-f-.-. .-_ _- . __, _- a -
M_ AM,_:".‘ ”-6-" ._-s *
O-J 'd‘ o ‘ _

HH H UH \l \ IH

  
  
  
  

 

THE EFFECT 0E ADHIXTHRES ON
THE THERMAL EXPAHSICH
0E CCHCRETE

THESIS m Innis“: or a. s.

W

JOSEPH C. LYNCH
1940

THESTS

 

 

llT t\l:..E Eu!

. iron.»

I! HIEETETEH

II

 

The Effect of Admixtures on the
Thermal Expansion of

Concrete

A Thesis Submitted to

The Faculty of
Michigan State College
of
Agriculture and Applied Science

by

“7
I t

Joseph Ctlézhch
Cendiflate for the Degree of

Bachelor of Science

December 1940

~ To the Civil Engineerfiyaculty

at Michigan State College,who have
ably assisted me in gaining my
education in Engineering,1 do
hereby dedicate this thesis on

thermal expansion of concrete--J.O.L-

Preface

Acknowledgement is made to the
Mr. Rothgery of the Civil Engineering
department and to Mr. Finney or the
Michigan Highway department for their
advice and kind criticism offered in

the develOping of this problem

 

1.
2.

5.
4.
5.
6.
7.
8.
9.
10.

Table of Contents

Page
IntrOdUCtion -‘Odnauuw-------‘-0-¢¢-n~-¢ vvvvvv g. 1

Materials --- ...... -------—---------—---------

A. Aetna Portland Cement--~-----.------------- 4
B. Aggregates --------—------—-~---- ...... ---- 6
0. Weight of Aggregates -- ....... -------—---- 11
Method of Preportioning ----~----------------- 13
Apparatus --------—-------------_ ...... ------- 15
Admixtures --————--------- ......... ----;------ 16

Experimental Procecdureu--~----------—a--~-~-- 19

Experimental Data Inc-nucoac-O—a-mbuaunmui— ...... -- 21
calwlation onuaoa-up-u—ao—n-u—oo-anon-omuuusn—u 25
CODOIUSIOD -Cfl-Wflfifl---~-O-u~~--lfluu-00------—-.§- ?4

Bibliography ---- ----------------------------- 95

Introduction

There is a keen interest in the engineering circles
at the present time in the effect of powdered admixtures
in Portland Cement concrete. Many types of inert admix-
tures have been used.designedly or due to their presence
in the aggregate. Admixtures have Various effects on the
preperties of concrete some of which may be more or less
benefical. An admixture should be considered only as one
way to secure desired results and its use should be com-
pared with other methods of obtaining similar results.

An increase in water content or in the porportion
of paste, which commonly attends use of admixtures, may
be expected to increase shrinkage. That a reduction in
water—cement ratio reduces volume change, 1.6., shrinkage
both during hardening and setting is less.

The earliest work recorded concerning the linear
thermic expansion of concrete is due to Bounioea, who
published his results in the "Annales des Ponts et
Chausses," in 1865. His work was performed on rectangular
prisms 65 to 94 inches long and about 7 inches on each
side, the blocks being placed in water whose temperature
varied from to to 95 degrees C. The apparatus used was
checked by measuring the determined coefficient of ex-
pansion for other materials. His results for concrete

(proportion not given) was 000000795 inches / degree F.

Professor W. D. Pence of Purdue University has made

a series of investigationswith Portland Cement concretes.
The values obtained with l cement to 5 gravel was . 000005SR/
degrees Fahrenheit. Other mixes are:
1-13-5 = .00000677
1-2-4' = .0000056
1-2g-5 3.00000558
1-5-6 :.00000557
The resistance of concrete to frost is of much
importance. hater expands abouti9peroent in volume
during freezing and when it is trapped in cavities the
pressuresthat may be exerted are very large. In a wet
concrete the water enclosed in the pores of the material
tends on freezing to force the particles
of Mortar apart or to set up severe internal stresses.
The resistance of concrete to freezing depends on
the density and impermeability of the material, for if
water can not penetrate into it no damage can be produced.
A neat Portland Cement on heating first expands
owing to the normal thermal expansion. This expansion
is Opposed, however, by a contraction due to the shrinkage
of the material as water is driven off from it. The con-
traction due to drying eventually becomes much larger than
the normal thermal expansion and the material then commences
to shrink. The actual temperature at which the maximum
expansion is reached varies with the size of the specimen

and the conditions of heating.
-2-

The foregoing considerations have led to an invest~
igation of the subject. First a method of making such
measurements was developed, then, experiments were made
to determine the effect of admixtures on the coefficient 0*

expansion.

Materials

The materials used consisted of Aetna Portland
Cement, sand, coarse aggregates, water, and nine ad-
mixtures.

The fineness of grinding of Portland Cement was deter-
mined. This is in terms of the percentage of cement
retained on a standard No. 200 sieve. A 50 gram sample
of cement was placed on a N6. 90 sieve and then placed
in a N6. 200 sieve. Hold the sieve in an inclined direction
and moved back and forth, tapping the sides of the sieve
about 150 times a minute. Every twenty-five strokes
rotate the sieve one sixth of a revolution. After several
minutes of shaking, remove the cement. Brush the pan to
make sure of getting all the particles. The percentage
of residue on the sieve with a correction factor for the

sieve applied, expresses the fineness of cement.

Data Sheet

I II
Trial
Material Portland Cement
Date 11/ 16/ 4o 11/ 16/ 40
Air Temperature 74° 74°
Method of Shaking Hand ' Hand
Weight of Sample 50 50
Eeight Retained on 200 # 88.09 V8.01
Weight Passing 200 # 41.91 41.96
height Lost 0.C0 0.05
a Lost 0.00 0.05
% Retained 16.18 16.02
Fineness ‘ 16.18 16.0?

Sample Calculations

Wt. lost 8 wt.passed 4. wt.retained_.wt.of sample
3 41.91 i 8.09 - 50 3 0.00 grams
% lost 2 Wt.lost x 2 : 0.00 x 2 - 0.00

% Retained a Wt.retained X 2

2 8.09 X 2 ' 16.18

Determination of Sieve Analysis of the Aggregates

Select a representative sample of the fine or coarse
aggregate by quartering a larger sample. Dry the sample
so as to have not less than 500 grams of fine or sand and-
about 5000 times the size of the largest sieve required
for the coarse aggregate.

Than arrange the sieves in order and placing in
sample of sand in the sieve. It was than placed in the
shaker for fifteen minutes and the amount retained on
each sieve was weight.

It was not possible to use the shaker on the coarse
aggregate so they were shaken by hand and the retained
amount was weight.

The percentage retained on each sieve for the sand
and coarse aggregate was than calculated. From this data
the fineness modulus for the sand and coarse aggregate

was found.

Data

Materials

Date

Method of Shaking
Height of Sample Used

Sieve Num~ Diameter

‘ Fineness Modulus

Sheet

Sand
11/ 15/40
Machine

510 grams

Total Wt. Total per Total per-

ber or Opening Retained cent Re-
tained
1.5 0 0
. 749 0 0
. 571 0 0
4 .135 7 7.5
_s .095 105 20.6
Y 14 .046 267 52.4
..28 ‘.0952 590 76.5
48 .0116 502 98.5
‘ 100 .005 508 99.4
. Totals ‘ 610 554.7

3.547

cent pass-
ing

100
100
100
92.6
79.4
47.6'
25.5
1.5
0.6

Data Sheet

Materials Fine Aggregates
Date 11/ 9/ 40
Method of Shaking Hand

Time of Shaking

Weight of sample used 2990 grams

Sieve Numr Diameter Total Wt.Tota1 Per- Total per-

ber of Retained cent 36- cent pass-

Opening tained ing
1.5 0 0 100
.74? 495 14.2 85.8
.371 9668 89.0 11.0

4 .185 2983 99.5 0.5

8 2990 100 0.0

14

28

.48

100

Totals 2990 702.7

Fineness Moduhus 7.027

Data Sheet

Materials Cenrse Aggregates
Date V 11/9/ 40
Method of Shaking Hand

Time of Shaking
height of sample 4420
Sieve Num- Diameter of Total Wt. Total fitb- Total per-

ber Opening Retained cent cent
Retained Passing

1.5 2277 6.5 93.7
.742 4312 97.5 ' 2.5
.571 4418 99.9 0.1

4 .185 442 100 0.0

8 .095

14 .046

28 .0252

48 .0116

100‘ .005

Totals 4420 803.7

Fineness Modulus 8.067

Data Sheet

Materials 50% Fine and 50% Coarse Aggregate
Method of Shaking Hand
Date 11/ 9/ 40

Time of Shaking

Weight of sample

Sieve Num- Diameter Total Wt- Total per- Total per-

ber of Retained cent cent

Opening Retained Passing
1.5 158 5.15 96.85
.742 2448 55.85 44.15
.571 4140 94.55 5.45

4 4280 100 2.25

8 4580 0.0

14

28

48

100

Totals 5705 755.2

Fineness Modulus 7.552

Note: The aggregate used in experimenting was made up

of 50% fine and 50% coarse aggregate.

To determine the unit weight of the aggregate

for concrete.

For aggregate whose maximum diameter 13%; inch or less,
use the 1/10 cubic foot measure. For aggregate whose
maximum diameter is over 1% inches in diameter use the
1 cubic foot measure. The sample of aggregate should be
room dry and thoroughly mixed. Room-drying at aggregate
should be done only until the surface moisture has dis-
appeared, inasmuch as;these values are used on that basis.
Fill the measure one-third full of aggregate and level
off with the fingers. Tamp the aggregate with the pointed
end of the tamper 95 times. Fill two-thirds full, now,
level off and temp. Fill the measure full to overflowing
and tamp, strike off the surplus with the tamping rod.
Determine the net weight of aggregate in the container.
From this you can determine the weight of the material

per cubic foot.

-11-

Data Sheet

Material Sand Coarse
(50-50)
Date 11/ 2/ 40 11/2/40.
Size or container 1/10 cu ft. 1 cu ft.
Weight of Container 2.62 # 12.05 #
Weight of material and container 13.80 # 123.07 #
Weight of material 11.18 # 111.02 #
neight per cubic foot 111.8 # 111.02 #

Sand: neight/ cu.ft.= ?t. material x 10 - 11.18 x 10 :
111.8 #

Coarse: Weight/ cu.ft.: Wt. material X 1 111.0? X 1 8

111.0? #

-12-

Method of Preportioning
a) Estimate a mix. that will give a ' given strength. Try
a 1.5 mix (combine fine and coarse)
1))Figure 5-4: page 89 Plain Concrete —- Bauer. Maximum
permissible fineness moduli is 5.75
c) Check in figure 5-3 Strength is low £3350 p.s.i try
mix 1:4.7 Strength is 2550 p.s.i. which is all right.

P percentage fine aggregate

fineness modulus of coarse

A
B 8 fineness modulus of combindi
C

3 fineness modulus of sand.

Percentage of coarse aggregate is equal to 100 minus the

percentage of fine aggregate

P - 7.53.9 .. 5.80 x 100 _ 1.758 x 100 = 45.7
WWW .. 5.547 15.985

percentage of coarse is 56.3%

d) Find weight per cubic foot of sand and aggregates.

Sand 3 111.8 d
Gravel 3 111-02;
Combination : 181.6 #

Calculations of Weight

Tatal 2 111.0

Shrinkage factor a 111.0 g 91.5
121.6
437 a: 5015
91.5
-13...

New proportion 1:5.15
CUbiO feet 88nd. 5015 X 4507

2.25 cubic feet

Eeight sand - 2.25 X 111.8

252; / cubic foot

Cubic feet coarse aggregate 5.15 X 56.5 3 2.80 cubic feet

Weight coarse aggregate 2.8 X 111.0 3 511 # / cu.ft.

Absolute Vol-

 

Water one cubic foot cement = .49
9.25 cu. ft. sand 2,?5 X 111.8 - 1.52
62.4 x 2.65 ’
2.80 cu ft. coarse aggregate
9.80 X 111,0 - - - 1.88
2.4 X 2.65 ‘
Total 4079

Weight materials per cylinder

 

Weight cement 3 .95 X 94 = 4.9 a
4.79
Weight water = .25 62.4 g 5.26 fi
4.7§“‘ X
Weight sand 3 .9 X 252 3 15.9 i
4.
height coarse aggregate 3 .25 X 511 ' 16.55
4.79
Proportions
Cement ' .212 g 1
4.9
4.9
Coarse Aggregate “ .lQLQ = 5'55
4.9

1 3 2.7 3 50

(A

-14-

Apparatus

The apparatus or it shall be called a comparator
was made for the experimenting. It consists of a resting
point for the cylinders and a federal dial to measure the
height of the cylinder at the start. The comparator is

better explained by the accompaning picture.

 

 

 

 

» ‘»-—-~, __

—— -—_.,«_1
‘“ --___

-15- a

Admixtures

The admixtures used in the experiment were the following:
Silica dust, Lime stone dust, Bentonite, Plastiment, Fly
Ash, Pozzolith, Orvus, Vinsol Resin, and Natural Cement.

Silica occurs in nature in all of the polymorphous
forms. The form that is most used is quartz. This raw
silica as brought from the quarry or mines requires to be
crushed or ground. This ground material is made extreukiy
fine and is called silica dust. Silica dust is used in the
ratio of 15 #/ sack of cement. The amount of silica
dust required for onccylinder is:

__y_s_ : sexes “.784 i?
94 ,
Limestone occurs in nature in the form of rocks.
These rocks brought from a mine are ground extremely fine.
The dust from the grinding is used as limestone dust. Lime-
stone dust is used in the ratio or 15 i / sack of cement.
The amount of limestone dust required for onecylinder is:

15 a .16 X 4.9 3 .784 #
94 ,

Bentonite, an extremely fine-grained clay formed by
the weathering or wind blown Volcano ash, was originally
observed in Fort Benton shales of the upper Missouri Valley,
In outward appearance, bentonite resembles a grey mud, -
but the very colloidal types exhibit considerable swelling
characteristics on additions or water. The ratio of bento-
nits is 5 #/ sack of cement. The amount required for one

-16-

cylinder is 5 z .0534
94
.0554 X 4.9 3 .262 §

 

Plastiment is an extremely fine grained material and
has a reddish-brown color. The ratio of Plastiment is l# /
sack of cement. The amount required for one cylinder is:

l P .0107

4
.0107 X 4.9 = .0524 #

Fly Ash, an extremely fine material, made from pul-
verized coal ash was used in the ratio of 15 # / sack of
.cement. The amount required for one cylinder is:

94 ’ .
.16 X 4.9 : .784 1;"

Puzzolith is a manufactured material which is produced

by the Master Builders. It is an extremely fine grained

material and is grey in color. The ratio used was 2# /

sack of cement. The amount required for one cylinder is:

7531“ 3 .0905
.0?05 X 4.9 3 .099 #

Orvus is a type of soap which has an extremely great
amount of carbon in its molecular formula. This soap will
act the same in salt water as in plain water. The ratio
used was .015 § / sahk of cement. The amount required

for one cylinder is .015 : .00016 #
94

.00016 X 4.9 2 .00078 #
The soap was mixed with water in the given amounts and 68

ml. was used.

Resin is either a natural or synflthetic product. The
resin used in this experiment was synflthetic Vinsol resins.
It is . light grey in color. The ratio used was full
amount. The amount required to make one cylinder is 4.9 #.

Natural cement is of two types:

1) With out grinding aid and 2) with grinding aid. The trade
name of the natural cement used was a Luxment and it is

made without grinding aid. The ratio used was i. sack to

6 sacks of cement. The amount required to make one cylinder

is: 4.9 g .897 £-

m

6

Experimental Proce¢dure

The mixing of the materials was done by hand. After
the sand, cement, coarse aggregate and water Um-acarefully
measured; the cement and sandwumr mixed first dry. Then the
coarse aggregate was spread over the cement-sand mixture and
remixed again. A crater was formed in the center, into
which the water was poured. When the water was absorbed,
the mixing was proceeded until the concrete is homegeneous
in color. Upon completion of the mixing, cylinders (6" x 12")
were filled in three increments, rodding at each increment
25 times with a round rod and bullet-pointed at the lower
end. Distribute the strokes in uniform.manner over the
cross section of the mold. Care is taken to place sufficient
strokes in order to prevent honeycombing. After the top
layer has been rodded, the excess is struck off. Steel
plugs i" in diameter and about 1" long mm inserted in the
ends of the cylinders before placed in the moist room for
curing. The samples were left in the molds for not longer
than twelve hours.

The conditions for curing the cylinders were the same
in all cases. There was a relative humidity at 100% and
the temperature of the moist room varied from.68° to 69° F.
They were left in the moist room for seven days before an
inital reading was taken. The test cylinder was then placed
in the electric oven, for twenty-four hours, at a temp-
erature of 110° C and again measured. They were then taken

to a refrigerator for twenty-four hours at a temperature of
-19-

_ 10° F before another reading was taken. The cylinders
were then allowed to stand at room temperature for 24 hours

before a compression test was made upon them.

 

Name of Admixture Used

None (plain)
Silica Dust
Limestone Dust

Bentonite

Plastiment

Fly Ash
Pozzolith
Orvus
Resin

Luxment

Experimental Data

ODE-‘Nli-‘PJHMHFQHNHI‘OHMHMHNH

Inital
Reading

50/1000
51 1000
60 1000
65/1000
59/1000
57/1000
40/1000
43/1000
48/1000
47/1000
52/1000
50/1000
45/1000
45/1000
69/1000
68/1000
50/1000
50/1000
55/1000
52/1000

Hot

61.5/1000
62.4/1000
71.5/1000
74.5/1000
70.6/1000
68.6/1000
52.5/1000
55.5/1000
59.6/1000
58.6/1000
6:5. 8/1000
61.5/1000
64.2/1000
56/1000

80.4/1000
79.4/1000
51.2/1000
61.2/1000
57/1000

65.9/1000

Cold

45.8/1000
44.9/1000
55.9/1000
56.9/1000
52.8/1000
50.9/1000
55.4/1000
56.5/1000
41.8/1000
40.9/1000
45.8/1000
45.8/1000
57/1000

59.1/1000
62.9/1000
63/1000

44.1/1000
44/1000

48.6/1000
45.7/1000

Change

17.6/1000
17.5/1000
17.6/1000
17.6/1000
17.8/1000
17.7/1000
19.1/1000
19.2/1000
17.8/1000
17.7/1000
18.0/1000
17.8/1000
17.2/1000
15.9/1000
17.5/1000
17.4/1000
17.1/1000
17.2/1000
18.4/1000
18.2/1000

Name of Admixture Used Compression Coefficient of

Strength Expansion
NODG 1 2600 p.801. 066666610
2 2605 p.s.i. .00000605
Silica Dust 1 9595 p.s.i. .00000611
2 2600 p.s.i. .00000611
Lime Stone 1 ?585 p0 8010 000000618
2 2580 p.8.i. .00000615
Bentonite 1 2590 p.s.i. .00000664
2 2510 p.801. 000000666
Plastiment l 2580 p.s.i. .00000618
2 2885 p. 3 c 1. 0 00000615
Fly Ash 1 2570 p.s.i. .00000626
2 2580 p. 30 10 000000618
Pozzolith l 2710 p.s.i. .00000596
2 2705 p.s.i. .00000591
OPVUS 1 2 30 p.3.1. 000000807
2 2625 p.s.i. .00000605
Resin 1 2715 p.s.i. .00000594
2 2710 p.s.i. .00000597
Luxment l 2540 p.s.i. .00000658
2 2550 p.s.i. .00000652
1

CEICUlation

110° 0 to ~10° F

 

K - e
'1"'x""'t 110° 0 a 250° F
K a coefficient ' (110° x 1.8) + 52 a
’ 250
e a elongation
l 1: length
t : temperature change 2500' to - 10° F
1) .0000061 : e e = .0176 '
2 X 240 '
K: 7.6
1000 X 12 X 940 ‘ '1756

2880660 3

.0000061 inches/ degree F

Conclusion

The applicability of these conclusion is limited to
concretes under conditions similar to those of this in-
vestigation with regard to type and proportions of materials,
fineness of cement, and the curing condiflions.

For each admixture two individual molds were made up
of the same mix and under the same curing and testing
conditions. The slight variations in the two cylinders is
probably due to the loss of heat or coldness in testing.
In general concrete mortar is subject to expansion, and
there are a number of admixtures on the market today which
supposedly reduce this expansion. But from the experimental
data none of the admixtures used gave any great effect in
the coefficient of expansiongmhe admixture that gave the
greatest expansion was Bentonite, while Pozzolith and
Vinso/Resin gave the least expansion. The experimental data
also shows that with a low coefficient of expansion
there is a high compression strength produced. From this
it can be said that the expansion causes tension in the
mortar rather than compression so that the bond is weakened.

0f the nine admixtures used only a few of these reduced
the coefficient of expansion , and even then, they do not .

completely eliminate the expansion.

-27..

1.

2.

4.
5.

7.
8.
9.

10.
ll.

15.

Bibliography

Cement and Concrete ----¢--L.c, Sabin

Practical Cement Testing 6-~---W.P. Taylor

Cement— Mortor- Concrete -—-----M.S.Falk

The Chemistry of Cement and Concrete ---F.fl. Lea
Portland Cement Association Information ---

Sheet Kc. St 24

structural Material Research Laboratory --—- D. A. Abrams
Concrete Its Manufacture and Use ------ Koehring

Plain Concrete_ ------¢~----------~-- E.E. Bauer

Master Builders Booklet less Water yet More

placeability
The properties of Silica .------------- R. B. Seaman
Properties of Mortars and Concretes Containing
Portland Pozzolan Cements --- R.E.Davis,
J.W. Kelly, G.E.Troxell, and H.E. Davis
Journal of American Concrete Institute
Volume 8 Mayv June 1937

Journal of American Concrete Institute

Volume 9 July- August 1957

'17

9 J 1
.T
w ......
I
S .. \
u.
, .1
u. . _
.‘t-

 

 

        

' ‘ ' _ . ‘-
é‘ ‘ E
l
I - .
_ . :2 3—..- -_'4-E._- H- v- A . _

M

314