e.
e
ws
=
oe
—
i
24
Pr

 

 

rod
4
£

ef

i

HESIS

Telemed Cee mi ms!) aan

C.F. MEANWELL C,J. MO LEAN

1920
A Study of the Permeability of

Concrete

A Thesis Submitted to
The Faoulty of

Michigan Agricultural College

by

Cc. F. Meanwell C. Je MoLean

Candidates for the Degree of

Bachelor of Science

June, 1920
INTRODUCTION.

The study of the permeability of concrete has
been quite a common form of researoh work. The results
so far have shown that concrete cannot hold water under
pressure. In this thesis a concrete that is theoretically
considered to be the densest possible has been used.
The densest conorete is the strongest and should be the
most imperveous. Upon this hypothesis we are testing the
densest conorete possible for its property of holding
water under pressure and if it fails it can be assumed
that a concrete without the aid of a water-proofing

compound is not water-proof.

To get the densest concrete possible, the gravel
was tested by Sieve Analysis and with the help of Taylor
and Thompson's Ideal Curve, a proportion of sand, gravel,
and cement was determined that would give the densest
concrete. It was decided to test four different cements
and a composite cement s0 that average results oould be

obtained.

The concrete was made into 6" diameter and 2"
high discs and tested under heads of 10, 20, 30, 40, 50,
and 150 feet.

MATERIALS

Cement. The cements used in these tests were
secured from local dealers in Lansing. Four Michigan
102681
cements were used and a composite cement composed of equal
parts of the other four was made. Some of the sements used
were rook cements and some were merl cem nts. The cenents

used were denoted as follows:

A Burt Cement (rook)
B Peninsular Cement (rock)
C Michigan Cement (marl)
D Wolverine Cement (mari )
E Composite Cement

Burt cement is a very dark almost biack cement
while Peninsular cement also a rook cement is very light
colered. The two marl cements were about the same oolor

being a medium light slate gray.

Sand and gravel. The sand and gravel used in these
tests was secured from a pit in the northeastern part of the
city of Lansing and wes of a very good grade having a good
sharp sand mixed with well graded stone. No stone larger

than one half inch was used however.
TESTS OF MATERIALS.

1. Specific Gravity Test.

Sixty five grams Of each cement was carefully
weighed and the volumes of these samples were determined
by placing them in bottles containing bensel and noting
the change in volume. From the weight and the volume
 
the speoifio gravity of each cement was determined. These
tests were carried on all at the same time and under the
same conditions so that a constant temperature was maintained.

fhe following results were obtained:

CEMENT WEIGHT QF CHMENT VOL. CHANGE SPECIFIC GRAVITY

A 65 gms. 20.75 GeG- 5.133
B 65 gms. 20.80 «a.c. 3.126
C 65 gms. 20.88 o.0. 3.112
D 65 gms. 20.90 a.a. 3-110
E 65 gms. 20.90 a.0. 5-110

The average of cements A, B, C, and D is 3.120 and
this compares with the oomposite Cement E. Standard
specifications require that the specifio gravity be 3.10.

The results obtained above are sufficiently close to standard
and the cements are therefore up to standard specific

gravity.
2. Bolling Test for Soundness.

Neat cement pets useing 25% water was made of each
kind of cement. Two pats of this neat cement were made for
each cement on glass plates. These pats were placed in
moist air for 24 hours and then in e steam beth for 5 hours

following.

when the pats were removed from the steam bath they
showed no signs of having raised at the edges or of having

cracked on the surface. These tests showed the cements to

 
be sound.
3. Fineness Test.

Kech oement was tested for fineness by placing
exactly 50 gms. of the cement on a number 200 screen and
shaking for 10 minutes in an automatio shaker. The residue
remaining on the soreen was carefully weighed and the per-
cent caloulated. ‘The following results were obtained:

CEMENT WEIGH? OF RESIDUE % OF THE TOTAL WEIGHT

6

A 9.56 15.12
B 10.77 21.6
C 15.83 27.7
D 6.56 135.1
E 9.61 | 19.23

Standard specifications for the fineness of cement
require that not more than 22% by weight shall remain on the
No. 200 screen. (See Hool end Johnson's Handbook page 833).
An allowable variation of 1.5% is permitted however but all
cements testing within this limit should be reported as 22%.
All the cements tested met the standard specifications

except cement C and this was very mah below standard.

Cement C was heated at a temperature of 212 degrees
Fahrenheit for one hour and then retested. The results
after heating were no different. Cement C wes not rejected

however but the tests with it were continued.

 
 

4. Normal Consistency Test.

The Vicat test for normal consistency was used.
500 gms, of cement was taken in each oase with a measured
quantity of water. The cement and water were mixed for one
half minute with a trowel ed then thoroly mixed and kneaded
by hand into a thick paste. The mass was then passed 6 times
from hand to hand and then pressed into the large end of a
tapered hard rubber ring. The ring was placed on a glass
plate and the top of the cement was smeothed off by a single
cut with a trowel.

.The glass plate and the ring were placed under a
rod having a diameter of 1 om. and weighing 300 gms. The
penetration in 30 sec. was determined by a scale graduated
in millimeters. For normal consistency the penetration
should be 10 mm for 30 seo. The results of the tests on
each cement are showm in Table [.

5. Tension end Compression Tests.

Standard briquettes 1 inch thick and with a@ cross-
sectional area of one square inch at the center were made

for all tension testa.

Neat Cement. Sight briquettes of neat cerent of
nermal consistency were made for each cement. Three were
tested at the end of 7 days, two were tested at the end
of 14 days and three were tested at the end of 26 days.
The forms containing the test pieces were placed in

 
 

TAOLE I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TABLE of SYorertaa. Corrisistavcr.
Cement BS es Bie Vay satel eg| hea
Ae a he a a Tec.
Burt a) Faney JOO Royce) FO
bd 47 Ze ok are) 2OlM., 177. 30
2 a) 28 Se Po} Lat as Ase GO
a os | 2s Kero) Nae ela)
Peninselar\ 3 oo Pe los~) 9.577.777.) BO
9 3 ae 300 41 .0P RM 2]
Maa sa Sy Ao & Neola] PAO haan FO
ae ae Fok ole a) oe JO
v9 ray Pa ole eee) S6.577.772. GO
id ey Pte ea O) Vea AEA FO
Pe) Ve Page & Yale) Acne JO
ad oi 23. role) arate Go
lai ehataiels Oo Pe Tara) aaa So
Composite! & om .F00 AIM mM Jo
o* L pag em Zoo “ee neae BO
Fesu.ts or Conereession TESTS of Cuses. AEM
Cement- Cue em aha “so. | we “gn
a #750 eho L ooh
4 Pry ca <1 tae Wee VEE
ae ch ood VOW Aa
ie Evaro Va
B Z2Z50 FIOO Velma} bala ey ad
ya) fe tele) VEWeXe)
7 FEIT Fase.
a Crem ees ae
cS ¥OOO 4/200
o ea Veer en
fa) Crm ee oe a eee
oO PER ie ee
Vx Cram Ea
‘ E 4975 | 5333 | /2/9 | 7933
i Pigs ceo | EE

 

 

 

 

 
 

 

 

GL7 27/4
dp a GEEZ ey 7A ae
S2P oe oh a ele
arg (oy aa
(7a a oS ae ea SIP oye
OS | (7d eyay~) eyoe

EZ bd
OLE | BE27 | GZ wea Wap
cass ord (eee aw
GS? fae a)
LOE | S77 |-oe2 Cae. el
| ge i tt Ly 47) 002
O07 On aa

Bria eo Ae Co AeA

— —

| OLS | meee 44 Way, Or“

c24 AL) leyrsay BAY leyncay tt il baled eee eee Med sey

Te dee A cnc Ao ee Ae IL 7 aw be/sq/ poor| qi bs/sq/ aE TOM aA
a ameter neal a BO ‘sPOOD2 lem aa yaa

FE 6a 92 a Ee] ee LLY ©

—$____

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sete sss eles

 

 

 

 

 

XYO BF 9 ‘P/ Y Yoav VOLVOLS C:/ P UAE eh:
FOS

LOSL MACNZ

 
 

 

 

 

MICHIGAN AGRICULTURAL COLLEGE

fy

i

 

 

 

 

ew a

 
 

 

COLLEGE

MICHIGAN AGRICULTURAL

ae)

77a

mmm

Ty

r

G,

 

oe poacorenss mane:

 

Tad

 
moist air for 24 hours and them the forms were removed
and the briquettes were placed in water at a temperature

approximately 70 deg. Fahrenheit until tested.

The results obtained from the tests are tabulated
in Table III and comparative curves are show in Graph I.
All the results of these tests were compared with the
average results obtained by the Technologic Branch of the
United States Geological Survey at the Structural Materials
Laboratory at St. Louis, Mo. These results are shown on
graphs in Hool and Johnson's Concrete Engineers Handbook
pages 217 to 219. The results for neat cement in tension
obtained by the U. S. G. S. are as follows:

7 days 655 lbs per sq. in.
14 days 700 " _ "
28 days 780 =" " " "

It will be noticed fromthe table that, for 7 days,
all cem@ts used tested below these results except cemnt C.
For 14 days the samples all tested abeve except cemnt B.
The 28 day test showed that the cements C & D and the
composite cement E were well above the average while the
rook cements A & B were far below the average of the
cements tested by the U. S. G. S. The general average of

these results is very good however.

Cement mortar 1:3. Eight briquettes of 1:3 cemnt
mortar were made and these tested on 7, 14 and 28 day

periods as above with the neat cement. The results are
given in Table III and oOmparative ourves show in

Graph II. ‘These briquettes were seasoned as the neat
cement briquettes above. The results of these tests were
sompared with the results obtained with 1:3 cement mortar
by the U. S. G. 8S. which are as follows:

7 days 265 lbs per sq. in.
14 days $15 * _ 8% "
28 days 425 " "_  FK

Cement D is the only cement that comes above these
results for 7 days, the rest fall below. For 14 days
cement B falls slightly below the U. S. average but the
rest are all well above the standard. The 28 day tests
show that the marl cemnte C & D and the composite censnt E
ere above the U.S.G.S. average while the rock cements
A & B are below. These tests bear out the results
obtained with the neat cement above and the general average
is very good.

Compression tests of 1:3 Cement mortar.

Two inch oubes were made of 1:35 cement mortar and
tested at the end of 26 days. Three cubes were made of
each cement and the average results taken. The results
are tabulated in Table II. <All the cubes were placed in
moist air for 24 hours and then immersed in water for the
rest of the 28 day period until tested. The results of
these tests were compared with the U. S. tests at St. Louis.

 

 

 
The cubes tested gave results that showed a unit
compressive stress of a little over 1000 lbs per sq. in.
while the U. S. samples tested over 4000 lbs per sq. in.
at the end of 28 days. fe were unable to acgoount for this
wide variation unless the cubes shonld have been embeded

in plaster of paris.

S$1i1t Test.

 

A silt test was made of the fine aggregate. A
Sample weighing 534.73 gms. was placed on a number 200
screen and a smll stream of water was allowed te run on
it while the soreen was slowly tipped from aide to side.
The residue was thoroly dried over a gas flame and then
reweighed. The resulting weight was 519.77 gms. The
difference in weight represented the weight cf the silt
removed by the water. ‘This was 14.96 gms. and was 2.88%
ef the original weight.

The Mechanical Analysis of the Aggregate.

In order to obbain the densest conorete possible
the sand and stone had to be carefully analised. Only
that aggregate which passed thru a 1/2 inch sereen was

 

used.

The screens used in this analysis were the 1/2",
1/4", 1/8" and numbers 10, 20, 30, 40, 50, 80, 100, 1650,
and 200. The samples were taken by quartering stook piles
and carefully weighing tre entire sample. The soreeming
was done with a mechanical agitator. The residue wes
weighed and these are the results tabulated in the
seocend colum in Tables IV, Vand VI. By dividing the
results in the second column by the total weight ef the
sample the percent caught in each screen is obtained md
these results are tabulated in colum three. The feurth
celum is a summation of the preceeding percents which
shows that if that specific screen was used that percent
of the total would remain on the soreen. The fifth colum
gives the results of the fourth oolum subtracted from
100 or is the amount that passes the soreen. The results
ef each of these samples conld have been plotted but they
would have been of no use in connection with the Ideal

Curve which was used as aoriterion.

To plot results the samples were divided into
coarse and fine aggregates. The coarse aggregate being
that which passed thrn a 1/2" soreen and remained on ea
1/8" screen and the fine aggregate was that which passed
the 1/8" soreen. Considering the two aggregates separately,
curves were plotted from the average resulta worked out in

Table VII and are shown in Graph III.

The ideal ourve was plotted from a set of rules
determined by William B. Puller end Sandford E. Phompson.
The principles by which the aggregates are proportioned
in order to approach the Ideal Curve were developed by

Mesers. Taylor and Thompson.
 

Experiments have siiow that wuere only two
aggregates are represented by arves that are to be
combined the best results are obtained wha the combined
curve intersects the Ideal Curve at the 40% line, at K.
AB is the sand or fine aggregate curve, CD is the stone
Or oc©arse aggregate curve and AKD is the Ideal Curve.

Scaling the ordinates at K, LK was found to be
40% and LG is 95.1%. The LEK:LG::40:95.1 and LG is equal
to 42.1% This means that 42.1% ef the dry materials by
weight sheuld be sand and cement and 57.9% should be stone

er cearse aggregate.

The Ordinates on the ourve AB were measured at
various points and 42.1% of the values takm and plotted
thus forming a new curve that approached the Ideal Curve
under the fine aggregate curve. Then vy useing the right
hand soale that is the esmount retained in peroent and
measuring from the top and taking 57.9% of these values
and plotting as above the rest of the aggregate ourve was
obtained. The resulting curve was a very close approxi-
mation of the Ideal Curve. Since this was the result that
it was desired to obtain these percents used in plotting
the curve were used in the proportioning of the aggregate

for the concrete for the test pieces.

A good average commercial mix of concrete is a

1:2:4 by volume which is approximately one part cement to
 

SIEVE ANALYSIS OF AGGREGATE (J amples.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Saniple £
Jotal Sa ale Oe Ea) ie tae ha
| Screeri Va latreli a Mae iT da) ya (Lice Sa Kinet Wold | S/ZE OF |
nd Ze % 18 RNG Reagan Pree
ye e e iS WN eae
“ig” VA Pf eke} 26.30 75,70 eae
ae x awa’, Ihe lors er fs O/EF"
10 eed a hs} avd IL L2G 0.065"
20 IG. 2S Z70 IS5.40 4460 EYL ae
AS i>, oe 12.60 68.00 JI2Z.00 en elh (oil
el FF. SZ owe) FESO PALS) ae eee
yd oo hoo Pes SE.GE A118 volo ALi
Fak) otek 2/4 ola 40d. ele de
/O0 pe Rm § ae 4 G7. 75 225 aN oo key ot
Yao te) Se (WS lashed fey) (aNeler Se
200 #/oO lo Me k pote dal Lay A loa lolas ae
Over 200 I-00 ioe 1) FS. 74 WA
Sample I
VE, (a) Ma Lek haa shied lg /_) a ane
Screen |Amovat Caught \Vanount-weu/d Vimoont would \ Size of
oe retained | pass. Se ava dtd
Ve ed Eibek Wo A rd teed
ws fe) Oo fe) 100.0 ome
a 25.29 | 28.40 cere W602.) O25"
cm Le /3, 08 Pye 5892 | 0.125"
a Re Yom od LA 4779 IZ. L/ 0.065"
ZO 6°.77 J.00 55.77 a 0032"
IO a 12.50 (oA ae F.7/ | 0.0/98"
a VE Mh GIF ai Ox Pd A) ro Wo) poxe
50 Vea ee oe ptt ae Views 2.0//0"
SO Cet ee esd GS. 56 $44 | 0.0069"
foo STRAE 48/ G 7.37 £63 0.0058"
ah] Ro dh 0.78 aoe ia ee 0.004/"*
Zoe AIP O.5F od tl Ve Pa Yee SL a
Crer 200. et My fro hee) 62

 

 

 

 

 

 

 

 

 
 

ra By / mr a
Eee] eK ee

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ioe ad : pela ih
P |
Screen |Amount Caught  \fmovunt would eee shedtah te
we g% nal Sieh 7 ATS ' :
i. S| 9 | - =. ae MSY asx
fe OQ
s ve. : : rd Oc r A? r
& . . «; beate, ; </- ; s : dl
a 10.8 14:73 SIFT 6425 i ian
a4 ek) ei aah ss YM oXel ree
de, 67.99 Fo ee IZ AS sh O92"
fee | /039 92 SG EA/ om JZ Of 0 OFS
pages or ar oy 55) OSD > 7
493 g iio ae 76.60 £3240 OO TO
IO 86.05 ee re SEO a oe ‘oN ol agom
i eh odted 750. | F550 eres eAerek ta
Vee a4 ha Aas mined ca tera
/5O 6.00 0. 80 98.26 {73 eloyo. tae
; ye
a Aotes ISE Neg 98.76 ; ficdoter
gate. = oA 0.GO IGG S56 | m
ad a AT

Data Yor Ldeal Curve dae ede of Three samp/es)

Ave. foto! wh* 7527 9ms3

(Oe gta (alee a aA ee ale, ee ee

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

plot a al —4) ge Amount Caug la adeeb Grigiirs lei ahel ea la
; wadld be retin eee Jcree
IW? Jns % 4 % (sJo le Tee koe
a O oO . 400.0 Pe We Tad
Gg | vdeo | és? 367 | oes”
oi . hed a a ea : / edd | , is a
x SO8. S7/ eh eel alan & e) OF fae
=e aa
S e) BS 2 Eye
/O SZOZ (2.902 Peele) 67, 5¢
oe
Zo ae ate) Pot had 7F.FO c
eo) 97978 J | 4 | ode So Ti (aM lh, = ie
. 4 .
Pe 76.50 | /6.78 y didi Macidadl
3O raf ae, V/s Mele, i P | c eae
Es eae Fee er 567 | 20069’
ere) i note Sioyd haan ida te ekoroky- Ma
450 ee ao FE.6T /. $- eNoler y
|
<0o alot O87 aera Rae AE ae
a Ia ole) eo Me Tot aA A are Me oO |

 

 
AGRICULTURAL COLLEGE

i /- |

 

EE tela eee tA

if the total
i. : his per-

ste for

or 25.4%.

fap cheat ebebeeaes teeta Aineen Saetooe

r dises 6
e moulded

nit drawing.

 

> monlds
ove in

termined

 

Pape ea

 

 

 

 

Oe er siguraits
nears rao Sf ae ois -Gussesee Seas Sees eee
Fea Be pone Gs 7278 ate eaaaected Sees

Fa) ba

 

 

Pissss': © Bssesssaee 5 asset aaesapeeesie Eimear neta : | the water was

 

 

 

rater end

 

Hae ee dies eetetees |
foc dgnarsazssssanasonsesogad laeetgead caase ezad Seatageeta sates Mecd in

eeu ua / nee ae /

 

. aap gel sessile eee eee eee Hsien Meet 8, and 89%.

Ste te He Siena Ba | | | ter and

: | as of an

 

ae

 
 

 

five parts sand and gravel by weight or 16.7% of the total

 

 

 

weight of the dry materinis should be cement. This per-
cent was used in the proportioning of the oonorete for
the tests. The sand used was 42.1% minus 16.7% or 25.4%.

The cearse aggregate used was 57.9% pO |

The Making of the Test Pieces.

The test pieces were made in the form of discs 6
inches in diameter and 2 inohes thick. They were moulded ty
in wrought iron rings which were tapered to permit drawing.

By experiment the amount of material to fill two moulds
was determined. From the percents determined above in
the sieve analysis the weighs to be used were determined
and they are as follows:

Cement 756 gms. - 16.7%:
Send 1200 gma. - 25.4%
Gravel 25765 gmp. - 57.9%

The materials were thoroly mixed before the water was
added. ‘The dry mixture was then formed into a crater and

 

the water added. The average amount of water used in

each batch was about 400 oc. which wee between 88% and 89...
The materials were yvradually worked into the center and
then thoroly mixed with a trowel until the mess was of an

even consistency.
The moulds were well oiled and oiled papers were
Placed on the glass plates upon which the discs were
moulded. The concrete was added in thin layers and
well rammed. When the mould was full it was jarred until
water appeared on the top. The mrface was then smoothed

eff by means of a trowel.

Seasoning of Test Pieces.

All disos were placed in vats ever water and
allowed to remain in this meist air for 24 hours. The
wrought iron rings were then removed by jarring the forms.
Each disc was marked with its letter and the date on which
it was made. They were then placed in damp sand, about
ene inch of sand being belew the concrete and the whole
covered by an inoh of the damp sand. They were allowed
to remain in this condition for the rest of the 28 day
period, the temperature remaining fairly censtant at
about 70 degrees F. The object of the damp sand was to
give the discos all the water necessary for a complete harden-
ing and also approach as nearly as possible the subsurface
oonditions that would be encountered in practice in ocon-
orete construction. At the end of the 28 day period

the disos were removed from the sand and tested.

fests of the Specimens.

Percent of water in the disos at the time of

 

testing.
13

An extra diso was made of composite cemént and
this disco was made and cured in the same manner as the
ether discs. At the emi of the 28 days the diso was
broken and three pieces weighing appreximately 100
grams each were secured. These were carefully weighed
to the nearest one hundredth of a gram. The pieces were
then heated over a gas flame for two hours and then re-
weighed. The difference in weight was the free or un-
combined water in the sample. The percentage was
computed in each case and the average of the three
Samples taken as the peroentage of free water in the
test specimens at the time of the beginning of the tests.
The following are the results obtained:

Sample Weight before Weight after Difference % of water

heating heating
1 92.11 86.07 6.04 6.556
2 94.64 89.11 5.49 §.801
3 88.19 83.09 5.10 5.783

Average percent 6.05

Description of the Apparatus.

The water pressure used in these tests was obtained
from the pressure tank in the hydraulic laboratory of the
College. <A two inoh pipe line was extended from the tank
and f's were placed at intervals of approximately 18 inahes.
To these were connected the s Pparatus for holding the test

(See Figs. ¥,If, Ml, ana
discos. The disos were held between cast iron plates
P41 keto Wis) Ti TV

 

 

a

ls ) -e-» -P el oe ee

fIYSOO

 

 

 

 

 

 

 

 

 

 

 

 
 

YT FlHY7d_

LL T/A YS LL TY

 

 

 

 
 

Testing Aipparatus. Fig. I

 

 

Fressure lipparatus. Fig I
 

Concrete Discs used jp Tests
F; 19 wz

 
clamped together by means of 6-3/8" belts. Both plates
had twe inoh holes at the center so that the water
pressure was applied on an area of 3.14 8q. inches. One
Plate was threaded and attached to the pipe line. 5 sets
ef plates were used and 5 discos were tested at the same
time. Leaks between the castiron and the concrete were

prevented by rubber gaskets.

Pressure Tests on the Specimens.

The first set of disos were tested under a head of
10 feet which is equivalent to a pressure of 4.33 peunds
per square inch. This pressure was maintained for a
period of four hours. The first difficulty which was
encountered was the cracking of the discs due to uneven
pressure in clamping the plates together. Those discs
that did not orack allowed no water to pass thru then.

These tests were repeated for heads of 20, 30,
and 40 feet with the same result. New discs were inserted
for each new head. When the 50 ft. head was applied it
was decided to allow this to remain for a longer period
se the pressure was kept at this head for 4 days or 96
hours. None of the discs allowed any water to pass thru

at this pressure in the 4 days.

Direct pressure from the oellege mains of about

150 ft. head or 64 peunds per sq. in. was next put en the

 

 

 
 

 

15

discs. The same discs that had withsteod the 50 ft.
head were need. This pressure was maintained for 12
heurs. All the specimens except the ene made with
cement C held back the water at this pressure for the
entire time of the test. The diso C allewed 2 oo. of
water to oome thru in the 12 hours. This test was re-

peated on the following day with the same results.

All the discs were weighed before and after the
test to determine the effect of the free water in the
discs on the permeability of the concrete. No water came
thru the whole discs however and the weight of the disos
were from ] to 6 grams lighter after the test than they
were before. This shows that if there was any permeation
the evaporation was much faster thus causing the disos to
weigh less instead of more. The floa@ of the laboratory
was kept flooded to prevent excessive evaporation. The

fellowing are the results of the tests for permeability.

10 ft. head.
Diso Weight before weight after Permeation in

test test 4 heurs
A 2359 gms. Cracked
B 2540 gms. 2336 gms. Cracked-64 oo.
C 2294 gms. Cracked
D 2576 gms. Cracked
E 2427 gms. 2423 gms. Cracked-9 oo.
Diss

wd wv a w Ff

Dise

wwe A we >

Dise

ww ows CP

£20 ft. head

Weight before
test

2300 gms.
2381 gms.
2545 gms.
2553 gms.
2350 gms.

SO ft. head

Weight before
test

£374 gms.
£342 gms.
2340 gms.
2386 gms.
2391 gms.

40 ft. head

Weight before
test

£564 gme.
2592 gms.
2590 gms.
2390 gms.
2506 gms.

Weight after
test

23585 gms.
2341 gms.

2327 gms.

Weight after
test

2570 gms.
2534 gms.
2335 gms.
2380 gms.
£390 gms.

Weight after
test

2360 gms.
2387 gms.
£383 gms.
£2384 gms.
23501 gms.

Permeation in
4 heurs

Cracked

Cracked-65 co.
00 oa.

Cracked

00 oa.

Permeation in
4 hours

00 ac.
00 oc.
00 oa.
OO ec.
OO oc.

Permeation in
4 hours

Crackeé-85 oc.
Cracked-3 oc.

 

Crackeé-6 ac.
00 oo.
OO aa.
17

50 ft. head
Disa Weight before weight after Permeation in
test test 4 hours
A 2581 ge. Not weighed 00 oc.
B 2426 gms. " " Cracked
C 2545 gms. " " 00 oo.
D 2399 gms. " " 00 oc.
E 2502 gms. " " 00 «ac.

At the end of the four hours the diso B was changed
for a new one and the test on the discs at the same pressure
was centinued for 92 more hours making a total of 96 heurs
for discs A. C. D. and E& At the emd of the 96 heurs there

was no trace of water coming thru any of the discs.

In order to determine the compressive strength of
the concrete discos used, a diso was embedded in Plaster of
Paris on the Riehle testing machine in the testing laboratory
and a load of 3000 lbs applied for 30 min. te permit a set.
The pressure was then continued up to the fall capacity of
the machine or 100,000 lbs. with no signs of failure in the
concrete. This is equivalent te a pressure of 3540 lbs. per

sq- in. which is very high for commeroial conorete.
 

 

L&

CONCLUSIONS

From the tests preformed it is to be conoluded
that water-proof concrete is a practical possibility if
the proper care is exeroised in the selection and grading
of the materials and in placing the concrete. The
proportions used in the making of the test pieces were
theoretical and were obtained by the use of Taylor and
Thompson's rules in connection with the Ideal Curve as
outlined above. However it is entirely possible to
analyse and grade stook gravel in the field and with the
proper mupervision amd care in the mixing, the laboratory
conditions can be very olosely duplicated. With oonarete
made in this way there is every reason to believe that
1% would be water-proof for any heads under which it would
be necessary to use it, even up to 160 ft.

The elight permeation thru the oonorete made
with the cement © oan probably be accounted for by the
coarseness of the cement. It will be observed that this
cement was very far below standard in the test on it for

fineness.

That such a method of proportioning concrete is
entirely possible is Glearly proven by the fact that large
construction ooOmpanies have followed it, not as a meens of
making a water-proof concrete, but as an economical measure

in the making of strong dense concrete.
APPENDIX I

The follewing is a list of the books, articles
and papers covering work en concrete from which meh
valuable information was obtained in the preparation ef
this thesis.

l. Professional Memeirs VYol. 7

Percdlation and upward pressure of water on
concrete dams. /

2. Standard Specifications and Tests for Portland
Cement of the American Society for Testing
Materials.

3. Proportiening of Concrete Mixtures.
Portland Cement Association.

4. Concrete Engineers Handbeok.
Hool and Johnson.

§. Reenforeed Concrete Vol. I
G. Ae Hool.

6. Effect ef Vibration, Jigging, and Pressure on
Fresh Concrete.
Duff A. Abrams.

7. Maseny Construction.
I. O. Baker.

el ND ease. -- :
we ce

 
APPENDIX II

The following is a ooOpy of the outline of the plan
fer the work on this thesis submitted previous to the

beginning of the work.

Outline of \Work.

1. Materials

1. Four different cements are to be used and a
mixture ef equal parts of the four
cements forming a composite cement,
making a test of five cement mixtures.

2. Gravel to be a good grade of sharp bank
gravel.

3. Water to be tap water from the college mains.

II. Mixtures

1. All mixtures of cenorete are to be determined
by a sieve analysis of the gravel used.

2. An analysis is to be made of the sand passing
the ene eighth inch screen and also of the
stene passing the one half inch screen and
remaining on the one eighth inch soreen.

ITI. Tests,

1. Conorete
a. Two discs are to be tested at each head for
each mixture. Tests te be made at the end
ef 28 days. (This part revised).

be Pension briquettes for tests at the end of
7, 14 and 28 days are to be made.

co. Compression cubes will be made for tests
of each mixture at the end of 28 days.

2. Heads.
Heads from 10 te 60 ft. in increments of 10
ft. will be used in making the tests.
(Later this was changed to inolude 150 ft.)
KUOM USE ONLY

 

 
 
          

 

aa = yp f SH
BER REERE ES &
ee

 

tl _ -

rome a

~ — ;
I mg = ~ — ae 7
SS y } F “Sa 9 ‘
- “ iH £

7
\ we :

aa pt ; od TIP Ne ends i 2)

i ae 1 EA St | Si SSS Se
MICHIGAN STATE UN

ry
3 12

93 03169 2654

 
 

a -

~~“ 3 Lae

ret
sn _

ee. ee

 

 

éo-

 

i
rh
—_ =n TED OP ge 4

&

at er = :

 

— 7 -——-+—.
ttt PB NS

9 -

ean. ee Aer os - +

~%

  

es =

 

 
LIBRARIES

(i

Mic HIGAN STATE UNIVERSIT
|
|
|

1293 03169 2

|

 

;
a
=

 
 

ene é =e ot ; mae ‘ =
SR ~ pa : ;
- +

   

 

a i

4 7
‘|
. vs "

ara

———— = em

-

    
MICHIGAN

3

 
 

        

= &
> i oe 7 . —e — inten = 5
.. . :
ee ne i me ‘
& ; eae
. . _ " aa 7
~ 9 s

 

- - : 3
4 t

WAP ec

sis _— sae for Se a

ee eo re —?/
MICHIGAN STATE UNIV LIBRARIES

i
HIN

4

ii iii

3 1293 031

 
    

  

™* &
_— ttle. ' — — =e ty
= ie :

ee ty By ae _.
_—_—— ~—-—-—
4 awe

 

     

   
  

  

 
 

 

gv
«_ —

_~

       
 

 

. —
“= ee

~