CEMENT and CONCRETE RESEARCH. Vol. 4, pp. 881-897, 1974.

Pergamon Press, Inc

Printed in the United States.

RELATIONSHIPS BETWEEN ELECTRICAL AND

PHYSICAL PROPERTIES OF CEMENT PASTES

Michael A. Taylor and K. Arulanandan

University of California
Davis, California 95616

(Communicated by R. Brady Williamson)

(Received April 15, 1974; in final form July 17, 1974)

ABSTRACT
Many aspects of mechanical behavior of engineering materials,
s p e c i f i c a l l y concrete and s o i l s , can be explained from a know-
ledge of physico-chemical and/or p a r t i c l e geometry factors. A
q u a l i t a t i v e measure of the above factors has been used to des-
cribe a modelled structure of the cement gel system, and attempts
have been made to correlate the engineering properties of hardened
cement pastes with t h e i r microstructureo This paper presents the
r e s u l t s of preliminary investigations which suggest that some of
the important mechanical properties of mature cement pastes may
be predicted q u a n t i t a t i v e l y from measurements of t h e i r e l e c t r i c a l
properties at early ages.

La repr6sentation qualitative de leurs propri6t6s physico-

chimiques, et de relations g~om~triques entre leurs particules
~l~mentaires, expliquant bien des aspects du comportement
m~canique des mat~riaux di construction industriels, a ~t~
appliqu6e aussi A la d~finition d'un "modUle" de repr6sentation
de structure pour les gels de ciment. On a m~me tent~ d'~tablir
une correlation entre leurs propri~t~s industrielles, et la
structure microscopique de diverses p~tes cimentaires apr~s
durcissement. Le present m~moire exposera les r~sultats d'~tudes
pr~liminaires qui donnent A croire que certaines caract~ristiques
m~caniques importantes de ces pStes, une fois arriv~es a un stade
de v i e i l l i s s e m e n t assez avanc~, pourraient se pr~dire ~galement
d'une mani~re q u a n t i t a t i v e , grace ~ la mesure de leurs propri~t~s
~lectriques aux premiers stades.

881
882 Vol. 4, No. 6
M. A. Taylor, K. Arulanandan

Introduction

Despite the fact that portland cement concrete is the most widely

used construction material in the world today the exact nature of its

hydration process is s t i l l not completely understood. The complexities

of the physics and chemistry of this process have tantalized even the most

capable of researchers so that the exact constitution ( i f indeed i t be exact)

and morphology of hydrated gel remain the subject of vigorous debate and

inquiry. The hydration process and the hydration products are believed

to exert major control over almost all the important engineering properties

of cements (and hence concretes) such as strength, e l a s t i c i t y , shrinkage,

cree~ permeability and durability.

At present there exist two widely disparate approaches to the

understanding of cement hydration. On the one hand there are the relatively

recently developed powerful tools such as X-ray examination, electron

microscopy and d i f f e r e n t i a l thermal analysis (DTA), by means of which basic

research is conducted on extremely small samples of material. These micro-

scopic and submicroscopic physical and chemical studies have greatly

increased our understanding of the structure and properties of cement gel.

The aim of these basic researches is to explain the resulting properties

of a gross sample of cement or concrete from its microscopic features and

thereby provide a fundamental approach to understanding concrete behavior.

On the other hand there is the familiar and traditional phenomenological

approach in which tests are made on samples of a much larger size in order

to measure engineering properties d i rec tly and from which empirical rules

are deduced for the observed correlations between parameters which influence
the hydration process (such as w/c r a t i o , time, temperature etc.) and the

resulting engineering properties. [ l ]

Both of these approaches have some disadvantages. For the modern basic
Vol. 4, No. 6 883
ELECTRICAL PROPERTIES, MECHANICAL PROPERTIES, CEMENT PASTE

research methods mentioned samples are often d i f f i c u l t , costly and time

consuming to prepare; the equipment is expensive; and only very small

samples can be studied at one time. Furthermore the observations are made

on a sample removed from i t s "natural" spatial and temporal environment.

Hence these sophisticated and elaborate techniques w i l l probably be

confined to the large research laboratory in the forseeable future.

The t r a d i t i o n a l approach suffers from the major shortcoming of a lack

of an underlying theory and hence for each new circumstance a new t e s t

must be devised and time taken to establish reasonable standards.

In addition to t h e i r individual advantages and disadvantages the

two approaches have two common practical disadvantages.

(a) Before any measurements can be taken using current engineering

t e s t s , one must wait for the concrete to harden.

(b) Most of the tests on hardened concrete are destructive in nature.

There is c l e a r l y a pressing need f o r , and great benefits to be derived

from, a nondestructive test which could predict the properties of hardened

concrete from those of the fresh mix.

This paper describes a r e l a t i v e l y simple technique, which may be of

great practical and theoretical value in studies of cement gel and offers

the following f i v e major advantages over the other techinques mentioned above

I. I t bridges the gap between sophisticated basic examination

techniques and the t r a d i t i o n a l empirical tests. The t e s t involves

straight-forward electrical measurements on samples of small (but

not microscopic) size which are direc tly and closely linked to the

physical and chemical properties of the microstructure of the gel.

2. No time is lost. The tests are performed at an early age (possibly

while the mixture is s t i l l unhardened thereby avoiding the great

expenses involved whenever delayed tests lead to l i t i g a t i o n and/or

removal of completed concrete works).

884 Vol. a, No. 6
M. A. Taylor, K. Arulanandan

3. The test is non-destructive.

4. The e l e c t r i c a l properties of the gel are controlled by parameters

which are impossible to measure in other ways. For example, the

permeability of a sample is controlled and influenced by the size,

shape, d i s t r i b u t i o n and c o n t i n u i t y of pores. There is no other

nondestructive technique which can monitor t h i s property rapidly.

Thus, e l e c t r i c a l measurements may provide new information unobtainable

by any other means.

5. The test method may be readily adapted for use in the f i e l d .

There have been many past investigations concerned with e l e c t r i c a l

properties of paste and concrete (Ref. 2 for example l i s t s 27 references)

but these have dealt almost exclusively with r e s i s t i v i t y measurements.

To date no investigators have measured the variations in d i e l e c t r i c

constant (function of capacitance) and c o n d u c t i v i t y with frequency.

Plots of these properties versus frequency are termed the d i e l e c t i c

and conductivity dispersions respectively, and these dispersions form

the basis of the technique presented here.

Gel Structure and i t s Relationship

to Engineerin 9 Properties of Concrete

The principal cementing agent in portland cement concrete is CSH gel,

an extremely fine aqgregation of calcium-silicate hydrates and other

c r y s t a l l i t e s formed as a result of the hydration of portland cement. This

hydration process begins at a rapid rate which then decreases (approximately

exponentially) with time. I f maintained at I00% relative humidity, about

75% of the cement hydrates within the standard 28 day test period but the

process has been observed to continue for 50 years and for practical purposes

is believed never to cease completely should water be present.

For the purposes of this paper i t is assumed that the properties of a

Vol. 4, No. 6 885
ELECTRICAL PROPERTIES, MECHANICAL PROPERTIES, CEMENT PASTE

concrete are determined almost exclusively by those of its component cement

gel. In recent years i t has been established that the behavior of any

material is intimately connected with its microstructure. In the case of

concrete i t appears that properties such as strength and elasticity are

controlled by the extent and degree of interlocking of calcium silicate

hydrate crystallites among themselves and their bond to the aggregate,

whereas the time dependent properties such as creep and shrinkage are

probably determined by the thermodynamic equilibrium of the physically

adsorbed water in the gel. Durability and permeability are governed by

the size, distribution and continuity of the gel and capillary poroses

[3, 4].

Table 1 outlines how the major characteristics of a concrete are

determined by microstructural and physico-chemical features of its

component cement gel. While the importance of some of these mechanical

factors has long been recognized there has been very l i t t l e direct

research into how they are influenced by the micro-physical features listed

in the Table and almost all work has, until very recently, been conducted

on the phenomenological rather than a more basic level.

The new technique presented here indicates that these microstructural

features of a gel may be characterized q u a n t i t a t i v e l y through the e l e c t r i c a l

properties of capacitance and conductivity of a sample by plotting and

interpreting dispersion curves. These e l e c t r i c a l properties may then be

used to characterize a physical model which includes the geometrical and

physical measures of the gel system. These parameters may then be

correlated with mechanical properties as measured by t r a d i t i o n a l methods.

The e l e c t r i c a l measurement technique has already proven highly

successful in studies of ion exchange resins (5), soils (6 and 9), ion

exchange membranes (7), polyelectrolytes (8). I t is easy to perform and

886 Vol. ~, No. 6
M. A. Taylor, K. Arulanandan

Table 1
M i c r o s t r u c t u r a l Features of Gel which Control
Important C h a r a c t e r i s t i c s of Concrete

Microstructural features of gel

which are responsible for
Concrete Characteristic Imoortance observed characteristic

(in approximate order

of importance)
I. Strengtn Structural load ~umber of solid-solid contacts
bearing capacity Gel-soace ratio
Best indicator
of overall quality

2. Shrinkage Unsightly Physically absorbed water

cracking,* content of gel
Leaks Capillary pore geometry
Redistribution of
stresses

3. Creep Excessive Latest theories attribute

deflections same mechanism to shrinkage
Cracking* and creep so that creed
Leaks could be termed "Icad induced
Redistribution of shrinkage".
stresses Hence, same factors as for
shrinkage govern, but with
the introduction of mechanical
stress the number of solid-
solid contacts is also
important

4. Elasticity Deflexions Number of solid-solid contacts

Stiffness Amount of physically absorbed
Stress water in gel (since this
distribution is now believed to ~e load
in reinforced bearing)
concrete

5. Durability Resistance to Chemical nature of gel

(chemical} corrosive Porosity of gel w i l l control
environments accessibility of corrosive
liouids

6. Durability Resistance to Size and distribution of Dores

(freeze thaw) thaw cycles Amount of free water in gel

7. Permeability Passage of water Size, distribution and

tortuosity of pores

Cracking causes problems in many forms. Apart from architectural unsightliness

the major problems concern leaking of water and corrosion of reinforcement.
Leaks in concrete structures have been estimates to be the leading cause of
l i t i g a t i o n in the construction industry in recent years.
Vol. 4, No. 6 887
ELECTRICAL PROPERTIES, MECHANICAL PROPERTIES, CEMENTPASTE

provides correlations between electrical and mechanical properties of

dispersible systems such as clays and other soils which have microstructures

similar to that of cement gel.

Preliminary results show good correlation between electrical parameters

and the best current knowledge concerning microstructure of cement paste

[4, lO]. I t is hoped that these preliminary results w i l l serve as a guide

for more detailed studies in the future.

Electrical Dispersion Characteristics of Cement Pastes

The application of an electrical impulse to a system produces a

response which can be measured in terms of resistance and capacitance

and expressed in terms of apparent dielec tric constant, ~, and apparent

conductivity, ~. The properties E, and ~ are frequency dependent, E

decreasing and o increasing with increasing frequency. The plots of

E and o against frequency are termed electrical dispersions. Figure l

shows the dispersion characteristics for a cement paste of water cement-ratio

0.35 (by weight) as a function of time. These dispersion characteristics

have been related to the structure of a system through the use of a simple

model as follows.

The microstructure of cement gel may be considered as being composed

of three main components:

A. Solid particles surrounded by an ion atmosphere (Guoy-Chapman

double layer) which are assumed to behave as integral units

B. Solid particles in contact with each other

C. Pore spaces mostly f i l l e d with water (often containing dissolved

compounds).

When a voltage is applied to such a system the current is considered

to be conducted along three paths:

888 Vol. 4, No, 6
M. A. T a y l o r , K. Arulanandan

120 I I L I I [ 0.0029
I

A
110 - 0.0027

"w 100 0.0025

Z
c~
CD
Z
° 90 0.0023 c~
o

<

-<
~_J

~ 80 S 0.0021 3
g

O_
Q_
< 70 0.0019

~/ Dielectric Conductivityo- I
Age Constante' Experimental /
Exoeri'-ental
60 23 Hrs. O • 0.0017
35 Hrs. A •
52 Hrs. O
Theoretical

50 0.0015
107 2 3 4 5 6 7

FREQUENCY IN HERTZ (Log Scale)

FIG. 1
Dielectric Dispersion of Cement Paste
(Water Cement Ratio of 0.35) as a Function of Age.

I. Through solution and conducting p a r t i c l e s in series

2. Through p a r t i c l e s in contact with each other

3. Through the s o l u t i o n .

A schematic representation of a model of t h i s system is shown in Fig. 2.

In Fig. 2 a, b, and c represent the f r a c t i o n a l cross-sections of paths

l , 2, and 3 respectively. For path l parameter d is a measure of the

length of the current path through the conducting solid ( p a r t i c l e

surrounded by i t s ion atmosphere).

Vol. 4, No. 6 889
ELECTRICAL PROPERTIES, MECHANICAL PROPERTIES, CEMENT PASTE

/ Liquid

_ ,

(l-d) •

~Solid

I_.a I_ b lc] 2 B. Electrical Analog

t-- YJ __1 of Model
[__
[-- 1 r I
FIG. 2
Schematic Model of Mechanical and
2A. Model of Current Paths Electrical Features ef Cement Gel

The parameter c is a measure of the continuous i n t e r s t i t i a l paths

and is dependent on porosity, p a r t i c l e shape, o r i e n t a t i o n , cementation
and also the extent of the absorbed layer; b represents the contact
area between " s o l i d " p a r t i c l e s ; a is a measure of the cross sectional
area of solution and conductive solids in series.

Each of the zones shown in Figure 2a can be represented by a parallel

c i r c u i t of resistors and capacitors as in Fig. 2b. The impedance of these
c i r c u i t elements are uniquely determined by the geometrical parameters
a, b, c, and d and by the d i e l e c t r i c constants Er' Cs and conductivities
Kr and Ks of the particles and the solution respectively. The d i e l e c t r i c
constant cs is a property of the pore f l u i d (eg. Dioxane = 2, Acetone =
21, Water = 80, Formamide = 109) and Ks is a measure of the concentration
of dissolved salts. The d i e l e c t r i c constant
and conductivity Kc
r r
of the conducting solid (which is the p a r t i c l e with i t s associated absorbed
layer) are dependent on the mineral solution interface characteristics and
the amount of water absorbed on the surface. Hence the parameters ar ,
Kr' ~s and Ks can be used to describe the compositional properties of
cement-water system and the geometrical properties of the system can be
described by a, b, c, and d.
!
The apparent d i e l e c t r i c constant Ctheor and the apparent conductivity
o for the composite model are evaluated by elementary e l e c t r i c a l network
890 Vol. 4, No.
M. A. T a y l o r , K. Arulanandan

analysis. The equations relating

~' and ~ to frequency are dependent
~theor
on the compositional and geometrical parameters of the system and are as
follows:

a i a
theor = d'(l - d)S LT-:-~ +

22
(2ErCs
2)]
ares
(I)
T-~ + ~ + ba r + ca s

: d(1
a
- d)S
+ k~ks
--~-- +

2~2 ~Cs r + + bk r + ck s (2)

w
ro S - l-d
s + + 2
~
2 ( ~
as ar
+ ~
) 2

is the capacitance of a unit capacitor in vacuum i.e.

.0885 X lO-12f.
is the angular frequency.

Using these techniques Spiegler, Arulanandan, et al. have shown how

the internal structure of several different materials (5 thru 9) can
be analyzed and classified by means of electrical measurements. Many of
the clay minerals they examined have structures which are similar
to calcium s i l i c a t e gel. By means of the developed technique several
correlations between electrical and physico-mechanical properties have
been found, for example:

ELECTRICAL MODEL PROPERTY CAN BE CORRELATED WITH

.Clays
Dielectric constant a Swell potential, compressibility
r
Continuous i n t e r s t i t i a l path c Permeability
Vol. 4, No. 6 891
ELECTRICAL PROPERTIES, MECHANICAL PROPERTIES, CEMENTPASTE

Cements
Solid-solid contacts b Strength, Elasticity
Continuous I n t e r s t i t i a l path c Permeability

Apparatus For Electrical Measurements

Impedances in the radio frequency range (.5 x lO6 to lO8 Hertz)

were measured with an "RX meter" (Boonton Radio Corp., Division of
Hewlett-Packard, Rockaway, New Jersey). A description of the cell and
the technique used to measure the electrical properties of the cement
are given in Ref. ~.

Results of Measurements on Cement Pastes

A small test program was completed in which variations in the values

of resistance and capacitance over the range IMHz to lO0 MHz were measured

for two pastes. Measurementswere made on 0.5" (12.7 mm) diameter by 1.5"

(38.1 mm) (approximately) long paste specimens. The water cement ratios

used were 0.30 and 0.35 by weight. Electrical measurements were made at

intervals up to an age of l week so that the influence of both age and

water-cement ratio could be examined.

In a second test series three samples of paste having water cement

ratios 0.30, 0.35, 0.40 by weight were cast and tested when approximately

24 hours old. This series was intended to substantiate and extend the

examination of the influence of water cement ratio.

Companion 2" (50.8 mm) diameter by 4" (lOl.6 mm) long specimens were

cast for each mix and tested for compressive strength at ages 3, 7, and lO

days. The results are given in table 2.

Dispersion curves for each series of measurements were made and then

the model parameters a, b, c, and d were selected so as to predict the

observed dispersions successfully. This was accomplished by means of a

computer optimization program in which the compositional and geometrical

parameters of the model were adjusted so as to produce a dispersion which

892 Vol. 4, No. 6
M. A. Taylor, K. Arulanandan

Tabl e 2
Compressive Strength of 2" Dia. by 4" Paste Specimens
3 days 5 days 7 days IO days
w/c r a t i o
0.3 6300 6800 7900 not
measured
6100 7000 7900

Average 6200 6900 7900

0.35 6200 8000 7800 9200

6100 6700 7400 7000
6000 7200 7400

Average 6100 7300 7500 8100

0.40 6100 5300 ~800" 6300

5600 5300 6900
4900 6500

Average 5600 5300 6700 6300

*Ignored in com~uting average

Table 3
Results of Optimization of Model Parameters a, b, c, d
for Various Pastes at Various Times.
(s-;) (s-s) (,)
w/c r a t i o AGE a b c d

.30 8 hrs. .04 .72 .24 .014

16 hrs. .08 .60 ,32 .003
27 hrs. .16 .54 .30 .039
37 hrs. .17 .62 .22 .004
2 ]/2 days .13 .76 .12 .018
3 I/2 days .16 .73 .11 .q]5
7 days .13 .80 .08 .026
lO days .15 .85 .00 .002

.35 4 hrs. ,20 ,51 ,30 .032

]0 hrs. .19 .<9 ,33 .049
23 hrs. .21 .53 .26 ,075
35 hrs. .28 .50 .22 .088
52 hrs. .27 .56 .18 .091
3 days .19 .65 .16 ,I14

.30 27 hrs. .16 .54 ,30 ,039

.35 23 hrs. .21 .53 .26 .075

.40 25 hrs. ,20 .21 .59 .064

(s = s o l i d L = liquid)
Vol. 4, No. 6 893
ELECTRICAL PROPERTIES, MECHANICAL PROPERTIES, CEMENT PASTE

matched the observed curve within a chosen accuracy. Two optimization

techniques were used - the Fletcher-Powell and the Simplex techniques.

Both methods give satisfactory results although studies are continuing

on the relative efficiencies of the two approaches.

These optimized results are given in Table 3 and interpretation in

terms of the model yields the following information, all of which is in

excellent agreement with current knowledge of the cement hydration process.

[IO]

I. For all w/c ratios the solid-solid contacts increase with time.

They increase greatly during the f i r s t day and then increase

at a slower, but steady rate throughout the f i r s t week. (Figs. 3

and 4).

2. At any given time the solid-solid part of the mass is greater

for lower water cement ratios. (Figs. 3 and 4, Table 3)

Lo! I
i
I
I Liquid c

I
o8 I
Solid - Liquid a
=
~..-J

=< 06 - I

o
~
k-- Solid - Solid b I/Liquid~
/
~0.4
0
c~
0.. ? i
Model
h--- ~ ~ .
-2.:-
0.2

0 I I,
0 I0 20 30 40 50 60 70 80 90
AGE IN HOURS

FIG. 3
Distribution of Current Paths for Cement Paste of Water-Cement Ratio 0.35.
894 Vol. 4, No. 6
M. A. Taylor, K. Arulanandan

I0 .....

•L__•u,d
IA
c
i
!

IASolid - Liquid a

08 w

L~
=
/ i
~_ 06 [ Llqudd~,
c:)

0= Solid - Solid b
i-. I
o..
o_ !
0Z
!i Mc~ei
i
i
J
i I I
20 ~0 60 80 tO0 lZO 140 160 180

AGE IN HOURS

FIG. 4
Distribution of Current Paths Versus Time for Cement Paste of Water Cement Ratio 0.3

3. For all w/c ratios the liquid path (through water-filled connected

capillaries and surface absorbed water) decreases f a i r l y steadily

with time. Since the amount of surface adsorbed water is proportional

to the amount of gel present the quantity of surface absorbed water

must increase, roughly in proportion to the volume of gel and, hence,

solids. Therefore the water f i l l e d connected capillaries are

decreasing by a very large amount to compensate for the increase

in surface absorbed water. The permeability of the mass must thus

be reducing steadily which is amply confirmed by previous investiga-

tions. [e.g. Ref. 3, P. 385]

4. I f , as just explained, connected capillaries decrease with time

then the mechaninism is probably as follows. The continuous

crystallization of gel converts a connected into an unconnected

space thus altering the proportions of a and c (decreasing c and

Vol. 4, No. 6 895
ELECTRICAL PROPERTIES, MECHANICAL PROPERTIES, CEMENT PASTE

increasing a). However, tests show that the value of "a" remains

roughly constant after l day so that as fast as spaces are being

converted from connected to unconnected status so are unconnected

spaces being converted to "solid" by complete f i l l i n g with gel.

This is in complete agreement with the spatial uniformity of gel

production as envisaged in current theories.

. Comparison of model parameters at age 25 hours for cements of

0.30, 0.35 and 0.40 w/c ratio (Table 3 and Fig. 4) shows that

the,parameter b (solid-solid contacts is very sensitive to w/c

ratio between 0.35 to 0.40 but appears to level off below 0.35.

This indicates that the parameter b is sensitive to "excess"

water in the mix since i t is well known that between about .30

and .35 by weight of water is needed for complete chemical

conversion of cement. The range w/c ratios used in the field

is between .35 and .60 so that b should be a sensitive indicator

of the potential strength of the gel under practical conditions.

[Fig. 5]

Conclusions

While i t is realized that many chemical features may modify the findings

presented here these preliminary investigations do indicate that radio

frequency dispersion measurements, when interpreted in terms of a reasonable

physical model, can provide insight and new knowledge concerning the

microstructure of fresh and hardened cement pastes. Further research

should be conducted. These studies may be extended in two directions. For

basic research the relationships between the physical, geometrical and

electrical properties could be examined and the simple model proposed here

extended in complexity. Practical applications of the method would begin

896 Vol. 4, No. 6
M. A. T a y l o r , K. Arulanandan

7 Day
1 Day • Strength
b psi

0.6

8000

0.4
--¢. \
\
\
\
\
\ \
\ \-- 7000
\ \
\
\
0.2 m
\

0 I I I 1 6ooo
0.30 0.32 0.34 0.36 0.38 0.40

WATER-CE~E,*IT RATIO BY WEIGHT

FIG. 5
1 Day Value of b and 7 Day Value of Strength plotted Against W/C Ratio

with further examination of correlations between early-age-electrical

and subsequent mechanical properties and the development of a field

measuring device.

ACKNOWLEDGEMENT

The authors acknowledge the help of two students from the Davis Campus

of the University of California; Mr. Gary Foster and Mr. Tom Linkhart who

assisted in the experimental program.

REFERENCES

I. Burks, S. D., "Will Concrete be the Leading Building Material of

the Future?" Journal of American..Concrete Ins titute Proceedings ,

Vol. 4, No. 6 897
ELECTRICAL PROPERTIES, MECHANICAL PROPERTIES, CEMENTPASTE

V 68, No. 5, May 1971, pp. 321-326.

2. Hammond, E. and Robson, T. D. "Comparison of Electrical Properties

of Various Cements and Concretes" The Engineer (London) January

21, 1955, pp. 78-80 and January 28, 1955, pp. ll4-115.

3. Neville, A. M., Properties of Concrete, p. 532, Wiley, 1963.

4. Neville, A. M., "Hardened Concrete - Physical and Mechanical

Aspects," A.CI., Monograph No. 6, A.C.I. Iowa State University

Press, 1971, Chapter 18.

5. Sachs, S. B. and Spiegler, K. S., "Radio Frequency Measurements

of Porous Conductive Plugs, Ion Exchange Resin Solution System,"

Journal of Physical Chemistry, V. 68, 1964, pp. 1214-1222.

6. Arulanandan, K. and Mitra, S. K., "Soil Characteriza'tion by Use

of Electrical Network," Proceedings of Fourth Asilomar Conference

on Circuits and Systems, Pacific Grove California, Nov. 1970.

7. Spiegler, K. S., Arulanandan, K., and Miller, D. G., "Transport

Phenomena in Composite Liquid and Solid Electrolyte Systems,"

Interim Report No. 3 to Office of Saline Water, U. S. Dept. of

the Interior, Washington, D. C., February 1971.

8. Sachs, S. B., Raziel, A., Eisenberg, H., and Katchalsky, A.,

"Dielectric Dispersion Properties of Aqueous Polyelectrolyte

Solutions,: Polymer Dept., Weizmann Institute of Science, Revohot,

Israel, 1970.

9. Arulanandan, K. and Smith, S. S., "Electrical Dispersion in

Relation to Soil Structure." Journal of the Soil Mechanics and

Foundations Division, ASCE, V. 99 # SMI2, Proc. Paper 10235,

December 1973, pp. Ill3-1133.

lO. Powers, T. C., "The Properties of Fresh Concrete" John Wiley and Sons,

1968, pp. 663.