Professional Documents
Culture Documents
Concrete
BY
G. E. Monfore
Principal Research Physicist
Applied Research Section
Research and Development Laboratories
Portland Cement Association
SYNOPSIS
Electrical properties of concrete are of importance concrete ties may, under certain condi-
in several applications: pf current interest is the elec- tions, be about one-tenth of the resistance
between rails on timber ties. This resis-
trical resistivity of concrete crosstier since this resistivity
affects the operation of railroad signal systems. An
investigation of electrical properties, including both tance between rails can be increased by in-
direct and alternating current studies, showed that sulating the rails from the concrete, a
moist concrete is essentially an electrolyte having a
method adopted by European railroads, or
r e s i s t i v i t y o f t h e o r d e r o f IO4 o h m - c m , a v a l u e i n t h e
range of semiconductors. Oven-dried concrete has a by increasing the resistance of concrete it-
r e s i s t i v i t y o f t h e o r d e r o f IO o h m - c m , a r e a s o n a b l y self. An investigation of the latter possibili-
good insulator. \ ties provided the basis for the present
Several admixtures that were studied made only
m o d e r a t e c h a n g e s i n r e s i s t i v i t y o f portland c e m e n t study. The literature on electrical proper-
pastes. ties of concrete was found to be meager;
Various coatings applied to dried concrete were the most thorough general study noted was
that of Hammond and Robson.( They
ineffective in preventing ingress of water and conse-
quent lowering of resistance.
The resistance of concrete crossties probably can- found that the electrical resistivity of con-
not be increased sufficiently by use of admixtures or crete was highly dependent upon the mois-
coatings on the ties. However, studies of methods of
ture content, and that the resistivity of
insulating rails from concrete crossties, to be published
soon, i n d i c a t e t h a t s a t i s f a c t o r y r a i l - t o - r a i l r e s i s t a n c e concrete made from calcium aluminate ce-
may be obtained by such methods. ment was about 10 times greater than the
The present studies suggest that electrical methods resistivity of concrete made from portland
might have practical applications in concrete research.
cement. These authors also included 27
KEY WORDS: admixtures (concrete): aggregates; refer_ences in their review of the subject;
reviews in the present paper, accordingly,
alkali content: coatings; concretes: conductivity: elec-
trical properties; moisture content; mortars; pastes;
railroad ties will be limited to those not covered by
Hammond and Robson. Nikkanen(2) con-
cluded that paste and concrete.are conduc-
INTRODUCTION tors similar in nature to electrolytes, and
Increasing use of concrete crossties by the listed data showing that the temperature
American railroad industry has emphasized coefficient of electrical conductivity of fresh
the need for additional information con- paste is of the same orde_r of magnitude
cerning the electrical properties of con- as that of common electrolytes. Henry(s)
crete. Such properties are of interest be- found the electrical resistivity of concrete
cause the railroad signal systems operate
through the rails. Field data indicate that *Numbersin parentheses designate references at
electrical resistance between rails laid on end of paper.
lyte. Other investigators, including Ham- Fricke(a) deduced a similar equation for
mond and Robson,( 1) consider the capaci-
tance to be in parallel with the resistance. uniform ellipsoidal particles, which for the
Also in laboratory measurements of AC special case of nonconductive particles, is
properties of electrolytes, the4 balancing
capacitor and resistor are connected in par- p=pm 1 +k-V,,,
kV . ..*.... (12)
allel. In the present paper the behavior of Ill
concrete subjected to AC potentials will be where k depends upon the shape of the el-
considered to be equivalent to a capacitor lipsoids. For spheres k is 2, and for a partic-
and resistor in parallel. The current through ular sand studied by Fricke k was 1.4.
such a combination is
Thus, for that sand,
I=+ . . . . . . . . . . . . . pzpm 1 4v
2.4 - V,
1.. . . . . . .(13)
- m
where
I is current in amperes, The resistivity of a composite consisting
E is potential in volts, and of nonconductive particles embedded in a
Z is impedance in ohms. conductive matrix may also be expressed in
For the particular case of a resistor and terms of an effective path length, L,. Since
capacitor in parallel the current must travel a tortuous path be-
cause of the obstructing particles, the effec-
tive path length is longer than the dimen-
sion of the composite in the direction of the
where
w = 2 ?r f, f being in hertz, Hz (cycles current. Noting that the cross-sectional
per second), and area of the matrix in a I-cm cube is numer-
C is capacitance in farads. ically equal to V,, the resistivity of the
Composiiion, yo by wt
where L is the apparent path length, i.e., Componenl
Cement A Cement B Cemenl C
the actual dimension of the composite in
the direction of current flow. SiO2 20.66 21.82 20.90
TEST SPECIMENS 5.62 4.72 5.42
AMA
One-inch paste cubes and 4-inch concrete
cubes were found to have electrical proper- , Fez01 2.17 2.57 2.64
Vacuum Tubs
Cuoncret
Voltmeter
n r l ,
r
Fig. 3 -Current-Potential-Time Characteristics
Cube of Concrete Specimen.
ages of 7, 28, and 90 days as listed in Table available and its use allowed very sensitive
3. Resistance increased with age, as later balancing of the bridge since harmonics
AC measurements substantiated, but E, from the oscillator were filtered out.
was essentially constant for a given cement. As the reference arms of the bridge of
E, for Concrete A was slightly higher than Fig. 4 are equal, the resistance and capaci-
that for Concrete B. tance of the specimen are equal to the
corresponding values of the variable resis-
AC PROPERTIES tor and capacitor at balance. It will be
Equipment shown later t t impedances of concrete
9
The bridge circuit used in studying AC specimens calculated by Eq. (9) were essen-
properties of pastes, mortars, and concretes tially equal to the resistances, and thus the
is shown in Fig. 4. All components were capacitances were of minor importance in
usual laboratory equipment except the de- the present studies. Nevertheless, the vari-
tector which was an audio frequency spec- able capacitor was necessary in order to ob-
tain a sensitive balance of the bridge.
trometer. This instrument includes a series
of narrow-band filters connected between Reproducibility
the bridge and the null indicator. Although The coefficient of variation of six com-
such an instrument is not required, it was plete determinations of the resistance of a
single plain l-inch paste cube was 1.1 per-
cent, and the coefficient of variation of the
OSCILLATOR
determinations of the resistances of six
companion l-inch paste cubes was 1.4 per-
cent. A somewhat higher coefficient of vari-
IOOO-OHM STANDARD
ation of 4.9 percent was found for the de-
terminations of the resistances of six 4-inch
concrete cubes having internal electrodes.
The first set of three of these cubes was cast
DETECTOR on one day and another set of three was
cast a week later.
VARIABLE CAPACITOR Testing Procedure Variables
Variations in the testing procedures for
ranges used in the present study, except for
temperature, affected the value of resist-
ance of concrete only slightly.
Frequencies of 100, 1009, and 10,000 Hz
(cycles per set) made minor differences in
resistance but major differences in capaci-
tance as shown in Table 4. However, values
of impedance Z, calculated from Eq. (9),
were essentially equal to the measured re-
Fig. 4- Bridge Circuit for. AC Messurements. sistances. Measurements were made at 90
days on a 4-inch Concrete B cube having inch plain paste cubes is shown in Table 7.
internal electrodes. Notice that the resistivity of paste having a
Since the resistivity of electrolytes de- water-cement ratio of 0.40 was about dou-
creases with increasing temperature, the re- ble that of paste having a ratio of 0.60.
sistivity of moist paste, mortar or concrete Note too that resistivity increased with
would also be expected to decrease with in-
time of continuous moist storage. Decrease
creasing temperature. Measurements of the
of evaporable water with continuing hy-
resistivity of a paste over the temperature
range from 40 to 100 F are shown in Fig. 5. dration would have likely been a factor in
For this specific range, an average increase this increasing resistivity, but other effects,
in temperature of one degree F caused an such as continuing release of alkalies and
average decrease in redstivity of one per-
cent.
The effect of potentials of 2 to 8 volts on
resistance or impedance was slight as shown
in Table 5. The specimen used for those
measurements was a 4-inch cube of Con-
crete A having internal electrodes, and was
tested after 7 days in limewater and 14 days
drying in an atmosphere of 50 percent rela-
tive humidity and 73 F.
Resistivities of 4-inch cubes having inter-
nal electrodes cannot be calculated directly 600~
from measured resistances because of un- I
4 0 6 0
I
8 0 100
I
6 0.00267 961.8
Paste and Concrete Variables
Water-Cement Ratio. The effect of wa- 8 961.4 0.00295 961.3
ter-cement ratio on the resistivity of l-
Resistance, ohms
I I Age.
(1000 hertz, A volts) External Resistance
Internal Resistance
days Externa Internal
Electrodes Electrodes (ratio)
7
28
483
616
380
481
I 1.27
1.28
90 844 621 I 1.36
-
5 0 0
o--Fm
01 IO
TIME. HR
100 i 000
leaching, might also influence the resistivi- stainless steel electrodes. As shown in Fig. 7,
tY. the early relationship was dilferent. The re-
Resistivities of paste having a water-ce- sistance of paste A was considerably higher
ment ratio of 0.41, the ratio used in the than that of Paste B during the first day, by
concretes, were computed from the data of 4 days the resistances were equal, and
Table 7. Resistivities of Paste A and Con- thereafter the resistance of Paste B was
crete A, having the same water-cement ra- higher, in pgreement with the data of Ta-
tio, are shown by the curves of Fig. 6. The ble 7. It should be noted that resistances of
resistivity of the concrete was about five both pastes were essentially constant fol
times that of the paste. about the first 5 hours and then began to
Alkalies. Paste A, which had a lower al- increase. The inflection in the curve for
kali content than Paste B, might have been Paste A is considered to be real, as it was
expected to have a higher resistivity than observed several times.
Paste B, but the opposite was observed at
In connection with the effect of alkalies,
ages of 7 days and longer, Table 7. To pro-
some measurements on solutions may be of
vide further comparison, additional meas-
urements on Pastes A and B were made at interest. Fig. 8 shows that the resistivities of
early ages using l-inch cubes* with internal sodium hydroxide solutions, which were
saturated with calcium hydroxide, de-
creased markedly as the concentration of
*These cubes were contained in plastic molds
during test, and water was available to the tops sodium hydroxide was increased.
of the cubes via small diameter syphons. By this Pure Compounds. Also of possible inter-
arrangement the cubes were kept continuously
moist without any appreciable leaching. est are the resistances of pastes of C,A and
CONCRETE E
Fig. IO- Effect of Type of Cure. Fig. I I - Effect of Drying et 50% RH end 73 F.
Table 7 that the resistivity of a paste hav- cement ratio was held constant at 0.50;
ing a water-cement ratio of 0.50 is only hence the resistivity of the paste matrix was
slightly greater than 1000 ohm-cm at an age the same for all of the specimens. Values of
of 3 months. The resistivities of the marble the resistivity of the mortars, calculated
and granite are essentially infinite as com- from Eq. (13), are listed in Table 10 and
pared to the resistivity of paste, and the are seen to agree very well with the mea-
resistivities of the sandstone and limestone sured values. Effective path lengths, L,,
are considerably greater than that of paste.
calculated from Eq. (14) and also listed in
However, the resistivity of some aggregates
when embedded in concrete, and thus ex- Table 10, should be constant for a given
posed to alkalies, would be lower than the specimen provided the particles remain
value obtained for specimens saturated nonconductive. It will be noted that the
with limewater. Even so, the resistivity of values of L, were indeed reasonably con-
the usual moist concrete must be consider- stant.
ably dependent upon the resistivity of the
paste matrix. TABLE 1 0-RESISTIVITIES AND PATH LENGTHS
OF MORTARS
Correlation between Measured and
Computed Values Re&ttvity, ohm-cm
(10,000 hertz, 4 volts) L
A series of l-inch mortar cubes were % Age,
made from Cement A and varying amounts days Measured Computed cz
of the 10-20 mesh quartz sand. The water- 1 3 757 - 1
14 1071 - 1
28 J154 - 1
T A B L E 9-RESISTIVITIES OF ROCKS
0.654 3 1430 1440 1.24
\ Absorption, Resbtivity, ohm-cm 14 2030 2040 1.24
% by wt (1000 hertz, 4 volts) 28 2200 2200 1.25
\ Type
I
0.486 3 2140 2130 lx37
14 3000 3010 1.36
28 3330 3240 1.40
ing implications. The value of 0.81, for in- Specimens taken directly from an auto-
stance, is an impossible value since the the- clave, or normal-cured or steam-cured speci-
oretical effective length can never be less mens that are air-dried for sufficient time,
than 1. This must mean that the aggregate have resistances manyfold greater than cor-
was conductive. Also the theoretical effec- responding moist specimens. Various coat-
tive path length of a given specimen cannot ings were tried on dry specimens to de-
change with time. The increase in values of termine if this high resistance could be
L, with time must therefore mean that the maintained.
conductivity of the particles was decreasing Dry 4-inch cubes with internal electrodes
with time. This might mean that reactions were given two applications of each mater-
occurred between Na+ and K+ ions and the ial tested, and after the coatings had
aggregate which reduced the number of hardened, the cubes were immersed in
conducting ions. This view has some sup- water. Resistances measured after several
port in the fact that values of L, were low- periods of immersion are shown in Fig. 12.
er for the high alkali Cement B than for
Cement A, and that the increase of L, with
time was greater for Cement B than for
TABLE 1 1-RESISTIVITIES AND PATH LENGTHS
OF CONCRETES
Cement A.
The foregoing discussions stemmed from
considerations of mortars and concretes as
composites of nonconductive particles in a
paste matrix. The paste itself may also be
considered as a composite of nonconductive
particles in a matrix of conductive evapor-
able water. The uncertainties, however, are
much greater in the latter case. The follow-
ing discussion is therefore of a speculative
nature, but is felt to be of value. The vari-
ous steps in the calculation of properties of TABLE 12-PASTE A, CONSIDERED AS A COMPOSITE
ties of the matrix pm were calculated by pm, resistivity of matrix 280 ohm-cm
a solution is 29 ohm-cm, Fig. 8. This value Resirtivity of 0.2 normal solution 29 ohm-cm
of 29 ohm-cm should be compared to the (Fig, 8)
value of 280 ohm-cm calculated from resis- Apparent fraction of matrix available 0.1
tance measurements on the paste. The in- for conduction
dications are thus that only about 10 per-
Development L a b o r a t o r i e s , M a y 1 9 6 8 4: i
Concrete A Concrete B
AUTOCLAVE - AUTOCLAVE -
w/c=0 47
\ 1
u
e+ n c o t _ e d , 0
IO
TIME AFTER IMMERSION IN WATER, DAYS
-A 20 30
The epoxy coating was the most effective, concrete is considerably dependent upon
but after 28 days of immersion in water, the paste matrix, is influenced only slightly
the resistance was only moderately greater by frequency in the audio range or by po-
than that of the uncoated specimen. Some tentials in the range of 2 to 8 volts, but is
materials actually decreased the resistance sensitive to temperature.
of the specimens. Resistivity of concrete may be increased
In another phase of the investigation, manyfold by air-drying, and upon oven-
coatings were applied to steel anchor nuts, drying at 105 C values in the range of good
Fig. 13, designed to be embedded in con- insulators are obtained. Several coatings
crete crossties to receive the hold-down that were applied to dry concrete were in-
bolts. These tests were part of an investiga- effective as barriers to the ingress of water
tion, to be published soon, by Hanson and and consequent decrease in resistance.
Hsu,(r3) of methods of insulating rails from Furthermore, the most effective admixture
concrete crossties. In contrast to the experi- used in the present study provided an in-
ence with coatings on concrete, several crease in resistivity considerably less than
coatings on steel anchor nuts embedded in needed. It seems, therefore, that insulation
concrete were found to be effective, Fig. 14. of rails from concrete crossties would be the
The intent of the study was to compare most successful way of obtaining adequate
the effect of the various materials when ap- rail-to-rail resistance. A paper to be pub-
plied in equal coating thickness. The act& lished soon(r3) will describe in detail vari-
coatings, however, varied considerably .in ous insulation methods.
thickness as indicated by the measured val- Some metals, particularly steel and alu-
ues given in Fig. 14, and resistances gener- minum, may be subject to electrolytic cor-
ally increased with increasing thickness. rosion under certain conditions when em-
The coatings on the steel anchor nuts were bedded in concrete.( 14,*5) The resistivity of
applied by a commercial firm using their the concrete is likely a factor in the rate of
own formulations, while those on the con- such corrosion that might be due to either
crete specimens were applied in the PCA stray or galvanic currents.
Laboratories. Since the conduction of electricity
CONCLUSIONS through moist concrete may be visualized
Conduction of direct current or alternat- as the movement of ions in the evaporable
ing current at audio frequencies through water of the paste matrix, and in some
moist concrete is essentially conduction by cases in the pores of the aggregate particles,
an electrolyte. Polarization effects that oc- any factor that affects the amount or
cur at the electrodes are evident with either properties of the liquid, or the kind or
direct or alternating currents. Resistivity of number of ions will influence the resis-
5
potentials E,,, E,, i IO 20 5 0
TIME, D A Y S
REFERENCES
I. Hammond, E., and Robson, T. D., Compari-
son of Electrical Properties of Various Cements
and Concretes, The Engineer (London), 199,
No. 5165, 78-80, and No. 5166, 114-115 (January
21 and 28, 1955).
2. Nikkanen, P., On the Electrical Properties of
C o n c r e t e a n d T h e i r A p p l i c a t i o n s , Valtion
Teknillinen Tutkimuslaitos, Tiedotus, Sarja III,
Rakennus 60 (1962), 75 pages. In Finnish with
English summary.
3. Henry, Robert L., Water Vapor Transmission
and Electrical Resistivity of Concrete, Final Re-
port, U. S. Naval Civil Engineering Laboratory,
Port Hueneme, California, Technical Report
Fig. 13 -Anchor Nut. R-314 (June 1964), 39 pages.