1

MIN

r
1

—
,___J
—
,_—__

,¥__ _
—
—
—
._—__
—
—
—
—
—

IN

 

A STUDY OF THE EFFECT OF
CEUTE 0N SEGREGATION

Thesis for the Degree of B. S.
\V» L. Davis
I 9 2 7

Ill l|.| . . . . .lll‘cll 1|.l“l]|
.43! \ft... {HI—II

if: c k

. .. . .m m h... thvu~vmruoti

.. .lpl‘r-

. l . .1 1 EV;
km .4. FEB ... :L\\..

. (2.31. n, :1 ”1...»qu

9.7!...1.u I.
.

 

w
s
E
H
T

 

 

 

. ... .‘ox 117 .{mv‘T-W Div.

1‘ r!.l~.,

A S

HO-“ CU“ r—gy. ‘1 TfifiTV'"‘-‘m 3"-1 C-‘L I._.-<I ,‘y.’ ’fiDfi‘fiT —'.-\. " MIC-T
‘ V -‘Y ‘5 .'. “.24 L: .‘.' -JU a. a.‘ -.a -u 4.. bad. “a or. .'. .L‘

Thesis

Submitted to tne acuity of the
Iichigan State College of

Agriculture and Appliefi Science
by

X. L. Davis

Candidate for

t e
7 Ad

TH E515

 

This thesis is a study of the effect of Celiye, ( a
special grade of diatonaceous silica of exceptionally
high purity manufactured by the Celite Products Co.
and obtained from Lompoc California ) on the segregation
of concrete when hauled.

A study of the effect of this material and various
other admixtures on the strength, workability, density,
yield, and such properties has been made, but to the
authors knowledge no studiEs on their effect on segre-
gation has been attempted. In view of the present aggi-
tation for and against centralized mixing plants and the
hauling of concrete to the jobs it seems as if a definite
studyr of the different admixtures on segregation would
be of value, Celite was chosen because according to
investagations made by J. C. Pearson and Frank A. Hitch-
cock of the U. 8. Bureau of Standards at Washington D. C.
Celite gives the better results along other lines than
the other admixtures so this admixture seemes to be the
most logical one to use providing it gives good results
in this case.

In running these experiments the concrete was mixed
and placed in a wash tub in the back of an Overland Coupe
The tub being filled from half to three fourths full to
allow the concrete to work around without losing any of

it, the speed maintained while hauling was from 20to

25 miles per hour. This prehaps was a little faster than
a truck would have been driven over the rough roads,
but it was thought that this extra speed would compen-
sate for the fact that the car used was eaiser riding
than a truck would be. ,

As there is no standard tests for segregation it was
desided to use the following method. After the concrete
was placed in the tub preparatory to hauling a sample
was taken by shoving a cylinder into the concrete and
working a trowel under the cylinder in such a manner as
to remove a portion of the concrete from the upper sur-

face of the tub, this sample was then placed on a stand—
ard No. 4 sieve and washed , the portion which was retain-
ed on the sieve was then dried and weighed. The load was
then hauled the resired distance and a sample was taken
from the top of the load, by the use of the same cylinder

( which was used befor) pushed down to the same depth as
befor and removed in the same manner. This sample was
also washed on a No. 4 sieve and dried and weighed. The
weight of this last sample devided by the weight of the
first sample and multiplied by 100 to obtain the result
in percentage. This result was then subtracted from 100%
giving the final result , which will be called in this

thesis the percent of segregation.

95464

.0
.
u . I.
a a
.. e .
. .
a 7 U
.
, \ w.
I c .
0..
...
a
I A. In
.
.. b
l
D a
. a Q

n
i l C:
.
.
.
|..
I I
. .‘ I
. .
I. u t
n
a . p
. .
. .-
a
I I I
D Q
.
a .
Q- I
'I .
. a .
I a
..
a
.
_ .
a .
.4 V
I .
a
_ .I
I.
m.
J
. 4 I. w
.

.f ; I .
I.
I
.
v
.a
6 a
Q a .
l
a
.
c
.I' l
u ..
u
I
. o .
.
a-
D I
.
I i
.
.
I
.v. U .-
c .
. .
.o .
.
u ; _
. .. r.
v . .1 a . .
a
u ‘
c .
1.. . . .
.. I
v Q
. ~i.I
r .
n
.
I
. I
. .
o\ .
.
n
a at
,5. . ..
a I
.1 4
Q
.
o . . .
I .
. l .
um
t
3.!
. I.
.. .
. 1
u .o
‘ -
. 6.
~ .
u
. .
\‘

.t
.- I
.

Us

.

1h .0
I.

.of

u .1.
u.

.. IR

.
.
.
a. ..
.
.
a
t D.
I u t
. Q
Q a.
..
9. I
. o
I
so .-
In.
. .l
a
.w
. I .
v
r .
0.
.
.
.
, . o
.
n .
‘ r u
. rt
0
.l
.
.
.
.
I I.
I in
I
l

. . ...
1 o
. f.
.
...
O
. .
. . t
.. . I
. ,
C .Q \
... A . v
\n. . . . .
. .

I. .
r
.‘s .
.O b
r l
n,
. i
a
. ..
. .
y o
.
' .
l
.
a
.
. 0'
. . .
i
l. .I
It
. n
_ .
I .
o .
I
v . . , O
I
r

 

. tw-
, . O -
I
.
. .
.. .. n t
, . c a
. o \- .
u .
.| I I
a I“
. .
. a
J . v
u l I I
. 1
n.
. U i
To . a . .. O
. 3. . ...
. u .I
r ‘ I
.. .
. I
u, . .
. . .
a v o u
. x .. .
. . .‘0
. A . x
v‘
. .. ..
a l
V
. I
‘. o .
. .
4 O . . . c
a (s
.. . .. .
,. .

Iwa

 

r
a.
. a .
o
.-
‘ .
. I . s
. I
I o u .
I
‘ l .. . .
. n
I. a . u
.. I I 1
, n O .
. . II...
\ . .

n, . .
.
. . I A
. ..
—
.
i ll
.
.. .- l . .
. . .
s .0. . . I
. . . .. .
1| n v
. i r
~ I . .l C
'-
I
.
. . n
I .1.
. .
. . . o
e . C O
. . .
~. ».V‘ _ no
.
.
I n O .o
.
. .. . i .
I .
.
. .
. u
. .. ..
. .
i . . a
.
.
.
e D
a . a
.
o .. . a
e .
r h
. h a

.
.0 . .
. l A D
I. . n
n: . .o .
.
. .
o I.
a .
v I
u .
. a . y S
.o . I
. . a
. .
t, ..

4...
. Q ..
. I .
a
Ca .
0
fl . .
. .
.I p .
v
a. I
u.
..
a,
.I.
u .
I.
.
. .
.
o C
.
I r
, .
w...
9. v. P .
I
.
. do
a
J n i
.
‘Q ,
x-
A
n
. ‘\,
. . .
.
..
O
v x x.
.l
I
. .- o
—I ‘t
0
xi
0
.
. .
c r
1
at
r
- N .

.-
A.
. .
t
r
n
s.
.
0
.w
o
.
r
. I
.
'3
C
n
.l
I
r
. .
u
.
u
w
.-
l
.n

2.

The following is a surple computation:

Wt. of material retained on no. 4 sieve befor hauling = 820 grams
" ” ” " w w . . after ” = 435 "

’

100” - 70% = 304 segregation.

The tests used in running these tests was a 1:2:4 mix_with
a 5 to 6 inch slump and a 4 to 5 inch slump.

It was desided that a wetter mix than this would very
seldom be used and in trying to use dryer mixes great difficulty
was experianced due to the fact that the concrete seemed to obtain
its initialflset and was hard to remove from the tub after hauling
it was practically impossiable toremove a fair sample from the top
in the usual manner. .

The amount of water used was not constant but was allowed to
vary to give the desired slump. It was found that bybadding Celite
more water was necessary to give the desired slump but it was con~
sirered that this was the only fair way in which to conduct the
expermints because by keeping the water constant and varying the
amount of Celite mixes would be obtained of varying consistenct
and workability which would cause results which would be untrue

in so far as segregation is concerned as a mix with a smaller
slump under ordinary conditions has less segregation after

hauling than one with a larger slump.

A sample mix without any Celite in it having a 4 tp 5 inch
or a 5 to 6 inch slump was hauled the desired distance and the
percent of segregation computed then the samples using Celite

were hauled and theprecent of segregation of these computed.

This did not give results which at aglance showed the im-
provement due to the use of iiillll-ala Celite so it was desided
that the precent of segregation as found for the samples which
contained Celite should be devided by the precent of segregation
as found for the corresponding sample without Celite and multipli-
cd by lpp which gave theresult in percent. Then this result was
subtracted from 1004 which gives the percent of improvement due to
Celite at a glance.

The following is a sample computation showing how the per-
cent of improvement due to Celite is obtained:

Percent of segregation of sample containing no Celite =30

w n w w n 0 3‘1 ” =5.5
365 * 100: 18.4¢

100_718.4= 81.6% inprovenent due to Celite.

o
. . l
. . r
‘-
i.
t a.
.o.
.
.I
.
r
. .
n . o
w
.
T.
..
. o
..
1‘
u.‘
. 0
co 0 .
o
‘
p.

f . C
. o .
I . a
. . .
i,
. fl \
o I I
.
‘ . .
U 1
2
.
1o
. u 4
.
n l
. O .
0 .
. ..
. . . . i o.
.. .
a. I
u C 1 . . .
x.
Q 1 . .
.
. .
. I.
o
'
o 1 . I
. J
“V v to
.1
.07 e
.
.Jv . u
a n
.0 . a
. I . _
5 . a
o \. a
. t 2..
a
u . I
o. .. ..
a . . . .
. O.
. V I
n I. .
.
. ..
p. . . . .
. h.
I .
.. .
4 h. .
a. . .I
. fl .
.. O.
. I
I . v . .
0 \
O a
1 | .
a
.
. .
p I
. h
a ..
.o ..
r o ..
. .
. 9 .
.
. .
..
I
0..
v
a I
. .
.‘ a
l . . v .o.
o o
o
n a
r d
> .9
. . A
o o \
.
o.» . .
.
. 1

.
..
i
.
.. .
.
a
a
c
.
.
o
s
.r
a .
.
O
x
. .t
r
.o
b e
.
.
O.
. t
. .-
1
r o
.l .
.
f.-
.
.
n
.
I

 

3.

Trips were made hauling the concrete two and five miles
over rough roads , trips were also made of ten miles on smooth

pavement. It was thought that in most cases where a central mix
ing plant would be used that the average length of hauling would
be between two and five miles and that by using these distances

over rough roads the worst conditions which would be met in
practice would at least be paralled in these experiments. The

ten mile distance over smooth pavement was chosen as a mnzimium
distance that concrete wouldprobably ever be hauled because if

hauled much farther the concrete would begin to set. It was also
thought that the ten mile trip over smooth pavement wouldgive
some idea of what results could be expected if the trips were

made over paved streets.
The Celite used was computed as a percentage of the cement

used. The measurements being based on weight instead of volume

All samples which seemed to be eratic were run twice and if
one of them was found to give results which did not check up
with the rest it was discarded but in some cases where small

differences were encountered the results were averaged.
The following list of data was computed in that manner.

.u

v

.-or,

.0

:4.

no

is

 

Pcent of

Precent of

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

e ation im rovement
Diggafgr g duepto Celite.
a
45.6 III-maauo—a-uuo-o-o—
5 milc
’°°‘ 35~ 23.4
, 03.: 48.5
40.6 -- ........
26.5 34.7
o-—————419.4 52.3
30 --H ------
10 mi?
pa‘5'fi 81,6
te.3 --_-_~----
.______‘15.7 51.2
2 mil
1“ 29.3 9
‘ 7.7 78.1

:h

.
w A
u n
c ‘
X
. 1 c .
. .
.
r v
.
u
I
.-
. o. .
a .
. a
I .
. . I.
U
Q
g
.
A . 6 .
e
o
A .. o
I . no I
‘
.
. I
n .
.
.
a
v
. It
.
O
o.
.‘
a \ u
" _

 

;.

I {I

'4.'/

b

!

It is readily seen from these experiments that Celite
helps to decrease segregation. It is also verv noticable that
the use of three percent of Celite gives the best results
except for the one test in which ten percent was used.

These results need careful thought and attention in this
connection or one is liable to draw the wrong conclusions.
This particular tests in which the ten percent of Celite was
used had only a four to five inch slump. No test using three
percent of Celite and a four to five inch slump was made.
However a test using five percsnt of Celite and having a
four to five inch slump was made, this test gave a percent

of improvement which was 11.3 percent higher than the result
obtained by using five percent of Celite with a five to six
-inch slump. From this it seems safe to conclude that the per—
cent of improvement due to a three percent test with a four

to five inch slump should be at least ten percent higher

than that due to a three percent test with a five to six

inch slump. At that rate it would make the percent of improv—
ment due to the use of three percent of Celite greater than

the percent of improvement due to the use of any of the other
precentages of Celite.

While discussing the difference between the results ob-
tained from the four to five inch slump. This leads to the

conclusion that segregation does not take place to such a
great extent in the dryer mixes.

The percent of improvement due to the use of Celitealso
was larger in the mix having a four to five inch slump, this
shows that better results can be obtained in the dryer mixes
by the use of Celite even than have been obtained in this
set of experiments.

The reason for spending most of the time making tests
for small percents es of Celite were that most of the other
tests ( for workab lity, strength, density, etc. ) which are
on record have been made in the lower percentages. The per—
centages which these other tests have used and found favor—
able lie from one percent to eight percent. With the best
results rangeing below five percent. .

The percent of segregation of tests ;hauled two miles
over rough roads were, as one would expect very much lower
than the percent of segregation obtained in the tests which
were hauled five miles. The percent of improvement due to
Celite also ran higher except for the one test in which four
percent of Celite ass used. This test should have been run

again but due to the lack of time it was impossiable.

The resultd obtained from the test “Tish was hauled ten
miles over smooth pavement are rather surprising . They show
a large decrease in segregation over the tests which were
hauled five miles oV(r rough roads and a slight decrease in
the precent of segtogaticn over the tests which were hauled
two miles over rough roads. The test hauled ten miles over
smooth pavement also showed a greater percent of improvement
due to Celite than tests hauled two or five miles over rough
roads. These results are rather surprising because even though
hauled over much better reads it was hauled twice as far as
the five mile test and five times as far as the two mile

 

‘ 1
c r
O
. A o
to t
.Q. o
f.
.
.o '-
.| .
t
_ ll —
r.
I A. .
. o.
r .
r
J

. o,
o
r
. \on
. 4..
. .
. ..
...
.o
o . .0
v- I.
u .o . .
\.I
4 .
. u . r
I. fl _
r I. o.
v
a
.. .\
.l o.
w. . I
u \
.
n . lb. o
\.
O
..3 x .
. .-
t
o o
‘
.
. :0.
. . _
'l v
a
.
.
.
.I
v‘
r
1 _
o
.
C
. o
C A. . a
. .

I . 4 I
s a .
a - D
I i
I 1
r .
s
-o .o
, I. . . V
s
a
o. I ,
.f u
.
n U
n In! .
u
.l
.
Q .h
o. . .-
II
I I .
l I»
. .
I I I
no . o y.
u
. o
o . .u‘ .
O C .
. . D
.‘ .o no.
. .
'0. o
. n .
u
. . .l
. I . \ . t _
n ' n
«a n I .u
. . I. W
e V\ 1 t
O
a .
.‘ a.
P. .
.. ..
o n .
. n
. a
. .
. . . I
nlo l
Io'xc ..
. . No
.. a ..
I.
.
\
K t .
. . . a
.. '
. \o ,
L s
u
.o.
I . .
. .
c .
. . .0 c
..
..V
.
I l. r
S
n.
o .. r

, r

.
7 .Q
‘-
.
.u 4
7
4
1‘ .oo
‘
u I
II ‘A
I pi
..
a J
I
If
u .
.. u .
..

.u
f
.I'
a... .
v. I
.5
.
i
o .
.I
I...
u.
.I . .
U .
n U
s
u. k
.
v U
u
5
a.
u
.
a . .
n

c .
-.
l
.
,b
;
I
n
I
.
0' .
a.)
.c
o
.
.
I

a
.l
. .o
1..
.
i
..
,II
0‘.
p
. J
.
.
. I
.
n u .
.3
.
.
.
I
.
_
\ .
r .
,‘O .
u. u
‘ .
o
.
.

 

,K

,1

tests but in spite of this better results were obtained.

This shows that if the tests were made by hauling over smooth
pavement much better results would be pbtained from these
additional tests. ,

In summerizing the results obtained from these tests and
the conclusions drawn from them , it is safe to say that
under all conditions of the tests the addition of Celite
helped to reduce the amount of segregation. It should also
be noted that most of the tests were very severe and that
probably never in practice would they be equaled for it is
hard to think of a case where a central mixing plant would
be used and the concrete hauled for five miles over rough
roads , it might in all probility be hauled five miles but
most of this would be either pavement or improved streets
of some type.

This last statment grows out of the theory that central
mixing plants will be more economical and profitable in

cities and communities where large amounts if concrete are
used every year for a number of years . At that rate it is
hardly possiable that concrete would be hauled over two miles
of rough road. With these supisitions and the results of the
test which was hauled ten miles over smooth pavement it is
very safe to assume that in the ordinary case of a.central
.mixing plant that the addition of three percent of Celite
would give a percent of improvement of around U5 or in other
words it would reduce the amount of aggregation 3/4 , this
is an improvement which it seems should be in favor of cent-
ralized mixing planed as well as the use of Celite. If seg-
regatoon can be reduced by 3/4 it will leave it small enough
so that it will cause no difficulty in most uses to which
concrete is put. Also it is apparent that if the addition
of three percent of Celite reduces the amount of segregation
by 3/4 it will give a better concrete than it Celite is not
used for the addition of this amount of Celite , ( according
to the experiments of Pearson and Hitchcock ) does not serious-
ly effect the strength of 122:4 or leaner concretes , while
it is general knowledge that segregation to any marked amount
does reduce the strength greatly.

‘

.r
. .
.
.
‘
o b
.
.
I
I l
‘o
I O
A
o
n
o
. .
I I no
a
a
. .
rt.
.
'
.
o
—
. 5
. o
.. n
.
,.u I
r
._
I

a .
l o . a.
u
a
.
q
.
.
a
.
. a . w.
n.
I. .
. .l
n . . u do
c Q x I Q -r
. o. r .
..
I a I
II
. . .
a v
n: O U. i
..
. o,
.
o ..
.l w
.
o r .
. o .
t. .
s.
n
.
t h
.
.
. .
.
Is. .
o
.
. .
.
. .
' Q-
“ a
.
r -
r
. .
O ,
.
u . n
o I
O

G.
or.
n

.o
31

 

 

Il.o. 1“. 4.}

nl Io.

mic. .\\ ... u .

 

‘ 1'7 fir-um—
‘ r" :n ‘-

 

h._-.__~'.— - -..,_-A___.._ ‘Laa—ZJ‘.£A——_ (‘5‘, _‘~___..._...L ..... h---_‘._-_

 

 

 

 

 

1i.-t_-
T620.1 2%??2 95464.
‘ D
t.ofu_i_'
Joellite on segregation.
T62 . '
o 1 D265 9

00p.2. 5454
Davis

 

 

 

"'Tliifilfflfilflflifliflifliflflffliflifljlmfifllfim