important than the
ence, the effect of the resulting electrical original picture on which it is based. The
flow potential can be included in the classic example of this is, of course, the
volume factor C2 of Eq 5. On the other Carnot equation which is still valid, al-
hand, electrical drainage is used in soils though its author believed that heat was
of low hydraulic permeability, and for a material substance.
practical purposes the main flow effect
is due to the applied electrical potential. CONCLUSIONS
However, at low moisture contents, the In all types of moisture transmission
Joule heat developed by the passing cur- through fine-grained, cohesive soils, the
rent often results
Copyright by ASTMinInt'la(all
thermal potential
rights reserved); interaction
Sun Jan 11 13:11:58 EST 2015 of the internal soil surface
superimposed on the
Downloaded/printed by electrical. In such with the water phase is of definite and
WINTERKORN ON WATER MOVEMENT THROUGH SOILS 35
often decisive importance. This interac- information. It is realized that much of

tion has been studied and well elucidated this information is in a form that is not
by soil physicists and colloid chemists. immediately usable by the soils engineer;
There is no valid cause for continued neg- the necessary transformation is, how-
lect by the soil engineer of the available ever, a job he must do himself.
REFERENCES
(1) Hans F. Winterkorn, "Surface Chemical (7) L. D. Baver and Hans F. Winterkorn,
Properties of Clay Minerals and Soils from "Sorption of Liquids by Soil Colloids, II,"
Theoretical and Experimental Develop- Soil Science, Vol. 40, pp. 403-419 (1935).
ments in Electroosmosis," Symposium on (8) Hans F. Winterkorn, "Mechanism of
Exchange Phenomena in Soils, Am. Soc. Water Attack on Dry Cohesive Soil
Testing Mats., p. 44 (1953). (Issued as Systems," Soil Science, Vol. 54, pp. 259-273
separate publication ASTM STP No. 142.) (1942).
_(2) Hans F. Winterkorn, "Potentials in Mois- (9) Hans F. Winterkorn and Henry Eyring,
ture Migration," Proceedings, Building "Theoretical Aspects of Water Accumula-
Research Conference, National Research tion in Cohesive Subgrade Soils," Pro-
Council of Canada, Ottawa (1953). ceedings, 25th Annual Meeting, Highway
(3) Hans F. Winterkorn and Robert B. B. Research Board, Vol. 25, pp. 422-435
Moorman, "A Study of Changes in Physi- (1946).
cal Properties of Putnam Soil Induced by (10) Hans F. Winterkorn, discussion of "Suc-
Ionic Subst tution," Proceedings, Highway tion Forces in Soils upon Freezing," by
Research Board, Vol. 21, pp. 415-434 Alfreds R. Jumikis, Proceedings, Am. Soc.
t
(1941). Civil Engrs., Vol. 80, separate No. 445
Hans F. Winterkorn, "Engineering Proper- (1954).
ties of Clay Soils," Bulktin No. 1, Winter- (11) Leo Casagrande, "Review of Past and
korn Road Research Institute, Princeton, Current Work on Electro-Osmotic Stabili-
N. J. (1950). zation of Soils," Report to Bureau of
)(S) Hans F. Winterkorn, "The Condition of Yards and Docks, Contract No. NOy-76303
Water in Porous Systems," Soil Science, (1954).
August, 1943; see also H. Freundlich* Hans F. Winterkorn, "Fundamental Simi-
"Kapillarchemie," Vol. 1, Leipzig, 1930. larities between Electro-Osmotic and
(6) Landolt-Bornstein, "Bodenkunde," Vol. 3, Thermo-Osmotic Phenomena," Proceed-
Section 3237, Springer Verlag, Heidelberg, ings, Highway Research Board, Vol. 27,
6th Ed., pp. 358-368 (1952). pp. 443^445 (1947).
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further
DISCUSSION
MR. S. J. Rosen.1—Frankly, I was closure).—The paper just presented

somewhat disappointed because I had treated in a very limited time the the-
anticipated getting more information oretical basis of water flow through
from this paper or possibly during its porous hydrophilic systems under several
presentation. We are constantly con- types of potentials. While flow under
fronted with the fact that the soils thermal potentials was included in the
throughout the country are radically treatment and while the problem of heat
different. When we consider that we conductivity is closely interrelated with
have soils containing extremely dry that of thermo-osmosis, it was impossible
sand, comparatively wet sand, clays, and to discuss, within the set frame of this
combinations of these, we know that paper, the problem of heat conductivity
these soils may compact differently and in soils in such a manner and in such
yield different values of thermal con- detail as to furnish the engineering in-
ductivity. formation wanted by Mr. Rosch.
Therefore I was wondering whether a With respect to the desired informa-
method has been adopted for determin- tion on heat conductivity, I should like
ing thermal conductivity of soils, and if to refer Mr. Rosch to the excellent works
so, whether any values have been of W. O. Smith. 2 - 3 - 4 Also, I would like to
standardized for the different types of say that I am well aware of Mr. Rosch's
soil. I have studied the reference material practical problem in which thermo-
in the paper but fail to find this type of osmosis plays as great a rdle as does
information. It is certainly needed for straight heat conductivity. The practical
the computation of current-carrying solution of this very important problem
capacity of cables buried directly in the will require a great deal of scientific and
soil, since the thermal conductivity of the engineering research, which can, how-
soil has an important bearing on this very ever, be based profitably on the theoreti-
important problem. cal foundations which have already been
A MEMBER.—I do not get the sig- laid.
nificance of measuring soil thermal con-
ductivity. Various results are secured de- 2
W.O. Smith, "Thermal Transfer of Mois-
pending on the investigator, varying in ture in Soils," Transactions, Am. Geophysical
ratio of two to one. That makes a lot of Union,
3
Vol. 25, pp. 511-523 (1944).
W. O. Smith and H. G. Byers, "The Ther-
difference. mal Conductivity of Dry Soils of Certain of the
MR. HANS F. WINTERKORN (author's Great Soil Groups," Proceedings, Soil Science
Soc.4 Am., Vol. 3, pp. 13-19 (1938).
1
Consulting Cable Engineer, Anaconda Wire W. O. Smith, "Thermal Conductivity of Dry
and Cable Co., Hastings-on-Hudson, N. Y. Soil," Soil Science, Vol. 53, pp. 435-459 (1952).
S R KUMAR KONATHALA (Indian Inst of Tech 36 Chennai) pursuant to License Agreement. No further reproductions
A LOW-HEAD PERMEAMETER FOR TESTING GRANULAR MATERIALS
BY E. G. YEMINGTON1
Highway drainage designs generally DRAINAGE-LAG PERMEAMETER

have been based upon empirical rules,
and only in recent years has the impor- A schematic drawing of this perme-
tance of obtaining permeability data on ameter developed for the measurement
granular materials been stressed for the of coefficients of permeability for open-
design of sand drains and blanket courses graded granular materials is shown in
used in embankments for assisting in the Fig. 1. The outer cylinder, or metal tank,
removal of water by consolidation of the is closed at one end and is equipped with
subsoils. In 1952, the Bureau of Public a quick-opening valve near the bottom.
Roads published a report2 on highway An inner cylinder, 6 in. in diameter, is
subdrainage. The report did not estab- used to hold the test sample, and has a
lish design criteria but it did present test wire mesh attached to the bottom to
methods, a new type of low-head drain- support the sample. Generally, No; 10
age-lag permeameter, and data on the wire mesh is used with No. 200 wire mesh
permeability and drainability of soils and inserted above. The sample cylinder is
their applications to highway subdrain- supported on a base ring which has
age. perforated walls. Provision is made, by
This paper presents additional test means of an Ames dial and a hook gage,
data and information on the above-men- to measure accurately the height of the
tioned drainage-lag permeameter for the water level in the smaller cylinder at the
measurement of permeability coefficients start and completion of the permeability
of open-graded granular materials. Com- test. Dimensional characteristics are such
parisons are made between the coeffi- that a hydraulic gradient of approxi-
cients of permeability obtained by this mately 0.2 can be readily maintained
device and an ordinary falling-head during testing.
permeameter for eight gradings of granu- The 6-in. cylinder with wire mesh at-
lar material. The results of these tests tached is placed on a solid plate; then a
indicate that the drainage-lag type of representative sample of the granular
permeameter minimizes turbulence in material to be tested is placed in the
permeability testing and can be used to cylinder and compacted to the desired
measure higher permeability values than density by either an impact or static
the ordinary falling-head type of per- method. The cylinder containing the
meameter.
1
compacted test sample is attached to the
Highway Physical Research Engineer, base ring in the metal tank and the hook
Physical Research Branch, Bureau of Public
Roads, Washington, D. C. gage assembly attached, as shown in
2
E. S. Barber and C. L. Sawyer, "Highway Fig. 1. Water is added to the tank slowly
Subdrainage,"
CopyrightPublic Roads, Vol. 26,reserved);
No. 12,Sun Jan 11 13:11:58 EST 2015
February, 1952.
by ASTM Int'l (all rights until the sample is completely inundated.
37 Chennai) pursuant to License Agreement. No further reproductions a
FIG. 1.—Drainage-Lag Permeameter.
After coming to an equilibrium, the derived3 by integration, assuming a con-

water level is determined with a hook
gage. The hook gage is then lowered an 8
Flow from tank,
arbitrarily selected amount, h. After
measuring the temperature of the water, Flow through sample,
the quick-opening valve at the bottom of
Continuing by eliminating d/t,
the tank is opened and the outflow
caught. A stopwatch is used to determine
the tune required for the inner water
level to reach the hook gage, at which Integrating, when
time the watch is stopped and the valve
closed simultaneously. The effective head
is a variable, H — h, where H is the Substituting
change in water level hi the metal tank
during the test. The tune, discharge,
change in water level inside the sample Letting
cylinder, and dimensions of the sample
and apparatus are used to determine the
coefficient of permeability, k. This de-
termination is facilitated by the use of where F is defined by hS/Q = 1
Fig. 2.Copyright
The formula for Int'l
by ASTM k in(all
Fig. 2 was
rights reserved); Sun Jan 11 13:11:58 EST 2015
YEMINGTON ON LOW-HEAD PERMEABILITY OF SOILS 39
stant rate of discharge, from the equa- function of temperature, the calculated
tion: permeability at an arbitrary temperature
(68 F (or 20 C)) is often reported as k& =
Ck, where C is the viscosity of water at
where: the test temperature divided by its vis-
q = volume of flow per unit time, cosity at 68 F. Values of C are plotted in
k = coefficient of permeability, Fig. 3.
FIG. 2.—Permeability from Drainage-Lag Device.

A = gross area of soil sample perpen- FALLING-HEAD PERMEAMETER
dicular to direction of flow, and
hd = head loss through a distance, d, of The ordinary falling-head permeam-
the soil in the direction of flow. eter used for comparison purposes, shown
Since the permeability depends upon in Fig. 4, was operated with an hydraulic
the viscosity of the water, which is a gradient of approximately 2.0. The co-
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No furth
FIG. 3.—Effect of Temperature of Water on the Viscosity Ratio.
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further repr
YEMINGTON ON LOW-HEAD PERMEABILITY OF SOILS 41
efficient of permeability is determined methods, it is desirable to eliminate the
from the formula: variable of air because of the difficulty of
controlling and measuring the degree of
saturation. Removal of the air was ac-
where: complished by placing each immersed
a = area of standpipe, sample in a large tank and subjecting it
d = thickness of sample, to a vacuum. In the drainage-lag device,
FIG. 4.—Ordinary Falling-Head Permeameter.

A = area of sample, the vacuum can be applied through a lid
t = time, fitting the top of the metal tank.
hi — initial head, and Compacted samples of each of eight
hz = final head. materials were tested alternately by each
Since it was necessary to pour water method. Table I shows the gradations of
rapidly into the device to build up hi, a the materials tested and their coefficients
baffle was placed above the sample. of permeability as determined through
use of the above-mentioned devices. It is
TEST RESULTS evident that the coefficients of perme-
ability of samples Nos. 1, 2, and 3 are
To simulate field conditions, it is often not influenced by the type of testing
necessary to measure the coefficients of apparatus used. The coefficients of per-
permeability
Copyrightof
by unsaturated
ASTM Int'l (all materials. meability
rights reserved); for13:11:58
Sun Jan 11 samples Nos.
EST 20154 and 5 vary
In thisDownloaded/printed
research to compare
by two test considerably for the two types of testing
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further r
apparatus. The higher values as de- during the immersion of the sample.
termined by the drainage-lag apparatus Additional loss of fines during testing was
are due to less turbulence than that considerably less in the drainage-lag
occurring when the ordinary falling-head permeameter than in the ordinary fall-
device is used. ing-head permeameter. Movement after
TABLE I.—GRADATION OF MATERIALS TESTED AND THEIR COEFFICIENTS OF
PERMEABILITY AS DETERMINED BY TWO TYPES OF PERMEAMETERS.
Sample Sample Sample Sample Sample Sample Sample Sample
No. 1 No. 2 No. 3 No. 4 No. 5 No. 6 No. 7 No. 8
Percentage passing:
1%-in. sieve 100
1-in. sieve 0
:Hj-in. sieve 100 100
%-in. sieve 96 0
No. 4 sieve 100 20
No. 10 sieve 100 100 100 100 2 1
No. 40 sieve 80 72 1 1 0 0
No. 200 sieve 14 1 0 0 .. •
Dry density, Ib per cu f t 90 113 113 102 102 101 101 104
Coefficient of permeability, ft per
day:
Ordinary falling-head perme-
ameter 18 129 624 1936 6380 a a a
Drainage-lag permeameter 18 129 629 2466 7775 9660 81 80S6 105 6256
0
Unable to determine permeability, fall too rapid.
b
No. 200 wire mesh removed.
Between 6000 and 7000 ft per day ap- immersion was minimized in the drain-
peared to be the maximum values that age-lag device by pouring water into the
could be measured with the ordinary tank at a constant rate to give an upward
falling-head permeameter, whereas much gradient.
higher values were measurable with the
drainage-lag permeameter. For samples CONCLUSION
Nos. 6, 7, and 8, it was necessary to in-
crease the porosity of the ring supporting The drainage-lag device provides a
the sample container. means of measuring the permeability of
Several tests were run with the No. 200 very coarse materials up to 100,000 ft per
wire mesh removed to measure the effect day. The effect of turbulence and the
of gradients on the movement of fines in movement of fines is minimized by the
poorly graded materials. It was observed use of small gradients comparable to
that the greatest loss of fines occurred those occurring in the field.
PERMEABILITY TEST FOR SANDS
BY T. Y. Cnu,1 D. T. DAVIDSON,1 AND A. E. WiCKSTROM1
SYNOPSIS
Information on the permeability of sand is often required in its utilization as
an engineering material. This paper presents a modification of Barber's falling
head permeability test for porous granular materials. The modified method is
simple to use and gives reproducible results.
The modifications recommended include: (1) a procedure for loading and
compacting the sample hi the permeameter tube, (2) a carbon dioxide treat-
ment to remove ah" from the sample, and (3) the repeated testing of the loaded
sample until permeability values obtained are within a recommended range of
variation.
The suggested permeability test was used to measure the permeability of
several Iowa sands at various densities. Compositional information pertinent
to the permeability of the sands is also presented.
Barber's falling head permeability In developing the modified perme-

test (I)2 has been widely used for the ability test for sands, experiments were
laboratory determination of the perme- made for the following purposes: (1) to
ability of sands.3 This method is simple evaluate different methods of loading and
to use and does not require complicated compacting the sample in the permeame-
equipment. However, test results ob- ter tube, (2) to determine the effect of
tained may not be readily reproducible, air entrapped in the loaded sample on
mainly because of the lack of a specific permeability, (3) to determine the effect
procedure for loading the sample in the of soaking the loaded sample in water for
permeability tube. This paper presents a varying periods of time prior to testing,
modification of Barber's test method as and (4) to compare the permeability
set forth in complete form in the Ap- values obtained by repeated tests on the
pendix, which method is easy to use and same loaded sample.
which gives reproducible results.
Loading and Compaction:
1
Assistant Professor of Civil Engineering,
Associate Professor of Civil Engineering, and In Barber's permeability test, the
Graduate Assistant, respectively, Iowa Engi- loading is done by pouring an air-dried
neering Experiment Station, Iowa State College, sample into a 2-in. diameter perme-
Ames, Iowa.
2
The boldface numbers in parentheses refer ameter tube to a depth of 6 in. Since the
to the list of references appended to this paper, tube is usually 48 in. long, the pouring of
see p. 50. dry sand in this manner may cause ap-
a
For definition, see Definition of the Term
Sand (C58-28), 1952 Book of ASTM Stand- preciable segregation of coarse and fine
Copyright
ards, Part by ASTM Int'l (all rights reserved); Sun Janparticles.
3, p. 988. Segregation
11 13:11:58 EST 2015 can be avoided by
S R KUMAR KONATHALA (Indian Inst of Tech 43
Chennai) pursuant to License Agreement. No further reproductions authorized
loading the sample in a moist condition, obtained by compacting the sample with
but it may be difficult to compact the a cylindrical weight as described in the
moist sand to a desired density. Appendix. Sands can be easily compacted
A comparison was made of different to Standard Proctor density by this
techniques for loading dry sand in the method.
permeameter tube. The procedure called
the "inverted method" was found to be Carbon Dioxide Treatment:
most satisfactory. In this method, the The permeability test can be per-
permeameter tube is inverted and the formed with the sample in either a com-
sample is loaded in what is then the upper pletely saturated condition or a partially
6 in. of the tube. After loading, the tube saturated condition. The condition of
is turned right side up for the perme- testing will usually depend upon the
ability test. The details of the loading purpose of the test. The simplest way to
procedure are given in the Appendix. obtain results on a comparable basis is to
TABLE I.—COMPARATIVE REPRO- use completely saturated samples. The
DUCIBILITY OF PERMEABILITY TEST0 permeability test for sands given in the
RESULTS OBTAINED WITH SAMPLES Appendix was developed primarily for
LOADED BY TWO DIFFERENT METH-
ODS. testing samples in the condition of com-
plete saturation. However, the procedure
Coefficient of
Method of Loading Experiment Permeability, for loading and compaction may also be
ft per day
used for preparing samples for testing in
Inverted method No. 1 28.2
the partially saturated condition.
No. 2 29.2 To obtain complete saturation, the
No. 3 29.4 sample should be free from entrapped
No. 4 30.9 air before being tested for permeability,
Sample poured from top No. 1 16.8 and no air should be introduced into the
to bottom of perme- No. 2 25.3 sample from the percolating water during
ameter tube No. 3 12.0
No. 4 19.6 testing. The entrainment of air in the
test sample can be prevented by evacua-
"The data are for sample No. 77-S in Ta- tion methods (2) or by passing carbon
ble III.
dioxide gas6 through the air-dry sample
The typical test data given in Table I after it has been loaded in the perme-
illustrate the high degree of reproduci- ameter tube and before it is compacted.
bility in test results obtainable by the The latter method is comparatively
inverted method of loading. The low simple to use and was found satisfactory
reproducibility of results obtained with by Christiansen (3) and others in per-
similar samples loaded by the method of meability tests with many types of soil.
pouring the sample from top to bottom It was the only method studied in the
of the permeameter tube are shown for present investigation.
comparison. All samples were tested for The two variables involved in the
permeability according to Barber's pro- carbon dioxide treatment are the rate
cedure. of flow and the duration of flow of carbon
The density of a sample loaded by the dioxide passed through the sample. Ex-
inverted method is close to the compact periments were performed with sand
unit weight.4 Higher densities may be samples to determine the satisfactory
4 6
Method of Test for Unit Weight of Aggre- The air in the sample is displaced by carbon
gate (C29-42),
Copyright1952 Book ofInt'l
by ASTM ASTM Stand-reserved);
(all rights dioxide Sun
whichJanwill
11 be quickly EST
13:11:58 dissolved
2015by per-
ards, Part 3, p. 975.
Downloaded/printed by colating water during the permeability test.
CHU, DAVIDSON, AND WICKSTROM ON SAND 45
quantities of the gas to be used. The meability due to the carbon dioxide
values recommended in the test method treatment. These observations are illus-
in the Appendix seem adequate. trated by the typical data given in Table
De-aired water may be used in perme- II.
ability tests to avoid the release of dis-
solved air as the water percolates through Soaking:
the test sample. The use of distilled water After the sample has been loaded in
at a temperature 5 to 10 F higher than the permeameter tube, treated with
room temperature, as recommended in
carbon dioxide and compacted to the
Barber's method, also serves to avoid the
desired density, the tube is immersed in
release of dissolved air. Because of its distilled water in a water bath (see Ap-
simplicity this method is recommended.
pendix). The permeability determination
TABLE II.—TYPICAL DATA SHOWING can be started either immediately after
EFFECT OF CARBON DIOXIDE TREAT- the water in the tube has risen nearly to
MENT ON THE PERMEABILITY OF
SANDS.- the water level in the bath or after the
sample has been soaked for an additional
Coeffi- period of time. Experimental results
Carbon cient of Rela-
tive
Sample Porosity, Dioxide Permea-
per cent Treat- bility,6 Permea- indicate that additional soaking up to 48
bility,"
ment ft per per cent hr does not appreciably change the per-
day
meability value. For this reason, a soak-
No 29.4 76 ing period prior to the permeability
39.6 (
Yes 38.8 100 determination is not considered neces-
No. 77-S sary.
36.6 |
No 12.0 49
Yes 24.6 100
Change in Permeability During Testing:
No 160.2 89
No.89-S 33.8 |
168.7
Yes 100 In Barber's method of test, the per-
0
Properties of samples used are given in
meability is determined on the basis of
Table III. the time required for the water in the
6
e
Data are the average of at least two tests. permeameter tube to drop from level A
Relative permeabilities are computed by
taking the coefficient of permeability obtained to level C as shown in Fig. 8 in the Ap-
with carbon dioxide-treated samples as 100 per pendix. Any change in the permeability
cent. of the sample during this time cannot be
Experiments were made to determine detected, since only one coefficient of
the effect of air removal by the carbon permeability is computed. If an inter-
dioxide treatment on the permeability of mediate level, level B in Fig. 8, is marked
several sand samples. The samples were on the permeameter tube and the time
loaded in the permeameter tube by the intervals required for the water to drop
inverted method (see Appendix) and from level A to level B and from level B
were compacted to various densities to level C are recorded, two coefficients
prior to permeability determinations by of permeability can be computed. If this
Barber's procedure. Test data showed is done, any change in the permeability
that the carbon dioxide treatment results of a sample during testing is indicated
in a higher permeability regardless of the by the difference between the two coeffi-
type of sand and the degree of compac- cients of permeability.
tion. The data also indicated that the To facilitate the comparison of the two
lower the permeability of the untreated coefficients of permeability, an intermedi-
sample Downloaded/printed
the greater the increase
by in per- ate level can be selected to conform to
46 SYMPOSIUM ON PERMEABILITY OP SOILS
the following relationship: readings is an indication of a change in

the permeability of the sample during
testing. Test results indicate that, when
a single permeability determination is
made on a sample, the first reading is
where: usually different from the second. It
h& = height6 of level A, was also found that the deviation be-
hb = height of the intermediate level tween the two readings often decreases
(level B), and with repeated permeability determina-
hc = height of level C. tions on the same loaded sample. This is
As shown in Fig. 8, h& = 36 in., hb = illustrated in Fig. 1. The deviation curve
FIG. 1.—Typical Curves Showing Variation in Permeability Test Results Obtained by Repeated
Testing of a Sand Sample (Sample No. 92-S in Table HI).
Per cent of deviation =
where:
Ti = time for water to drop from level A to level B, and
Ti = time for water to drop from level B to level C.
20f-f in., and he = 12 in. Equation 1 and shows that after several successive per-
the formula for computing the coefficient meability determinations the amount of
of permeability given in the Appendix deviation between the two readings be-
show that, if the permeability of the comes relatively small and remains
sample is constant during the test, the nearly constant.
time intervals required for the water in Figure 1 further shows that the perme-
the tube to drop from level A to level B ability increases with repeated testing
and from level B to level C will be the but approaches a uniform value as the
same. For simplicity the tune required deviation curve flattens. The perme-
for the water to drop from level A to abilities obtained from tests having small
level B will be referred to as the first
deviations between first and second read-
reading, and the time from level B to
level C the second reading. Any differ- ings are, probably, more representative
ence between the first and the second of the test sample than the permeabili-
ties obtained from tests having large
6
AllCopyright
heights by
areASTM Int'l (all
measured from rights
the reserved);
water Sun Jan 11 13:11:58 EST 2015
level inDownloaded/printed
the bath. by deviations. For this reason, it seems de-
TABLE III—MECHANICAL COMPOSITION OF FOUR IOWA SANDS.
Textural Composition0 Surface Texture6 Aggregation Characteristics
Sorting Specific Average
Sample Coefficient, Gravity Sphericity
Sand, Silt, Clay, So Cement
per cent per cent per cent Markings Coatings Amount Size, mm
No. 77-S 87.6 9.6 2.8 1.6 2.66 0.71 Dull and Partially coated by Common 1-3 Calcareous
rough calcareous clay clay
(faceted)
No. 79-S 84.7 11.0 4.3 1.4 2.65 0.73 Dull and Partially coated by Abundant H-iH Ferruginous
smooth ferruginous clay clay
No. 89-S 98.1 1.6 0.3 1.2 2.67 0.75 Dull and Partially coated by Absent
rough iron oxide
No. 92-S 92.9 4.3 2.8 1.3 2.68 0.61 Dull and Completely coated Common 1-2 Slightly cal-
rough by slightly calcareous
(pitted) careous clay clay
«b Sand—2 to 0.074 mm; silt—0.074 to 0.005 mm; clay—less than 0.005 mm.
Descriptions apply to most sand-size particles in the sample.
FIG. 2.—Relationship Between Permeability Value and Deviation Between First and Second
Readings.
Per cent of'permeability -= ^ X 100
£0
where:
*» = coefficient of permeability corresponding to a deviation greater than ±0.5 per cent, and
to = coefficient of permeability of the same test sample corresponding to a deviation less than ±0.5 per cent.
Per cent of deviation is defined in Fig. 1.
CopyrightFIG. 3.—Particle-Size
by ASTM Accumulation
Int'l (all rights Curves
reserved); Sun Janfor
11 the Four Iowa
13:11:58 Sands.
EST 2015
sirable in routine permeability tests to cent variation hi the permeability value
repeat the determination until the devi- seems to be a reasonable tolerance, ±2.0
ation between the first and second read- per cent is recommended as an allowable
ings is small and then to report the limit of deviation between first and
coefficient of permeability obtained by second readings. In many tests, this
this determination. requirement was met in the second per-
To determine the permissible deviation meability determination.
of first and second readings, an analysis
was made of permeability and deviation PERMEABILITY OP FOUR IOWA SANDS
data for the four sand samples (see Table The test method presented in the
III) at various densities. Figure 2 shows Appendix was used hi determining the
the relationship between permeability
values and deviations. The permeability TABLE IV.—COEFFICIENTS OF PER-
MEABILITY OF FOUR IOWA SANDS AT
VARIOUS DENSITIES.
Dry Density
Poros- Coefficient
of Perme-
Sample Per cent ity,
of ability,
Ib per cu Standard per cent ft per day
ft Proctor
Density
100 92 39.6 39.5

No. 77-S 105 97 36.6 24.5
109 100 34.5 18.4
106 90 35.6 6.9

No. 79-S 113 96 31.5 2.4
119 101 27.8 1.4
103 96 38.1 258.5

No. 89-S 104 97 37.5 237.5
Porosity, per cent 108 100 35.5 199.4
FIG. 4.—Permeabilities of the Four Iowa 110 102 34.0 175.4
Sands at Various Porosities.
97 90 42.0 77.1
values are expressed on a percentage No. 92-S 99 92 40.8 66.3
103 96 38.3 54.5
basis. The. coefficient of permeability 107 100 35.7 34.1
corresponding to a deviation less than
±0.5 per cent is taken as 100 per cent
permeability and is represented by the permeability of four Iowa sands. The
small triangle in the upper left of the source and properties of the sands were
figure. The per cent of permeability cor- reported in a previous paper (4). Com-
responding to deviations of more than positional information pertinent to their
±0.5 per cent is computed by the equa- permeability is given hi Fig. 3 and hi
tion given in Fig. 2. The computed values Table III. The sorting coefficient, SQ in
for the four sand samples are plotted as the table, is computed by the formula
dots in the figure; each dot represents one (5):
permeability determination. As shown by
the dots within the rectangular dashe.d-
line area, when the deviation is less than
±2.0 per cent, the permeability value is The values of Q* and Qb are determined
within Copyright
about ±5by per
ASTM cent
Int'lof(all
therights per from
100reserved); the
Sun Jan particle-size
11 13:11:58 EST 2015accumulation
cent permeability value. Since a 5 per curve: Q& is the maximum diameter of
the smallest 75 per cent by weight of the Each curve in Fig. 4 shows the de-
soil particles, and Q\, is the maximum crease hi permeability that accompanies
diameter of the smallest 25 per cent. A a decrease hi porosity. The figure also
well graded sand will have a higher sort- illustrates that porosity is not the only
ing coefficient and, usually, a lower important factor affecting permeability.
porosity than a poorly graded sand. The For example, sample No. 79-S at a poros-
average sphericity values are for sand- ity of 35 per cent has a coefficient of
size particles in the sample, as deter- permeability of about 6 ft per day, but
mined by the Rittenhouse chart (6). A sample No. 89-S tested at a similar
perfect sphere has a sphericity of 1.00. porosity has a coefficient of about 200 ft
The predominant kind of clay mineral in per day. This great difference hi perme-
each of the four sands is illite. ability is largely due to the differences in
Permeability test results for the four mechanical composition as shown hi Fig.
sands at various densities are sum- 3 and Table III.
marized hi Table IV and plotted in Fig.
4. The data show that the sands differ A cknowledgment:
greatly in their coefficients of perme- The subject matter of this paper was
ability and the coefficient varies with the obtained as part of the research being
density for each sand. done under Project 283-S of the Iowa
The differences in permeability of the Engineering Experiment Station of Iowa
sands are related to their void charac- State College. This project, entitled
teristics. The term 'Void characteristics" "The Loess and Glacial Till Materials of
as used here refers not only to the amount Iowa: An Investigation of Their Physical
of voids in a sample but also to other and Chemical Properties and Techniques
related variables, such as the size dis- for Processing Them to Increase Their
tribution and continuity of the voids. All-Weather Stability for Road Con-
The number of voids in a sample can be struction," is being carried on under
determined easily and is commonly ex- contract with the Iowa State Highway
pressed hi terms of the porosity of the Commission and under the sponsorship
sample. The other variables are difficult of the Iowa Highway Research Board.
to measure quantitatively, but they can The project is supported by funds sup-
be evaluated qualitatively from composi- plied by the Commission and the U. S.
tional information. Bureau of Public Roads.
REFERENCES
(1) E. S. Barber, "Suggested Method of Test for (4) W. W. Williams, D. T. Davidson, and T. Y.
Permeability of Porous Granular Materials," Chu, "Properties of Five Iowa Fine Sands,"
Procedures for Testing Soils, Am. Soc. Test- Proceedings, Iowa Acad. Sci., Vol. 60, pp.
ing Mats., p. 177 (1950). 442-464 (1953).
(2) T. W. Lambe, "Soil Testing for Engineers," (5) W. C. Krumbein and F. J. Pettijohn, "Man-
John Wiley & Sons, Inc., New York, N. Y., ual of Sedimentary Petrography," Appelton-
pp. 52-62 (1951). Century-Crofts, Inc., New York, N. Y.
(1938).
(3) J. E. Christiansen, M. Fireman, and L. E. (6) G. Rittenhouse, "A Visual Method of Es-
Allison, "Displacement of Soil-Air by COs timating Two Dimensional .Sphericity,"
for Permeability Tests," Soil Science, Vol. Journal of Sedimentary Petrology Vol. 13,
61, No. 5, pp. 355-360 (1946). No. 2, pp. 79-81 (1943).
APPENDIX
METHOD OF TEST FOR PERMEABILITY OF SANDS
Scope: Ruler.—A ruler reading to A in.

Thermometer.—A thermometer accurate
This method describes test procedures for
to 1 F or 0.5 C.
determining the coefficient of permeability
Stop Watch.—A stop watch reading to
of sands as defined in the ASTM Standard
i sec.
Definition of the Term Sand (C 58) .3
Water Bath.—A water bath with at-
Apparatus: tachments for holding permeameter tube
as shown in Fig. 8.
The apparatus used in this test consists of
the following: Loading of Test Sample:
Permeameter Tube.—A rigid trans-
parent plastic tube approximately 48 in. A representative air-dry sample of suffi-
in length and about 2 in. in inside di- cient quantity, usually 500 g, is prepared
ameter. The tube is marked as shown in and its weight is recorded. The sample is
Fig. 8. mixed until uniform. To obtain the oven-dry
Brass Screens.—Two circular brass weight of the sample, the air-dry weight is
screens. One is of No. 200 sieve cloth, the corrected for hygroscopic moisture.
other of No. 10 sieve cloth. The diameters The sample is loaded in the permeameter
of the screens are the same as the outside tube in four layers, each approximately 1.5
diameter of the permeameter tube. in. thick. The loading funnel and supporting
Screen Holder.—A holder to attach the device shown in Fig. 5 are used in loading the
screens tightly to the permeameter tube. sample by the following procedure:
Supporting Device.—A supporting de- 1. Invert the permeameter tube, and as-
vice made of rubber stoppers, a brass tube, semble loading apparatus as shown in Fig. 5.
etc., as shown in Fig. 5. For loading the first layer, the stem of the
Loading Funnel.—A special funnel as funnel rests on the top of the supporting
shown in Fig. 5. device.
Tamping Rod.—A straight f-in. round 2. Pour a portion of the sample sufficient
metal rod, approximately 8 in. long. to make a layer approximately 1.5 in. thick
Attachment for Carbon Dioxide Treat- into the stoppered funnel. If segregation oc-
ment.—An attachment made of a rubber curs as the sample is being poured, remix to
sleeve, rubber stopper, etc., as shown in a uniform condition in the funnel.
Fig. 5. 3. Pull the stopper and allow the sand to
Bottle of Carbon Dioxide and Flow Meter. flow down into the stem of the funnel. The
—A 50-lb tank of liquid carbon dioxide small hole in the neck of the funnel permits
with a flow meter, accurate to 0.01 cu ft the escape of air from the stem.
per min, attached. 4. Carefully remove the funnel. The re-
Cylindrical Weight.—A 4-lb steel or sulting layer will usually be free from segre-
brass weight having a diameter of about gation.
If in. 5. Use the tamping rod to rod the layer
Balance.—A balance of 1 kg capacity five times. The rodding is distributed uni-
Copyright by
and accurate to ASTM
0.1 g.Int'l (all rights reserved); Sunformly
Jan 11 13:11:58
over theEST 2015
layer and extends through it.
51 Chennai) pursuant to License Agreement. No further reproduction
The above procedure is followed in load- the top of the permeameter tube, the one of
ing each of the other three layers. After all No. 200 sieve cloth being in direct contact
four layers are loaded, any excess sand above with the loaded sample. The permeameter
FIG. 5.—Âpparatus for Preparing Sample for Permeability Determination.

(a) Apparatus for loading sample in permeameter tube.
(6) Apparatus for carbon dioxide treatment.
the t6p of the permeameter tube is removed tube is then turned right side up and the
and saved for weighing. The weight of sand supporting device removed. In removing the
in the permeameter tube is determined by supporting device, the following procedure
subtracting the by
Copyright weight
ASTMofInt'l
all (all
excess
rights sand is Sun
reserved); used:Jan 11 13:11:58 EST 2015
from theDownloaded/printed
initial weight of by
the sample. 1. Lift the rod inside the brass tube until
Two Sbrass screensKONATHALA
R KUMAR are attached tightly to ofthe
(Indian Inst short
Tech crosspursuant
Chennai) bar is clear of theAgreement.
to License slots, andNo further r
Downloaded/printed by FIG. 7.—Arrangement of Apparatus for Testing a Group of Samples.
FIG. 6.—Carbon Dioxide Treatment of Sand Samples.
then rotate it about 90 deg until the cross cylindrical weight is placed on the sample in
bar can be rested on the end of the brass the permeameter tube. The portion of the
tube. This step is necessary to prevent dis- tube containing the sample then is patted
turbance of the sample during removal of the with the palm of the hand until the desired
supporting device. density is obtained.
2. With the permeameter tube hi a verti-
cal position, remove the whole supporting NOTE.—Since both the weight of the sample
device slowly from it. in the tube and the inside diameter of the tube
are known, the depth to which the sample is
Carbon Dioxide Treatment: compacted can be computed from the desired
density.
The sample in the permeameter tube is
treated with carbon dioxide gas in the man-
Permeability Determination:
The sample in the permeameter tube is

immersed in distilled water at room tem-
perature by one of the following procedures:
Method A.—For samples loaded hi the
permeameter tube without additional com-
paction.—After the carbon dioxide treat-
ment, the sample is gently immersed hi the
water bath. The permeameter tube is posi-
tioned as shown in Fig. 8.
Method B.—For samples loaded and
further compacted in the permeameter tube.
—The sample is very gently immersed in the
water bath with the weight used for com-
paction on top of the sample to prevent ex-
pansion during immersion. The weight is re-
moved when the rising water nears the top
of the sample. The position of the perme-
ameter tube in the water bath is as shown
in Fig. 8.
After the water in the penneameter tube
FIG. 8.—Arrangement of Apparatus for Per- has risen to the level of the water bath, the
meability Determination. tube is filled with distilled water at a tem-
peraturen betwee 5 and 10 F above room
ner shown in Figs. 5 and 6. The rate at which temperature, without disturbing the sample
the carbon dioxide gas is passed through the in the tube. The time intervals required for
sample is 0.03 cu ft per min. During the first the water in the tube to drop from level A
| min of the carbon dioxide treatment, valve to level B and from level B to level C are
A is left open to permit the escape of air from recorded with a stop watch.
the chamber below the sample. The valve If the difference between the first and
is then closed, and the treatment is con- second tune intervals is more than 2.0 per
tinued for an additional 15 min. The amount cent of the average value, the preceding step
of carbon dioxide passed through the sample is repeated one or more tunes until the
is about 0.45 cu ft. difference is less than 2.0 per cent. The depth
and the temperature of the sample hi the
Compaction (Optional):
permeameter tube are then measured.
After the carbon dioxide treatment, the
sample is compacted, if necessary, to the de- NOTE.—Figure 7 illustrates the assembly of
sired density before being tested for perme- apparatus for testing a group of samples simul-
Copyright by ASTM Int'l (all rights reserved);
ability. In compacting the sand, the 4-lb taneously. Sun Jan 11 13:11:58 EST 2015
Calculation: ha, = height of level A from the water level

The coefficient of permeability, k, in feet in the bath,
per day, is calculated as follows: he = height of level C from the water level
in the bath, and
t = time in minutes required for water to
c drop from level A to level C during the
where: last permeability determination.
C — temperature correction (viscosity of
water at the temperature measured in NOTE.—For h& = 36 in. and hf = 12 in.,
the above formula can be simplified to:
the sample after testing divided by
viscosity of water at 68 F),
d = depth of sample in inches measured
after the test,
THE PERMEABILITY OF FINE-GRAINED SOILS
BY T. WILLIAM LAMBE*
Importance of Permeability Measure- adsorbed water (7).2 A look at some of the
ments: factors influencing the permeability of
The permeability of a soil is one of its fine-grained soils, therefore, is timely.
most fundamental and important proper- Theoretical Permeability Equations:
ties. It enters into nearly all seepage,
settlement, and stability problems con- From a comparison of flow through
fronting the soil engineer. The amount of soils with flow through capillary tubes,
leakage through and under dams, the the following equation (8) was developed:
rate at which a building settles, and the
rate at which the strength of a deposit
increases after it has been subjected to a where:
consolidating pressure are typical of the k = the Darcy coefficient of perme-
many problems in which the permeability ability, or simply permeability,
of a soil can be a critical factor. DB = some effective particle diameter,
The importance of evaluating the 7 = unit weight of permeant,
permeability of a pervious soil has been p = viscosity of permeant,
long recognized and test techniques for e = void ratio, and
measuring it have been well developed C — shape factor.
and are widely used. The permeability An expression for the permeability of
of fine-grained soils, however, has not porous media, proposed by Koêny (2)
received extensive study. Soils with and improved by Carman (1) is:
permeabilities of less than 1 /t per sec are
often considered "impervious" and are
not subjected to permeability testing.
Soil permeabilities less than 1 n per sec where:
are becoming increasingly important to k0 = constant depending on pore shape
soil technologists from both practical and and ratio of -length of actual flow
theoretical considerations. More use is path to soil bed thickness, and
being made of "impervious" soil to line S = specific surface area.
canals and reservoirs and to construct Since Z>8 is denned as the diameter of
cores for earth dams. Research workers particle having specific surface S, Eq 1
are studying the permeability of fine- may be considered a simplification or
grained soils to learn more of the nature extension of the Kozeny-Carman equa-
of particle surfaces and the thickness of tion.
1
Associated Professor and Director of the * The boldface numbers in parentheses refer
Soil Stabilization Laboratory, Massachusetts to the list of references appended to this paper,
Copyright by ASTM Int'l (all rights
Institute of Technology, Cambridge, Mass. reserved); Sun
see Jan 11 13:11:58 EST 2015
p. 67.
S R KUMAR KONATHALA (Indian Inst of56 Tech Chennai) pursuant to License Agreement. No further reprodu
LAMBE ON FINE-GRAINED SOILS 57
These equations have been found to ing fine-grained soils; one can also argue
express accurately the permeability that the equations are sound but that
characteristics of saturated sand. Labora- the knowledge of soils is not extensive
tory tests suggest that a plot of degree of enough to permit proper interpretation
saturation versus permeability of sand of the equations.
FIG. 1.—Void Ratio versus Permeability. From Cornell University report (12).
approximates a straight line (4). The The purpose of this paper is to present
various factors affecting the permeability and discuss briefly the major factors
of cohesionless soils are, therefore, rela- affecting the permeability of fine-grained
tively well known. soils and to show that these factors are
The permeability equations are of not covered in a correct, or at least an
very limited use to the soil engineer for interpretable, manner by the Carman-
fine-grained soils for two reasons: (1) the Kozeny equation. The factors to be ex-
difficulty of selecting the effective "con- amined in this paper are:
stants" and soil characteristics, and (2) 1. Soil composition.
the fact that these various terms are not 2. Permeant characteristics.
independent, but interrelated hi a very 3. Void ratio.
complexCopyright
manner.byOne
ASTM Int'l (all rights reserved); Sun Jan 11 13:11:58 EST 2015
can well argue for 4. Structure.
discarding the equations when consider- 5. Degree of saturation.
EFFECT OF SOIL COMPOSITION is one of the least permeable soil minerals

and is therefore widely used by the
The influence of soil composition on
engineer as an impermeabilizing additive
permeability is generally of little im- to other soils.
portance (mica and organic matter are
The magnitude of permeability vari-
two of the exceptions) with silts, sands,
ation with soil composition ranges
and gravels; it is of major importance widely. Figure 1 shows the ratio of
with clays. Clay "composition" includes permeability of calcium montmorillonite
minerals, exchangeable ions, and im-
to that of potassium montmorillonite at
purities (free iron oxides and organic
a void ratio of 7 is approximately 300. It
matter, for example). further shows that the permeability of
At Cornell University consolidometer kaolinite can be 1000 times that of
test specimens were subjected to variable montmorillonite. The lower the ion ex-
head permeability measurements and the change capacity of a soil, the lower, of
data in Fig. 1 obtained. The permeabili-
course, the effect of exchangeable ion on
ties of the various ionic forms of mont- permeability.
morillonite at a void ratio of 8 vary as:
There is no term3 in the permeability
Ca > H > Na > K equation that directly covers composition
or indirectly covers exchangeable ions.
At a void ratio of 1.5 for kaolinite: A composition term with a value range
for each soil mineral group could be
H > Ca > K > Na added to the equations. Since the
mineralogy of a soil is seldom known, the
A comparison of permeabilities at use of such a term might be limited.
constant consolidation pressure is prob-
ably more indicative than one at constant EFFECT OF PERMEANT
void ratio. Such a comparison can be
approximated from the coefficients of Michaels and Lin (7) studied the
consolidation listed by the Cornell report permeability of saturated kaolinite to
for kaojinite, namely, various fluids. They plotted their results
in terms of the absolute permeability,
H > Fe > Mg > Ca > Na which is equal to the Darcy coefficient
From the Cornell and other data (for of permeability times viscosity over
example, Winterkorn (10) and Winter- density. If the permeability equations
korn and Moorman (11)), the following were correct, the absolute permeability
order for permeability at a given con- of the kaolinite at a given void ratio
solidation pressure appears to be usually should be the same for all fluids. Figure 2
clearly shows that this is indeed not true:
true: the permeability of kaolinite is con-
Ca > Mg > K > Na siderably different for the various fluids.
The data in Fig. 2 were obtained from
Figure 1 shows that the permeabilities tests in which the kaolinite was molded
of the three minerals are hi the following in the liquid which was to be used as the
order: montmorillonite < attapulgite < * The specific surface reflects, to some extent,
kaolinite. It is seen from Fig. 1 that, at a composition. The effective surface area is more
void ratio as high as 15, sodium mont- nearly represented by the total area available
morillonite had a permeability less than than for hydration, as indicated by glycol adsorption,
7 by the external area measured by nitrogen
10~ cm per sec.
Copyright by Sodium
ASTM Int'lmontmorillonite
(all rights reserved); Sunadsorption.
Jan 11 13:11:58 EST 2015
FiG.52.—Permeability of Kaolinite to Various Fluids as a Function of - (e = void ratio).

FromfMichaels and Lin (7).
FIG. 3.—Effect of Desolvation on Kaolinite Permeability; Initial Permeant, Water. From Michaels
and Lin (7).
permeant. Figure 3 presents results from ous one. Figure 3 shows that, although
tests hi which water was used as the different permeabilities were obtained for
molding fluid and initial permeant; each different permeants, the differences are
succeeding permeant
Copyright displaced
by ASTM the previ-
Int'l (all rights reserved);smaller than
Sun Jan 11 thoseEST
13:11:58 in 2015
Fig. 2.
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further repro
Michaels and Lin's work shows that molding fluid has an effect on perme-
the major effect of permeant on permeability; data are presented in the next
ability (corrected for permeant density section of this paper to show the influence
and viscosity) is caused by differences in of amount of molding fluid. In other
soil structure (see section on Effect of words, permeability depends not only on
Structure on Permeability). If the re- void ratio but also on the method by
maining permeant effect were attributed which it is obtained.
to a layer of immobilized fluid, the
kaolinite had an effective thickness of EFFECT OF STRUCTURE
adsorbed water of 50 to 100 A. Most of
Permeability depends to a considerable
this effective thickness was more properly
extent on the arrangement of soil parti-
explained by electro-osmotic counter-
cles, or "structure." The importance of
flow.
structure on almost all soil properties has
One must conclude that viscosity and
been recognized, and theoretical explana-
density are not the only permeant
tions for the role of structure on behavior
characteristics, as indicated by the
have been proposed (5). Data on the rela-
theoretical equations, \ha,t influence the
tionships between permeability and
permeability of fine-grained soils. Since
structure have been obtained and are
the electro-osmotic backflow and thick-
presented here.
ness of immobilized fluid increase with
fluid polarity, some measure of polarity Compaction-Permeability Test:
might well be included in the equations.
Further research is required to determine To obtain permeability data on com-
whether such improvement in the equa- pacted soils, the apparatus shown in Fig.
tions is feasible. 4 was constructed. The following test
procedure was developed and employed:
EFFECT OF VOID RATIO 1. The soil sample was brought to de-
sired water content, mixed with a stand-
Figure 2 shows that k versus is
ard amount of work, and allowed to
equilibrate for at least 24 hr.
not a straight line as the equations indi- 2. It was compacted in a Harvard
cate. The normal soil testing procedures Miniature Compaction size mold, and
consider adsorbed fluid as normal pore then weighed.
fluid rather than part of the soil particle. 3. The mold was mounted in the ap-
Since the absolute amount of immobi- paratus (Fig. 4), the permeant chamber
lized fluid probably depends on pore size filled, and desired gas pressure applied.
and water content (among other things), 4. Time and flow measurements were
the commonly measured void ratio is not taken at various times until a constant
equal to the effective one nor is it a rate of flow was reached.
constant percentage of it. 5. The apparatus was disassembled
Considerable data have shown that and the sample swell or shrinkage meas-
generally the plot of void ratio versus log ured (the measurement must be made as
fpermeability approximates a straight soon as possible after the pressure has
Mine. This relationship holds, of course, been released to minimize the effect of the
only when all other soil characteristics are expansion due to pressure release).
kept constant, since void ratio is a de- 6. The specimen was extruded,
pendent variable. Data have been pre- weighed, dried, and reweighed.
sented Copyright
to showbythat
ASTMtheInt'l
nature of the
(all rights reserved);The
Sun above
Jan 11 procedure has2015
13:11:58 EST proved to be
Pa. For soil No. 1 (a silty sand), the mini-

mum permeability occurs at water con-
tent at, or slightly greater than, the op-
timum for compaction. This relationship,
rather than the reverse exhibited by soil
No. 2 (a sandy silt), is most common (see
Wilson (9) and Lambe (3)). The data in
Fig. 5 are the molding water content, the
as-molded density, and the permeability
reached after permeation. Figures 6 and
7 present data on three soils which show
how the water content, density, degree of
saturation, and permeability change
with permeation.
Samples compacted drier than the op-
timum (Fig. 6) picked up moisture,
swelled, and became more nearly sat-
urated upon permeation; samples com-
pacted near optimum showed little
change; samples wetter than optimum
did not behave consistently. Permeation
caused a reduction in permeability for all
samples in which fines were not washed
out (Fig. 7). Figures 8 and 9, in which
data are cross-plotted, show the moisture
and density changes caused by permea-
tion.
Figure 6 illustrates a fact not fully
FIG. 4.—Permeability Test Set-Up. appreciated by many experimenters,
namely, that the degree of saturation af-
a very rapid and dependable test to ter permeation is usually considerably
measure the compaction and permeabil- below 100 per cent.
ity characteristics of a fine-grained soil. Sample Mixing:
The main objections are that deaired
water is not used (the use of gas pressure In general, the more nearly homogene-
precludes it) and the small sample size ous an isotropic soil, the lower its perme-
leads to data scatter. These disadvan- ability. The fines in a soil have a higher
tages of the test are considerably more impermeabilizing effect (excluding the
than outweighed by its convenience and case of lenses of fines normal to flow) if
general dependability. Unless otherwise they are well distributed so they can
noted, it was employed to obtain the data most effectively plug voids among the
presented and discussed in the remaining larger particles. Mechanical mixing "dis-
portion of this paper. tributes fines.
Mixing also breaks down some of the
Influence of Molding Water Content: soil aggregates ("mechanical disper-
sion"),4 thereby supplying fines for void
Figure 5 presents compacted density 4
and permeability molding
versus Int'l water Remolding a Boston blue clay sample caused
Copyright by ASTM (all rights reserved); Sun Jan 11
a reduction of13:11:58 EST to
permeability 2015
a value 1/200 of
contentDownloaded/printed
for two soils from
by New Castle, that of the undisturbed value.
FIG. 5.—Density and Permeability as a Function of Molding Water Content.
FIG. 6.—Changes in Density and Water Content Caused by Permeation.
8
FIG. 7.—Changes in Permeability Caused by Permeation. FIG. 8.—Water Content Change Caused by Permeation.
plugging and destroying large voids. curves in Fig. 11 show that dispersion re-
Mechanical mixing, because of its blend- duced the permeability of a soil contain-
ing and dispersing, effects a reduction in ing only 8 per cent by weight finer than
soil permeability. Figure 10 illustrates 0.07 mm.
the pronounced influence mixing can \Smce dispersion causes a permeability
have on permeability. The data in the reduction, aggregation should cause an
lower curve were obtained on a sample increase^Figure 12 shows that such an in-
that was initially mixed thoroughly, then crease does, indeed, occur. The data in
reused from test to test. The upper curve Fig. 12 were obtained from two series of
FIG. 9.—Density Change Caused by Permeation.

was obtained from tests in which a differ- tests, one series on an untreated sandy
ent sample was employed for each de- clay and one on the same clay to which
termination and each sample was given 0.05 per cent (of clay dry weight) of
the same amount of mixing time. sodium polyacrylate had been added.
Only one sample was used for each series.
Effect of Dispersion and Aggregation: Since the sample was densified in the
The importance of mechanical disper- permeameter by vibration, the altera-
sion has already been pointed out. Data tion of structure with densification was
have been presented elsewhere (3) to show minor.
that the minimum permeability of a soil Discussion of Data:
compacted in the presence of a chemical The data presented in this section as
dispersant can be less than one tenth of well asSun
Copyright by ASTM Int'l (all rights reserved); thatJanin11Effect of EST
13:11:58 Permeant
2015 on
the minimum of the natural
Downloaded/printed by soil. The Permeability emphasize the importance
FIG. 10.—Effect of Mixing on Permeability. Jamaica clay.
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No furtherFIG. 11.—Effect
reproductions of Dispersion on Permeability.
authorized.
of structure, as the following compari- complete aggregation is emphasized to

sons illustrate: some extent at present, even though the
Dry Density or Void Permeability, Figure

Soil Ratio Degree of Saturation cm per sec Furnishing
Data
Jamaican clay 115 Ib per cu ft Approximately same 4 X 10-« 6

116 Ib per cu ft 7 X 10~8
Virginia sandy clay 1.3 100% 2.7 X 10~4 12
1.3 100% 1 X 10~s
The first comparison, between a sample best method of determining such num-
compacted dry of optimum and one wet bers has yet to be established.
of optimum, shows two samples at essen- The permeability of fine-grained soils
tially the same void ratio and degree of varies as some power of this "structure
saturation having a permeability ratio of coefficient."
approximately 60. The second compari-
son, also between samples at the same EFFECT OF DEGREE OF SATURATION
void ratio and degree of saturation, shows Soil pores filled with entrapped air are
a permeability ratio of greater than 3. not serving as channels for flowing water.
The reduction in permeability that oc- A partially saturated soil is not, there-
curs with permeation (Fig. 7) is caused fore, transmitting its maximum amount
by a change in structure. As flow through of liquid during permeation. While this
a soil occurs, particles tend to move to reasoning is commonly accepted, data
positions of greater stability to seepage correlating the degree of saturation with
forces. This particle shifting always re- the permeability of fine-grained soils are
sults in lower permeability, if particles not readily available.
are not washed/out of the soil. While this The influence of degree of saturation
particle shifting can result in either an on permeability is relatively minor in
increase or decrease in density or satura- comparison with composition, structure,
tion degree, it usually causes a densifica- and void ratio. Thus, while there are test
tion and higher degree of saturation. data available, the effects of degree of
The permeability reduction has to be saturation are masked, as the data in
explained, therefore, by changes in struc- Figs. 6 and 7 well illustrate. For example,
ture rather than incidental changes in the sample of Maine silt compacted at
density or degree of saturation. 20.4 per cent moisture decreased in per-
To evaluate directly a "structure" meability during permeation even though
term for the permeability equations will a decrease in density and an increase in
be exceedingly difficult. Attempts to meas- degree of saturation occurred. The per-
ure the extent of aggregation have been meability decrease caused by the altera-
made (for example, see Michaels and tion of structure more than outweighed
Lambe (6)), but no simple way of giving the combined effects of degree of satura-
soil a number to indicate accurately its
tion and density alteration.
position in a structure scale has been de-
veloped. One is not likely to be developed SUMMARY
because of the complex nature of struc-
ture. The concept of a scale ranging from The major factors influencing the per-
0 for Copyright
completebydispersion to rights
ASTM Int'l (all for meability
100 reserved); of 13:11:58
Sun Jan 11 fine-grained soils are: (1)
EST 2015
soil composition, (2) permeant characteristics of soil and permeant are of
teristics, (3) void ratio, (4) structure, and limited practical use in fine-grained soils.
(5) degree of saturation. Data are pre- These equations can be improved by the
sented to indicate the magnitude of the
addition of terms and by reinterpretation
influence each of these variables can
have. In their present state, the theoreti- and evaluation of the present equation
cal equations showing the relationships terms. Considerable research is required
between permeability and the charac- to accomplish this goal.
so 100
FIG. 12.—Effect of Aggregation on Permeability.
REFERENCES
(1) P. C. Carman, "Fluid Flow Through Gran- neering Chemistry, Vol. 46, June, 1954. pp'
ular Beds," Transactions, Inst. Chemical 1239-1246.
Engrs. (London), Vol. 15, p. 150 (1937). (8) Donald W. Taylor, "Fundamentals of Soil
(2) ]. Kozeny, Berichte Wien Akademie, Vol. Mechanics," John Wiley and Sons, Inc.
136a, p. 271 (1927). (1948).
(3) T. William Lambe, "The Improvement of (9) Stanley D. Wilson, "Effect of Compaction
Soil Properties with Dispersants," Journal, on Soil Properties," Proceedings, Confer-
Boston Soc. of Civil Engrs., April, 1954, , ence on Soil Stabilization, Massachusetts
pp. 184-207. Institute of Technology, June, 1952, pp.
(4) T. William Lambe, "Soil Testing for Engi- 148-158.
neers," John Wiley and Sons, Inc., New (10) H. F. Winterkorn, "Engineering Properties
York, N. Y. (1951). of Clay Soils," Bulletin No. 1, Winterkorn
(5) T. William Lambe, "The Structure of Road Research Inst. (1950).
Inorganic Soil," Proceedings, Am. Soc. (11) H. F. Winterkorn and R. B. B. Moorman,
Civil Engrs., Vol. 79, pp. 315-1-315-49. "A Study cf Changes in Physical Proper-
Separate No. 315, October, 1953. ties of Putnam Soil Induced by Ionic
(6) A. S. Michaels and T. William Lambe, Substitution," Proceedings, 21st Annual
"Laboratory Evaluation of Polyelectro- Meeting of the Highway Research Board,
lytes as Soil Flocculants and Aggregate December, 1941, pp. 415-433.
Stabilizers," Agricultural and Food Chemis- (12) "Final Report, Soil Solidification Research,
try, September, 1953, pp. 835-843. Vol. 2, Fundamental Properties, Clay-
C7) A. S. Michaels and C. S. Lin, "The Perme- Water Systems," Cornell University,
ability of Kaolinite," Industrial and Engi- Ithaca, N. Y., September, 1951.
THE PERMEABILITY AND SETTLEMENT OF LABORATORY
SPECIMENS OF SAND AND SAND-GRAVEL MIXTURES
BY CHESTER W. JoNES1
SYNOPSIS
This paper describes the procedure and results of laboratory permeability-
settlement tests on samples from a river deposit containing about half sand
and half gravel. The tests were conducted (1) on specimens containing the
sand fraction passing the No. 4 sieve, (2) on specimens containing the sand
fraction plus various percentages of gravel with f-in. maximum particle size,
and (3) on specimens containing the sand fraction plus various percentages of
gravel with 3-in. maximum particle size. The tests of (1) and (2) above were
made on specimens of 8-in. diameter and 5-in. height, whereas those of (3)
were made on large-scale specimens of 19-in. diameter and 9-in. height. The
specimens were placed at different densities, and settlements were measured
during a definite loading schedule during the permeability test.
The results show that the permeability of both the small and large specimens
containing 20 to 50 per cent gravel was considerably less than the permeability
of the sand fraction alone. In addition, the permeability of specimens contain-
ing the f-in. material was somewhat greater than those containing the 3-in.
material. The permeability decreased with increase in placement density
and, with the exception of the specimens containing f-in. material, decreased
with increasing load. The settlements increased with load and generally
increased with gravel content.
The study shows that the differences between the results of the small and
large-scale tests are sufficient to justify the use of the latter where a reasonably
close value of the permeability is required.
It is often good engineering practice ment of a loaded specimen can be found

to use soil-gravel mixtures in earth by laboratory tests. In Bureau of Recla-
structures where such mixtures occur mation work prior to 1940, permeability
in deposits near the structure because testing was conducted only on the soil
of the high stability of these materials, fraction (minus No. 4 material) and,
In some cases, permeability and settle- when soii_gi.avei mixtures were encoun-
ment are important functions of the teredj the effect of the presence of the
structure, and values of these properties
, ' r j - . M. ^ ' gravel was either disregarded or at best
are needed for design and construction
purposes. The determination of the roughly estimated.
permeability and the associated settle- About 1940> the Bureau laboratories
1 —— „ „ „ , „ , . constructed some large-scale laboratory
Engineer, U. S.byBureau of Reclamation,
Denver, Copyright
Colo. ASTM Int'l (all rights reserved);
apparatus Sun Jan loaded
for conducting 11 13:11:58
perme- EST
S R KUMAR KONATHALA 68
(Indian Inst of Tech Chennai) pursuant to Lic
JONES ON SAND AND SAND-GRAVEL MIXTURES 69
FIG. 1.—Gradation of j-in. Maximum Size Material Used in Small Permeability-Settlement

Test Specimens.
FIG. 2.—Gradation of 3-in. Maximum Size Material Used in Large Permeability-Settlement Test
Specimens.
ability tests on specimens containing in the pervious zones of an earth dam or

soil-gravel mixtures, and the apparatus in drainage filters. The ultimate program
has since been in nearly constant use.2 will include similar tests on other soil-
Large-scale types of tests such as these gravel types. The permeability-settle-
are necessarily more expensive to make ment tests reported herein can be
than smaller tests, and questions have conveniently divided into several series
often arisen as to the differences in as follows:
results of the two types of tests and
whether or not any correlation between
them exists.
This paper describes the first phase of
a Bureau of Reclamation research pro-
gram involving the conducting of
permeability-settlement tests on soil
and soil-gravel mixtures in small- and
large-scale laboratory apparatus. The
FIG. 3.—8-in. Diameter Laboratory Permeability Apparatus.
materials used were predominantly 1. Tests on the sand fraction passing

sand and sand-gravel mixtures and re- a No. 4 sieve in specimens of 8-in.
sulted in tests of high permeability. diameter and 5-in. height.
Such material would find common use 2. Tests on the sand fraction plus
2
W. G. Holtz and V. S. Meissner, "Sug-
various percentages of gravel of f-in.
gested Method of Test for Permeability and maximum particle size hi specimens of
Settlement of Earth Materials Containing 8-in. diameter and 5-in. height.
Particles Up to Three Inches in Size," Pro-
cedures Copyright
for Testing 3. Tests on the sand fraction plus
by Soils,
ASTMAm.
Int'lSoc.
(all Testing
Mats., July, 1950, p. 184. various percentages of gravel of 3-in.
JONES ON SAND AND SAND-GEAVEL MIXTURES 71
maximum particle size in specimens of in Fig. 2. In both series the gravel con-
19-in. diameter and 9-in. height. tent was varied by percentage increments
Each of the above tests was performed of 0,20,35,50, and 65. In a few instances
at three different density conditions, of the larger specimens, 80 per cent
with all specimens in each test series gravel contents were used.
being placed at similar relative densities.
The specimens were subjected to a TEST APPARATUS
definite system of increasing loads during A drawing of the permeability-settle-
the permeability tests. The resulting ment apparatus for the 8-in. diameter
recorded settlements are presented. specimens is shown in Fig. 3, and a
photograph of the large-scale appa-
ratus for 19-in. diameter specimens is
seen hi Fig. 4. Both tests are identical
in principle of operation, but they differ
in design to provide the necessary ca-
pacity for maximum particle size and of
water flow to each size of specimen.
The water flow through the specimen is
determined by timing the discharge
from a volumetrically calibrated head
tank which operates on the mariotte
principle to provide a constant head of
water for the permeability specimen.
Practically constant loads were applied
to the specimens during the permeability
tests by the springs which could be
compressed to give the desired loading.
Since the materials tested were rela-
tively pervious, it was necessary to
eliminate excess piping of water between
FIG. 4.—Large 19-in. Diameter Laboratory the specimen and the inside cylinder
Permeability Apparatus. wall. After trying various methods of
overcoming this, a liner of sponge rubber
MATERIALS TESTEDD cemented to the cylinder was found to
The source of the materials used was be most convenient and most successful.
a deposit at a commercial sand and Also because of the perviousness of the
gravel plant located on the South Platte soil, accurate measurements of low
River hi Denver, Colo. The sand and water heads were necessary and, at
gravel particles were generally of igneous tunes, a piezometer with a vernier gage
types. The particles were subangular to (shown in the photograph of the large-
subrounded with the angularity becom- scale apparatus of Fig. 4) was used to
ing more pronounced with the degree of enable readings to be made to the near-
fineness. est 0.01 in.
The grading of the sand fraction, to- TEST PROCEDURE
gether with the gradings of the materials
used in the f-in. series of tests, is shown Density Control:
in Fig. 1. The grading of the materials The minus No. 4 material used in
used inDownloaded/printed
the 3-in. series ofbytests is shown these tests was too granular to yield a
regular-shaped moisture-density curve placement of desired relative densities

from a standard compaction method. between the two limits. The maximum
For this reason the relative density density is found by vibrating to constant
method of control for placement densi- density the saturated material in a
ties was used. This method involves the container of known volume and de-
FIG. 5.—Maximum and Minimum Densities of Sand and Sand-Gravel Mixtures.
TABLE I.—SOIL PLACEMENT DENSITIES AND VOID RATIOS OF

PERMEABILITY TEST SPECIMENS.
M-in. Material 3-in. Material
Gravel, 50% Relative 60% Relative 70% Relative 50% Relative 60% Relative 70% Relative
per cent Density Density Density Density Density Density
Den- Void Den- Void Den- Void Den- Void Den- Void Den- Void
sity" Ratio sity Ratio sity Ratio sity Ratio sity Ratio sity Ratio
0 103.8 0.575 106.6 0.534 109.7 0.490 103.8 0.57 106.6 0.534 109.7 0.490
20 111.7 0.464 114.0 0.435 116.3 0.406 111.9 0.462 114.2 0.433 116.7 0.401
35 115.4 0.416 117.8 0.387 120.4 0.359 117.2 0.395 119.5 0.368 122.0 0.340
50 118.8 0.376 121.5 0.346 124.2 0.316 121.7 0.344 124.2 0.316 126.8 0.289
65 118.3 0.381 121.0 0.351 123.9 0.319 127.2 0.285 129.7 0.261 132.4 0.235
80 112.2 0.458 115.2 0.418 118.4 0.381 125.8 0.300 128. 0.276 130.8 0.250
0
Density in pounds per cubic foot.
determination of maximum and mini- termining the dry unit weight. The
mum densities3 by definite laboratory minimum density is found by deter-
procedures and the computation and mining the dry unit weight of material
3
Designation E-ll, U. S. Bureau of Recla- poured in a dry condition into a con-
mation Earth Manual (1951). A manual on the tainer of known volume. The results of
use of earth materials
Copyright for foundation
by ASTM and reserved);
Int'l (all rights con- Sun Jan 11 13:11:58 EST 2015
structionDownloaded/printed
purposes. by these tests are shown in the plot of Fig. 5.
JONES ON SAND AND SAND-GRAVEL MIXTURES7 73
The placement densities used for the kept separate until time for specimen
test specimens were 50, 60, and 70 per placement, was soaked in pans of water
cent relative density as computed from overnight and then surfaced-dried prior
the following formula: to combination with the sand (minus
No. 4 fraction). The sand fraction was
moistened sufficiently to facilitate speci-
men compaction. The sand and gravel
where: fractions were combined and tamped to
7 = placement density, desired density hi layers in the perme-
7max = maximum density, ability cylinders. The specimens con-
7m m = minimum density, and taining no gravel were placed hi five
Dr = percentage relative density de- 1-in. layers, those with f-in. maximum
sired (expressed as a decimal). size gravel in three 1.67-in. layers, and
FIG. 6.—Relationship of Permeability to Gravel Content for Specimens with Variation in Density.
The specimen placement densities those with 3-in. maximum size gravel
computed as shown above and the cor- in two 4.5-in. layers.
responding void ratios are listed hi
Table I. These densities are in the same Specimen Loading:
range of values as those one could
reasonably expect of similar material The 8-in. diameter permeability speci-
in a compacted earth structure. mens were loaded by exerting a force on
the springs over the specimen by means
of a hydraulic jack between the plates
Specimen Preparation:
above the specimen (see Fig. 3). The
For each specimen of sand or sand- load was maintained by tightening the
gravel mixture, material sufficient to nuts on the tension rods against the
make the desired density and volume steel plate between the jack and the
was first weighed. The gravel, which was springs. The 19-in. diameter permeability
specimens were loaded in a large uni- were made at intervals during the per-
versal testing machine. meability test by means of dial gages
The sequence of loads for the speci- between reference points located on
mens during the permeability test was opposite sides of the permeability
representative of 20, 40, 80, and 160 ft cylinder (see Figs. 3 and 4).
of overburden. For this purpose the The water used for the permeability
density of overburden was assumed to be tests was taken from the cold tap of the
FIG. 7.—Relationship of Permeability to Load for Specimens with Variation in Gravel Content
and Density.
130 Ib per cu ft; thus the overburden Denver city water system. The water
figures are approximately equal to the for the tests was drawn into a reservoir
load expressed in pound per square inch. located near the laboratory ceiling and
was allowed to remain there sufficiently
Permeability Testing: long so that the water temperature was
The permeability test was continued slightly above the temperature of the
under each loading from one to several soil specimens which were located near
weeks until the flow of water became the laboratory floor. This procedure
reasonably constant. Settlement readings was followed to cause a slightly falling
temperature gradient through the soil permeability by the addition of 20 per
specimen to avoid deposition of air in cent gravel. The permeability of the
the soil voids from that dissolved in the specimens with the f-in. maximum
water. size material containing 35 and 50 per
cent gravel generally increased some-
TEST RESULTS
what over those containing 20 per cent
Permeability Tests: gravel, then increased greatly for the
The range of permeabilities for the specimens containing 65 per cent gravel.
tests reported herein extend from about The permeability r>f the specimens with
5000 to 200,000 ft per year. In all cases the 3-in. maximum size material con-
duplicate specimens placed under as taining 35 per cent gravel decreased
nearly similar conditions as possible were somewhat from those containing 20
tested, and the results shown in Figs. per cent gravel and then, with the ex-
6 and 7 are the averages of the duplicate ception of the specimens placed at 70
specimens. The percentage difference of per cent relative density, increased
permeability values of individual speci- considerably for the specimens contain-
mens from the average of duplicates ing gravel percentages over 35 per
varied from 0 to 72 per cent with the cent. For the specimens containing 3-in.
over-all difference of 18 per cent. There particles and placed at 70 per cent
is a slight tendency for this percentage relative density, the permeability de-
difference to decrease with increase in creased uniformly with the increase in
load but no definite tendency for it to gravel from 20 to 65 per cent, then
increase or decrease with the specimen increased greatly with those containing
density or gravel content. In perme- 80 per cent gravel.
ability determinations on material as The variation of permeability with
pervious as that used in the tests re- relative placement density for specimens
ported herein, some anomalies are having different gravel contents is shown
certain to occur. A slightly different in Fig. 6. In nearly all cases the per-
arrangement of the particles composing meability of specimens decreased with
specimens may easily be reflected in increase in density and this was particu-
permeability and settlement. Therefore, larly noticeable for the specimens
only general trends of the tests are containing no gravel.
discussed. The variation of permeability with
The variation of the permeability with load for specimens containing different
gravel content for specimens placed gravel contents and placement densities
at 50, 60, and 70 per cent relative is shown in Fig. 7. The permeability of
densities and subjected to a 20-ft over- the specimens containing 3-in. maximum
burden load are shown in Fig. 6. The size gravel showed a slight tendency to
most significant fact apparent is the decrease with increases in load applied
comparatively lower values of perme- during the permeability. However, for
ability for the specimens containing 20, the smaller specimens, both for the
35, and 50 per cent gravel compared to specimens without gravel as well as
those of the specimens containing no those with the f-in. maximum size
gravel. For both the specimens with gravel, this tendency was not in evi-
f-in. maximum size gravel and the dence. The smaller specimens showed a
specimens with 3-in. maximum size tendency for the permeability to increase
gravel, the permeability was reduced to somewhat for an increase in load from
about Copyright
10 to 20byper
cent of the sand 20- to 40- and 80-ft loads and then to
decrease for the 160-ft loads, but the ment of the specimens with f-in. maxi-
permeability at the highest load was mum size gravel was greater than (in
almost invariably higher than that for some cases about twice) that of the
the lowest load. specimens with 3-in. maximum size
gravel.
Specimen Settlement: In Fig. 9 the plot shows the relation-
The settlement of loaded permeability ship of settlement to gravel content for
specimens made from the granular specimens varying in load and density.
pervious material tested was compara- This shows that the specimens contain-
tively low. This fact may explain why ing no gravel settled considerably less
the permeability of some of the speci- than those containing gravel. In general,
FIG. 8.—Relationship of Settlement to Load for Specimens with Variation in Gravel Content
and Density.
mens increased with increase in load, for the specimens containing f-in.
as mentioned in the preceding para- gravel, the highest settlements occurred
graph. The decrease in void content of in the specimens containing 20 and 35 per
the specimens was so small that even a cent gravel. For the specimens contain-
slight rearrangement of the particles ing 3-in. gravel, the settlements generally
in the specimens resulted in an increase increased with the gravel content up to
in permeability, rather than a decrease the highest content used.
as might be expected. A previous study4 on consolidation of
The plot in Fig. 8 shows the variation sand-gravel mixtures made in the Bureau
of settlement to load for specimens with 4
H. J. Gibbs, "The Effect of Rock Content
different gravel contents and densities. and Placement Density on Consolidation and
The settlement Related Pore Pressure in Embankment Con-
Copyright byincreased
ASTM Int'lapproximately
(all rights reserved); Sun Jan 11
struction," 13:11:58 EST
Proceedings, Am.2015
Soc. Testing
linearly with increase inbyload. The settle- Mats., Vol. 50, p. 1343 (1950).
Downloaded/printed
of Reclamation laboratory showed re- per cent of relative density in Fig. 5.

lationships of 1-in. maximum size Therefore, this basic feature of the two
material to 3-in. maximum size. These studies has relationship when considera-
comparisons were made at the same tion is given to the different methods of
placement density for a given percentage initial placement control. There was
of gravel. The 3-in. maximum size considerable difference hi the degree of
material consolidated a greater amount consolidation obtained on the two studies
because it was better graded in the because the previous study involved a
FIG. 9.- -Relationship of Settlement to Gravel Content for Specimens with Variation in Load and
Density.
coarse particle sizes and was more plastic soil matrix, whereas the present
adapted to densmcation. Although the study involved granular soil only. In the
settlement for the present study showed previous study, therefore, gravel had a
greater consolidation for f-in. maximum considerable effect in reducing consolida-
than for the 3-in. maximum, the feature tion, whereas the present study which
of better adaptability to densmcation involved only granular soil showed less
of the better graded 3-in. maximum effect of gravel content and even resulted
material is also
Copyright by shown by (all
ASTM Int'l therights
higher in Sun
reserved); some
Jan increase in EST
11 13:11:58 consolidation
2015 with
resulting compacted density
Downloaded/printed by for a given gravel content.
CONCLUSIONS do not interfere with one another to

From the results of the laboratory any extent, but they cause the mixture
permeability-settlement tests made on to become better graded and greatly
the pervious South Platte River sand reduce permeability. For about 20 to
and sand-gravel mixtures, the following 65 per cent gravel, although the grading
conclusions can be drawn: is improved, there is a variable inter-
1. The permeability of sand-gravel ference among the coarser particles,
mixtures containing 20 to 50 per cent and the permeability may either increase
gravel is considerably lower than that or decrease a moderate amount. For
for the sand fraction (minus No. 4 gravel contents over about 65 per cent,
material) alone. there is major interference of the
2. The permeability of sand-gravel coarser particles, the density decreases,
mixtures containing 3-in. maximum size and the permeability rapidly increases.
material in large 19-in. diameter speci- 6. The settlements of the permeabil-
mens was approximated by small 8-in. ity specimens increased approximately
diameter specimens containing f-in. linearly with load, with the settlement
maximum size material for 20 per cent of the specimens containing f-in. gravel
gravel, but for the higher gravel contents higher, in most cases, than the settle-
the difference between the two types of ments of the specimens containing 3-in.
specimens was generally large. gravel.
3. The permeability of sand and sand- 7. The settlement of specimens con-
gravel mixtures decreases with increase taming 3-in. gravel increased more or
in placement density in a more or less less uniformly with gravel content,
linear fashion. whereas, for the specimens containing
4. The permeability of specimens with f-in. gravel the settlement was highest
3-in. gravel showed a slight tendency to for the 20 and 35 per cent gravel.
decrease with increase in load. The 8. For determinations of permeability
permeability of small specimens, both and settlement of soils containing gravel,
with and without gravel, showed a the use of the 3-in. maximum size gravel
tendency to increase in permeability in 19-in. diameter specimens is justified
with load. and recommended. The use of the 8-in.
5. The effect of the proportion of diameter specimens from which material
gravel on the permeability of the sand- larger than f-in. maximum size gravel
gravel mixtures can be stated in the has been removed will generally result
following general terms. For 9 to about in values of permeability and settlement
20 per cent gravel, the gravel particles which are too high.
DISCUSSION
MR. DON KiRKHAM.1—The curves ance of any part of a conductor is ac-

of the relationship of permeability to companied by an increase... of the
gravel content for mixtures of sand and resistance of the conductor as a whole."2
gravel (Fig. 6) are, at least in part, of As further impermeable pieces of gravel
the shape one would expect them to have are added to the conducting sand, it
on theoretical grounds. For 100 per cent should be expected that the permeability
sand the permeability is a maximum. would further decrease, whereas, ac-
With the addition of impermeable pieces tually, the experiments now show an in-
of gravel, the permeability then falls crease in the permeability. This apparent
off until about 35 to 65 per cent of the discrepancy from theoretical expecta-
impermeable pieces of gravel have been tions may be explained on two grounds.
added. So far the curves are in accord- (1) At the higher gravel contents the
ance with the hydraulic analogue of a conducting medium (the sand) did not
theorem in electricity which states that completely—or at least not so densely as
"any increase ... in the specific resist- originally—fill the voids between the
1
Professor of Soils and Physics, Iowa State
gravel. (2) The theorem referred to ap-
College, Ames, Iowa. plies to components of a composite
2
See, for example, Sir James Jeans, "The medium on a volume basis, whereas the
Mathematical Theory of Electricity and Mag- components in the experimental medium
netism," Cambridge University Press, Cam-
bridge, England, Fifth Edition, p. 349 (1933). here are given on a weight basis.
MEASUREMENT OF THE HYDRAULIC CONDUCTIVITY OF SOIL IN
PLACE*
BY DON KIRKHAM1
SYNOPSIS
Several recently developed and two proposed methods for measuring the
hydraulic conductivity of soil in place are described. Particular attention is
given to the so-called auger hole, piezometer and tube methods. In the auger
hole method an auger hole is bored to a depth below the water table and the
water in it is pumped out. After the hole has been emptied, the ensuing rate of
rise of water in it is determined and this rate is then converted to the hydraulic
conductivity of the soil by use of a suitable formula. In the piezometer and
tube methods the procedure is the same except that the auger hole is now ef-
fectively cased, and there is only a small cavity at the base of the cased hole
into which the rate of water entry is noted. The latter methods have the ad-
vantage that they permit the measurement of the hydraulic conductivity at
virtually a point in the soil. The auger hole method, on the other hand, gives
an average conductivity for the whole length of the hole. Results from the
auger hole method are less variable. The problem of soil anisotropy is dis-
cussed, and it is pointed out that the anisotropic components of the hydraulic
conductivity may be determined by use of the methods.
A knowledge of the hydraulic conduc- to predict rationally the proper depth

tivity (I) 2 - 3 of soil in place is important and spacing of farm drain tiles and drain-
in many activities. In agriculture, a age ditches; to determine the amount of
knowledge of this soil coefficient is needed seepage in and around soil conservation
dams; and to predict whether a site for a
* Journal Paper No. J-2505 of the Iowa farm pond will be suitable. More money
Agricultural Experiment Station, Ames, Iowa, has been spent on land drainage in Iowa
Project 998.
1
Professor of Soils and Physics, Iowa State alone than was spent to build the
College,
2
Ames, Iowa. Panama Canal. In activities not neces-
The boldface numbers in parenthese refer sarily agricultural, questions like the
to the list of references appended to this paper,
see8p. 96. following are asked. How rapidly will an
The term "hydraulic conductivity" is used, aquifer produce water? How fast will
rather than "permeability," etc., in keeping
with a recommendation of a committee of the water drain through the subsoil below a
Soil Science Society of America (1). The term proposed highway or a proposed airfield?
hydraulic conductivity may be defined as the At what rate should ground water be dis-
coefficient k in Darcy's law, v = ki, v being the
velocity of seepage and » the hydraulic gradient. posed of about a foundation or in an em-
Values of k depend on properties of the fluid, bankment? The answers to all the above
as well as of the porous medium, and reflect any questions depend on a knowledge of the
interactions of the fluid with the porous medium
such as swelling of soil and its attendant reduced hydraulic conductivity of the soil.
porosity. The term permeability is reserved to It has been common practice, in get-
apply to that (linear) impedance characteristic
of porous media which is independent of the ting answers to the above questions, to
Copyright
viscosity, etc., of by
theASTM takeJansoil
fluids.Int'l (all rights reserved); Sun cores orEST
11 13:11:58 bulk soil to the labora-
2015
S R KUMAR KONATHALA (Indian Inst of80Tech Chennai) pursuant to License Agreement. No further reprod
KJRKHAM ON HYDRAULIC CONDUCTIVITY 81
tory for permeameter measurements. impedance of the tube to flow, or block
Such measurements, because of the dis- the flow almost completely (4). Vari-
turbed nature of the samples, may not ability in hydraulic conductivity meas-
indicate the true value of the hydraulic urements, to the extent of 3000 per cent,
conductivity of the soil hi place. In the has been attributed to the presence of
present paper, methods for measuring the trapped air in cores (5).
hydraulic conductivity of soil in place 7. The chemical nature of the water
are presented which are intended to ob- used in the laboratory may differ from
viate problems associated with core and that of the ground water of the soil in
bulk samples, such as: place. The resulting deflocculation or
1. Core and bulk samples are small flocculation of soil material may produce
compared to the mass of soil of interest. channel sizes in the cores, quite different
Large numbers of the samples must be from the sizes in the natural soil (6, 7, 8).
taken for any intended true representa- The above list, although not complete,
tion of conditions. The smallness of core shows that there is a real place for field
samples may be especially significant in methods. In the following, the so-called
soils with much clay. In these soils, drain- auger hole, piezometer and tube field
age may occur largely through cracks and methods will receive major attention.
channels (2) whereas core samples, since These methods may all be designated as
they are small, might not include the single-cavity methods. A recent two-
cracks and channels at all. In fact, cavity (two-well) method will be de-
workers often discard core samples with scribed. Mention will be made of some
cracks as being unrepresentative of true older field methods. A proposed four-well
soil conditions. method and a proposed single-well
2. Core samples may be compressed method, both as yet untried experi-
when taken, thereby lowering the true mentally, but of possible interest for
hydraulic conductivity of the soil as further study and development, will be
found in place. described; a dry auger hole method and
3. If the soil is too dry, core samples some methods depending on the dis-
will shatter. The natural, water-conduct- charge of drain tubes or ditches will be
ing channels in the soil will no longer noted.
exist.
4. If the sample is taken too wet, the SINGLE-CAVITY METHODS
soil will puddle. Natural channels will be The auger hole method, the piezom-
sealed. eter method, and the tube method are
5. Even if a "good" core is taken—a similar in that the following procedure
core obtained when the soil is not too wet applies to all of them (0).
or too dry—a conductivity determina- A cavity is bored into the soil to a
tion made in the laboratory may still not depth below the water table. The water
represent the water-saturated, field con- seeps into the cavity and rises to an equi-
dition because air may become trapped librium level, which is the water table.
in the laboratory samples to result in the This level may be reached in a few min-
explosive (3) break down of natural utes or a number of hours, depending on
channels in the soil. the tightness of the soil. The equilib-
6. The trapped air in laboratory sam- rium level is recorded. The cavity is
ples, aside from, or in addition to, the pumped out a few times to permit any
explosive effects just mentioned, may puddled-over pores along the wall of the
block channels andASTM
Copyright by greatly
Int'l increase
(all rights the cavity
reserved); Sun to
Janbe
11flushed
13:11:58out
ESTby2015
the in-seeping
ground water. After a last pumping (2 this condition is unity and the soil water
or 3 pumpings generally suffice) the time pressure just atmospheric.
for the water to rise a certain distance Aside from the need of a water table,
in the cavity is observed. The time and there is another, more immediately ap-
the distance of rise are used finally in a parent, limitation on the cavity meth-
suitable formula to yield the hydraulic ods: the soils in which the methods are
conductivity of the soil in place. used must support a cavity. That is, the
The three single-cavity methods differ cavity must be fairly well defined. The
in several respects. In the auger hole cavity walls may not break down. In
method the cavity is uncased. In the sandy soils, unless the cavities have a
piezometer and tube methods, the side screened liner, or its equivalent, the
walls of the cavity are partially or com- methods may fail. It has been found
pletely cased. For the piezometer method that a number of sandy soils have suffi-
a 4-in. long cavity at the base of a cased cient clay to support a cavity without
hole 2 in. in diameter is generally used. the use of liners. Cavities will not tend
For the tube method the side walls of the to break down if they are only partially
cavity are completely cased. Water can
seep in only at the base of the cavity.
Since flow can occur along the whole
length of the side walls of the auger hole,
the auger hole method yields an average
value of the hydraulic conductivity for
the depth of the soil profile below the
water table. The piezometer and tube
methods yield essentially "point" values
of the hydraulic conductivity in the soil.
If the soil is anisotropic, the tube method
tends to measure the vertical component FIG. 1.—Auger Hole Method for Determining
of the hydraulic conductivity. The auger Hydraulic Conductivity of Soil.
hole and piezometer methods tend to
measure the horizontal component of the pumped out. It is not necessary to pump
hydraulic conductivity. out the cavities completely in order for
It is emphasized that all three cavity the methods to work. Hvorslev (12) has
methods will work only if there exists a pointed out some problems associated
water table, that is, a ground water sur- with cavities.
face where the pressure is atmospheric.
This point, which is not at all trivial, has AUGER HOLE METHOD
been enunciated in the form of a law by The simplest method one can probably
L. A. Richards (10). The existence of a imagine for measuring the hydraulic
water table is a necessary but not suffi- conductivity of soil in place is to bore a
cient condition. Luthin and Day (11) post hole into the soil below the water
have shown experimentally that water in table and observe the rate at which
water-saturated soil will not flow into an ground water will enter the hole. Diserens
auger hole penetrating the soil, if the (13) was apparently the first to report
water table moves vertically downward this method. Hooghoudt (14) used it
a-t a velocity equal to the quotient of the further and gave an approximate analy-
hydraulic conductivity and the soil sis of the problem. His name is associated
porosity. The hydraulic
Copyright by ASTM gradient for
Int'l (all rights with theSun
reserved); method
Jan 11in13:11:58
the Netherlands,
EST 2015
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreemen
KIRKHAM ON HYDRAULIC CONDUCTIVITY 83
where the method is extensively used.

Hooghoudt assumed in his analysis that
water entered the hole horizontally from
the sides and vertically from below. The
assumptions obviously do not represent
the truth. Nevertheless, they yielded ap-
proximately correct results.
Figure 1 illustrates the auger hole
method. The hole is bored to a depth d
below the water table. The time A^
needed for the water to rise a distance
A& in the hole, and the average level h
of the water in the hole for the time
interval At, are recorded. The hydraulic
conductivity k is then determined from
the relation (15):
where S is a geometrical function given

by
It has seemed worthwhile to present

the above Eqs. 3a and 36, as the author
knows of no place in the literature where
these functions, except in compressed
notation, are displayed. Only the three
terms shown in the right-hand side of
hi which K0(irr/2d') and Ki(irr/2d), etc., Eq 2 are needed for practical work.
are Bessel functions denned by the To aid in the practical use of Eq 1,
infinite series (16): Spangler (17), utilizing values of K0(x)
and KI(X) tabulated in the British As-
sociation Tables (16), has prepared a
chart of values of S. This chart is repro-
duced as Fig. 2.
Strictly speaking, Eq 1 is valid only
if an impervious layer coincides with the
bottom of the auger hole. If such a layer
does not so coincide, then Eq 1 yields
an apprpximate value of the hydraulic
conductivity, the approximation be-
coming better as the ratio d/r becomes
greater. If the distance 5 of the imper-
meable layer below the base of the auger
hole is known, then 5 in Eq 1 should be
Copyright by ASTM Int'l (all rights replaced
reserved);bySuna Jan
geometrical factor
11 13:11:58 EST S'2015
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agr
which would be a function of s, as well factor S' cannot now be given except for
as of r, d, and h. Unfortunately this a few values of r/d, h/d, and s/d. Values
which can be given are based on work
with electric models (18)4and are therefore
subject to experimental error. When
further values of S' are available (theo-
retical values would be preferable as
they are not subject to experimental
error), charts like Fig. 1 for s/d = 0.25,
0.5, 1, 2, and <x>, should be completed.
Figure 1 covers the case s/d = 0. The
case s/d = <x> should not depart much
from that of s/d = 2. L. F. Ernst
of the Soil Science Inst.,T.N.O., Gronin-
gen, The Netherlands, has prepared some
unpublished charts from which some
values of S' could be obtained.
Figures 3 and 4 show flow nets for the
4
To obtain values of S' from reference (18),
multiply values of A given in Fig. 4 there by
Gr/16)(l - h/d)(A/5). If the worker using the
auger hole method will agree to keep the value
of h/d in Fig. 1 less than"0.2 and will also agree
to use an auger hole of 4-in. diameter, then the
effect of an impermeable layer can be taken into
account by utilization of a chart prepared by
Johnson, Frevert, and Evans (19, Fig. 4). The
FIG. 2.—Values of S in Eq 1 from Spongier (17). A in this chart is the A just noted.
FIG. 3.—Lines of Equal Hydraulic Head (Equipotentials) and Streamlines in Soil about an
Emptied Auger Hole.
The hole extends from a water table to an impervious layer, the soil extending horizontally to infinity. The hydraulic
head, referred to the level of the impervious layer, is taken 0 at the bottom of the hole and 100 at the water table. The
zero-streamline (point) is taken at the top of the hole and the 100-streamline (above which 100 per cent of the water
enters the hole) at the impervious layer. Distances shown are in arbitrary units. From Kirkham and van Bavel (15).
FIG. 4.—Equipotentials and Streamlines, as in Fig. 3 Except Hole is Wider and Half Full of Water. From Kirk-
ham and van Bavel (15).
auger hole problem. In Fig. 3 the hole measurement is cased, except for a small
has been pumped out completely; in Fig. cavity at its end. The rate of entry into
4 it has been pumped out to half its this cavity is a measure of the hydraulic
depth. The flow nets show at a glance conductivity of the soil immediately
how water enters the hole and how the around the cavity—not over the whole
driving hydraulic head is dissipated. The length of the hole.
streamlines (arrows) show the paths of A specific example of the piezometer
water flow. The equipotentials, the lines method is of interest. After trash and the
everywhere perpendicular to the stream- surface sod have been removed from the
lines, show how the hydraulic head is location hi question, a hole l^f in. in di-
dissipated. With the aid of these flow ameter is bored to a depth of 6 in. below
nets it has been possible to show that if the ground surface. A length of 2-in.
Ah or the sum of several A/f's in Fig. 1 inside diameter, thin-walled, electrical
is kept such that h stays less than J d, conduit, hereafter called the piezometer,
then the effect of a cone of depression sharpened on one end, is driven into the
(which will always develop to a greater hole to a depth of 5 in. with light blows
or lesser extent about the top of the from a maul. A driving head is used on
auger hole as the hole fills) will have neg- the top of the piezometer to prevent its
ligible influence on the validity of Eq 1. damage. The lrf-in. soil auger is then
Aside from the effect of the cone of inserted into the piezometer and the soil
depression to introduce error in Eq 1, is removed to a depth 6 in. below its
there is a further, theoretically more ob- sharpened end. The piezometer is again
vious, error in the equation, namely, driven 5 in. deeper. This procedure is
use of Ah/At for dh/dt. But calculations continued until the sharpened end of the
show that very little error, less than 2 or piezometer is at the desired depth of
3 per cent, will result through use of measurement. The procedure thus far is
finite values of Ah and At, instead of the designed to prevent compaction of the
limiting form dh/dt, provided Ah is kept soil. At the same tune, because of the
less than the value J d, as described differential diameters of auger and pie-
above. zometer, the necessary tight fit of the
piezometer with the soil is obtained.
PIEZOMETER METHOD Leaks along the outside of the piezom-
It has been pointed out that the auger eter would vitiate the method.
hole method gives an average value of A cavity 4 in. in length and 1 jf in. in
the hydraulic conductivity of the soil diameter is now carefully augered below
over the length of the hole below the the bottom end of the piezometer. A stop
water table. Many soils are stratified. on the auger is provided to make this
One stratum may be highly permeable 4-in. depth precise. The auger is also
compared to the others. For this reason provided with a hollow shaft through
the auger hole method will fail to give a which air can enter to prevent the de-
correct picture of the way the hydraulic velopment of suction at the base of the
conductivity varies with depth. The auger. Suction, if not released, would,
piezometer method (Fig. 5) overcomes upon withdrawal of the auger, pull the
the difficulty. In this method one ob- soil on the auger thread back into the
tains, as has been mentioned, essentially cavity. Also, without the release of suc-
the hydraulic conductivity of the soil at tion, a larger cavity than the intended
a point. This is possible because the hole one could result. It should be recognized
wHICHCopyright
is bored into the
by ASTM soilrights
Int'l (all forreserved);
the here Sunthat,
Jan 11when boring
13:11:58 EST is done below the
2015
water table, atmospheric air cannot enter pouring lead" inside a further encasing, insu-
the bored out cavity, except by artificial lating, plastic cylinder. The other end of the
means. wire makes contact, through the metal reel
After the cavity has been prepared, clutch, with one side of a current-limiting
water seeping into it is pumped out two resistor. The other side of the resistor is
series-connected to a milliammeter, a 33-v
or three times to remove the effect of hearing aid battery, and to "ground."
puddled soil on the cavity walls. A small Ground can be the inverted (brass) cup,
pitcher pump with a flexible hose to go since it makes contact with the top of the
to the tube bottom is used. When, after piezometer. When the weighted end of the
a pumping, the rate of rise of water in the reel wire touches the water, current flows
conduit is the same as for a preceding through the circuit, and the needle of the
pumping, the effect of puddling is as- milliammeter, serving as a "yes-no" meter,
sumed to be absent. deflects. The depth of the probe is then
established by means of two scales. One
scale is marked on the insulated copper wire
at foot intervals. The other scale, which is
a 1-ft section of a steel surveyor's scale and
is mounted on the reel post, provides for
reading the former scale to i^-y ft.
Thin, flat, flexible wire for the reel has
been found preferable to round wire, mainly
because a numbered scale can be engraved
on it. Flat 300 ohm television wire is par-
ticularly suitable for this purpose.
A battery of fairly high voltage is used
with the probe, so that in the case of rela-
tively salt free soils, the value of the resist-
ance through the water part of the circuit
FIG. 5.—Piezometer Method for Measuring will be small compared to that of the limiting
Hydraulic Conductivity of Soil (20, 21). resistor. Also use of the 33-v battery makes
increased resistances in the circuit due to
Soil water is now allowed to rise in the polarization negligible. Use of a 1^-v battery
piezometer until the equilibrium (refer- has not proved generally satisfactory. A
ence) level LT, Fig. 5, is reached. To hearing aid battery is used because it is
light.
determine the level LT, and also to de-
termine required values of L\ and Z2 At the same time that the electric
(Fig. 5), a reel-type electric probe similar probe is used to determine the levels LI
to the one described by Luthin (22) is and LZ for Fig. 5, a stop watch is used to
used. obtain the corresponding tz — h . The for-
mula for converting the measured quan-
The electric probe (Fig. 6) consists of a
rubber-insulated, copper wire wound upon a tities shown in Fig. 5 to hydraulic con-
clutch type (Pflueger Sal-Trout No. 1558) ductivity, for a 4 in. long, lyf-in.
fishing reel secured to a post on a stand, the diameter soil cavity at the base of a 2-
stand hi turn resting stably on the piezom- in. inside diameter cavity, is:
eter. An inverted cup arrangement on the
bottom of the stand fits over the top of the
stand to provide for the stability. The end
of the insulated copper wire leading from the
reel into the piezometer is encased, for where k will be in inches per hour if /2
weight, Copyright
in a sleeve of lead,
by ASTM Int'lprepared
(all rightsby and t\Sun
reserved); areJan
hi11seconds,
13:11:58and
EST L\ , Z,2 , and
2015
Li are each in the same units, inches; the + Lz — 2Lr)/2 is the average head,
factor 665 has been obtained with the aid
of reference 21, Fig. 2.
Equation 4 is more complex than it
need be for most field work. A simpler, Even if the quantity
very approximate expression, to be de-
FIG. 6.—Clutched, Fishing Reel Type, Elec- FIG. 7.—Electric Probe of Piezometer
tric Probe for Determining Water Levels in Method as Adopted for Reading Water Levels
Piezometer Method. After Luthin (22). in Auger Holes.
rived below, which is valid for almost all Eq 6 is as much as T77, use of Eq 6 results
practical cases is, with units as in Eq 4: in an error of only 10.4 per cent.
After measurements have been made
at the desired number of depths, the
piezometer may be pulled out by a Veih-
meyer soil tube jack or by a block and
(in. per hr).. (5) tackle on a tripod or by the hydraulic
liftSun
Copyright by ASTM Int'l (all rights reserved); of Jan
a tractor.
11 13:11:58 EST 2015
or in more easily remembered
Downloaded/printed by form, since It has been repeated that the piezom-
eter method yields essentially the hy- Since the approximation loge hi/hz =
draulic conductivity at a point in the 3H/(average head) can be used, not only
soil. This implies that a highly per- in the piezometer method but also in the
meable layer or very impermeable layer tube method and in a number of other
near the piezometer cavity should have cavity problems, for example, those
negligible effect on a determination, and catalogued by Hoorslev (12, p. 31), it is
this is, in fact, so. If an end of the cavity desirable to compute a few values of the
is closer than its length to such a layer, errors associated with the approximation.
the error may be ±8 per cent for a cavity Let Ah/(average head) = /. Then, hi the
1 in. in diameter and 4 in. long. For f ollowing pairs of numbers, the first num-
layers, at further distances, which are ber is/ and the second number is the per
highly permeable or very impermeable, cent error: 0.1, 0.09; 0.2, 0.4; 0.3, 1.3;
the error will be less than 8 per cent (21). 0.4, 2.0; 0.5, 3.8; 0.6, 6.4; 0.7, 10.4.
Figure 7 shows the electric probe for
the piezometer method as adapted for
use with auger holes. A pointed rod to be
pushed into the soil for a "ground" is
not shown in the figure.
Equation 5 remains to be derived.
From the notation in Fig. 5:
But the identity can be written: FIG. 8.—Tube Method for Measuring Hy-
draulic Conductivity of Soil (23).
TUBE METHOD
Also from calculus: A special case of the piezometer
method, where the piezometer is of large
diameter and the length of the cavity
at the piezometer base is zero, has, as
so that taking x = (hi — hz)/(hi + hz) indicated earlier, been called the tube
and neglecting the terms in s3, x6, etc.: method. This method, illustrated in Fig.
8, has been used primarily with tubes of
8-in. diameter, but theoretically any
diameter tube could be used. As with the
piezometer method, one measures (Fig.
or since (hi -f A2)/2 is the average head, 8) the levels hi and h2 and the tune inter-
and hi — hz is Ah: vals tz — h corresponding to hi — h%.
The formula for hydraulic conductivity
is (23):
or hi terms of Li,Lz, and LT :
where k is in inches per hour; r, hi, hz,

and E in inches; and tz — ti in minutes.
which when putbyinASTM
Copyright Eq 4 Int'l
yields(all Eq 5.reserved);The
rights Sun Jan 11 is
time 13:11:58 EST 2015
measured in minutes becaus
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further repr
D£ the inherently slower time of water sociated with the three methods, one
rise associated with the method due to the may arbitrarily take, as the size of the
smaller ratio of entry area to tube cross- sample, that volume of soil about the
section. cavities through which 80 per cent of the
The factor E in Eq 7 is a shape factor available hydraulic head is dissipated. If
depending on the depth and diameter of this is done, it may be computed for a
the tube; numerical values for it may be 2-in. diameter piezometer cavity 4 in.
taken conveniently from a table by long, for an 8-in. diameter tube, and for a
Spangler (17). 4-in. diameter auger hole, all penetrating
As an example of use of Eq 7 for a to a common depth 30 in. below the water
tube 8 in. in diameter and 16 in. below the table, that the sample sizes will be re-
water table, E =• 20.8 in. Equation 7 be- spectively in the proportions 100 to 270
comes: to 1400. Therefore, it may be estimated
that the sampling errors due to soil
heterogeneity for the three methods will
be as 100~1/2 to 270~1/2 to 1400~1/2 = 1
or putting tz — t\ = At, and hi — hz = to 0.6 to 0.4. Field practice bears out
AA and using the approximation derived this proportionality in a general way.
at the end of the last section: But it should be remembered that in
uniform isotropic soils the errors associ-
ated with the methods, other than those
due to soil heterogeneity, should nor-
A limitation on the tube method for mally be about the same, the auger hole
8-in. diameter tubes is that it can be method being perhaps somewhat more
used with presently developed equipment subject to inherent errors than the others.
only to a depth of about 30 in. below the
Reproducibility of Results:
soil surface. The reason that the tube
method tends to measure the vertical Data have been obtained which indi-
component of the hydraulic conductivity cate the reproducibility of determina-
is that all the streamlines which enter the tions for each method by itself. For the
tube must do so in a vertically upward auger hole method the following values
direction. of k in centimeters per day have been
reported (19): 403 ± 13, 488 ± 12, 98.5
COMPARISON OP THE AUGER HOLE, ± 3.9, 207 ± 11, 206 ± 5.2. The erro
PIEZOMETER, AND TUBE METHODS ±13, ±12, etc., when reduced to coeffi-
If the soils for the three above de- cients of variation are respectively, 3.2,
scribed methods are homogeneous and 2.5, 4.0, 5.3, 2.5, the average being 3.5
isotropic, the methods should all yield per cent. This 3.5 per cent is a lower
the same values of k. If the soils are coefficient of variation than one might
heterogeneous and anisotropic, the re- expect. The soil at the sampling sites here
sults will not be the same. Ignoring, reported was unusually uniform. On less
however, the effect of anisotropy for uniform sites, coefficients of variation of
the moment, it seems clear that the re- 10.7 and 14.4 have been reported (9).
sults for any one of the methods should Some rather complete experimental
be the more reproducible, in heteroge- data (21) (see Table I) are presented for
neous soil, the larger the soil sample. the piezometer method for three highly
permeable Iowa prairie soils and one
Size of Copyright
Samples:by ASTM Int'l (all rights reserved); Sun
Iowa clay soil.
Jan 11 13:11:58 The prairie soils are
EST 2015
To determine the size of samples as- highly permeable, presumably because of
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further reproductio
TABLE I.—HYDRAULIC CONDUCTIVITY MEASUREMENTS WITH PIEZOMETERS (21).
Hydraulic Conductivity, in. per day
Inside Pipe Depth of Depth of
Site0 and Soil Diameter, Water Sample,6 ft
in. Table, ft Coefficient of
Pipe 1 Pipe 2 PipeS Pipe 4 Pipe 5 Pipe 6 Average, in. per day Variation
No. 1 0.756 1.0 2 725 514 586 648 618 ± 90 13

Webster silty clay loam 0.756 1.0 3 1150 784 1222 1069 1056 ± 192 5.5
0.756 1.0 4 1144 1569 1355 1081 1287 * 221 5.8
No. 2 1.50 1.1 2 788 532 622 860 767 748 728 * 116 6.3
Webster silt loam 1.50 1.1 3 878 1140 888 955 878 813 925 ±114 8.1
1.50 1.1 4 840 810 814 862 1218 1150 949 * 207 4.6
No. 3 1.0 0.6 3 1.79 1.78 1.57 1.92 1.92 1.97 1.82 ± 0.147 8.1
Luton clay (not typical) 1.0 0.6 4 61.8 64.2 61.8 62.6 * 1.3 2.1
No. 4 2.0 0.4 2 159 153 103 95 127 ± 33 26

Marion silty clay loam 2.0 0.4 3 232 103 114 78 132 ± 68 52
1.0 0.4 3 134 159 80 124 =*= 40 32
a
6
The locations, except site No. 3, are on highly permeable prairie soils.
From surface of soil to top of cavity. Length of cavity is 4 in.
cracks, root channels, worm holes, and tained on small test sites, of areas gener-
inherently good agricultural structure. ally much smaller than an acre. On
The clay soil is of relatively poor struc- large test sites variability may be large.
ture. Notice that the inside diameters of On a 30-acre field, considered to be uni-
the piezometers varied from soil to soil, form, measurements (25) of the hydraulic
so that comparison of diameter effects conductivity varied from about 2 to 10
cannot be assessed from this table. The in. per day at 18-in. depth and from
worst coefficient of variation in the table about 2 to 6 in. per day at 30-in. depth.
is 52; the least, 2.1. Sand or clay lenses in a soil can of course
Ayers (24), using piezometers with cavi- cause extremely large variations in meas-
ties of 1-in. diameter and of 4-in. length, ured values.
reports a number of values of hydraulic
conductivities, together with their stand- Man Hours for the Three Methods:
ard deviations, for a glacial outwash
Some figures for the man hours for
area in Canada. From his data one com-
the piezometer and tube method, but
putes an average coefficient of variation
not for the auger hole method, are avail-
of 50. At one depth on one site he found
able. The auger hole method is faster
a standard deviation in excess of the
than the others.
mean (the coefficient of variation being
For the piezometer method (21), one
115). He attributes this to "variation in
man working in loam soil with 1-in.
permeability within the test area." In
piezometers can install four piezometers
considering a coefficient or variation such
at 2, 3, and 4 ft; obtain readings at each
as 115, one should remember, besides
Ayers' comment, the fact that values of of the depths; and remove the piezom-
eters, all in 6 hr. On heavy clay soil the
measurements of hydraulic conductivity
time is doubled. If two men, one un-
by common (core) methods often deviate
skilled, work together the number of
by more than 1000 per cent (5).
readings in a given time can be doubled.
For the tube method, reproducibility
Experience shows that the times for 1-
of results has been found to be good, if
in. piezometers apply equally well to 2-
measurements are made below the sur-
in. ones.
face 1-ft layer of soil (23). Considerable
For the tube method, under suitable
variability is found at about the plow-
conditions, two men working 8 hr can
layer depth. Some sample coefficients of
make determinations at 6, 12, 18, 24,
variation for 6-, 12-, and 18-in. depth
for an Iowa peat soil are respectively 47, and 30-in. depths with 12 tubes, each
tube at each depth. The tube method
10, and 12. For this particular field of
soil it was computed that 32.3, 1.1, and thus appears to require considerably
less labor than -the piezometer method.
1.5 tubes (that is, about 32, 1, and 2
But note that these figures for the tube
tubes) would be needed at the 6-, 12-,
method are for depths to only 30 in.
and 18-in. depths, respectively, to obtain
an average value of the hydraulic con- Soil Anisotropy:
ductivity of the soil which would be
within ±10 per cent of the true value It has already been stated that the
two thirds the times that these sets of auger hole and piezometer methods tend
tubes would be used hi an extensive pro- to accentuate the horizontal hydraulic
gram of determinations. conductivity of the soil. The possibility
The data given above on reproducibil- of sorting out the horizontal and vertical
ity of results refer
Copyright by to measurements
ASTM hydraulic
ob- reserved);
Int'l (all rights Sun conductivities
Jan 11 13:11:58byESTuse2015
of the
methods was pointed out by Reeve and FURTHER FIELD METHODS

Kirkham. Childs has considered the
problem in some analytical detail (26). M. Excluding the three methods described
Maasland and the author have done above (which were not yet developed),
further work on the problem and have Wenzel (28), hi 1942, described a number
obtained experimentally some, as yet of field methods for determining "per-
unpublished, values of the shape factors meability." Two of the most common
E needed for anisotropic cases. Some methods are based on discharge and
approximate values of E for anisotropic draw-down measurements for a single
cases can be obtained from the paper by well. These methods are well known
Hvorslev.5 and will not be mentioned here.
Use of any of the methods described
above for determining anistropy should Child's Two-Well Method:
ordinarily yield higher values of hori- Quite recently Childs(26) has proposed,
zontal hydraulic conductivity than the and he and co-workers (29) have utilized,
vertical (27). As an exception to the rule, a method employing two wells rather
H. P. Johnson6 found vertical values than one. The two wells are of equal
higher than horizontal, for a certain diameter and penetrate to the same
loessal soil. The general subject of soil depth below the water table, preferably
anisotropy, as it is related to seepage to an impermeable layer, if one exists.
problems, has been considered by Dutch Water is pumped at a steady rate out of
investigators. M. Maasland6 has gathered one well and carried by a hose into the
this work together, as well as that of other. A small but distinct hydraulic
Austrian and German investigators, into head difference is thereby created be-
the English language. tween the levels of water in the wells. If
Q is the pumping rate, Ah the hydraulic
Effect of Roof Channels, etc.: head difference, L the length of each well,
N. Kadir7 considers the effects of root r the radius of each well, and d the dis-
channels and worms holes on the meth- tance between their axes, then, the hy-
ods, also the effect of the use of im- draulic conductivity k is given by:
pervious disks at the base of piezometer
cavities. He concludes that a cone of
draw-down, rather than worm or root
channels, accounts for anomalies some-
times observed in the methods and that The units of k will be in inches per hour if
impervious disks placed at the bases of Q is in cubic inches per hour and L, d, r,
cavities have relatively small effects. and Ah are all in inches. Equation 8 is
correct only if the wells penetrate to an
8
For Hvorslev's case No. 4, p. 31 of reference impermeable layer. If an impermeable
(12), it is believed the factor 2.75 D is probably layer is not reached, a correction is to
too large; a better value may be 2.45 D. [See be applied.
Fig. 3 of reference (23), noting that the last two
plotted points, both of value A or E = 19.40 in., Some field results obtained with Eq 8
pertain to an 8-in. diameter tube; thus, E = have been given. It is not clear from the
(19.40/8) D = 2.45 D. See also Fig. 3 of reference data whether more than one pair of wells
(21) where the same result, 2.45 D, was obtained
independently.] was established at any one site. There-
6
M.S. thesis on file in the library of Iowa fore reproducibility of results from
State College, Ames, Iowa.
7
Ph.D. thesis on file in the library of Utah pairs of wells cannot be ascertained.
Copyright by
State Agricultural ASTMLogan,
College, Int'l (all
Utah. There
rights reserved); SunisJan
good
11 agreement
13:11:58 ESTfor results ob-
2015
tained by interchanging the pumped-out especially if sand liners are not used. By
and pumped-into wells. placing two more cased wells of the
Labor involved in making the two- piezometer type between the two outer
well measurements is not reported. Quite wells, a measurement of hydraulic con-
a little labor may be involved, as each ductivity should be obtainable which, in
well is cased with a screen and gravel theory, would be independent of any
liner. clogging effect at any of the four "wells."
The size of soil sample for the two-well This four-well method would be entirely
measurement will be quite large. This analogous to the four electrode method
volume was computed for a pair of wells, for measuring the electrical conductivity
each of 6-in. diameter, and extending to of soil (30, 31, 32, 33). The rate of water
an impermeable layer 30 in. below the movement between the outer two wells
water table. The calculation shows that and the difference in head between the
the size of sample, in which 80 per cent inner two wells would be measured. The
of the head will be lost, will be approxi- ratio of the rate of water movement to
mately equal to the corresponding vol- the difference in head between the inner
ume of soil for a single auger hole 4 in. in two wells would then be, to within a
diameter and 30 in. deep. The calculation determinable geometric factor, the hy-
thus raises the question as to whether a draulic conductivity.
single well, to which the auger hole
method would be applied, might not A Proposed Single-Well Method:
yield results as good as, or better than, The four-well method would probably
the two wells. But note that if an hori- have practical application only in con-
zontal anisotropy is in question the two fined artesian aquifers of well-known
wells might better be used in either geometry. The following proposed single-
event. well procedure might be of more general
In the two-well method, the well into use. Bore a cavity, as with a post hole
which the water is pumped will always auger, into the soil a foot or two below
be subject to a certain amount of clogging the water table. After puddling effects
by suspended material carried into the have been removed and the true ground
soil pores. Frevert and Kirkham found water level has been established, main-
that when a head was applied to a tain a small, pumped down distance Ah
cavity to force water into the soil in the well and record the pumping rate
rather than vice versa, the measured hy- Q. Let E be the appropriate shape factor
draulic conductivity was reduced by a for the well. Then,
factor of five. But Childs et al. worked
primarily in sandy soils. Also the screen
Q = kE&h (9)
and gravel liners, which they used, in which all quantities but k, and hence
must have reduced the clogging effect. k itself, will be known.
In the auger hole, piezometer, and tube The method will be clearer if it is as-
methods, the seepage is from the soil sumed that the well is of hemispherical
into the hole; accordingly clogging of the shape, its flat surface to coincide with
pores by suspended material is not a the water table as it exists before pump-
problem. ing. When the pumping is underway the
streamlines will be essentially horizontal,
A Proposed Four-Well Method: because it was agreed to make Ah small
It has appeared that clogging of pores compared to the depth or radius r of the
can be Copyright
a problembyinASTM
a two-well method,
Int'l (all well. Therefore,
rights reserved); Sun Jan 11the flow rate
13:11:58 EST into
2015 the
well will be (34, pp. 258-263; note espe- as simple as that for the proposed single-
cially near bottom p. 262): well method, can be used (see reference
17, p. 134 or reference 28, p. 78). The
Q = /K2*r)A& (10) procedure utilizes, besides the original
Comparing Eqs 9 and 10, it is seen that pumped well, two additional observa-
the shape factor for the hemispherical tional wells, rather than a single well as
well is E = 2irr. For a post hole, E will for Eq 10.)
involve, besides the hole's radius r, also If one follows through the mathemati-
its depth d. cal derivation of Eq 10, he will observe
Notice that the streamlines for the pro- that a distance corresponding to re hi
posed single-well method will all (except Eq 11 drops out of the analysis, when re
for a few at the top of the well, which goes to infinity. This observation is, in
exist when pumping is begun) originate short, the basis of the proposed single-
theoretically at infinity. This is an im- well method.
portant observation in the method be- Dry Auger Hole Method:
cause, at infinity, the hydraulic head will
always be Ah higher than it will be in the In order to determine the amount of
well because the supply of water at in- seepage to be expected when a canal is
finity is infinite and hence cannot be constructed in an arid area, the Bureau
drawn down by a finite amount of pump- of Reclamation has developed a dry
ing. Hence it is seen that Q will remain auger hole method for determinhig the
constant, as given by Eq 10, to within hydraulic conductivity of soil in place.
less than [2irrAh/(2Trr* - 2*rAA)] X 100 They designate the procedure as a "well
per cent. permeameter test" (35). The method
Notice further that the proposed single- consists of (a) boring an auger hole in
well method, as described, will not work dry soil, (6) casing the upper portion with
if the well penetrates to an impermeable a screen, (c) filling the lower portion with
layer. In that case, the flow rate would sand, and (</) observing the rate at which
be (34, p. 153): the hole will accept water. The diameter
of the hole may be 4 to 8 in. and the
depth 5 ft or more. Detailed procedures
and nomographs have been supplied for
the field worker.
where h is the depth to the impermeable
layer, re is the radial distance to an ex- Methods Depending on Discharge Meas-
ternal boundary where the hydraulic urement of Drain Tubes or Ditches:
head is Ah higher than at the well; the Seepage formulas have been developed
other symbols are as before. Since re in for flow into tile drains (36 to 40) and
Eq 11, to be comparable with the case ditches (36, 41), etc., in terms of the
of Eq 10, would be infinite, it follows hydraulic conductivity k. If all quanti-
that Q hi Eq 11 would be zero. Actually ties hi these formulas are known except k,
the flow to the well would not be zero. then k can be computed. Such a proce-
The flow hi excess of zero would be that dure would probably not have very much
originating in the layer of thickness Ah, use. Installations of the type hi question
flow not accounted for in the proposed are generally made, or should be, after
method. (To handle the case where the a knowledge of the hydraulic conductiv-
well penetrates to an impermeable layer, ity of the soil has been obtained by
a known procedure, but one which is not other methods.
REFERENCES
(1) L. A. Richards, "Report of the Subcom- (13) E. Diserens, "Beitrag zur Bestimmung der
mittee on Permeability and Infiltration, Durchlassigkeit des Bodens in Natiirlicher
Committee on Terminology, Soil Science Bodenlagerung," Schweizerische Landw.
Society of America," Proceedings, Soil Monathefte,, Vol. 12, Nos. 7, 8, pp. 188-198,
Science Soc. Am., Vol. 16, No. 1, Jan., 204-212 (1934).
1952, pp. 85-88. (14) S. B. Hooghoudt, "Bijdragen tot de Kennis
(2) S. B. Hooghoudt, "Tfle Drainage and Sub- van Eenige Natuurkundige Grootheden
irrigation," Soil Science, Vol. 74, No. 1, van den Grond, No. 4," Verslagen van Land-
July, 1952, pp. 35-48. bouwkundige Onderzoekingen, No. 42(13)
(3) R. E. Yoder, "A Direct Method of Aggre- B, Algemeene Lansdrukkerij, The Hague,
gate Analysis of Soils and a Study of the Holland, pp. 449-541 (1936).
Physical Nature of Erosion Losses," Jour- (15) Don Kirkham and C. H. M. van Bavel,
nal, Am. Soc. Agronomy, Vol. 28, No. 5, "Theory of Seepage into Auger Holes,"
May, 1936, pp. 337-351. Proceedings, Soil Science Soc. Am., Vol. 13,
(4) M. J. Jamin, "Memoire sur 1'Equilibre et pp. 75-82 (1948).
le Mouvement des Liquids dans les Corps (16) British Association Mathematical Tables,
Poreux," Comptes Rendus, Vol. 50, pp. Vol. VI, Part I, Bessel Functions of Orders
172-176, Jan. 23, 1860. See also J. H. Zero and Unity, University Press, Cam-
Poynting and J. J. Thomson, "A Text- bridge, England (1937).
Book of Physics," Vol. 1, Properties of (17) M. G. Spangler, "Soil Engineering," Inter-
Matter, Charles Griffin and Co. Ltd., national Textbook Co., Scranton, Pa.,
London, 7th Edition, pp. 142-143 (1919). Chapter X (1951).
(5) J. E. Christiansen, "Effect of Trapped An- (18) C. H. M. van Bavel and Don Kirkham,
on Soil Permeability," Soil Science, Vol. "Field Measurement of Soil Permeability
58, No. 5, Nov. 1944, pp. 355-365. Using Auger Holes," Proceedings, Soil
(6) A. Evan Harris, "Effect of Sodium on Science Soc. Am., Vol. 13, pp. 90-96 (1948).
Permeability," Soil Science, Vol. 32, No. 6, (19) H. P. Johnson, R. K. Frevert, and D. D.
Dec., 1931, pp. 435-446. Evans, "Simplified Procedure for the
(7) J. F. Lutz, "The Physico-Chemical Proper- Measurement and Computation of Soil
ties of Soils Affecting Soil Erosion," Re- Permeability Below the Water Table,"
search Bulletin 212, Missouri Agricultural Agricultural Engineering, Vol. 33, No. 5,
Experiment Station, July, 1934. May, 1952, pp. 283-286.
(8) Milton Fireman, "Permeability Measure- (20) Don Kirkham, "Proposed Method for
ments on Disturbed Soil Samples," Soil Field Measurement of Permeability of Soil
Science, Vol. 58, No. 5, Nov., 1944, pp. Below the Water Table," Proceedings, Soil
337-355. Science Soc. Am., Vol. 10, pp. 58-68 (1945).
(9) R. C. Reeve and Don Kirkham, "Soil (21) J. N. Luthin and Don Kirkham, "A Pie-
Anisotropy and Some Field Methods for zometer Method for Measuring Permeabil-
Measuring Permeability," Transactions, ity of Soil in Situ Below a Water Table,"
Am. Geophysical Union, Vol. 32, No. 4, Soil Science, Vol. 68, No. 5, pp. 349-358
Aug., 1951, pp. 582-590. Discussion, Vol. (1949).
33, No. 3, June, 1952, pp. 461-462. (22) J. N. Luthin, "A Reel-Type Electric Probe
(10) L. A. Richards, "Laws of Soil Moisture," for Measuring Water Table Elevation,"
Transactions, Am. Geophysical Union, Vol. Agronomy Journal, Vol. 41, No. 12, Dec.,
31, No. 5, Oct., 1950, pp. 750-756. 1949, p. 584.
(11) P. R. Day and J. N. Luthin, "Pressure (23) Richard K. Frevert and Don Kirkham, "A
Distribution in Layered Soils During Con- Field Method for Measuring the Perme-
tinuous Water Flow," Proceedings, Soil ability of Soil Below a Water Table,"
Science Soc. Am., Vol. 17, No. 2, April, Proceedings, Highway Research Board,
1953, pp. 87-92. Vol. 28, pp. 433-142 (1948).
(12) M. Juul Hvorslev, "Tune Lag and Soil (24) H. D. Ayers, "Soil Permeability as a Factor
Permeability in Ground-Water Observa- in the Translocation of Salts on Irrigated
tions," Bulletin No. 36, Waterways Ex- Land," Scientific Agriculture, VoL 31,
periment Station,
Copyright Corps ofInt'l
by ASTM Engineers, U. S.reserved);Sept.,
(all rights Sun 1951, pp.13:11:58
Jan 11 383-395.EST 2015
Army, Vicksburg, Miss., April,
Downloaded/printed by 1951. (25) Don Kirkham and J. W. deZeeuw, "Field
Measurements for Tests of Soil Drainage (34) M. Muskat, "The Flow of Homogeneous
Theory," Proceedings, Soil Science Soc. Fluids Through Porous Media," J. W.
Am., Vol. 16, No. 3, July, 1952, pp. 286- Edwards, Inc., Ann Arbor, Mich. (1946).
293. See especially Fig. 5 and compare (35) C. W. Jones, "Comparison of Seepage
Fig. 2. Based on Well-Permeameter and Ponding
(26) E. C. Childs, "The Measurement of the Tests," Earth Materials Laboratory Report
Hydraulic Permeability of Saturated Soil in No. EM-264, U. S. Dept. of Interior,
Situ. I. Principles of a Proposed Method," Bureau of Reclamation, Denver, Colo., Au-
Proceedings Royal Soc., A, Vol. 215, pp. gust, 1951, multilithed. (See especially the
525-535 (1952). appendix, which is taken from Bureau of
(27) C. G. F. Vreedenburgh, "De Parallelstrom- Reclamation Earth Manual, tentative
ing door Grond Bestaande uit Evenwijdige edition, reprinted Feb., 1952, Denver, Colo.,
Regelmatig Afwisselende Lagen van Ver- pp. 255-266 (1951).)
schillende Dickte en Doorlaatbaarheid," (36) S. B. Hooghoudt, "Bijdragen tot de Kennis
De Ingenieur in Ned. Indie, Vol. 4, No. 8, van Eenige Natuurkundige Grootheden
Aug., 1937, pp. 111-113. van den Grond. No. 7," Verslagen van Land-
(28) L. K. Wenzel, "Methods for Determining bouwkundige Onderzoekingen, No. 46 (14)B,
Permeability of Water-Bearing Materials," Algemeene Landsdrukkerij, The Hague,
Paper No. 887, U. S. Geological Survey Holland, pp. 515-707 (1940).
Water Supply (1942). (37) V. S. Aronovici and W. W. Donnan, "Soil-
(29) E. C. Childs, A. H. Cole, and D. H. Ed- Permeability as a Criterion for Drainage-
wards, "The Measurement of the Hydraulic Design," Transactions, Am. Geophysical
Permeability of Saturated Soil in Situ. II," Union, Vol. 27, No. 1, Feb., 1946, pp. 95-
Proceedings Royal Soc., A, Vol. 216, pp. 101.
72-89 (1953). (38) Don Kirkham, "Flow of Ponded Water
(30) F. Wenner, "A Method of Measuring Earth into Drain Tubes in Soil Overlying an
Resistivity," Bureau of Standards Science Impervious Layer," Transactions, Am.
Papers 258, pp. 469-478, Oct. 11,1915. Geophysical Union, Vol. 30, No. 3, June,
(31) Don Kirkham and G. S. Taylor, "Some 1949, pp. 369-385. See especially p. 383.
Tests of a Four-Electrode Probe for Soil (39) Don Kirkham, "Reduction in Seepage to
Moisture Measurements," Proceedings, Soil Soil Underdrains Resulting from their
Science Soc. Am., Vol. 14, pp. 42-46 (1949). Partial Embedment in, or Proximity to,
(32) R. Woodward Moore, "Earth-Resistivity an Impervious Substratum," Proceedings,
Tests Applied to Subsurface Reconnais- Soil Science Soc. Am., VoL 12, pp. 54-60
sance Surveys," Symposium on Surface
(1947).
and Subsurface Reconnaissance, Am. Soc.
Testing Mats., p. 89 (1951). (Issued as (40) J. J. van Deemter, "Results of Mathe-
separate technical publication ASTM STP matical Approach to Some Flow Problems
No. 122.) Connected with Drainage and Irrigation,"
(33) H. LeRoy Scharon, "Electrical Resistivity Applied Scientific Research (Holland), Vol.
Geophysical Method as Applied to Engi- A2, pp. 33-53 (1949).
neering Problems," Symposium on Surface (41) Don Kirkham "Seepage into Ditches in the
and Subsurface Reconnaissance, Am. Soc. Case of a Plane Water Table and an Im-
Testing Mats., p. 104 (1951). (Issued as pervious Substratum," Transactions, Am.
separate technical publication ASTM STP Geophysical Union, VoL 31, No. 3, June,
No. 122.) 1950, pp. 425-430.
MEASUREMENT OF PERMEABILITIES IN GROUND-WATER
INVESTIGATIONS
BY W. O. SMITH1 AND R. W. STALLMAN1
SYNOPSIS
Certain physical conditions must be considered in the measurement of per-

meabilities of soils and sediments. Among these are sediment structure (par-
ticularly whether the sample is disturbed or undisturbed), the occlusion of air,
and, finally, the effects of colloidal material. To meet some of these conditions,
a special sampling procedure has been developed for obtaining undisturbed
samples suitable for permeability measurement. The particular sampling
device is of the piston type and contains an inner barrel, in which an undis-
turbed soil sample is taken. This inner barrel, with its undisturbed sample, is
removable and serves as the permeameter tube in subsequent permeability
measurements.
The techniques for permeability measurement are described.
It is emphasized that the test liquid must be one that is in equilibrium with
the soil sample, so far as colloidal chemical properties are concerned. For
example, in the determination of ground-water permeabilities, actual ground
water, which is assumed to be in equilibrium with the sediments of the aquifer,
must be used. If fluids such as distilled water or ordinary laboratory tap
water are used, then severe washing of colloidal material may occur, owing
to a change in the exchangeable ions, with a consequent erroneous measure-
ment of the permeability.
Finally, the alterations in permeability due to trapped air and other gaseous
material are discussed. They result in a decreased value of permeability.
SIMPLE TRANSMISSION OF WATER The basic ideas involved in under-

THROUGH SEDIMENTS standing permeability phenomena in
The principal items that characterize sediments are simple. They have been
the response of a sediment to the trans- given by Hazen, Darcy, Kozeny, Slichter,
mission of a given fluid through it are and others. The vast majority of sedi-
the nature of the sediment, whether ments contain channels that consist of
gravel, sand, or clay, the material ini- networks of multiple conical capillaries.
tially occluded in its pores, such as air, ln nature, the flow through such tubes is
and finally the kind of permeating fluid, generally laminar. In such flow there is
1
Physicist and Hydraulic Engineer, respec- no turbulence, and the energy associated
tively, Ground Water Branch, Water Resources wit]j the driving head is used almost en-
Division,Copyright
U. S. Geological
by ASTM Survey, Washington,
Int'l (all rights . Jan 11 13:11:58 EST 2015
reserved); Sun
D. c. Downloaded/printed by tirely to overcome viscous drag between
S R KUMAR KONATHALA (Indian Inst of98
Tech Chennai) pursuant to License Agreement. No further reprodu
SMITH AND STALLMAN ON MEASUREMENT or PERMEABILITIES 99
flowing elements of liquid. For example, total flow. A second correction, necessary
in the case of a simple cylindrical tube, in very fine capillaries, is for the decrease
the flow is in a series of concentric an- in channel size caused by adsorption of a
nuli, each one of which has a different film (4, 5) of the wetting liquid. The film
velocity; the drag is between adjacent thickness constitutes an appreciable frac-
annuli, and all the energy of the driving tion of the size of the flow channel. There
head is used to overcome it. The flow is are also abnormal viscosity effects.
steady once the motion is established.
According to Darcy's law, the velocity EFFECTS OF ION EXCHANGE IN CLAYS
for purely viscous flow, VK , through an The influence of the permeant fluid
element, for a pressure differential, dp, must be considered, especially as it af-
between faces of unit area and a distance, fects the condition of the sample. The
dx, apart, is given by: composition of the water has little or no
v* K(dp/dx) (1) effect on the structure of sands or silts,
The quantity K is called the coefficient provided there is no clay present. Ordi-
of permeability and is determined by the nary gases likewise have little or no effect
on structure.
geometry of the pore system and the
However, when clays are a part of the
nature of the fluid. The minus sign indi-
sediment, many complications arise in
cates that the flow is opposite to the di-
the permeability measurement. Ordinary
rection of pressure increase. The pressure
gradient is dp/dx. If uniform, the flow, gases, for example, may be adsorbed by
Q, through a section of area A is AV* , or the clay complex. If water is used as the
Q = AV-H, and thus: permeant fluid, the physical properties of
the clay are altered, according to the
Q = -KA (dp/dx) (2) chemical composition of the water. If
Equation 2 may be written K = water is distilled to a high degree of
— (Q/'A)/(dp/dx) so that the perme- purity and then passed through a cal-
ability coefficient K is seen to be simply cium-saturated clay, the calcium will be
the quantity of fluid driven through unit leached out, and a hydrogen-saturated
area by a gradient of unity. This equa- colloid formed by an exchange of ions.
tion is not applicable to extremely coarse This colloid is generally freely mobile and
sediments, such as gravel, in which water begins to migrate. The hydrogen clay will
is moving at moderately high velocity. generally pack mechanically into a layer
The larger openings give rise to turbu- somewhere else in the sediment, and the
lence under moderate gradients, and the permeability will be characteristic of the
analogy is with pipe rather than capillary sheet of closely packed hydrogen clay,
flow. rather than of the original sediment. This
example illustrates the importance of
Corrections for Fine Capillaries: ion-exchange phenomena. These effects
Equations 1 and 2 are subject to cor- depend on the type of clay also. Mont-
rections when extremely fine capillaries morillonite, for example, has several
are used. One cause for correction arises times the ion-exchange capacity of
from the so-called "slip flow" (l, 2, 3),2 a kaolinite, and therefore is capable of
condition that results when molecular greater changes in physical properties be-
diffusion contributes appreciably to the cause of ion exchanges.
2
Related effects were observed in 1940
TheCopyright
boldface numbers
by ASTMinInt'l
parentheses
(all rightsrefer in the
reserved); Sunlaboratory of theEST
Jan 11 13:11:58 U. S. Department
2015
to the list of references appended to this paper,
see p. 114. of Agriculture, Division of Soil Chemis-
try and Physics, by the senior author of cient calcium to be in equilibrium with a
this paper, in collaboration with C. Niki- sediment containing calcium clay is
foroff and M. Drosdorf, although the passed through that clay, disturbance
results of their tests were never pub- should not occur. If, however, the sedi-
lished. It was planned at the time to do ment is leached with water containing
more extensive work, but war conditions sodium ions, an ion exchange ought to
prevented. At that' time three columns occur, with the liberation of a freely
of homogeneous medium quartz sand, mobile and more highly dispersed sodium
about 12 in. long, were set up. Three clay, which doubtless will migrate and
permeants were prepared and passed up- elsewhere seal off the sediment.
ward through these columns: (1) a water The above considerations seem to in-
suspension of a natural colloid from the dicate that in permeability measurements
B horizon of the Dayton silt loam, a clay it is necessary to use water that is similar
pan; (2) a water suspension of calcium- in composition to that which occurs in,
saturated colloid from the same Dayton or will ultimately be passed through, the
soil; and (3) a sodium-saturated colloidal sediment. For example, permeability of
suspension, also from the same Dayton the sediments lining an irrigation ditch
soil. All colloids initially moved freely should be measured with water typical
through the respective sands, but finally of that used for the irrigation. Perme-
sealed them. The natural and calcium- abilities of sediments below the water
saturated colloids sealed off their respec- table should be measured with ground
tive sands much more quickly, but much water taken from those sediments, or at
less completely, than the sodium colloid least water that is closely similar in
which actually took several times as long chemical composition, and hence is in or
as the other suspensions to accomplish close to equilibrium with the sediment.
this. The tightness of the seal and the These examples serve to illustrate the
time required to achieve it is explained fact that the permeabilities, to be useful,
by the difference in the size of the ulti- must be measured with the permeant
mate particles of each: the particles were that is to be used in the actual applica-
much smaller in the case of the sodium- tion. They further illustrate the im-
saturated colloid. portance of the theory of ion exchange,
When the columns were later examined, in clays, where water is involved. Fur-
it was found that the sodium-saturated ther information is available in an ex-
clay packed densely in a sheet about £ in. tensive amount of literature on the sub-
ject (6, 7).
thick, to form the seal near the outflow
(upper) end of the columns. Below this, INFLUENCE OF PHYSICAL CONDITION
small amounts of colloid were found for a OF SAMPLE
distance of about 8 in.; these amounts Another item of importance in perme-
were not sufficient to close the pores ap- ability measurements is the condition of
preciably. Similar conditions prevailed the sample—that is, its structure. In
in the case of both the calcium-saturated order to understand clearly the role
and natural colloid. played by structure, consideration of the
If water that is in equilibrium with common soils is helpful.
the sediment is used as the permeant Several different types of structure
fluid, changes in the mechanical arrange- generally are recognized by pedologists.
ment ofCopyright
the solids are not to be expected. In the
by ASTM Int'l (all rights reserved); Sun authors' opinion,
Jan 11 13:11:58 ESTthe most useful
2015
For example, if a water containing
Downloaded/printed by suffi- classification is that suggested by Niki-
SMITH AND STALLMAN ON MEASUREMENT OF PERMEABILITIES ior
foroff (8). In the soil there are primary It is well to bear in mind again that
and secondary structures, each of equal each single aggregate, such as a prism,
importance. The arrangement of the has a characteristic porosity, the primary
single soil particles is called the primary porosity formed by the spaces included
structure; hi turn, the single particles between the primary particles. When the
are generally combined into aggregates, soil has no secondary structure, but con-
which are the units of secondary struc- sists merely of one continuous mass of
FIG. 1.—Schematic Diagram of Some Common Soil Structures.
ture. The aggregates fall naturally into primary particles, it is said to be massive.
four simple patterns: granular, platy, If the soil possessing a natural secondary
prismatic, and blocky. In each of these, structure is crushed, there results an as-
the aggregates have the approximate semblage of fragments. Each fragment
geometrical shape indicated by the retains the porosity of the original aggre-
name: grains, plates, prisms, and cubes. gate of which it was a part, while a new
There are characteristic pore spaces be- system of secondary channels or openings
tween the secondary units. These rela- between the fragments is formed. Con-
tions are illustrated schematically in sequently there is a new secondary
Fig. 1.Copyright
Figure 2byis aASTM
photograph
Int'l (all of a pris-
rights porosity.
reserved); Sun Jan In
11 this connection
13:11:58 EST 2015it is to be
matic Downloaded/printed
arrangement. by observed that fragments slake easily in
water, whereas particles in their natural estimate is to be made on a natural soil,

structural condition do not, because of then the permeability of a natural and
protective coatings. undisturbed sample must be made, be-
If water-transmission properties are cause the greater part of the infiltration,
required in soil that is to be used in a especially in the initial stages, occurs in
crushed state, then the permeability the secondary pore space (9). When
FIG. 2.—Semidesert Brown Soil, Found Near Akin, Colo., Showing Prismatic Structure.
must be determined on the sample when swelling occurs, the secondary pore space
it has reached a porosity that is charac- gradually closes, and the ultimate ability
teristic of its new use. Slaking of frag- to transmit water is determined by the
ments when water is first introduced primary pore—that is, the space between
results in a changing apparent perme- the primary particles that compose the
ability which eventually stabilizes, pro- aggregates.
vided the soil is one that cannot be other- It is thus seen that considerable care
wise altered by the water. must be exercised in taking samples. To
If, on the other hand,
Downloaded/printed by an infiltration meet this condition, a drive sampler has
PIG. 3.—Piston
Copyright by ASTM Int'l (all rights reserved); Sun Jan 11 13:11:58 EST 2015Drive Sampler.
been designed and applied in a technique anchorage plate with the attached
for making permeability measurements. vertical guide cylinder, 1, is the basic
support of the sampler during the
APPARATUS FOR TAKING UNDISTURBED sampling operation. It is held in position
SEDIMENT SAMPLES by three steel pins, 5, that are driven into
Many of the sampling devices in cur- the soil through holes in the base plate.
rent use have been reviewed in detail by Part 13, an adjustable piston fitted with
Hvorslev (10). a rubber piston ring is fastened to the
A sampler suitable for ground-water vertical guide cylinder by means of a
investigations in which it is generally de- crosspiece, 12, attached with four screws,
sired to determine the natural rate of 16. The piston mount is different from
transmission of water in the ground must that of the orthodox piston-drive sam-
take samples that are. adequate for use plers. In most devices the piston is per-
in determining the hydraulic properties mitted to ride freely on top of the sample
of sediments. Specifically, the sample as the sampler barrel is forced into the
must retain its natural structure during ground. This allows the sample to ex-
(a) the sampling operation, (6) transport pand and thus to increase in length as it
to the laboratory, and (c) placement of enters the barrel. In this sampler, axial
the sample hi a permeability apparatus. expansion of the sample is prevented, and
The senior author of this paper has the sample preserves its original con-
developed a sampler that meets these figuration. (Compaction of the sample is
requirements (Fig. 3(a)). With this minimized by the technique used.)
device, undue disturbance of the sample The barrel, 3, which is driven into the
during the driving operation is prevented ground, has a steel drive shoe, 11, at-
by means of a piston that can be ad- tached to the bottom. The drive shoe has
justed to fit tightly against the sediment a 2-in. inside diameter, and the barrel a
face. The sample is caught in a rigid core 2j-in. inside diameter. Three pieces of
retainer of clear, smooth plastic. The clear plastic tubing, 4 and 8, with a wall
sample is not disturbed as it passes into thickness of f in., are inserted into the
the tube. The clear plastic case contain- barrel from the bottom before the drive
ing the core can be made to serve as the shoe is attached. The center tube (5J in.
barrel of a specially designed perme- long) is the core retainer, 4, and ulti-
ameter, without removal of the sample. mately becomes the barrel of the perme-
Hence, the disturbance of the natural ameter. The rings, 8, f in. long, shown at
structure of the sediments is kept to a each end of the center tube, are parts of
minimum from the beginning of the the core retainer, and their use forms an
sampling operation to the end of the important part of the sampling tech-
permeability test. The present equip- nique, as explained fully in the section
ment is designed for surface sampling in on Obtaining the Sample.
excavations. Additional equipment is Use of the clear plastic core retainer
contemplated to apply the present permits inspection of the sample in the
sampler design to sampling inside cased 'field. If the sample is not a good one, as
or uncased borings at considerable may be determined by inspection, an-
depths. other can be taken immediately, thereby
eliminating the cost of visiting and pre-
DESCRIPTION or SAMPLER paring the sampling site more than once.
The Copyright
unassembled sampler is shown in The plastic parts are held at the top
by ASTM Int'l (all rights reserved); Sun Jan 11 13:11:58 EST 2015
Fig. 3(6). The triangular
Downloaded/printed by base, 2, or by a retainer ring, 9, screwed into the
top of the barrel. A cap wrench, 10, hav- driving head, base plate, and core re-
ing keys to fit the slots of part 9, fits tainer and end rings are made of stainless
freely on top of the retainer ring. The steel. The base plate and driving head are
cap wrench transmits the driving force of brass; the core retainer and end rings
from the bottom of the driving fork, 6, are polished, clear plastic tubing, 2 in.
to the top of the sampler barrel. The in inside diameter, cut to length. The
FIG. 4.—Adjustment of Piston to Make Good Contact with Sediment Face.
FIG. 5.—Placing of Driving Fork in Position.
driving fork is slotted to clear the cross- commercial plastic tubing used has an
piece supporting the adjustable piston. outside diameter slightly less than 2| in.,
The detachable driving head, 7, made of just sufficient to allow a free fit in the
brass, fits into the top of the driving core barrel.
fork, which is made of stainless steel in
order to reduce the total weight. (This OBTAINING THE SAMPLE
is an important factor in horizontal
sampling.) A 24-oz rubber mallet is used In obtaining undisturbed samples,
for driving. properSuntechnique is as important
Copyright by ASTM Int'l (all rights reserved); Jan 11 13:11:58 EST 2015 as the
All Downloaded/printed
parts of the sampler
by except the use of proper apparatus. The procedure
106 YMPOSIUM ON PERMEABILITY OF SOILS
now followed by the authors for taking much care is required to keep the core
undisturbed samples is outlined below. undisturbed during its removal from the
Sites are selected in the area under in- site and its preparation for storage. The
vestigation, from which to obtain repre- size of the excavation needed for sam-
sentative samples of the materials. These pling varies with the depth from which
sites should be staked or otherwise samples are required and the structural
clearly described for the sampling crew. properties of the earth materials forming
As excavation is required to obtain sam- the walls of the pit. In general, the pit
ples of materials below the surface, all should have a minimum cross-section of
construction excavations or borrow pits about 4 by 6 ft, and have straight,
being dug at the time of sampling, or vertical walls. The rectangular pit is dug
FIG. 6.—Further Removal of Sediments Around Sampler Barrel.
made not long before, should be exam- with a hand spade. Care must be taken
ined to determine whether suitable throughout the spading to reduce to a
samples can be taken from the walls. The minimum the disturbance of the ma-
cost of sampling can be reduced materials at the wall face. A complete de-
terially by thus taking advantage of scription of the materials encountered hi
existing excavations. After a network of the digging is recorded, to be a part of
the best sites is described, the sampling the notes describing the samples.
crew collects the samples. All surfaces to which the sampler is
Samples of all unconsolidated ma- fastened are dressed flat with a spade or
terials above the water table can be taken spatula. One end of the pit is used for
if the particle size of the material is small taking vertical samples. This requires a
in comparison to the diameter of the ledge about 18 in. wide extending across
barrel. Some have been taken a short the pit, with its top near the land surface.
distance below the
Copyright water table,
by ASTM butrights
Int'l (all if the After Sun
reserved); a satisfactory sample
Jan 11 13:11:58 EST is obtained
2015
saturated by and granular, at the initial elevation of the ledge, the
material is clean
Downloaded/printed
SMITH AND STALLMAN ON MEASUREMENT OF PERMEABILITIES 107
ledge is cut down to the elevation at driven through the holes in the base
which the next sample is desired. Hori- plate into the sediment. The piston is
zontal samples are taken from the wall of adjusted by freeing the lower lock nut
the pit. In taking horizontal samples, the that rests against the crosspiece and ap-
top edge of the base plate of the sampler plying a wrench to the upper of the two
(2 in Fig. 3) should be horizontal to ob- top lock nuts on the piston shaft (see Fig.
tain maximum support from the steel 4). The shaft is turned slowly until the
pins. piston rests firmly on the sediment face;
The sampler is "loaded" with a core then the bottom lock nut, 14, is tightened
FIG. 7.—View of Barrel Through "Relief" Hole.
retainer and end rings as follows: (a) the and the unit is ready for the driving
drive shoe, 11, is removed; (6) the inside operation.
of the sampler barrel, 3, is thoroughly The sampler must be rigid against the
cleaned with a brush; (c) a plastic end sediment face to obtain a perfect sample;
ring, 8, a core retainer, 4, and another if mounted loosely, it may wobble during
end ring are inserted in that order; (c£) the driving operation, causing the sample
the drive shoe, 11, is fastened into posi- to be fractured.
tion; and (e) the sampler barrel, 3, is in- The driving head is fitted into the top
serted in the guide cylinder, 1. of the driving fork, which is inserted into
The barrel of the sampler is retracted the sampler, straddling the crosspiece
so that the shoe is above the base plate, (Fig. 5). Successive blows of the rubber
and theCopyright
samplerbyisASTM
held against the pre- hammer
Int'l (all rights reserved); Sun Janon
11 the driving
13:11:58 EST head
2015 force the
pared sediment face as the
Downloaded/printed by steel pins are barrel into the sediment. Generally, the
FIG. 8.—View of Barrel Just Before Removal from Sediments.
force of the hammer blow should be

sufficient to move the barrel about^ in.
per blow. It is desirable to strike the
driving head with the rubber mallet as
near center as possible. Glancing blows
cause the sampler axis to deviate from
the principal direction of driving, which
fractures the core.
After the barrel has been driven about
2 in., a long screwdriver or similar "re-
lieving" tool is used to remove material
from around the barrel as far into the
sediments as the junction of the barrel
with the shoe (Figs. 6 and 7). This "re-
lieving" is done to reduce the amount of
compaction caused by entry of the
sampler barrel. Care must be taken to
leave sufficient bulk around the steel
pins for support of the sampler during the
FIG. 9.—Removal of Core from Sampler Bar-
remainder of the sampling operation. The rel. In this particular sampling," shoe' remains at-
driving and relieving process is con- tached to core during removal from barrel.
tinued until the plastic core retainer and
end rings are filled. The steel pins are re- should not be pulled from its location,
moved, and the outer part of the sampler because the forces involved may fracture
is removed carefully from the sediment the sample and possibly destroy it. The
face. The sampler barrel will remain sampler barrel should be cut free by ex-
lodged in the sediments (Fig. 8). It cavating around it, and the end of the
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further r
FIG. 10.—Severing of Top End Ring from Core by Insertion of Spatuk Between Parts and Cutting
Across End of Core.
core at the edge of the drive shoe should

be severed with a spatula.
The drive shoe is carefully removed
from the barrel, but remains intact on
the core, keeping to a minimum the
disturbance of the part of the core in the
lower end ring. The end rings and core
retainer are then easily removed as a
unit (Fig. 9). A broad spatula is inserted
between both end rings and the core re-
tainer, and the core is severed by slowly
forcing the spatula through it (Fig. 10).
A small amount of the material taken
from the end rings is heaped on the ends
of the core and rounded off before the
vinylite caps are placed over them. The
addition of this loose material insures a
tight fit of the core in the core retainer
during transport; the loose material is
carefully scraped away in the laboratory. FIG. 11.—Inversion and Preparation of Other
A band of waterproof tape seals the End Cap.
of Core, Just Before Sealing with Vinylite
cracks between the lucite core and
vinylite caps and holds the caps in place. for the drive shoe end of the core (Fig.
If long storage is contemplated, samples 11). The sample, ready for transport to
should be sealed in wax in addition to the laboratory, is shown in Fig. 12.
being taped. This procedure is repeated The lucite sample tube is readily made
to serve as the barrel of the permeameter. If the permeant is unsaturated, the gas
The top and the bottom parts required will be reduced or removed by solution.
to accomplish this are shown hi Fig. 13. This factor has been observed to produce
The jig shown in Fig. 14 is used hi drill- erratic results hi the measurement of
ing the barrel for insertion of head gages. permeability. In order to be consistent
As was previously stated, the sampler for comparison purposes, it is preferable
described was designed primarily for to measure permeabilities at a common
sampling above the water table. It is to fluid-saturation level. This is best
be expected that ah- or other gases par-
tially fill the interstices of the sample.
Even if the sample is taken below the
FIG. 12.—Completely Sealed Core Ready for

Transport.
FIG. 13.—Drawing of Permeameter Parts Re-
quired for Permeability Measurements on Sedi-
water table, it is difficult to maintain ment Core Shown in Fig. 12. An empty core re-
tainer is shown.
100 per cent saturation up to the tune
the sample is placed hi the permeameter.
The gases locked hi the sample pores hi achieved in practice by maintaining 100
effect decrease porosity and hence reduce per cent fluid saturation throughout the
the cross-section available for fluid trans- test.
mission. As the permeability test is be-
gun, liquids initially circulating through AIR REMOVAL
the sample by-pass the larger pores, seal- Air locked in a sediment sample can
ing off quantities of gas. Capillary forces, generally be removed by one of several
as shown schematically in Fig. 15, hold procedures. An application of vacuum
the gas hi place. If the permeant is sat- followed by an application of atmos-
urated with gas, there is the possibility pheric pressure will remove trapped air
that the gas bubbles
Copyright willInt'l
by ASTM grow largerreserved);
(all rights and ifSunrepeated sufficiently
Jan 11 13:11:58 EST 2015(11). The proce-
furtherDownloaded/printed
reduce the cross-section
by of flow. dure has the disadvantage that it tends to
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreement. No further repro
disturb the sample, particularly by al- by a high-capacity pump. The vacuum

tering the porosity to that of a tighter chamber used in the hydrologic labora-
packing. It is much more useful as a tory of the Ground Water Branch, U. S.
means to degas sediments when grain Geological Survey, is a large rectangular
density is being determined. copper tank 1 ft by 4 ft, and 9 in. deep.
A better procedure, which is less dis- A float controls the intake of water
turbing to the sample, is to circulate gas- through valve B (see Fig. 16). An electri-
free water through the sample until all cal timer, C, controls the vacuum ap-
trapped air is removed by solution. This plication valve, D. The same timer
may require several weeks hi extreme controls the valve, E, that permits
cases. But approximate checks on atmospheric pressure to prevail within
the tank. The timer, F, controls the out-
FIG. 15.—Locking of Air in Sediment Pore by

FIG. 14.—Brass Jig Used to Tap Lucite Core Capillary Forces.
Barrel for Head Gages Required for Permeabil-
ity Measurements.
flow of water into the sand filter and
trapped gases have shown that prac- storage reservoir. The operation (refer-
tically no gas remains in most sand sam- ring to Fig. 16) is as follows:
ples after the circulation of deaired water As the water level hi the tank is low-
for a period of 8 hr. Studies to determine ered, a mercury switch, A, operated by a
residual gas by more refined means are float in the tank closes the circuit through
contemplated for the future. The gas- the solenoid valve, B, allowing water to
free water must be in equilibrium, chem- pass into the tank. Water flow into the
ically, with the clay fraction of the tank continues until switch A is opened
sample. and solenoid valve B shuts off the water.
The apparatus that the authors have The process timer, C, operates with a
used for some tune to degas water is 7-min cycle. When it closes the circuit,
shown hi Fig. 16. In it water is degassed D (a solenoid valve) opens, allowing
by theCopyright
application of a Int'l
by ASTM vacuum of ap-
(all rights vacuum
reserved); to be
Sun Jan 11 applied
13:11:58to the2015
EST water in the
proximately 28 in. of mercury
Downloaded/printed by produced tank, while E, the air-intake valve, iso-
FIG. 16.—Schematic Diagram of Device for Obtaining Air-Free Water for Use in Permeability
Measurements.
FIG. Copyright by ASTM

17.—Schematic Int'lof
Diagram (all rights reserved);
Equipment Used forSun Jan 11 Permeabilities
Measuring 13:11:58 EST in2015
Ground-Water
Investigations.
lates the system from the atmosphere. wells, for measuring the temperature of
After vacuum has been applied for about entering and outgoing water, are placed
2 min, timer C closes solenoid valve D at K and K'.
and opens valve E, and the tank returns
to atmospheric pressure. Then timer F SUMMARY
closes the circuit to solenoid valve G,
which opens to permit gravity flow from The problem of determining the per-
the evacuation chamber to the sand meability of sediments in the laboratory
filter and reservoir below. After about has been discussed as it relates to ground-
4| min of gravity flow, solenoid valve G water investigations. Some of the sig-
is closed by F. Timer C then takes over nificant factors in permeability measure-
and starts the cycle again. ments are discussed briefly. These are (a)
This device, constructed as described, influence of physical condition of the
permits the rapid degassing of water and sample, (6) the effect of ion exchange on
provides a continuous supply of about permeability of samples containing
100 ml per min of practically air-free clays, and (c) the effect of occluded air
water, sufficient to operate several per- on the observed permeability.
meameters simultaneously. It is emphasized throughout that re-
liable permeability investigation can be
MEASUREMENT OF PERMEABILITY accomplished only if all aspects of such
studies are considered. These begin with
The actual procedures involved in field sampling and end with laboratory
permeability measurement are simple. A measurement. Only 'by such careful
degassed water, preferably one in chem- procedure can the erratic results so often
ical equilibrium with the sediment, as seen in permeability measurements be
suggested earlier, is used to saturate the eliminated.
sample and to maintain saturation dur- It is suggested that the anticipated ap-
ing the test. The tube in which the plication of the data to be obtained be
sample is taken serves as the barrel of given detailed consideration when de-
the permeameter without removal of the signing a permeability test.
sample.
The schematic diagram of the con- A cknowledgments:
stant-head apparatus for measurement
of permeability is shown in Fig. 17. A The authors acknowledge the help of
is the deairing equipment, E is the per- A. H. Frazier of the Water Resources
meameter, D and F are manometers. B Division Equipment Laboratory, U. S.
and H are constant-head control tanks, Geological Survey, Columbus, Ohio. Mr.
intake and discharge, respectively; C is Frazier built the prototype of the sam-
a constant-level funnel for the intake or pler described in this paper and made
discharge; G is a collecting vessel for the many helpful suggestions during con-
discharge from A. The form of H is sub- struction. George H. Taylor, of the Sur-
ject to modification, depending on the vey's Ground Water Branch and Re-
sample under test and particularly its gional Engineer in charge of the Missouri
discharge rate. Measurements of head River Basin program, made many sug-
differential, discharge rates, and dimen- gestions. Gordon Andreason and S. J.
sions of the permeameter are sufficient to Spiegel, both of the Geological Survey,
calculate the permeability.
Copyright Thermometer
by ASTM Int'l gave much
(all rights reserved); assistance
Sun Jan to the
11 13:11:58 ESTauthors.
2015
REFERENCES
(1) H.Adzumi, "The Flow of Gases Through "Electroviscosity IV. Some Extensions of
a Poms Wall," Bulletin, Chemical Soc., the Theory of Flow of Liquids in Narrow
Japan, Vol. 12, p. 304 (1937); P. C. Car- Channels," Vol. A198, p. 581 (1949).
man, "Diffusion and Flow of Gases and (6) Walter P. Kelly, "Cation Exchange in
Vapours Through Micropours," Proceed- Soils," Rheinhold Publishing Co., New
ings, Royal Soc. (London), Vol. A203, p. York, N. Y. (1948).
55 (1950). (7) L. D. Baver, "Soil Physics," John Wiley
(2) P. C. Carman and J. C. Amell, "Surface and Sons, Inc., New York, N. Y., 2nd Ed.
Area Measurements of Fine Powders Using (1948).
Modified Permeability Equations," Cana- (8) C. C. Nikiforoff, "Morphological Classifi-
dian Journal of Research, Vol. 26A, p. 128 cation of Soil Structure," Soil Science, Vol.
(1948). 52, pp. 193-211 (1941).
(3) F. M. Lea and R. W. Nurse, "Permeability (9) W. O. Smith, "Pedological Relations of
Methods of Fineness Measurements," Sup- Infiltration Phenomena," Transactions,
plement to Transactions, Inst. Chemical Am. Geophysical Union, Vol. 30, pp. 555-
Engrs., Vol. 25, p. 47 (1947). 562 (1949).
(4) J. C. Henniker, 'The Depth of the Surface (10) M. J. Hvorslev, "The Present Status of
Zone of a Liquid," Reviews of Modern Obtaining Undisturbed Samples of Soils,"
Physics, Vol. 21, p. 322 (1949). Harvard University, Cambridge, Mass.
(5) G. A. H. Elton, "Electroviscosity I. The (1940).
Flow of Liquids Between Surfaces in Close (11) W. O. Smith, "Density of Soil Solids and
Proximity," Proceedings, Royal Soc. (Lon- Their Genetic Relations," Soil Science, Vol.
don), Vol. A194, p. 259 (1948); ibid., 6, pp. 263-272 (1943).
DISCUSSION
MR. ARNOLD I. JOHNSON.'—Smith The test fluids used in permeability

and Stallman describe very well some of measurements should be considered
the important factors to consider when carefully. Water of almost any composi-
determining the permeability of soils tion probably will react differently, in
and sediments. The Hydrologic Labora- flow through a sample, from oil, gas, or
tory of the Geological Survey, formerly other test fluids. Likewise, water of one
located at Lincoln, Nebr., and recently chemical composition may react dif-
moved to Denver, Colo., has studied ferently from that of another composi-
many of these factors and for a number tion. Therefore, it is seen that the test
of years has used sampling equipment fluid should be similar in all ways to
and permeability apparatus similar to the fluid that is, or will be, intimately
that described by Smith and Stallman, associated with the soils or sediments
but modified by the writer. It seems being tested. For ground-water investi-
pertinent to emphasize some of the gations it would be ideal if the test fluid
limitations of equipment of this type. were of the same chemical composition
The need to use undisturbed samples as the water found in the formation
cannot be overemphasized. The per- from which the sample was obtained.
meability of the secondary structure can This idea is recommended for small re-
be many times more than the permea- search projects, but when samples are
bility of the primary structure. shipped in, in quantity, from all over
Considerable field experience with the the United States, such as they are at
piston sampler described has shown that the Denver Laboratory, it is not eco-
it is effective under many conditions. It nomically feasible to have enough of the
generally must be used at exposed loca- actual water also shipped in, or prepared
tions such as the land surface, test-pit in quantities sufficient for routine use,
wall, exposed faces of road cuts, walls unless desired for a special study. How-
at stream channels, etc. It has been used ever, it is feasible to use a de-aired water
in the zone of saturation in shallow pits. system for all permeability tests.
However, modifications would be needed The Denver Laboratory uses a per-
to make it suitable for use at depth in meability system similar in principle
a borehole. This—like other existing to that described in this paper, but be-
samplers—does not take completely cause a large number of permeameters
undisturbed samples of all materials must be kept operating and a wide
under all conditions. In our experience, range of materials must be tested, the
very moist materials were commonly system differs considerably in physical
compacted during sampling, and very features. One 30-gal tank for the vacuum
dry silt or clay samples usually were
chamber and another 30-gal tank for
badly fractured.
storage of the de-aired water are used.
1
Chief, Hydrologic Laboratory, Ground The timing cycle is set at 3£ min on
Water Branch,
Copyright Water
by ASTMResources Div.,
Int'l (all rights U. S. Sun Jan 11 13:11:58 EST 2015
reserved);
Geological Survey, Denver,
Downloaded/printed by Colo. vacuum and 3^ min on atmospheric
S R KUMAR KONATHALA (Indian Inst of Tech 115Chennai) pursuant to License Agreement. No further reproductions a
pressure to provide a longer period for generally applicable to all sampling

drawing the air out of the water. The problems. Rather, it was designed to
tap water is sprayed in the tank near the incorporate many of the features con-
top of the vacuum chamber. The per- tained in other individual sampler de-
meameters themselves have had to be signs, such as the adjustable piston,
redesigned to adapt them for testing transparent core retainer, core barrel
either disturbed or undisturbed samples guide, triangular base to provide access
of consolidated or unconsolidated ma- to barrel for relieving compaction, etc.
terials. For samples of very low per- The combination of these design features
meability, the permeameters, as de- tends to make this unit more versatile
scribed, have not been found satisfac- than many other sampler designs which
tory in our laboratory and have had to contain only one or two of the most de-
be adapted for use as variable-head sirable components.
permeameters. Most soil technicians hesitate to take
The use of mechanically de-aired water drive samples when soil moisture is very
as test fluid offers some difficulties, as high or very low. Such an attitude has
well as advantages. The dissolved min- been common since the tunes of Doku-
erals, which are held in solution in the chaev and Glinka some 70 years ago. It
water by the dissolved carbon dioxide is well recognized that, under drive
gas, may precipitate out upon removal sampling, very wet samples are usually
of this gas in the de-airing process. The compacted, and very dry ones generally
precipitation most likely may occur upon fractured. It is possible, however, to
entrance of the water into the sample or obtain good undisturbed samples of
within the entrance filter disk. This may very dry sediments if extreme care is
cause progressive plugging of the sam- used, but even then, only one out of three
ple by the colloidal precipitates, and an attempts may yield a good sample.
apparent progressive decline in per- Samples of very moist or very dry
meability of the sample. However, from soils can be obtained if a piston sampler
the investigations being conducted in is slowly jacked into the sediment. This
the laboratory, it is known that the use procedure requires a modification of the
of de-aired water may give results for techniques described in the paper. Jack-
permeability 2 to 5 times greater than ing equipment has been designed for
those obtained by use of ordinary tap this purpose, and the results will be
water on like samples. released after field trials are completed.
In summary, it should be emphasized The need for a piston-type core barrel
that the determination of permeability for use in boreholes is great. We have
some preliminary designs for a piston
for soils and sediments is affected by
sampler for use on a rotary drill stem, as
many factors. Some of these still await
well as jacking equipment for shallow
definitive investigation, and the Denver boreholes. No construction has been
Hydrologic Laboratory hopes to con- attempted as yet.
tribute to this end in the ensuing years. Permeant fluids similar in composi-
MESSRS. W. O. SMITH AND R. W. tion to the water found in a formation are
STALLMAN (authors).—The drive sampler indicated only when clay-bearing sedi-
described in the paper was designed for ments are under consideration (see Miss
use hi exposed locations such as land Foster's discussion). Water of a composi-
surface,Copyright
test-pitbywall, exposed faces of tion similar to that which is passing or is
road cuts,
Downloaded/printed by channels, etc.
walls of stream to pass through the sediment in the field
It wasS not, of course,
R KUMAR considered
KONATHALA as Instis ofnot
(Indian required
Tech Chennai)aspursuant
a permeant in clay-
to License Agreement. No
DISCUSSION ON MEASUREMENT or PERMEABILITIES 117
free or nearly clay-free sands, for exam- clay. Another factor that may affect
ple. Some exceptions exist—for example, permeability measurements of clay-
a sand containing a small amount of Na- containing samples is the tendency of
montmorillonite. In many cases ordinary some clays to swell in water. Differences
de-aired water is satisfactory as a per- in the swelling characteristics of dif-
meant. If, however, its use is contra- ferent clays may be related to their
indicated, any results obtained will crystal structure, chemical composition,
likely be of questionable value. to the kind and degree of isomorphous
The use of de-aired water generally replacements in their structure, and to
renders all pore space in a sample avail- the amount and nature of their asso-
able for the transmission of the per- ciated exchangeable cations.
meant. It eliminates the locking of air in
CRYSTAL STRUCTURE AND CATION-EX-
pores by capillary forces, and the con-
CHANGE or CLAY MINERALS,
sequent reduction of pore space. It was
recognized early in our investigations The three great groups of clay min-
that some of the minerals initially held erals are the kaolins, the montmorillo-
in solution in water by dissolved COa may nites, and the hydrous micas, or so-called
precipitate out upon removal of this gas illites. In the kaolin group the unit
in the de-airing process. Accordingly, a structure consists of a silica sheet and an
sand filter was inserted (unit—shown in alumina sheet bound together. The sum
Fig. 16) at the outlet of the storage tank of the positive charges of silicon and
to remove these precipitates. The sand fil- aluminum in these sheets exactly equals
ter did not prove to be effective for remov- the sum of the negative charges of the O
ing colloidal silica. However, MacQueen2 and OH ions; the structure has no un-
found that the carbpn disks at the sample satisfied electrical forces by which other
ends did act as filters for the silica. If ions can be held to it. The small cation-
tap water contains appreciable amounts exchange capacity of the kaolin group of
of silica, it can be removed by the inclu- minerals—usually less than 15 milli-
sion of a carbon disk filter at the outflow equivalents per 100 g—is attributed to
end of the degassing chamber. These broken bonds on the edges of the parti-
filters have proved to be very satisfac- cles. Clays of the kaolin group exhibit
tory. little or no swelling in water.
Permeability investigations have been The montmorillonite group of clay
a subject of investigation by many minerals has the fundamental structural
hydrologists, engineers, and physicists pattern of the micas—two silica sheets
since the times of Darcy and Hagen, a bound together by an alumina sheet.
century ago. There are still problems to Isomorphous replacement of trivalent
be solved, particularly those relating to aluminum for some quadrivalent silicon
the effect of clay minerals. in the silica sheets, or of magnesium or
Miss MARGARET D. FOSTERS (by let- some other bivalent or univalent cation
ter).—Some factors of significance in for trivalent aluminum in the alumina
determining the permeability of sedi- sheet, gives rise to net residual charges
ments Jiave been discussed by Messrs. on the structure, which are neutralized
Smith and Stallman of the U. S. Geolog- by cations held electrostatically. These
ical Survey. Among these was the effect constitute the exchangeable cations.
of ion exchange in samples containing Thus, whereas kaolinite has a balanced
Copyright by ASTM Int'lU.(all
structure and a low cation-exchange
* I. S. MacQueen, Engineer, S. Geological capacity, montmorillonite has an un-
Survey Downloaded/printed
(oral communication,
by 1950).
3
U. SS. R
Geological
KUMARSurvey, Washington,
KONATHALA D. C.
(Indian balanced
Inst of structure,
Tech Chennai) and,
pursuant becauseAgreement.
to License of the No furth
cations held electrostatically to effect a erable divergence in the values given by

balance, a high cation-exchange capacity different investigators for the size of the
—from 70 to 110 milliequivalents per hydrated cations and for the number of
100 g. molecules of water associated with each
Whether or not a montmorillonite ion. Bernal and Fowler5 presented data
swells greatly in water depends to a that created doubt as to hydration of
large extent on the character of the ex- some ions. Hendricks, Nelson, and Alex-
changeable cations. If the exchangeable ander6 concluded that (a) the sodium
cations are mostly calcium and magne- ion, assumed to be most highly hydrated,
sium ions, the montmorillonite swells is not hydrated at all, and (b) magnesium
only to about twice its dry volume; but and the alkaline-earth cations have 6
if the exchangeable cations are largely molecules of water of hydration.
sodium ions, the montmorillonite swells These and other differences of opinion
in water to 20 or more times its dry as to the degree of cation hydration, and,
volume. in particular, as to whether the sodium
The hydrous mica group of clay min- ion is hydrated at all led Grim7 and
erals—by some called illites—has, like others8 • 9 to conclude that the concept of
the montmorillonites, the fundamental cation hydration as responsible for the
structural pattern of micas. The hydrous thickness and other characteristics of the
micas have an even higher degree of water film was not justified and that some
isomorphous replacement in the struc- other cause for the development of the
tural sheets than the montmorillonites water film must be sought.
and consequently higher charges. How- As data from investigations of the
ever, these higher charges in the hydrous relation of exchangeable bases to plastic
micas are to a large extent neutralized properties indicate that the character of
by potassium, which is fixed and not the adsorbed cation does affect the thick-
exchangeable. The cation-exchange ness of the water film, Grim10 concluded
capacity of the hydrous micas is usually that the adsorbed cations primarily
greater than that of the kaolinites but control the thickness of the water film
less than that of the montmorillonites. by means of the force with which they
They do not tend to swell greatly in
water. 5
J. D. Bernal and R. H. Fowler, "A Theory
of Water in Ionic Solution with Particular
THE CAUSE OF SWELLING Reference to Hydrogen and Hydroxyl Ions,"
Journal of Chemical Physics, Vol. 1, p. 515
A number of theories have been pro- (1933). 6
S. B. Hendricks, R. A. Nelson, and L. T.
posed to account for the swelling charac- Alexander, "Hydration Mechanism of the Clay
teristics of clays. According to the theory Mineral Montmorillonite Saturated with Vari-
generally accepted for many years, swell- ous Cations," Journal, Am. Ceramic Soc., Vol.
62, p. 1457 (1940).
ing is a function of the hydration of the 7
R. E. Grim, "Properties of Clay, Recent
exchangeable cations, and the great London, Marine Sediments," Thos. Murby and Co.,
p. 466 (1935).
swelling of sodium montmorillonite as 8
R. Houwink, "On the Structure of the
compared with calcium montmorillonite Hydration Hull of Inorganic Soil Colloids,"
Kottoid-Zeitschrift, Vol. 93, p. 110 (1937).
is due to the greater hydration of the 9
S. Spiel, "Effect of Adsorbed Electrolytes
4
sodium ion. However, there is consid- on the Properties of Monodisperse Clay-Water
Systems," Journal, Am. Ceramic Soc., Vol. 23,
4
G. Wiegner, "Some Physico-Chemical p. 33 (1940).
10
Properties of Clays. I. Base Exchange or Ionic R. E. Grim, "Modern Concepts of Clay
Interchange II. Hydrogen
Copyright by ASTMClays,"
Int'l (all rights Soc.
Journal, Minerals,"
reserved); Sun Jan Journal of Geology,
11 13:11:58 Vol. 50, p. 225
EST 2015
Chemical Industry, Vol. 60,
Downloaded/printed by pp. 65, 103 (1931). (1942).
DISCUSSION ON MEASUREMENT OF PERMEABILITIES 119
hold the sheets together. "According to which are made up of a definite ionizing
this concept the great swelling of sodium- compound that is in true equilibrium
montmorillomte is due, not to the hydra- and whose charge is due to ionization.
tion of Na+ serving as a wedge to force When such a clay is dispersed in water,
the layers apart, but to the absence of a each unit can come to equilibrium with
strong bond holding the layers together the outer solution; the exchangeable or
because of the univalent character of electrostatically held cations tend to dis-
sodium and because its small ionic size sociate and the clay unit can be pic-
enables it to fit well into the hole of the tured as a complex anion. The sus-
silica sheet. Calcium-montmorillonite pended clay unit may be likened to a
does not expand greatly because of the dissociated electrolyte with one of its
slight hydration of Ca++ but because the ions in the colloidal range of dimensions.
bivalent character of the ion tends to Dissociation of the adsorbed cations
hold the sheets together so that a thick leaves some of the structural units
series of water sheets cannot form be- negatively charged. Thus charged, the
tween them." units repel each other, and the clay
The high swelling of (sodium) ben- appears to swell. The more complete
tonites in comparison with other clays the dissociation and the greater the num-
strongly suggested to Baver and Winter- ber of units carrying a charge, the greater
korn11 that the former attract large is the swelling; the less complete the
amounts of water as a result of forces dissociation, the fewer are the units
associated with the inner layer of the carrying a charge and the less the swell-
colloidal surface. In colloid chemical ing.
terms the nature of the inner layer and Marshall13 has presented data showing
its effect on adsorption of ions and water the fraction active, that is, ionized, hi
molecules determine to a great extent the course of titration of an acid mont-
the character and amount of hydration morillonite with NaOH and Ca(OH)2.
in aqueous clay systems, with the kind For sodium the values for active frac-
of ion adsorbed on the surface playing tion, calculated from different parts of
an important role. the titration curve, range from 0.007
In addition to these, a number of other to 1.20; for calcium the range is from
theories have been proposed to explain less than 0.001 to 0.066. Near the point
the development of water films and the of neutrality there is almost a 20-fold
great swelling of sodium montmorillonite, difference in active fraction, 1.20 for
but none seem as satisfactory as the sodium as against 0.066 for calcium.
concept of cation dissociation. This great difference in the degree of
CLAYS AS COLLOIDAL ELECTROLYTES dissociation of sodium and calcium
montmorillonites can be correlated with
From considerations of structure, the great difference in their swelling
Marshall and Krinbill12 concluded that characteristics.
the clays of the montmorillonite group
may be classed as colloidal electrolytes; EFFECT OF IONIC SUBSTITUTION ON
that is, they are colloids the units of THE SWELLING OF SODIUM
11
MONTMORILLONITES
L/D. Baver and H. Winterkorn, "Sorption
of Liquids by Soil Colloids: II. Surface Behavior Most studies on the swelling of mont-
in the Hydration of Clays," Soil Science, Vol. morillonite have been on the effect of
40, p. 403 (1935).
18
« C. E. Marshall and C. A. Krinbill, "The C. E. Marshall, "Ionization of Calcium
Clays Copyright
as Colloidal Electrolytes,"
by ASTM Journal
Int'l (all rights from Sun
reserved); Soil Jan
Colloids and Its EST
11 13:11:58 Bearing
2015on Soil Re-
of Physical Chemistry, Vol. 46,
Downloaded/printed by p. 1077 (1942). lationships," Soil Science, Vol. 65, p. 57 (1948).
the exchangeable cation, although Kel- volumes; those with the greater sub-
ley14 observed that the swelling of stitution had the lower swelling volumes.
sodium-saturated montmorillonites from The depressing effect of substituted
different sources is not always propor- ferrous iron was greater than that of
tional to the cation-exchange capacity substituted ferric iron, and the depress-
and, furthermore, those having approxi- ing effect of substituted ferric iron was
mately equal cation-exchange capacity about the same as that of substituted
may swell quite differently. Grim7 magnesium. Thus the composition of
pointed out that the physical properties montmorillonite, at least with respect
of clays may be ascribed to two factors to the amount of iron and magnesium
—the character of the exchangeable present, is an important factor in swell-
cation and the composition of the clay ing.
mineral. By composition Grim referred If swelling is assumed to be related
to the makeup of the clay mineral, to dissociation of adsorbed or exchange-
particularly with regard to substitutions able cations, the decrease hi swelling
within the structure. found to accompany increase in substitu-
In a study by^the writer15 on the swell- tion of other cations for aluminum hi
ing of montmorillonite, a range from the alumina layer suggests that these
21 to 66 ml per g was found in the swell- substitutions affect the degree of disso-
ing volumes of twelve sodium-saturated ciation of the exchangeable cations;
montmorillonites. This range hi the in other words, the substitution hi some
swelling volumes of samples that had the way affects the strength of the clay
same exchangeable cation—sodium— unit as an anion.
strongly suggested that the swelling Because of differences in polarizing
volume of montmorillonite is greatly in- power, any ion substituting for another
fluenced by some other factor or factors must cause a change hi the polarization,
than the character of the exchangeable particularly of neighboring ions. How-
cation. The lack of correlation between ever, it is reasonable to assume that
swelling volume and total exchange not only is this effect felt by the nearest
capacity indicated that differences in neighbors but it is reflected throughout
the amount of sodium present due to the whole structure, thus altering the
differences in exchange capacity have energy relationships of the whole.
little influence on swelling volume. On The greater the substitution, the more
the other hand, there was good corre- the internal energy relations of the struc-
lation between swelling volume and the ture will be affected. Such an alteration
amount of substitution—by iron and hi the energy relationships of the struc-
magnesium for aluminum—in the alu- ture may well affect the bonding energy
mina layer, hi the direction of decrease and the degree of dissociation of the
in swelling volume with increase in exchangeable cations on dispersion of the
substitution. Samples with the lesser montmorillonite in water. The decrease
substitution had the higher swelling hi swelling volume found to accompany
increase in substitution in the alumina
14
W. P. Kelley, Review and Comments on layer seems to indicate that the effect
Mattson's Papers on "The Laws of Soil Col- of the substitution is hi the direction of
loidal Behavior," Soil Science, Vol. 56, p. 443
(1943). decrease in the degree of dissociation
16
Margaret D. Foster, "Geochemical Studies of the montmorillonite. Although a
of Clay Minerals (II): Relation Between Ionic colloidal electrolyte like montmorillo-
Substitution and Swelling
Copyright by ASTM in Montmorillonites,"
Int'l (all rights reserved); Sun Jan 11 13:11:58 EST 2015
American Mineralogist, Vol. 38,
Downloaded/printed by p. 994 (1953). nite cannot be strictly compared with a
DISCUSSION ON MEASUREMENT OF PERMEABILITIES 121
soluble electrolyte, the change in the APPLICATION OF THE SWELLING CHAR-

strength of acetic acid brought about by ACTERISTICS OF CLAYS TO PER-
the substitution of one or more of the H MEABILITY MEASUREMENTS
atoms by positive or negative atoms or
groups—such as Cl, OH, or CHs—and From the foregoing discussion it may
the increase in the strength of ammonium be seen that the significance of swelling
hydroxide caused by the substitution as a factor affecting permeability meas-
of one or more of the H atoms by CHs urements of clay-containing sediments
groups are examples of the way in which depends on the kind of clay present.
the degree of dissociation of a compound If the clay is a kaolin or a hydrous mica,
can be changed by substitutions. the effect of swelling would be of little
Although the hydrous micas have the importance, as these clays swell little,
same structural pattern as montmorillo- if any, in water. The presence of calcium
nites and have an even greater amount montmorillonite may affect the meas-
of isomorphous substitutions in their urement to some extent, as calcium
constituent layers and a greater charge montmorillonites swell to about twice
on the structural units, they cannot be their dry volume in water. But sodium
regarded as colloidal electrolytes. Most montmorillonite, with its characteris-
of the charge on the structural unit is tically great swelling in water, would
usually balanced by potassium, which be expected to have considerable effect
is bound in a different way from sodium on the permeability, the extent of the
and calcium. Potassium does not disso- effect depending not only on the amount
ciate when the hydrous mica is dispersed of sodium montmorillonite present but
in water; it cannot be exchanged as can also on its composition. Thus swelling
sodium and calcium. Potassium holds as a factor affecting permeability meas-
the units making up the particles firmly urements is of great importance only
together; exchangeable cations in- the when the clay 16
present is a sodium mont-
interior of the particle are inaccessible morillonite.
to water and cannot dissociate—the MESSRS. SMITH AND STALLMAN.—
number of cations available for disso- Miss Foster has given a very clear out-
ciation depend on particle size and not line of ion-exchange phenomena which
on the total number present as hi mont- play an important part in the trans-
morillonite. As relatively few of the mission of water in clay-bearing sedi-
electrostatically held cations present are ments. Its role in the swelling of clays,
available for dissociation, the degree of and the consequent effect on permeabil-
dissociation is correspondingly low re- ity, are of great importance.
gardless of the character of the ex- If dry air or organic liquids were used
changeable cation, and the swelling is of as permeants the observed permeabil-
a very low order. ities would be relatively high. The
13
Marshall found that the exchange- action of water is quite different be-
able cations of kaolinite are dissociated cause of ion-exchange phenomena; and
to a greater extent than those of mont- thus it is apparent that, if the per-
morillonite, except in montmorillonite meability to water is desired, water is
near the point of neutralization. How- generally the only permeant that can
ever, the number of exchangeable be used.
cations present is so small that this In this connection it should be recog-
greaterCopyright
dissociation does not lead to 16
by ASTM Int'l (all rights reserved); Publication
Sun Jan 11authorized by the
13:11:58 EST 2015Director,
any great amount of swelling.
Downloaded/printed by U. S. Geological Survey.
nized that all clays usually swell almost Air removal is best accomplished by
instantaneously once water is in contact subjecting the sample to a high vacuum
with them. The amount of swelling prior to the introduction of water. A
depends upon the type of clay. For ex- mercury head can be used to drive water
ample, Na-montmorillonite swells to through a clay sample in a variable-
about 15 or 20 times its initial volume, head permeameter. Its use eliminates
whereas Ca-montmorillonite swells to evaporation of the inflow water. With a
about twice its original volume. Kaolin- sediment containing limited amounts of
ite swelling is a matter of several per Ca-montmorillonite, a constant-head
cent and may be significant hi some permeameter probably would be more
tests. suitable because of the existence of a
A knowledge of the clay present in a more permeable pore space.
sediment is often useful when selecting In connection with the ion-exchange
permeameter equipment suitable for a mechanism outlined by Miss Foster, two
given sediment. If, for example, the sedi- additional items should be mentioned.
ment consists of a sand matrix in which First, the type of exchangeable ions
the clay embedded hi the pores contains affects the migration of clays in sedi-
large amounts of Na-montmorillonite ments and the subsequent packing.
and the permeant is pure water, a vari- Second, there is the role of organic ions
able-head permeameter properly pro- in exchange phenomena. Many of the
tected for evaporation is better suited organic ions show a behavior similar
to a measurement of permeability than to that of the inorganic ions. Hendricks17
is a constant head device. This is be- has studied the exchange phenomena in
cause the excessive swelling of the Na- montmorillonite when organic ions are
montmorillonite will almost completely present.
seal off the pores, resulting in very low 17
Sterling B. Hendricks, "Base Exchange of
permeability; and the consequent low the Clay Mineral Montmorillonite for Organic
flows cannot be measured accurately Cations and its Dependence Upon Adsorption
Due to Van der Waals Forces," Journal of
with the constant head device described. Physical Chemistry, Vol. 45, pp. 66-81 (1941).
DETERMINATION OF PERMEABILITY OF GRANULAR SOIL BY AIR
SUBJECTED TO A DECREASING PRESSURE DIFFERENTIAL
BY ARTHUR S. WEAVERI
SYNOPSIS
The object of this paper is to describe the design and operation of

an apparatus for determining the coefficient of permeability of granular
soils, utilizing air as the percolating fluid, under so-called falling-head
conditions. Also, the results of a few typical tests are presented, and
comparisons made with the results obtained using water as the test
fluid.
The majority of investigators agree necessary only to dry the sample and to
that, in determining the coefficient of dry the air to be passed through the
permeability by the commonly used sample of soil. Furthermore, the tend-
water-test procedures, it is necessary to ency to settle is less, since the pressure
dry the sample, evacuate it, and then gradient increases toward the low-pres-
saturate it with distilled, deaerated sure end of the sample. The seepage
water to prevent clogging of the void forces tending to produce settlement are
spaces with air or solid contaminants. therefore greatest at the low-pressure
If it is desired to determine the perme- end, where a given particle may move
ability at several different values of void the shortest distance, and least at the
ratio, it is therefore necessary either to high-pressure end where .the possible
have a large amount of the material on displacement is greatest.
hand or to dry out the sample between If a constant pressure differential is
tests. When attempting to determine the applied to the sample, the coefficient of
coefficient of permeability by the use of permeability may easily be calculated
water, it is found that if the material is by the use of Darcy's law, upon taking
in a relatively loose state, the seepage suitable measurements of the absolute
forces produced by the flow of the water pressures at the ends of the sample, of
may produce settlement of the sample. the temperature of the air flowing, and of
This action decreases the void ratio and, the volume passed during a given time.
consequently, the coefficient of perme- However, the problems involved in main-
ability. Since the computation of the taining constant pressures, and especially
void ratio is based on constant pre- in measuring the volume of air with
determined sample dimensions, the sufficient accuracy, are such as to nullify
correction to be applied is difficult to any advantages this procedure has over
determine accurately. the usual water test.
If air is used as the test fluid it is
1
TESTING APPARATUS
Assistant
CopyrightProfessor
by ASTM of
Int'lMechanical Engineer-
(all rights reserved); Sun Jan 11 13:11:58 EST 2015
ing, University of Maine,
Downloaded/printed by Orono, Me. In the author's apparatus (Fig. 1), the
air used is supplied by a compressor and volume of the tank was determined by
passes through a dryer to a receiver tank. filling it with deaerated water and meas-
When the pressure in this tank has uring the weight and temperature
reached a predetermined value, the thereof.
valve hi the compressor line is closed, Two manometers are provided to
and the air is allowed to flow through a measure the pressure in the reservoir.
suitably prepared sample to the sur- The tubes are of j-in. outside diameter
rounding atmosphere. by i^-hi. inside diameter transparent
The size and capacity of the compres- plastic, approximately 8 ft tall, and are
sor may be varied within wide limits, connected to the metal piping system
and, in fact, in certain installations it with flare type fittings. One manometer
might prove desirable to use a cylinder contains mercury and will measure pres-
Air Line to Triaxial Test

FIG. 1.—Schematic Diagram of the Air Permeameter.
of commercial compressed air as the sures up to approximately 50 psig; the

source of the air for testing. other contains water, and is used to
Silica gel is used as the drying agent, obtain more accurate determinations of
as it may be re-used upon heating it pressures hi the range from 0 to 4 psig.
periodically to drive off the adsorbed A Bourdon-tube pressure gage is also
moisture. A 41-cu in. refrigeration type installed to indicate the reservoir pres-
unit is used in the present apparatus and sure before opening the valve to either
is installed by means of flare type fittings, manometer, as a safeguard against blow-
which simplifies its removal for heating. ing the liquid out of the tube.
No means of cooling the air after the The details of the sample holder are
compression and drying processes are shown in Fig. 2. Originally, all parts were
necessary, as the heat thus liberated is to have been made of brass; however,
dissipated through the walls of the piping due to the then current unavailability of
before reaching the pressure tank. seamless brass tubing of the proper size,
The tank used hi all testing mentioned seamless steel tubing was substituted for
in this paper is a cylindrical steel drum, the holder barrel, with very little diffi-
16 hi. Copyright
in diameterby ASTM Int'l (all rights
arid 36 in. long. The reserved); Sun Jan 11 13:11:58 EST 2015
culty experienced due to rusting. The
S R KUMAR KONATHALA (Indian Inst of Tech Chennai) pursuant to License Agreemen
WEAVER ON PERMEABILITY OP GRANULAR SOIL 125
sample is held in place by circular screens by thermistors placed immediately be-
of 200-mesh brass cloth, backed by fore and after the sample. However, this
circular brass plates perforated with proved to be an unnecessary refinement,
small holes. One screen and plate combi- since a change of i C in the temperature
nation is made to fit snugly inside the of the air produces a corresponding
barrel against the locating pins while the change of only 0.3 per cent hi the co-
efficient of viscosity. The temperature
may be satisfactorily measured with a
mercury-in-glass thermometer clamped
directly below the sample holder outlet.
The atmospheric pressure is measured
by means of a mercury column barom-
eter.
The pressure in the tank, which is
determined at the beginning and end of
a suitable tune interval, the atmospheric
pressure, the temperature of the air
flowing, and the various dimensional
constants of the equipment provide suffi-
cient information for calculating the
coefficient of permeability.
THEORY
The following derivation shows the
development of the equation used hi
such calculations. The soil will be con-
sidered to be isotropic and homogeneous,
and the velocities and flow rates involved
will be assumed to be so small that iso-
thermal flow conditions will exist. Then,
if the fluid flowing is a nearly perfect gas,
Muskat2 indicates that the flow re-
lationship is
I
in which y represents the density of the
FIG. 2.—Detail of Sample Holder. fluid, n the porosity, p. the viscosity, and
jo the density at atmospheric pressure.
other fits into the annular space ma- In this expression, kp is the so-called
chined in the bottom cap. Leakage of air physical permeability, which is related
past the screw threads joining the upper to the coefficient of permeability for any
cap to the barrel is prevented by a soft specific fluid by the relationship kp =
rubber ring, placed as shown. The unit
is connected to the main piping by means The kp remains constant for a soil of
of a quick coupler. 2
M. Muskat, "The Flow of Homogeneous
Originally the temperature of the air Fluids
Copyright by ASTM Int'l (all rights reserved); Through
Sun Jan 11 Porous Media,"
13:11:58 McGraw-Hill
EST 2015
flowingDownloaded/printed
through the sample
by was measured Book Co., Inc., New York, N. Y. (1937).
given porosity and temperature and is rate of flow measured at the mean sample
independent of the properties of the fluid pressure may be written
flowing. In the proposed testing appara-
tus, the flow is one-dimensional, so that
The time to discharge a volume of air

AF at a constant rate Q equals A V/Q. Or,
Although this equation accurately repre-
sents the flow process, it is believed in-
solvable, at least in any usable form, and
further simplification is necessary.
If the pressure is to be measured only
at the extremities of the sample, it
becomes unnecessary to consider the Now, if the duration of each step is
manner in which the density or pressure assumed to be infinitely small, Px —
varies along the length of the sample. PX' ~ dP, and
In the following discussion, the actual
pressure-time relationship is replaced by
a series of constant-head "steps." Ana-
lyzing the air flow during one so-called By integration, the total time to reduce
"step" shows that the tank pressure the tank pressure from an initial value
actually drops from Px to PX', and a PI to some lower value P% may be found:
certain quantity of air is discharged. It
is then assumed that during this "step,"
the tank pressure remains constant and
equal to the average pressure, P =
Since, during the actual testing pro-
while discharging the same cedure, the pressures are obtained by
quantity of air. The pressure then drops means of manometers, this equation may
instantaneously to the average pressure be written, after rearranging terms,
of the next "step" and so on. At the
beginning of the "step," PXF = w*RT,
and at the end, /V V = w*'RT, express
the conditions hi the tank. The weight of where Pi and P2' are gage pressures.
air lost, therefore, Aw = wx — wx', is Finally, since the results of this investi-
V gation will be of use primarily in soils
equal to ^(P* — Px')- Expressing Aw; engineering applications, the foregoing
Kl
hi terms of volume measured at the mean expression should be changed to read hi
terms of the coefficient of permeability
pressure hi the sample, with respect to water flow, or,
in which Pa represents the pressure of where /v and juw are based on standard
the atmosphere at the sample exit. conditions of temperature and pressure,
Writing Darcy's expression applied to and n& is based on the temperature in the
the constant
Copyrighthead "step,"
by ASTM Int'l the
(all volume sample.
rights reserved); V is11the13:11:58
Sun Jan volume of 2015
EST the pressure
WEAVER ON PERMEABILITY OF GRANULAR SOIL 127
tank and piping leading to the sample; riod, then, the effective length of the
L and A are the length and cross-sec- sample increases from zero to the actual
tional area, respectively, of the sample. length and is reflected by a coefficient of
Certain other factors must be con- permeability which decreases from in-
sidered in conjunction with the use of finity to the true permeability of the
this equation in permeability determina- material.
tions. One may not merely apply an
arbitrary initial pressure and measure TEST PROCEDURE
the time for it to decrease a certain The procedure used hi testing with the
amount, because it is necessary for air permeameter follows.
laminar flow to exist in the flow pas-
sages for the foregoing analysis to be Sample:
valid. It is suggested that the presence or The sample was oven-dried to constant
absence of laminar flow be determined weight and mixed until uniform.
experimentally, rather than by an at-
tempt at mathematical analysis. If the Procedure:
sample is compacted to a certain density 1. Preparation of the Sample in the
and several consecutive tests are made, Sample Holder.
the second having a lower initial pressure
than the final pressure of the first, and The arrangement of the components
so on, the coefficients of permeability as of the sample holder is shown in Fig. 2.
computed from these tests should be The material tested was placed hi the
approximately equal. If turbulence exists barrel dry and hi layers of approximately
at the higher pressures, a markedly lower equal thickness. The layers were com-
value of the coefficient will be noted. pacted with a tamping rod, the amount
At the beginning of a test, when the of compaction depending upon the void
valve between the tank and the sample ratio desired. After each layer was com-
is opened, a transient flow condition of pacted, the surface was scarified to insure
somewhat uncertain duration is created, uniform blending of the material at the
during which no valid permeability de- boundary. Great care was taken to
terminations may be made. It was found prevent stratification or pocketing of
experimentally that, in all cases, the particles of uniform size.
transient condition had essentially dis-
appeared when the tank pressure had 2. Testing.
decreased to approximately 70 per cent (a) After compaction of the sample,
of its original value. Tests made during the weight of the sample was determined
this period of transition indicate an and the brass caps screwed on tightly.
apparent coefficient of permeability that (b) The assembled sample holder was
is much greater than the true value. The attached to the test panel by means of
equations given above are based on the the quick coupler, / (Fig. 1).
presence of falling-head flow throughout (c) Valves A, C, F, and G were opened;
the sample, whereas at the instant of valves B, E, H, and 7 were closed.
opening the valve, flow exists only at the (d) Either valve H or I was opened,,
face of the sample nearest the source of corresponding to the manometer to be
air. As the time interval increases, flow used.
is established further along the sample, (e) Compressed air was admitted to
until Copyright
falling-head conditions prevail
by ASTM Int'l (all rights reserved); pressure
the tank
Sun Jan 11 by connecting
13:11:58 EST 2015the rub-
throughout. During this
Downloaded/printed by transition pe- ber hose to the spring-loaded valve, D.
(y) When the manometer indicated more nearly constant, and to facilitate
the desired pressure, the hose was re- assembly of the unit. The temperature
moved, closing D. of the water was determined by thermis-
(g) The reservoir pressure was allowed tors located at both ends of the soil
to drop to approximately 75 per cent of sample, as well as by placing a thermom-
its original value. eter in the water supply bottle. Since the
(h) After thus eliminating the tran- coefficient of viscosity of water varies by
sient effect, the time for the liquid hi the approximately 2.5 per cent per deg Cent
manometer to fall a measured distance change in temperature, it is necessary to
FIG. 3- -Grain Size Distribution Curves for Samples Tested in the Air Permeameter.
was determined by means of a stop effect a more accurate determination of

clock. the water temperature than of the air, in
(t) The air temperature and the order that the results of the two types
atmospheric pressure were measured as of test may be compared. It was found
described previously. that the water supply temperature did
The procedure followed in determining not accurately represent the temperature
the coefficient of permeability of the of the water actually percolating through
same materials using water is found in the sample.
the ASTM publication, "Procedures for
Testing Soils,"3 with one modification to TEST RESULTS
the equipment: plastic spacer rings were Several samples of granular soil of
substituted for the Ottawa sand, so that varying description were tested in ac-
the length of the sample could be kept cordance with the foregoing procedure
3
Procedures forASTM
Testing Soils, with most satisfactory results. The grain-
Copyright by Int'l (all rightsAm. Soc. Sun Jan 11
reserved); 13:11:58 EST 2015
Testing Mats., July, 1950, p. 179. size distribution relationships of four
WEAVER ON PERMEABILITY OF GRANULAR SOIL 129
typical samples are illustrated in Fig. 3. tests on both samples 1 and 4. The
Sample 1 is an artificially prepared ma- discrepancy, evidenced in the water and
terial consisting of particles passing a air values for sample 2 is due to mal-
No. 100 sieve and retained on a No. 200 functioning of one of the thermistors used
sieve of the U.S. Standard sieve series. to measure the water temperatures. The
Samples 2, 3, and 4 are natural soils ob- water test could be performed on sample
tained from the Soils Laboratory of the 3 only at the lowest void ratios obtain-
Maine State Highway Commission. Void able, due to the marked instability at
ratio versus coefficient of permeability lesser densities. The high vacuum used
relationships obtained by testing these to deaerate the samples, combined with
FIG. 4.—Void Ratio versus Coefficient of Permeability.
four soils in the apparatus described in the large capillary rise naturally occur-
this paper are illustrated in Fig. 4. It is ring in a silt, drew the saturating water
noted that in each case the experimental into the pores of this sample faster than
relationship conforms very closely to the it was supplied. Because of the extremely
high velocity of the water, "boiling" and
expression, suggested by complete structural rearrangement re-
4
Taylor. Also shown are values of the sulted, with ensuing segregation of coarse
permeability coefficient obtained by the and fine particles and subsidence. Two
water test specified previously. It is seen consolidation tests were performed hi an
that there is very close correlation be- attempt to determine the permeability
tween the results of the two types of by other means, but due to the extreme
4
sensitivity and rapid consolidation rate
D. Copyright
W. Taylor,by ASTM
"Fundamentals of Soil
Int'l (all rights of
reserved);the
Sunmaterial, no valid
Jan 11 13:11:58 results were
EST 2015
Mechanics," John Wiley andbySons, Inc., New
Downloaded/printed
York, N. Y. (1948). obtained.
CONCLUSIONS ratio may be utilized in the testing of

each material. It must be emphasized
The relationship between the void that the results of air permeameter tests
ratio and the coefficient of permeability represent actual subsurface conditions
is found to correspond very closely to the no more nor less accurately than do
empirical relationships proposed by water-test results, and should not be used
other investigators as representing the without careful evaluation and compari-
average behavior of granular materials. son of laboratory and field conditions.
This fact is taken as experimental veri- The air permeameter is simpler, less ex-
fication of the theory that the flow of air pensive, and less time-consuming to
or other gases through a porous material operate and eliminates the nuisance
is subject to the same laws that govern attendant upon the use of water. It
the flow of liquids. The use of a gaseous would, furthermore, be more suited to
fluid rather than a liquid fluid for testing installation in a field or mobile testing
purposes requires a slightly more in- laboratory.
volved application of the basic theory, Acknowledgment:
which is compensated for by simplifica-
tion of certain aspects of the testing This investigation was performed as a
portion of the requirements for a Master
procedure and, in certain instances, in-
of Science degree at the University of
creased accuracy. Maine, under the guidance of Dr.
If the procedure described above is Hamilton Gray, Soils Engineer, Maine
adhered to, the values of the coefficient State Highway Commission. Acknowl-
of permeability obtained by means of the edgment is gratefully made to the
air test will exhibit less deviation from Technology Experiment Station of the
the mean curve than will those resulting University of Maine for supporting the
from the water test. A greater variety of cost of this investigation, and to those
materials may be tested in the air members of the faculty who contributed
permeameter, and a greater range of void their suggestions and criticisms.
SELECTED REFERENCES ON PERMEABILITY
COMPILED BY ARNOLD I. JOHNSON1
FLOW THEORY
(1) L. D. Baver, "Soil Physics," John Wiley & (12) J. L. M. Poiseuille, "Recherches Experi-
Sons, Inc., New York, N. Y., pp. 221-251 mentales sur le Mouvement des Liquides
(1940). dans les Tubes de Tres Petit Diametre
(2) H. Darcy, "Les Fontaines Publiques de la (Experimental Investigations upon the
Ville de Dijon (The Water Supply of Flow of Liquids in Tubes of Very Small
Dijon)," Dalmont, Paris (1856). Diameter)," Memoires presentes par divers
(3) G. H. Fancher and J. A. Lewis, "A Note on Savans, 1'Academic (Royale) des Sciences
the Flow of Fluids Through Porous Media," de 1'Institut de France, Vol. 9, pp. 433-543
Science, Vol. 75, p. 468 (1932). (1846). Translated by W. H. Herschel,
(4) V. C. Fishel, "Further Tests of Perme- Rheological Memoirs, Vol. 1, No. 1, 101 pp.,
ability With Low Hydraulic Gradients," Easton, Pa., June, 1940.
Transactions, Am. Geophysical Union, Part (13) Osborne Reynolds, "An Experimental
2, pp. 499-503 (1935). Investigation of the Circumstances Which
(5) C. V. Givan, "Flow of Water Through Determine Whether the Motion of Water
Granular Material," Transactions, Am. Shall be Direct or Sinous and of the Law
Geophysical Union, Part 2, pp. 572-579 of Resistance in Parallel Channels," Trans-
(1934). actions, Royal Soc. (London), Vol. 174, pp.
(6) G. Hagen, "Ueber die Bewegung des 935-982 (1883).
Wassers in engen cylindrischen Rohren (14) Osborne Reynolds, "On the Dynamical
(Movement of Water in a Narrow Cylindri- Theory of an Incompressible Viscous Fluid
cal Tube)," Annalen der Physik und and the Determination of the Criterion,"
Chemie, Vol. 46, pp. 423-442, Leipzig Transactions, Royal Soc. (London), Vol.
(1839). A186, pp. 123-164(1896).
(7) L. P. Hatch, "Flow Through Granular (15) Osborne Reynolds, "On the Equations of
Media," Journal of Applied Mechanics, Motion and the Boundary Conditions for
Vol. 5A, pp. 85-86 (1938). Viscous Fluids," Papers on Mechanical and
(8) L. P. Hatch, "Flow Through Granular Physical Subjects, Vol. 2, Cambridge Uni-
Media," Journal of Applied Mechanics, versity Press, pp. 132-137 (1901).
Vol. 7A, p. 109 (1940). (16) L. A. Richards, "Concerning Permeability
(9) M. K. Hubbert, "The Theory of Ground Units for Soils," Proceedings, Soil Science
Water Motion," Journal of Geology, Vol. Soc. America, Vol. 5, pp. 49-53 (1940).
48, pp. 785-944(1940). (17) C. S. Slater, "The Flow of Water Through
(10) C. E. Jacob, V. C. Fishel, and M. K. Hub- Soil," Agricultural Engineering, Vol. 29,
bert, "Report of the Committee on Ground pp. 119-124(1948).
Water, 1944-45," Transactions, Am. Geo- (18) C. S. Slichter, "Theoretical Investigation of
physical Union, Vol. 27, Part 2, pp. 245-273 the Motion of Ground Waters," V. S.
(1945). Geological Survey Annual Report, Part 2,
(11) Morris Muskat, "The Flow of Homogene- pp. 295-384 (1899).
ous Fluids Through Porous Media," Mc- (19) W. O. Smith, "Capillary Flow Through an
Graw-Hill Book Co., Inc., New York, N. Y. Ideal Uniform Soil," Physics, Vol. 3, pp.
(1937). 139-146(1932).
(20) Karl Terzaghi, "Theoretical Sofl Mechan-
1
Chief, Lincolnby
Copyright Hydrologic Laboratory,
ASTM Int'l (all rights U. S.
reserved); Sunics,"
Jan John Wiley &
11 13:11:58 Sons,
EST Inc., New York,
2015
Geological Survey, Lincoln, by
Downloaded/printed Nebr. N.Y., pp. 235-344(1943).
S R KUMAR KONATHALA (Indian Inst131
of Tech Chennai) pursuant to License Agreement. No further
METHODS
(1) V. S. Aronovici and W. W. Donnan, "Soil- (15) D. W. Kessler, "Permeability of Stone,"
Permeability as a Criterion for Drainage- Technical Paper No. 305, Nat. Bureau
Design," Transactions, Am. Geophysical Standards(1926).
Union, Vol. 27, pp. 95-101 (1946). (16) P. W. Ketchum, A. E. R. Westman, and
(2) D. F. Barnes, "Flow and Percolation R. K. Hursch, "Measurement of Perme-
Studies Abroad; Current Experiments at ability of Ceramic Bodies," Circular No. 14,
the Hydraulic Institute of the Technical Univ. of Illinois Engineering Experiment
University of Berlin," Civil Engineering, Station (1926).
Vol. 3, No. 7, pp. 389-391 (1933). (17) T. T. Knappan and R. R. Philippe, "Prac-
(3) T. W. Bendixen, M. F. Hershberger, and tical Soil Mechanics Muskingum," En-
C. S. Slater, "A Basis for Classifying Soil gineering News-Record, Vol. 116, No. 15,
Permeabilities," Journal of Agricultural pp. 532-535 (1936).
Research, Vol. 77, pp. 157-168 (1948). (18) A. W. Marsh, "The Collection and Study
(4) G. J. Bouyoucos, "A New Method of of Natural Soil Cores for Determining Ir-
Measuring the Comparative Rate of Per- rigation Properties," Proceedings, Soil
colation of Water in Different Soils," Jour- Science Soc. America, Vol. 13, pp. 515-518
nal, Am. Soc. Agronomy, Vol. 22, pp. 438- (1948).
445 (1930). (19) F. T. Mavis and T. P. Tsui, "Percolation
(5) A. Casagrande, "New Facts hi Soil Mechan- and Capillary Movements of Water
ics from the Research Laboratory," En- Through Sand Prisms," Iowa University
gineering News-Record, Vol. 115, pp. 320- Engineering Bulletin No. 18 (1939).
323 (1935). (20) A. F. Melcher, "Apparatus for Determin-
(6) G. M. Fair, "Flow of Water Through ing the Absorption and the Permeability of
Sand," Civil Engineering, Vol. 4, p. 137 Oil and Gas Sands for Certain Liquids and
(1934). Gases Under Pressure," Bulletin No. 9, No.
(7) C. R. Fettke and R. D. Mayne, "Perme- 3, Am. Assn. Petroleum Geologists, pp.
ability Studies of Bradford Sand," National 442-450(1925).
Petroleum News, Vol. 22, p. 61 (1930). (21) T. V. Moore, "The Determination of Per-
(8) M. Fireman, "Permeability Measurements meability from Field Data," Proceedings,
on Disturbed Soil Samples," Soil Science, Am. Petroleum Inst., Vol. 14, No. 4, pp.
Vol. 58, pp. 337-353 (1944). 4-13(1933).
(9) Glennon 'Gilboy, "Soil Mechanics Re- (22) P. G. Nutting, "Movements of Fluids in
search," Proceedings, Am. Soc. Civil Porous Solids," Journal, Franklin Inst.,
Engrs., Vol. 57, No. 8, pp. 1171-1177 Vol. 203, p. 313 (1927).
(1931). (23) P. G. Nutting, "Physical Analysis of Oil
(10) Glennon Gilboy, "Improved Soil Testing Sands," Bulletin, Am. Assn. Petroleum
Methods," Engineering News-Record, Vol. Geologists, Vol. 14, No. 10, pp. 1337-1349
116, No. 21, pp. 732-734 (1936). (1930).
(11) Wi E. Goode and J. E. Christiansen, "Ob- (24) F. B. Plummer, Sidon Harris-, and John
taining Soil Cores for Permeability Tests," Pedigo, "A New Multiple Permeability
Agricultural Engineering, Vol. 26, pp. 153- Apparatus," Technical Publication 578, Am.
155 (1945). Inst. Mining and Metallurgical Engrs.
(12) H. H. Hatch, "Percolation Tests for Hy- (1934).
draulic Fill Dams," Proceedings, Am. Soc. (25) R. R. Procter, "Field and Laboratory
Civil Engrs., Vol. 58, No. 8, pp. 1301-1342 Verification of Soil Suitability," Engineer-
(1932). ing News-Record, Vol. Ill, No. 12, pp.
(13) W. L. Homer, "A Rapid Method for De- 348-351 (1933).
termining Permeabilities of Consolidated (26) J. C. Russel, "The Movement of Water in
Rock," Petroleum Engineering, May, 1934, Soil Columns and the Theory of the Con-
pp. 25-27. trol Section," Proceedings, Soil Science
(14) R. Hulburt and Douglas Feben, "Hydrau- Soc. America, Vol. 11, pp. 119-123 (1946).
lics of Rapid Filter Sand," Journal, Am. (27) P. C. Rutledge, "Recent Developments in
Water Works Assn., Vol. 25, No. 1, pp. Soil Testing Apparatus," Journal, Boston
19-65 Soc.Sun
(1933). by ASTM Int'l (all rights reserved);
Copyright Civil
JanEngrs., October, EST
11 13:11:58 1935.2015
JOHNSON—SELECTED REFERENCES 133
(28) C. S. Slater and N. G. Byers, "A Labora- (32) C. F. Tolman, "Ground Water," McGraw-
tory Study of the Field Percolation Rates Hill Book Co., Inc., New York, N. Y.,
of Soils," Technical Bulletin No. 232, U. S. pp. 200-213 (1937).
Dept. Agriculture, pp. 1-23 (1931). (33) R. N. Traxler and L. A. H. Baum, "Perme-
(29) R. M. Smith and D. R. Browning, "In- ability of Compacted Powders," Physics,
fluence of Evacuation upon Laboratory Vol. 7, pp. 9-14(1936).
Percolation Rates and Wetting of Undis- (34) L. K. Wenzel, "Methods for Determining
turbed Soil Samples," Soil Science, Vol. Permeability of Water-Bearing Materials,
62, pp. 243-253 (1946). with Special Reference to Discharging-Well
(30) R. M. Smith and D. R. Browning, "Some Methods." U. S. Geological Survey Water-
Suggested Laboratory Standards of Sub- Supply Paper 887, pp. 1-192 (1942).
soil Permeability," Proceedings, Soil Science
Soc. America, Vol. 11, pp. 21-26 (1947). (35) R. D. Wyckoff, H. G. Botset, Morris
(31) N.. D. Stearns, "Laboratory Tests on Muskat, and D. W. Reed, "Measurement
Physical Properties of Water-Bearing Ma- of Permeability of Porous Media," Bulle-
terials," U. S. Geological Survey Water- tin, Am. Assn. Petroleum Geologists, Vol.
Supply Paper 596-F (1928). 18, No. 2, pp. 161-190 (1934).
FLUIDS
(1) B. A. Bakhmeteff and N. V. Feodoroff, Industrial and Engineering Chemistry, Vol.
"Flow Through Granular Media," Journal 25, No. 10, pp. 1139-1147 (1933).
of Applied Mechanics, Vol. 4A, pp. 97-104 (11) G. R. Free and V. J. Palmer, "Interrela-
(1937). tionship of Infiltration, Air Movement, and
(2) L. D. Baver, "Soil Characteristics Influenc- Pore Size in Graded Silica Sand," Proceed-
ing the Movement and Balance of Soil ings, Soil Science Soc. America, Vol. 5, pp.
Moisture," Proceedings, Soil Science Soc. 390-398 (1940).
America, Vol. 1, pp. 431-437 (1936). (12) C. C. Furnas, "Flow of Gases Through
(3) G. E. Bertram, "An Experimental Investi- Beds of Broken Solids," Bulletin No. 307,
gation of Protective Filters," Soil Mechan- U.S. Bur. Mines (1929).
ics Series No. 7, Harvard Univ. Graduate (13) W. H. Greene and G. A. Ampt, "Studies on
School of Engineering (1950). Soil Physics, Flow of Air and Water
(4) G. B. Bodman, "The Variability of the Through Soils," Journal of Agricultural
Permeability 'Constant' at Low Hydraulic Science, Vol. 4, pp. 1-24 (1911).
Gradients During Saturated Water Flow in (14) W. H. Greene and G. A. Ampt, "The Per-
Soils," Proceedings, Soil Science Soc. meability of an Ideal Soil to Air and Wa-
America, Vol. 2, pp. 45-53 (1938). ter," Journal of Agricultural Science, Vol.
(5) H. G. Botset, "The Measurement of Per- 5, pp. 1-25(1912).
meabilities of Porous Alundum Discs for (15) R. E. Horton, "An Approach Toward a
Water and Oils," Review of Scientific In- Physical Interpretation of Infiltration
struments, Vol. 2, No. 2, pp. 84-95 (1931). Capacity," Proceedings, Soil Science Soc.
(6) S. P. Burke and W. B. Plummer, "Gas America, Vol. 5, pp. 39SM17 (1940).
Flow Through Packed Columns," Indus- (16) N. Johnston and C. M. Beeson, "Water
trial and Engineering Chemistry, Vol. 20, Permeability of Reservoir Sands," Trans-
pp. 1196-1200 (1928). actions, Am. Inst. Mining and Metallurgical
(7) J. Chalmers, D. B. Taliaferro, and E. L. Engrs., Vol. 160, pp. 43-55 (1945).
Rawlins, "Flow of Air and Gas Through (17) M. C. Leverett and W. B. Lewis, "Steady
Porous Media," Oil Weekly, Vol. 64, No, Flow of Gas-Oil-Water Mixtures Through
12, pp. 19-30 (1932). Unconsolidated Sands," Petroleum Tech-
(8) J. E. Christiansen, "Effect of Entrapped nology, Vol. 3, No. 2, Tech. Pub. 1206
Air Upon the Permeability of Soils," Soil (1940).
Science, Vol. 58, pp. 355-365 (1944). (18) R. C. McCurdy, "A Study of the Petroleum
(9) J. E. Christiansen, M. Fireman, and L. E. Drainage Problem." Unpublished thesis
Allison, "Displacement of Soil-Air by CO2 Stanford University (1933).
for Permeability Tests," Soil Science, Vol. (19) Morris Muskat and H. G. Botset, "Flow of
61, pp. 355-360 (1946). Gases Through Porous Materials," Physics,
(10) G. H. Fancher and J. A. Lewis, "Flow of Vol. 1, No. 1, pp. 27^7 (1931).
Copyright
Simple Fluids by ASTMPorous
Through Int'l (all rights reserved);
Materials," Sun Muskat,
(20) Morris Jan 11 13:11:58 EST 2015
R. D. Wyckoff, H. G.
Botset, and M. W. Meres, "Flow of Gas- (24) R. R. Sullivan, "Further Study of Flow of
Liquid Mixtures Through Sands," Trans- Air Through Porous Media," Journal of
actions, Am. Inst. Mining and Metallurgi- Applied Physics, Vol. 12, p. 503 (1941).
cal Engrs., Petroleum Div., Vol. 123, pp. (25) R. R. Sullivan and K. L. Hertel, "The
69-96(1937). Flow of Air Through Porous Media,"
(21) A. F. Pillsbury and David Appleman, Journal of Applied Physics, Vol. 11, p. 761
"Factors in Permeability Changes of Soils (1940).
and Inert Granular Material," Soil Science, (26) S. A. White and E. C. Steinbremer, "De-
Vol. 59, pp. 115-123 (1945). termination of Air Permeability of Soil by
(22) W. L. Powers, "Soil-Water Movement as Means of Sphygmomanometer," Journal of
Affected by Confined Air," Journal of Forestry, Vol. 48, pp. 840-841 (1950).
Agricultural Research, Vol. 49, pp. 1125- (27) F. A. Wickersham, "Gas Permeability of
1134(1934). Refractory Brick," Iron Age, Vol. 119, pp.
(23) R. M. Smith and D. R. Browning, "Per- 1521-1522 (1927).
sistent Water-Unsaturation of Natural Soil (28) B. G. Zimmerman, "Determining En-
in Relation to Various Soil and Plant Fac- trapped Ah- in Capillary Soils," Engineer-
tors," Proceedings, Soil Science Soc. ing News-Record, Vol. 117, pp. 186-187
America, Vol. 7, pp. 114-119 (1943). (1936).
POROSITY
(1) B. A. Bakhmeteff and N. V. Feodoroff, (10) G. M. Fair and L. P. Hatch, "Fundamen-
"Flow Through Granular Media," Journal tal Factors Governing the Streamline Flow
of Applied Mechanics, Vol. 4A, pp. 97-104 of Water Through Sand," Journal, Am.
(1937). Water Works Assn., Vol. 25, No. 11, pp.
(2) B. A. Bakhmeteff and N. V. Feodoroff, 1551-1565(1933).
"Flow Through Granular Media," Pro- (11) J. L. Fowler and K. L. Hertel, "Flow of a
ceedings, Fifth International Congress of Gas Through Porous Media," Journal of
Applied Mechanics (1938). Applied Physics, Vol. 11, p. 496 (1940).
(3) C. F. Barb, "Porosity-Permeability Rela- (12) J. B. Franzini, "The Effect of Porosity on
tions in Appalachian Oil Sands, Mineral Permeability in the Case of Laminar Flow
Industries Experiment Station Bulletin No. Through Granular Media." Unpublished
9, The Pennsylvania State University, pp. thesis for Ph.D. degree, Stanford Uni-
47-59 (1930). versity (1949).
(4) L. D. Baver, "Soil Permeability in Relation (13) H. J. Fraser, "Experimental Study of the
to Noncapillary Porosity," Proceedings, Porosity and Permeability of Clastic Sedi-
Soil Science Soc. America, Vol. 3, pp. 52-56 ments," Journal of Geology, Vol. 43, No. 8,
(1938). pp. 910-1010 (1935).
(5) L. D. Baver, "Soil Porosity in Relation to (14) L. C. Graton and H. J. Fraser, "Systematic
Gaseous and Water Movement," Transac- Packing of Spheres With Particular Rela-
tions, Am. Geophysical Union, Part 2, pp. tion to Porosity and Permeability," Journal
414-419 (1940). of Geology, Vol. 43, No. 8, pp. 785-909
(6) F. C. Blake, "The Resistance of Packing (1935).
to Fluid Flow," Transactions, Am. Inst. (15) L. P. Hatch, "Flow of Fluids Through
Chemical Engrs., Vol. 14, pp. 415-421 Granular Material: Filtration, Expansion,
(1922). and Hindered Settling," Transactions, Am.
(7) P. C. Carman, "Fluid Flow Through Geophysical Union, Vol. 24, Part 2, pp.
Granular Beds," Transactions, Inst. Chemi- 536-547 (1943).
cal Engrs. (London), Vol. 15, pp. 150-166 (16) G. H. Hickox, "Flow Through Granular
(1937). Materials," Transactions, Am. Geophysical
(8) A. S. Gary, B. H. Walter, and H. T. Har- Union, Part 2, pp. 567-572 (1934).
stad, "Permeability of Mud Mountain (17) W. L. Howe and C. J. Hudson, "Studies in
Dam Core Material," Transactions, Am. Porosity and Permeability Characteristics
Soc. Civil Engrs.,, Vol. 108, pp. 719-737 of Porous Bodies," Journal, Am. Ceramic
(1943). Soc., Vol. 10, p. 443 (1927).
(9) T. H. Chilton and A. P. Colburn, "Pressure (18) P. F. Jones, "Porosity, Specific Perme-
Drop in Packed Tubes," Industrial and ability, and Geometry of Spacing," Oil
Copyright Chemistry,
Engineering by ASTM Vol.
Int'l (all rights
23, pp. reserved); Sunand
913-918 Jan Gas
11 13:11:58
Journal,EST
Vol.2015
41, No. 3, pp. 45-47
(1931).
Downloaded/printed by (1942).
JOHNSON—SELECTED REFERENCES 135
(19) E. G. W. Lindquist, "On the Flow of Water Mechanics: Determination of Permeability
Through Porous Soil," / Congres de Grands of Clay," Engineering News-Record, Vol.
Barrages, Vol. 5, pp. 81-102, Stockholm, 95, pp. 832-836 (1925).
Sweden (1933). (24) F. G. Tickell, O. E. Mechem, and R. C.
^20) W. R. Purcell, "Capillary Pressures, Their McCurdy, "Some Studies on the Porosity
Measurement Using Mercury and the and Permeability of Rocks," Transactions,
Calculation of Permeability Therefrom," Am. Inst. Mining and Metallurgical Engrs.,
Journal of Petroleum Technology, Vol. 1, Vol. 103, pp. 250-260 (1933).
pp. 39-48 (1949). (25) A. Westman and H. Hughill, "The Packing
(21) W. Schriever, ''Law of Flow for the Passage of Particles," Journal, Am. Ceramic Soc.,
of a Gas-free Liquid Through a Spherical- Vol. 18, pp. 767-769 (1930).
Grain Sand," Transactions, Am. Inst. Min- (26) R. Woodburn, "The Effect of Structure
ing and Metallurgical Engrs., Petroleum Changes in Houston Clay on Plant De-
Div., pp. 329-336(1930).' velopment and Water Relationships,"
(22) R. M. Smith, D. R. Browning, and G. Agricultural Engineering, Vol. 26, pp. 193-
G. Pohlman, "Laboratory Percolation 195 (1945).
Through Undisturbed Soil Samples in (27) F. C. Zeisberg, "The Resistance of Absorp-
Relation to Pore-Size Distribution," Soil tion Tower Packing to Gas Flow," Trans-
Science, Vol. 57, pp. 197-214 (1944). actions, Am. Inst. Chemical Engrs., Vol.
(23) Charles Terzaghi, "Principles of Soil 12, pp. 231-236 (1919).
OTHER FACTORS
(1) L. E. Allison, "Effect of Microorganisms on (9) J. E. Christiansen, "Some Permeability
Permeability of Soil under Prolonged Sub- Characteristics of Saline and Alkali Soils,"
mergence," Soil Science, Vol. 63, pp. 439- Agricultural Engineering, Vol. 28, pp. 147-
450 (1947). 150 (1947).
(2) B. A. Bakhmeteff and N. V. Feodoroff, (10) M. Fireman and G. B. Bodman, "The
"Flow Through Granular Media," Journal Effect of Saline Irrigation Water Upon the
of Applied Mechanics, Vol. 5A, pp. 86-90 Permeability and Base Status of Soils,"
(1938). Proceedings, Soil Science Soc. America, Vol.
(3) T. W. Bendixen and C. S. Slater, "Effect 4, pp. 71-77 (1939).
of the Time of Drainage on the Measure- (11) M. Fireman and O. C. Magistad, "Perme-
ment of Soil Pore Space, and Its Relation ability of Five Western Soils as Affected by
to Permeability," Proceedings, Soil Science the Percentage of Sodium of the Irrigation
Soc. America, Vol. 11, pp. 35-42 (1946). Water," Transactions, Am. Geophysical
(4) G. B. Bodman, "Factors Affecting Down- Union, Vol. 26, pp. 91-44 (1945).
ward Movement of Water in Soils," Bul- (12) M. Fireman, O. C. Magistad, and L. V.
letin, Am. Soil Survey Assn., Vol. 17, pp. Wilcox, "Effect of Sodium Nitrate and
35-38 (1935). Ammonium Fertilizers on the Permeability
(5) G. B. Bodman and M. Fireman, "Changes of Western Soils," Journal, Am. Soc.
in Soil Permeability and Exchangeable Agronomy, Vol. 37, pp. 888-901 (1945).
Cation Status During Flow of Different
Irrigation Waters," Transactions, Fourth (13) R. Gardner, "Some Soil Properties Related
to the Sodium Salt Problem in Irrigated
International Congress of Soil Scientists,
Soils," Technical Bulletin No. 902, U. S.
Vol. 1, pp. 397^00 (1950).
Dept. Agriculture (1945).
(6) G. B. Bodman and F. F. Harradine, "Mean
Effective Pore Size and Clay Migration (14) J. H. Griffith, "Physical Properties of
During Water Percolation in Soils," Pro- Earths," Bulletin No. 101, Iowa Engineer-
ceedings, Soil Science Soc. America, Vol. 3, ing Experiment Station (1931).
pp. 44-51 (1938). (15) A. E. Harris, "Effect of Replaceable
(7) W. L. Butcher, "Water Flow Increased Sodium on Soil Permeability," Soil Science,
With Rise of Temperature," Engineering Vol. 32, pp. 435^46 (1931).
News-Record, Vol. 76, Part 2, p. 326 (1916). (16) Allen Hazen, "Some Physical Properties of
(8) A. J. Carlson and M. C. Eastman, "Factors Sands and Gravels with Special Reference
Influencing Permeability Measurements," to Their Use in Filtration," 24th Annual
Petroleum Technology, Vol. 3, No. 2, Tech. Report, Massachusetts State Board of
Pub. 1196, 6 pp.
Copyright (1940). Int'l (all rights reserved);Health,
by ASTM Sun Janpp.11541-556
13:11:58(1892).
EST 2015
(17) D. S. Hubbell, "Effect of Soil Aggregates Petroleum Geologists, Vol. 16, No. 4, pp.
on Water Movement in Two Calcareous 373-384 (1932).
Soils," Journal, Am. Soc. Agronomy, Vol. (31) A. M. O'Neal, "Soil Characteristics Sig-
39, pp. 762-770 (1947). nificant in Evaluating Permeability," Soil
(18) D. S. Hubbell and T. M. Stubblefield, "The Science, Vol. 67, pp. 403-409 (1949).
Effects of Sqil Amendments on Soil Perme- (32) A. F. Pillsbury, "Effect of Particle Size and
ability and on Water Movement," Pro- Temperature on the Permeability of Sand
ceedings, Soil Science Soc. America, Vol. 13, to Water," Soil Science, Vol. 70, pp. 299-
pp. 519-522 (1949). 300(1950).
(19) M. R. Huberty and A. F. Pillsbury, "Solid (33) J. A. Putnam, "Unidirection, Transient
Liquid, Gaseous Phase Relationships of Flow of Compressible Fluids in Porous
Soils on Which Avocado Trees Have De- Media." Unpublished thesis, University of
clined," Proceedings, Soc. Horticultural California (1943).
Science, Vol. 42, pp. 39-45 (1943). (34) L. A. Richards, "Capillary Conduction of
(20) S. Iwanani, "On Resistance of Layer of Liquids Through Porous Mediums,"
Small Balls to the Flows of Viscous Fluids," Physics, Vol. 1, No. 5, pp. 318-333 (1931).
Transactions, Soc. Mechanical Engrs. of (35) H. B. Roe and J.VK. Park, "A Study of thi
Japan, Vol. 4, No. 16 (1938). Centrifuge Moisture Equivalent as an In-
(21) F. H. King, "Principles and Conditions of dex of the Hydraulic Permeability of
the Movements of Ground Water," U. S. Saturated Soils," Agricultural Engineering,
Geological Survey Annual Report, Vol. 19, Vol. 25, pp. 381-385 (1944).
No. 2, pp. 61-294 (1899). (36) H. E. Rose, "An Investigation into the
(22) C. H. Lee, "Sealing the Lagoon Lining at Laws of Flow of Fluids Through Beds of
Treasure Island with Salt," Transactions, Granular Materials," Proceedings, Inst.
Am. Soc. Civil Engrs., Vol. 106, pp. 577- Mechanical Engrs., Vol. 153, pp. 141-168
607 (1941). (1945).
(23) J. F. Lutz and R. W. Learner, "Pore-Size (37) Chanan Singh, H. R. Luthra, and V. I.
Distribution as Related to the Permeability Vaidhianathan, "On the Transmission
of Soils," Proceedings, Soil Science Soc. Constant of Water in Subsoil Sands," Re-
America, Vol. 4, pp. 28-31 (1939). search Publication, Punjab Irrigation Re-
(24) F. T. Mavis and E. F. Wilsey, "A Study search Inst., Vol. 5, No. 9, Lahore (1939).
of the Permeability of Sand," Iowa Uni- (38) F. G. Tickell, "Permeability of Unconsoli-
versity Engineering Bulletin No: 7 (1936). dated Rocks," Bulletin, Am. Assn. Pe-
(25) F. T. Mavis and E. F. Wilsey, "Filter Sand troleum Geologists, Vol. 19, No. 8, pp.
Permeability Studies," Engineering News- 1232-1238 (1935).
Record, Vol. 118, pp. 299-300 (1937). (39) F. G. Tickell and W. N. Hiatt, "Effect of
(26) O. E. Meinzer and V. C. Fishel, "Tests of Angularity of Grain on Porosity and
Permeability with Low Hydraulic Gradi- Permeability of Unconsolidated Sands,"
ents," Transactions, Geophysical Union, Bulletin, Am. Assn. Petroleum Geologists,
Part 2, pp. 405^09 (1934). Vol. 22, pp. 1272-1274 (1938).
(27) Felix Menichikovsky, "Effect of Nature of (40) Hubert J. Tracy, "The Effect of the Addi-
Exchangeable Bases on Soil Porosity and tion of Sodium Chloride upon the Flow of
Soil-Water Properties in Mineral Soil," Water Through a Standard Ottawa Sand,"
Soil Science, Vol. 62, pp. 169-181 (1946). Unpublished thesis, University of Arkansas
(28) R. E. Moore, "Water Conduction from (1952).
Shallow Water Tables," Hilgardia, Vol. 12, (41) H. A. Wadsworth and A. Smith, "Some
pp. 383-426 (1939). Observations upon the Effect of the Con-
(29) W. R. Nelson and L. D. Baver, "Movement tainer upon the Capillary Rise of Water
of Water Through Soils in Relation to the Through Soil Columns," Soil Science, Vol.
Nature of the Pores," Proceedings, Soil 22, pp. 199-211 (1926).
Science Soc. America, Vol. 5, pp. 69-76 (42) J. Wityn, "On the Permeability of Loam
(1940). Soils," Proceedings, Internat. Soc. Soil
(30) C. M. Nevin, "Permeability, Its Measure- Scientists, New Series, Vol. 2, pp. 209-243
ment and Value," Bulletin, Am. Assn. (1926).

Stp163-Eb 31066

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Stp163-Eb 31066

Hochgeladen von

Copyright:

Verfügbare Formate

Copyright by ASTM Int'l (all rights reserved); Sun Jan 11 13:11:58 EST 2015

Reg. U. S. Pat. Off.

ASTM Special Technical Publication No. 763

Copyright by ASTM Int'l (all rights reserved);

The papers and discussions in this Symposium on Permeability of Soils

In principle, determination of the permeability tests are compared, de-

The permeability of soils is a most im- CONTROLLING ENVIRONMENTAL AND

that control in the environment: whether and varied character of environmental

GRAVEL SAND SILT

the mechanics of flow of water through predominance of the coarse, medium or

FIG. 2.—Distinguishing Characteristics of Grain Size Curves.

FIG. 5.—Permeability - Relative Density Relationships.

FIG. 6.—Relations Between Permeability and Hazen's Effective Size, Z?ie.

MR. E. D'APPOLONIA.1—The reliability consideration: the first deals with the

characteristic limiting values in which ^ to f Ib per cu ft higher. For practical

were from 1 to 2 Ib per cu ft higher or information needed for making practical

The movement of liquids through WATER TRANSMISSION UNDER

ij = coefficient of viscosity in equation, not more and not less. This

Soil Grain Size,

Quartz 2-1 0.055 Fine quartz sand 0.03

HYGROSCOPICITY OF DIFFERENT ALTJMINO SILICATES (7)

H-Montmorillonite ... 32,7

temperature. The constant, Ci, contains h = height of capillary rise,

Soil Type Provenience Sand Minimal

Sand Greenlo, Holland 69.6 28.0 1.0 1.4 0.8 7.3

Since TE is numerically equal to TV fore hi the total heat capacity between

often decisive importance. This interac- information. It is realized that much of

MR. S. J. Rosen.1—Frankly, I was closure).—The paper just presented

Highway drainage designs generally DRAINAGE-LAG PERMEAMETER

FIG. 1.—Drainage-Lag Permeameter.

After coming to an equilibrium, the derived3 by integration, assuming a con-

FIG. 2.—Permeability from Drainage-Lag Device.

FIG. 3.—Effect of Temperature of Water on the Viscosity Ratio.

FIG. 4.—Ordinary Falling-Head Permeameter.

BY T. Y. Cnu,1 D. T. DAVIDSON,1 AND A. E. WiCKSTROM1

Barber's falling head permeability In developing the modified perme-

the following relationship: readings is an indication of a change in

100 92 39.6 39.5

106 90 35.6 6.9

103 96 38.1 258.5

METHOD OF TEST FOR PERMEABILITY OF SANDS

Scope: Ruler.—A ruler reading to A in.

FIG. 5.—^Apparatus for Preparing Sample for Permeability Determination.

The sample in the permeameter tube is

Calculation: ha, = height of level A from the water level

EFFECT OF SOIL COMPOSITION is one of the least permeable soil minerals

FiG.52.—Permeability of Kaolinite to Various Fluids as a Function of - (e = void ratio).

Pa. For soil No. 1 (a silty sand), the mini-

FIG. 5.—Density and Permeability as a Function of Molding Water Content.

FIG. 9.—Density Change Caused by Permeation.

of structure, as the following compari- complete aggregation is emphasized to

Dry Density or Void Permeability, Figure

Jamaican clay 115 Ib per cu ft Approximately same 4 X 10-« 6

FIG. 12.—Effect of Aggregation on Permeability.

It is often good engineering practice ment of a loaded specimen can be found

FIG. 1.—Gradation of j-in. Maximum Size Material Used in Small Permeability-Settlement

ability tests on specimens containing in the pervious zones of an earth dam or

FIG. 3.—8-in. Diameter Laboratory Permeability Apparatus.

materials used were predominantly 1. Tests on the sand fraction passing

regular-shaped moisture-density curve placement of desired relative densities

FIG. 5.—Maximum and Minimum Densities of Sand and Sand-Gravel Mixtures.

TABLE I.—SOIL PLACEMENT DENSITIES AND VOID RATIOS OF

of Reclamation laboratory showed re- per cent of relative density in Fig. 5.