113
043
THS

 

A STUDY OF THE EFFECT OF

DIFFERENT ADMIXTURES ON THE

WORKABILITY OF CONCRETE

Thai: for thn Dam of B. 3.
MICHIGAN STATE COLLEGE

William E. Johnson
1947

 

A STUDY OF THJ EFFLCT
OF DIEFARANT‘ADLIXTUHnS 0H Th; WORLABILITY

<33 001703; T411

‘3 Thesis Submitted to
The faculty of
MICHIGIQI 5314.11; COIL-LG;
of
AGRICpLTURE AID AEELILD SCIQICE

n2

‘William.3. Johnson
h
Candidate for the Degree of

Bachelor of Science

August 1947

A» m: or 00131141: :5

Ia”
Acknowledgment; ------------------------ --- ..... 1.
Introduction ----------------------------------- 2,
Discussion .................................... 5,
Erocedure -------- - ............................ 9;
Test Data and Curves -------------------------- 12.
Conclusion --- -------- -- ----- ---- ----------- -—— 27.
Eibliogreghy --------J-------- ------- ---------— 29.

110111-10: .1..41o;.;_;:z T

The author of this paper takes this ollortunity
to thank lrofessor C.L. Allen, head of the Degartment
of Civillngineering at Michigan State College, and G.C.
westfall, Instructor in Civil Lngineering at Ziohigan
State College for their coogeration and assistance in

the prelaration of this thesis.

William.E. Johnson

III TRCDUC T101?

Introduction

It was the purpose of this study to find the effect
of several of the common admixtures on the wcrkability
of the concrete mix. the number of admixtures used, however,
was limited by the time permitted for tle preparation
of this report.

‘With different text books and authorities the concept
of an admixture varies. Therefore,to make clear what is
meant by an admixture in this report, a definition will
be stated that agrees with the author's concept. "An
admixture is any material added to a concrete or mortar
mix other than water, cement, fine aggregate, or coarse
aggregate to impart a primary desired condition." In
some cases, however, one desired condition may be sacrificed
to a small extent to favor another..An example of this
would be adding an air—entraining agent to increase
percentage of air voids, but perhaps, decreasing strength
slightly.

Admixtures for use in concrete are classed on their
ability to promote certain basic actions in concrete
mixes. These classifications are as follows:

1. Dispersing Agents

2. Accelerators

3. water Proofing Eastes
4. Elasticizing Agents

5. Finely Divided Eowders

6. Air Bntraining Agents

7. Retarders

8. Miscellaneous

At one time a number of materials classed as admixtures
for concrete were widely promoted as being capable of
ridding concrete of all kinds of ills, and as a result
admixtures became little thought of as being useful
because few of the materials actually produced what was
claimed in their favor. Recent advances in the development
of admixtures and in the understanding of their proper
use have, however, led to the acceptance of the fact that
if selected properly and used in the right proportions,
they offer a satisfactory means of producing specific,
desired results. A knowledge of what an admixture will or
will not do should permit design of concrete for the desired
end to the best advantage.

For this report five admixtures were selected for
investigation as to their effect on workability. The
admixtures used were all common ones used to promote
varied primary desired conditions. ho attempt, whatsoever,
was made to test these admixtures for any other conditions
that they might create. In this experimentation the only
objective in mind was the effect upon the degree of work-
ability.

Also, no distinction was made as to qualification
in any certain class. The first two, Vinsol resin and

Orvus, are known chiefly as air entraining agents. The

4.

next, Eozzolith, Is a dispersing agent. The accepted
use for Calcium Chloride as an admixture is for decreasing
the time of set. Hydrated Lime is known mostly as a

plasticizing agent.

if”
ioC'

VIHSOL BASIN

Vinsol resin is a by-product of the naval stores
industry, and was originally used as an insoluble resin
interground with the Portland cement clinker. Vinsol resin
may, however, if neutralized with a caustic solution,be
added at the mixeras well as ground with the cement clinker.
It is added by either method usually in amounts varying from
one to three one hundredths of one percent. In this research
the neutralized Vinsol resin was added in varying amounts
up to one percent to find if there was any continued
effect on workability at the upper limits, realizing however,
that other conditions whollyunsatisfactory may be in-
troduced.

Vinsol Resin, either when intergrouad or added at

the mixer, tends to entrain a considerable amount of air
and causes the concrete or mortar to be more cohesive

or fatty..A difficulty withVinsol resin is that it causes

a variable behavior. The same percentage added to identical
mixes may produce widely different results in some cases.
This variation implies that the effectiveness of the Vinsol
resin withrespect to air-entrainment, bleeding, segregation,
fattiness,durability and strength will also vary. In

the present state of knowledge results show that these
variations are unpredictable. This means that the precise
desired properties are hard to obtain with.Vinsol resin.
The reason for this variation seems to lie in the fact

that the air-entraining action of the resin is

dependent upon neutralization or saponification of the
resin by the alkalies of the cement. This has led to
pre-neutralization of the Vinsol resin with a caustic
soda solution. While this measure has improved performance,
it has by no means eliminated variation. In any case
property variations depend upon‘degree of saponification.
ORVUS

Orvus is a product made by Proctor and Gamble Co.
Its greatest constituent is sodium lauryl sulphate. As
compared with Vinsol resin laboratory tests show that
addition of Orvus to the mix have similar effects on
most properties. Unlike Vinsol resin, it does not impair
strength, but shows an increase at all ages. Orvus is
a definite chemical compound produced synthetically,

and does not cause variability such as Vinsol resin.

IOZZOlITH

Pozzolith is a product made by Master Builder's
Company. It ia.a dispersing agent . Like the two air
entraining agents already discussed, it is an approach
to the problem of the improvement of properties of
Portland cement. Application of the dispersion principle
was developed in the early 1930's. It is today used
both in building construction and pavement work.

When lortland cement in mixed with.wataa the in-
dividual particles tend to gather together and stick to each

other; afloo is formed.This floc is due to lack of mutually

repellent electrostatic charges on cement particles. It

7.

is the purpose of a dispersing agent introduced into the
:mix to break up this floc, and makethe cement act as
individual particles.

Dispersion of the cement produces three important
effects. The water which.was trapped among the floc is
liberated to become part of the mixing water. The surface
area of the cement particles, through dispersion,becomes
greatly increased..A certain amount of air is entrained,
and thus, a dispersing agent also becomes an air-entraining
agent.

Dispersion, through those three important effects,
would indicate production of a higher strength concrete,
better workability, greater water tightness by making
possible a reduction in water-cement ratio, greater
durability with.respect to freezing and thawing and

resistance to corrosion.
CALCIUM CHLORIDE

Ca1cium.Chloride is used as an accelerator. Often
it is necessary to accelerate setting or hardening or both.
A.great many inorganic electrolytes will serve this need.
The most widely used which is typical of the group of
inorganic electrolytes is Calcium Chloride. This compound
is used because it is most economical and available.
It is commonly used in quantities up to three percent

of the weight of the cement.

8.

EXDRATED LIME

Oneof the admixtures most commonly used for improving
workability is hydrated lime. It is used in amounts up.
to five pounds per bag of cement. Contractors using
hydrated lime as an admixture report that they get,
especially with the loaner mixes, more uniform.concrete
from.batch to batch than when they do not use it. Because
of the extreme fineness of lime, it reduces to a minimun
the effects of small variations in amounts of mixing

water on the consistency of the concrete.

."E'ROCIZDUPE

9.

PROCEDURE

In a research problem.like this there can be only
two variables. One of the variables is one of the constituents,
and the other is a result which.may or may not vary.

In this report the water-cement ratio, the amount of

cement, andthe gradation of the aggregate was kept constant.
The kind of cement and aggregate was also kept the same
throughout. Only the admixture was varied.

Tswas decided at the beginning of the experimentation
to take five common admixtures and thoroughly investigate
their flOW'properties when varying percentages of admixtures
were added to the mix. For these tests the slump and flow
table were used according to A.S.T.M; method of tests.

For slump tests the fine and coarse aggregates were
combined in proportions so that the combination would
approach the ideal grading based upon Weymouthis theory
of particle interference. Several trials showed that this
ideal migfmost closely approached with a fifty percent
fine and fifty percent coarse aggregate combination than
with any other.

A mix of 135.5 by weight with a water cement-ratio
of eight tenths by volwme or fifty three one hundredths by
weight was used for the slump tests. Any pebbles in the
aggregate coarser than three-fourths inch.were graded out.
For nach.different percentage of admixture incorporated
in the min; three separate batches were made. One slump

test was taken from.each batch. The curves were plotted

10.

from.an average of these.

Only fine aggregate was used for the flow table test,
as this is chiefly a test for mortars. Consistent results
could not be obtained when using a combination of both
aggregates. Although.a different gradation will, perhaps,
effect the mix differently as to degree of workability,
the general effect should be the same regardless of
gradation. The mix was the same in proportion of cement
to aggregate as in the slump test, but with a water-cement
ratio of seventy three one hundredths by weight. This
increase in water was necessary to obtain a workable
consistency because, perhaps, of the greater percentage
of air voids in a finer graded aggregate.

'With each.percentage of admixture three separate
batches were made. From,each.batch.two flow table tests
were taken, and from.an average of these results the
curves were plotted.

It is known that the degree of workability varies
with.the time of mixing to some extent. It was therefore
important to keep the mixing time as near constant as
possible. Each batch.was mixed from two to two and one
half minutes by hand. At the end of this period of mixing
the workability seemed to remain fairly constant.

All the aggregate, fine and coarse, was dried by
gas burner until the water content was reduced as much
as possible.None of the aggregate was used the same
day that it was dried so as to be sure not to use any

above room.temperature. In small batches such.as were

11.

used, it would be impossible to calculate the amount of

water present and allow for it in the mix.

Below is tabulated the seive analysis data for the

fine and coarse aggregates. This represents an average

of two tests.

 

 

Seive' Percent Eassing 50% 505- Combined

Size ‘Sand Eebbles Sand Iebbles

100 2 o 1 o l
50 14 o 7 o 7
30 59 o 19.5 o 19.5
16 59 o 29.5 o 29.5
8 80 4 40 2 42.0
4 95 5 45.5 2.5 49.0
i 95 6 47.5 5.0 50.5
5/8 100 15 50 0.5 54.5
t 100 42 50 21.0 71.5
5/4 100 100 50 50.0 lOG>0

 

Tad T Lg. TA
gem

CUR ATS

Flow Table

Percent Admixture

VIESOL RQSIH

Percent Flow

 

 

 

 

0.00 120,116 116,116 120,112
.05 120,120 125,116 125,125
.09 125,120 152,152 125,150
.12 125,125 125,150 150,150
.20 142,157 142,157 142,152
.30 152,148 145,150 157,115
.50 148,150 158,150 145,155
1,00 155,162 160,155 162,160
Slump Cone
Percent.Admixture Inches‘Slump
0.00 l i 1%
.05 1 1% 1%
.09 2 2% 2%
.12 2 2% 2%
. 50 5 5 5/4 53,
1.00 5 4 4%

 

 

 

 

ORVUS
Flow Table
Eercent Admixture Percent Flow
0.00 150,125 125,125
.05 145,145 152,125 142,157
.09 140,157 140,157 142,142
.12 150,155 150,150 155,152
.20 167,169 167,165 160,167
.50 170,166 159,186 178,175
.50 180,182 182,180 165,180
1.00 182,185 168,170 180,185
Slump_Cone

15.

 

PercentyAdmixture

Inches Slump

 

0.00
.05
.09
.12
.50

1.00

C»! 0) 03
.5!-

C 1
.42

.p

N. ".4

14.

LOZZOLZTH

Flow Table

 

 

 

 

 

Percent.Admixture Percent Flow
0.00 125,125 151,155 151,151
.50 155,157 157,150 25,152
1.00 157,157 140,155 145,140
1.50 150,150 156,155 148,155
2.00 162 ,156 56,156 152,158
5.00 155,165 166,169 165,165
4.00 177,177 184,187 168, 75
5.00 177,188 187,187 180,188
Slumngone
Percent Admixture Inches Slump
0.00 1% 1i 1—1
.50 2% if 5
1.00 5 5 5%
1.5 4 4 Si
5.00 4; 4 5/4 4 5/4

 

HYDRATJD LIES

Flow Table

Percent Admixture

Percent Plow

 

 

0.0 150,125 127,125

1.0 120,127 125,127 125,127

2.0 150,125 125,127 127,125

5.0 150,150 150,125 150,150

5.0 150,155 125,125 155,155
Slump Cone

 

Percent.Admixture

 

0 . 0 1 11 1;-
1.0 1 2 2

2 . 0 2 2 2
5 . 0 2 2 2-3,-
5.0 2.;- 2-5;- 5

 

Flow Iable

Percent Admixture

CALC I LEI CI-ZLOR IDLJ

Percent 1

‘1ow

 

 

 

 

0.0 125,127 125,127 127,127
1.0 150,125 125,125 127,127
5.0 150,150 150,125 150,127
5.0 155,155 150,150 125,150
Slump Come

Percent Admixture Inches Slump
0-0 1 1 5/4 14
1-0 5 15 14—
5.0 2 2 2
5.0 2% 5 2

 

1'7.

 

8-

 

19.

 

Mia-M,"

20.

 

21.

 

22.

 

25.

 

24.

 

25.

 

’ 26.

 

COUCLUS I 011

27.

CONCLUSION

The results secured in these investigations seem
satisfactory in their indications of the effect of these
different admixtures on theworkability of a concrete.mix.
It is, howevwr, admitted that the results of this
research are hardly more than an indication, leaving room.
for much.more investigation in search of more complete
knowledge of theeffect of these admixtures on workability.

Vinsol resin and Orvus, air-entraining agents,
probably increased workability due to replacement of water
by entrained air, thereby having the effect of increasing
the amount of mixing water. Also, the additional air
incorporated probably acted in some measure as water in
making the mix mobile.

The effect of Pozzolith on workability seems the
greatest of any admixture used. A dispersing agent like
Pozzolith has the effect of breaking down the cement floc,
andin doing so releases water which has been trapped
within the particle clumps. Then this water becomes a
part of the mixing or placing water..Also, Pozzolith does
entrain some air. This, too, adds to mobility. These,it seems
logical, were the actions that resulted.

Hydrated Lime and Calcium.Chloride effected the work-
ability of the mix very little, as indicated by the
flow tests. The slump increased to greater extent, however,

with the percent of admixture added. It is the author's own

28.

conclusion that this was due to the greater ratio of fines
in the gradation used for the slump tests. In this test
coarse aggregate was used, where in the flow test the mix
included fine aggregate only, and adding more fine material

had less effect.

29.

BIBLIOGRAPHY
Master Builders Research Papers: No.
Economics of Cement Dispersion 56
Application of the Principle of -
Dispersion to Portland Cement 55
CementDispersion and Air
Entrainment 59
Use of Admixtures 58
Cement Dispersion and the Water
Cement Ratio Law. Special
By --------------- --Edw. W. Scripture Jr.
Highway Engineering ------------------ John H. Bateman
Plain Concrete --------------- - ------ Edward E. Bauer
Pozzolith ---------------- -- Master Builders Bulletin
Applied Siil Mechanics ------------- William S. Housel

A.S.T.M. Standards on Concrete Products

1“ "

ft~

 

 

 

.I‘

MW'IW'V‘U'" O"o*~ 4 - o o ~ ' t-O‘I"'"'¢b"‘""""" ‘I' ‘ " 'I c-IV‘V"I "I-v .‘vl 'vv- ' ' ' ' .. -

MICTH'IGAN ST'ATE UhliAVERSiTY LIIB’ReBIES
1l 11W ”'1 J [H WW
\ 1293 0'3'062 30614“

 

H3.

 

O i._. -__-