1

TEMPERATURE CONTROL OF CONCRETE-Sec. 7-12 115

exposure temperatures as sinusoidal waves
with applicable periods of 1 day, 1 week or 2
weeks depending upon the severity of the
weather to be used for the design, and 1 year.
Solar radiation is then added to obtain the
final range of mean concrete temperature.
For average (mean) weather conditions, the
ambient air temperatures are obtained from a
plotting of the mean monthly air temperatures
on a year scale. For usual and extreme
weather conditions, the above ambient air
temperatures are adjusted for a 7-day period
and a 14-day period, respectively, at the high
and low points of the annual curve. The
amount of the adjustment for these weather
conditions is described in subsection (a)
below.
The thickness of section for these studies is
measured along lines normal to the exposed
surfaces, the intersection of the normals being
equidistant from the two faces.
(a) Ambient Air Temperatures.
-When
computing the range of mean concrete
temperature, mean daily, mean monthly, and
mean annual air temperatures are used. The
theory applies the daily and annual air
t em peratures as sinusoidal variations of
temperature, even though the cycles are not
true sine waves. The annual and daily
amplitudes are assumed to be the same for all
weather conditions.
To account for the maximum and minimum
recorded air temperatures, a third and
somewhat arbitrary temperature cycle is
assumed. This temperature variation is
associated with the movements of barometric
pressures and storms across the country. Plots
throughout the western part of the United
States show from one to two cycles per
month.
Arbitrarily, this third temperature variation is
assumed as a sine wave with either a 7-day or
14-day period for usual weather conditions
and extreme weather conditions, respectively.
For extreme weather conditions, the
amplitudes of the arbitrary cycle are assigned
numerical values which, when added to the
amplitudes of the daily and annual cycles,
will account for the actual maximum and
minimum recorded air temperatures at the
site. For usual weather conditions, these
amplitudes are assigned values which account
for temperatures halfway

between the mean monthly maximum

(minimum) and the maximum (minimum)
recorded. When computing the mean concrete
temperature condition, no third cycle is used.
(b) Reservoir Water
Temperatures. -The
reservoir water temperatures used in
determining the range of mean concrete
temperature for a proposed dam are those
temperatures which will occur after the
reservoir is in operation. These reservoir
water temperatures vary with depth, and for
all practical purposes can be considered to
have only an annual cycle. For preliminary
studies, the range of mean concrete
temperature with full reservoir is the normal
condition. For final designs, stage
construction should be taken into
consideration and the design reservoir
operation used. When the reservoir is to be
filled or partially filled before concrete
temperatures have reached their final stage of
temperature equilibrium, further studies are
needed for the particular condition.
(c)SolarRadiationEffec
t . -The
downstream face of a dam, and the upstream
face when not covered by water, receives an
appreciable amount of radiant heat from the
sun. This has the effect of warming the
concrete surface above the surrounding air
temperature. The amount of this temperature
rise above the air temperature was recorded
on
the faces of several dams in the western
portion
of the United States. These data were then
correlated with theoretical studies which took
into consideration varying slopes, orientation
of the exposed faces, and latitudes. The
results of these studies are presented in
reference [ 11 .l These theoretical temperature
rises due to solar
radiation should be corrected by a terrain
factor obtained from an east-west profile of
the site terrain. This is required because the
theoretical computations assumed a horizontal
plane at the base of the structure, and the
effect of the surrounding terrain is to block
out some hours of sunshine. This terrain
factor will vary with elevation and from
abutment to abutment.

Numbers in brackets refer to items in bibliography, sec.

7-31.

116
(d) Amplitudes of Concrete
Temperatures.-

Therangeoramplitudeofconcr
e t e temperatures is determined by applying
the above-described external sinusoidal air
andwater temperatures to the edges of a
theoretical flat slab, the width of the slab
being equal to the thickness of the dam at the
elevation under consideration. The problem is
idealized by assuming that no heat flows in a
direction normal to the slab. The law of
superposition is used in that the final
amplitude in the concrete slab is the sum of
the amplitudes obtained from the different
sinusoidal variations.
To apply the theoretical heat flow in a
practical manner, unit values are assumed for
the several variables and a curve is drawn to
show the ratio of the variation of the mean
temperature of the slab to the variation of the
external temperature. Figure 7-1 shows the
relationship thus derived for temperature
variations in flat slabs exposed to sinusoidal
variations for h* = 1.00 square foot per day, a
period of 1 day, and a thickness of slab of II .
A correlation equation is given to take into
account the actual thickness of dam,
diffusivity constant, a n d p e r i o d of time.
The computations are shown in figures 7-2
and 7-3.* For the actual thickness of dam, I,,
a value of I, is obtained from the correlation
equation for each of the air temperature
cycles.
For each value of I, , a ratio of the variation of
mean concrete temperature to the variation of
external temperature is obtained. The sums of
the products of these ratios and their
respective
amplitudes are algebraically added to and
subtracted from the mean annual air
temperature to obtain mean concrete
temperatures for the condition of air on both
faces. Mean concrete temperatures are then
obtained in the same manner for a fictitious
condition of water on both faces, and the two
conditions are simply averaged together to
obtain the condition of air on one face and
water on the other. Solar radiation values are
then added to obtain the final range of mean
concrete temperatures.
2 These and sev~al other figures and tables in this chapter
were reprinted from Bureau of Reclamation Engineering
Monograph No. 34, listed as reference ] 11 in the bibliography,
sec. 7-3 1.

DESIGN OF GRAVITY DAMS

7-13. Temperature Gradients.
Temperature distributions in a mass where
boundary conditions vary with time are easily
determined by the Schmidt method. (See
references [ 11, [23, [3], [4] .) This method is
generally used for temperature studies of
mass concrete structures when the
temperature gradient or distribution across the
section is desired. The depth of freezing, and
temperature distribution after placement are
typical of the solutions which can be obtained
by this step-by-step method. Different
exposure temperatures on the two faces of the
theoretical slab and the autogenous heat of
hydration are easily taken into consideration.
An early objection to the Schmidt method
of temperature computation was the time
required to complete the step-by-step
computation. This has been overcome by the
use of electronic data processing machines
which save many man-hours of work.
Programs have been developed which will
take into consideration any thickness of
section, varying exposures on the two faces of
the slab, variable initial temperatures, a
varying heat of hydration with respect to time,
and increasing the thickness of slab at regular
intervals as would occur when lifts of
concrete are placedon previously placed lifts.
Ase c o nd m e th o do ft e mp e ra t ur e
computation in mass concrete which is
particularly adaptable to thick walls and
placement lifts near the rock foundation was
devised by R. W. Carlson. This method is
described in reference [ 51. It, like the
Schmidt
method, is essentially a step-by-step
integration
which can be simplified by selection of
certain
variables. Conditions such as initial
temperature distributions, diffusivity, and
adiabatic temperature rise must be known or
assumed. Carlsons method can also be
modified to take into account the flow of heat
between different materials. This would be
the
case where insulated or partially insulated
forms are used, or where concrete lifts are
placed on rock foundations.
Thevariationintemperaturein
a
semi-infinite solid at any particular point can
also be estimated from figure 7-4. This
illustration gives the ratio of the temperature