ect i^f Heat on the
' 1 1
■ 1 Q^-^ .-^^v-^-'
Digitized by the Internet Archive
EFFECT OF HEAT ON THE STRENGTH OF
Herbert Clinton Beck
Bachelor of Scienci:
COLLEGE OF ENGINEERING
UNIVERSITY OF ILLINOIS
UmVERSITY OF ILLINOIS
May 25, 1911
I recommend that the thesis prepared under my supervis-
ion by HERBERT CLIFTON BECK entitled Effect of Heat on tne Strength
of Cement Mortar be approved as fulfilling this part of tne require-
ments for the degree of Bachelor of Science in Civil Engineering ►
Instructor in Civil Engineering.
Head of the Department of civil Engineering,
Ei^TEOT OF IlEAT ON TIlE STRENGTH OF CSiO]NT MORTAR
Table of Contents.
Conditions of Tests, 2.
Discussion and Interpretation, . 4-.
Tables and Plates, 7-15.
EFFECT OF HEAT ON THE STRENGTH OP CEl^OT MORTAR.
Intro ducr. lon.
¥lth the present extensive use of mortar and concrete as
building materials and as flreprooflng materials their action un-
der high temperature has hecome of much Interest. If a concrete
structure he subjected to the Intense heat of a fire It Is Impera-
tive to Know to what extent Its walls and floors have been damaged.
The object of the tests herln described was to discover
If possible the Initial temperature at v/hlch mortar begins to lose
its strength and also to discover whether or not there Is a defi-
nite relation between loss of strength and temperati:re. as a
matter of convenience In heating '.ne tests were made on small
cubes of cement mortar rather tnan on larger ones of concrete.
When the tests were started practlcailly th' only ex-
periments along this line were those performed by Mr. I. H. Wool-
son In 1907. These experiments are of much value but leave
som^ethlng to be desired, since temperatures below 700° F. (4-00° 0. )
are not considered and the variation of age of the test pieces Is
small. Since the beginning of the tests described in the follov;-
Ing page several articles on the subject have been published. The
most notable of these are; a paper entitled, "Some Thermal Proper-
ties of Concrete," by Mr. Chas. L. Norton of the Massachusetts
Institute of Technology (Proceedings of the National Association
of Cement Users), and papers by l/ir. ifoolson before the American
Society for Testing Materials. Further reference to these exper-
iments and comparison with the results there obtained will be made
in the following discussion.
The experiments herein described consist of series of
tensile and compressive tests on "briquettes and cubes respect-
ively. iSeveral sets of each v/ere m.ade. Half of each set was
heated before testing and half were tested without heating. The
age of the test pieces and the degree of temperature ivere varied
In the different tests.
It was thought probable that there would be a somewhat
sudden drop In the strength of the mortar at the temperature at
which the v;ater of hydration was given off but the results do not
seem to justify this belief.
Conditions of T ests.
The worKlng conditions for the tests here described will
now be given In detail so that in comparing the results ?/lth those
of other experiments the personal element may be reduced to a mini-
The cement used, Chicago AA Portland, was of good quali-
ty, fine, and of moderate rate of setting. It was of uniform
grade as shown by Vlcat tests of plasticity. The sand was natural
banK deposit commonly Known as Wabash sand. It was screened to a
maximum size of one tenth of an Inch. Plate I shows a sieve anal-
ysis curve for the sand.
The test pieces consisted of ordinary briquettes and
3-ln. cubes, made of mortar or normal plasticity. The briquettes
were compacted by pressure of the fingers, but for the cubes a
wooden rammer was used.
The test pieces v/ere Kept In the molds under damp cloths
for twenty- four hours and were then placed in wnte^.v, ThPi .spnnl-
mens were dried in air at room temperature for one weeK before
In the tests the specimens were heated in a gas drying
oven and for the higher temperatures in a gas Kiln. In no case
was the flame allov/ed to touch the pieces. The temperature was
measured with a mercury pressure thermometer and was raised at
rate of about 100 C. in ten minutes. The maximum temperature
was maintained for one hour and the specimens were then cooled
slowly to room temperature, after which they were tested for
strength, either at once or after standing for some time.
The compression pieces were bedded in plaster of Paris
to insure a plane bea^^ing surface. The load was applied with a
Riehle testing machine of one hundred thousand pounds capacity.
The load was centered by a ball bearing head and the specimen was
compressed at the rate of l/l6 inch per minute.
Each set of cubes consisted of six specimens and each
set of briquettes of ten specimens.
Interpi^tatlo n and Pis cu ssl on of Tests.
One of the points V7hlch these experiments "bring out Is
the lacK of uniformity of results of tests on heated specimens, or
the unreliability as regards strength of mortar which has been
heated to a high temperature. This effect of heat Is shown In
both the tension and the compression tests but appears to be the
greater In tension. The variation of the Individual specimens
from the mean of the set Is much greater tn the heated than In the
unheated sets. The value of mortar or concrete as a fire-proof
building material therefore probably lies In the low thermal con-
ductivity of the material rather than in Its frequently supposed
Immunity from loss of strength by fire.
Prof. iToolson's tests show no reduction In strength for
a temperature of less than 750 F. (4-00 c). The experiments
performed by the writer shovi a reduction for lov7 temperatures for
ages of four weeKs or under, although above this age the loss in
strength is not great. Prof, woolson's experiments on limestone
cubes showed that the cubes disintegrated on standing after heating
The tests described In this thesis, performed on 1 - 3 mortar vary
somewhat from this rule. For low temperatures there was a gain in
strength on standing after heating, the specimens being weaKest
while still warm. For higher temperatures the specimens were wealc
while warm, gained slightly In strength immediately after cooling,
and then gradually lost strength with increasing age.
For cubes over twenty-eight days old there was little
reduction of strength for temperatures under 300° 0. Above this
age the strength in some Instances was increased by application of
heat at low temperature. For ages over twenty-eight days and
temperature between 300° and 500° C. the loss In strength was very
marKed, the reduction being as high as 20 per cent In some cases.
It Is probable that above forty-two days an Increase In the ago
v^ould effect the results very little, while an Increase In temper-
ature over 500 0. would cause a marKed reduction in strength.
In no case did the heating cause cracKs In the specimens
or otherwise change their appearance. When heated to temperatures
above 300 C, hov/ever, the specimens when strucK together gave out
a metallic sound llKe that made by well burned brlcK. One Interest-
ing feature of the experiments was the apparent change In Internal
structure of the heated cubes. Instead of being somewhat crystal-
line as In an unheated specimen, the Interior of the heated cubes
seemed to have disintegrated, leaving a surface from which granular
particles could be crumbled as from rotten sandstone. It Is sug-
gested that this may account for the loss of strength, the varying
amounts of thermal expansion of the components of the mortar des-
troying the bond between the unit particles.
In Interpreting the results of these experiments it
must be remembered that the tests are of necessity too few in
number and cover too limited a range of temperatures to prove
definitely the conclusions given above. The short time of
heating also may give results from which wrong conclusions
might be drawn. However, the results may safely be taKen as
indicators of the probable action of mortar or concrete when
subjected to high temperatures.
TABLE I. COi^iPKESSIVE STRMGTIi OF 3-IN. CUBES - HEATED AW UN-
1-3 Portland Cement Mortar.
1 — ■
, Unheated Cubes
It . per
Vo . per
sq , In.
s t re ng t h
sq . In.
T "7 c r\
1 ( "7 r\
TABLE 2. C0.\iPRE8SIVE STRENGTH OP 3-IN. CUBES HEATED.
1-3 Portland Cement Mortar.
c c m p r
e s s 1 V ?
1/2 day after Heating
7 days after heating
TABLE 3. COiAPRESSIVE STRENGTH OF 3-IN. CUBES. HEATED' AND UimEATED.
1-3 Portland Cement Mortar.
Unheated Cubes. '
C en t .
^Heated cubes tested before cooling.
TABLE 4-. TENSILE STRENGTH OF STANDARD BRIQUETTES - HEATED AITO
1-3 Portland Cement Mortar.
J IC II
T emp *
S t r e
n g t h
T . "I r"i
Q f" <ar)
U IlilC d CU.
1 >i 10 V
1 "K "n ra'io
X U . CI
N.B. Each value given Is the mean or five tests.
u. or I. 8. s. roHM 3
u- or I. 8. %. rOHM 3
L'- OF I. 8. S. rONM 3
U. Of- I. ti. S. FOHM a
U ur I S. S. FORM 3