strengths of and · comparisonofcompressivestrengths ofmoldedandsawedconcrete specimens...
TRANSCRIPT
COMPARISON OF COMPRESSIVE STRENGTHS
OF MOLDED AND SAWED CONCRETE
SPECIMENS
MOHAN SITALDAS JETHMI
B. S, , Kansas State University, 196k
A MASTER'S THESIS >
Sulanitted in partial fiilfillment of the
requirements for the degree
MASTER OF SCIENCE ,
. Department of Civil Engineering
KANSAS STATE UNIVERSITY
MANHATTAN, KANSAS
196lf
'>.
LD
! ' / TABLE OF CONTENTS
Page
SYNOPSIS lii
INTRODUCTION 1
£XPERlHa<THL iVORK 3
^ope ••••• ••••••• 3
i^eparatlon of Specloens •••••••••••• h
Materials ••••••••••••••••• h
Fresh Concrete •••••••••••••• 6
Molding and Curing • •••••• 6
Methods of Test ••••• 7
Results • •••••••• 8
Test of Significance •••••••• 10
DISCUSSlOH OF RESULTS 11
CONCLUSION 12
ACKNO/^iiuXiMENT. . . • • 13
BIBLIOGR..HIY Ih
APPENDIX • . . • » 16
/
ill
COMPARISON OF COMPRESSIVE STRENGTHS OF
MOLDED A.ND SAWED CONCRETE SPECIMENS
MOHAN SITALDA5 JETIIV/ANI^
SYNOPSIS
When standard concrete test cylinders fail to
jrield the specified 28-day ccsnprcsslve strength, it
Ic the usual practice to test cores or prisms taken
froa the concrete in place* So far, there is no
generally accepted method of correlating the
strengths of cores or sawed prisms end molded
specimens. It is therefor© essential that the
relationship between their strengths be established,
so that the results can be interpreted correctly.
This experiment was designed to compare the
CfMsporessive strengths of molded and saw0d prisms
that were cast, cured, and tested under «iinilar
corKlitions* Opportunity also existed to investi-
gate the relationship between the ratio of priM
•"•Graduate Student, Civil Engineering Department,Kansas State University, Manhattan, Kansas, U. S* A*
widths, cl , to maximum sizes of coarse aggregate,
a, and the strength of test specimens. In all,
three series of test v/ere made. Since series
III was an improvement over the first two, all
conclusions are based on the results o: tained
from this series.
The results indicate that the compressive
strength of ^-in* sawed specimens is significantly
less than that of M—ln. molded specimens. The
statistical analysts does not show any significant
difference in the compressive strengths of 2-in,
prism. For 1.33- in. specimens the difference
is significant at c<-.05 and not significant at
^= .025.
Though no simple relationship between the
ratio of prism width to maximum size of coarse
aggregate and the strength of the test specimen
is apparent, the results do indicate that th6re
is increasing influence of coarse aggregate on
the variance of strength properties of specimens
as the edge dimension of the specimen approaches
the diameter of the maximum size of coarse aggregate.
IBTHODUCTION
The cc»Qpressive strength of standard eoncreto oyXioders
is the foundation of our concrete design and Is usually used
ac the criterion of concrete quality* The strength of almost
every concrete structure is computed on the hasls of the
apparent strengths obtained from test specimens. Therefore,
It is important that this property be determined as correctly
as possible*
Qofortunatelyi there are many factors Influencing the
indicated strength of molded concrete test specimens* Such
factors sometines lead to erroneous results* If the Indicated
28-day strength of molded cylinders fails to yield the speci-
fied mlnlmuB strength, it is often necessary to test cores
or prisms taken fros the concrete in place to confirm that
results previously obtained were not chance events*
l^ere is only a limited amount of data on the relation-
ship betwe^ compressive strengths of cores and molded cylirvlers
in the literature. Thla limited amount of available data shovs
considerable conflict*
The Biureau of Hoclamation Concrete Manual^^^ states that
tests of drilled cores taken from structures almost invariably
show strengths greater than those obtained fro® control cylin-
ders \iAiich are standard cured for 28 days*
numbers in parenthesis—thus^^^—refer to correspondingitem in the bibliography*
H« Ppice^*^' observed tiiat core strengths were almost
all higher than control strengths in the case of cores taken
from structures which had some curing; and he states that if
cores were taken from columns of buildings (which were ci^red
and were protected from moisture), the cores might show higher
strengths than the moist-cured control specimens. Core
strengtiis ranged from about 70 to 180 percent of control
cylinder strengths, but were for the most part higher.
F, E, Legg, Jr,,(^) compared compressive strengths of
standard molded cylinders and cores drilled from pavement slabs
at ages 28 days, 90 days, one year, and five years. With core
data corrected to a ratio of length to diameter of 2.00 (using
a factor of 0.9^ for a ratio of 1.25), cores tested 121, 113,
101 and 100 percent of standard cylinders at the four ages
respectively*
y. K. Wagner^ found that cores yield only about 6? per-
.«ent of the 28-day cylinder strength and 8^ percant of the 7
day cylinder strength. He compared the results of standard
cylinders at 7 and 28-days with h by 8-inch diamond-drilled
cores taken from the same concrete after field ciiring.
Average core strengths at 28 days, as obtained by
B. A. Lapinas^^\ ranged from 61 to 7^ percent of standard
28 day cylinders for concrete containing type I cement, i'lhere
type IV cement was u^ed and the peak temperature ranged from
110 to 120 F, cores gave substantially the same strengths
as companion standard cylinders at 28 days.
3
Bryttxit Mather and rf. 0* Tynes^'^^ observed that standard
eylladers and cores yielded approximately equal strengths for
Xovstrength concrete, but cores yielded lower results than
cylinders for high-strength concrete.
Research conducted by R. D. Gaynor^^^ indicated that at
95 days, cores from well cured slabs gave strengths ten percent
less than 28 day standard cylinders*
Aj^Mrently, a greater understanding of the significance
of tests of cores is needed. The work reported here was
designed to compare the coapressive strengths of sawed and
molded concrete prisms that were cast, cured, and tested under
similar conditions*
Opportunity also existed to investigate the relationship
between the ratio of pirisBi width, d, to laaxiffiiBn size of coarse
aggregate, a, and the strength of the test specimen*
ieXPERIMENTAL rfORK
Scop^
The main purpose of this experiment was to compare the
compressive strengths of molded speeioens with those of cor-
responding sawed specimens* Care was taken to ensure similar
casting, curing and testing of both kinds of specimens.
Since the speciii^ns compared had the same ratio of d/a,
an opportunity existed to observe the relationship, if any,
between the ratio d/a and the cc»npressive strength.
Ibe experiment was conducted in three series as shown in
Table !• Since about nine specimens are required to make
proper statistical inferences, nine specimens of each size
were cast wherever possible •^•'•^^ However, this could not be
achieved in every case, due to various handicaps. The concrete
mixer used for series I did not have aiough capacity to mix the
required batch. For series III, only four ^-inch specimens ^
could be cast because the vibrating table was not big enough
to accanodate a bigger mold. But according to Tucker, as
quoted by Mather^^)^ njf we reduce the diameter of the test
specimen, we need only make the number of specimens proportional
to the ratio of the areas to obtain equal statistical informa-
tion." Thus it is inferred that four ^-inch cubes are statis-
tically equivalent to sixteen 2-inch cubes.
I^eparation of Specimens
Materials . Portland Cement Association' s^9) recommenda-
tions were followed in designing the mix. Six gallons of
water, 2^-5 pounds of Kaw River sand (Fineness Modulus 3.00),
and 255 pounds of crushed pseudo-quartzite were used per sack
of normal Portland cement.
Series I had 3A- in. maximum aggregate size, and the
maximum size of aggregate was 3/8 in. for the other two
series. This change was made to keep the least specimen size
at least three times the maximum naninal size of the aggregate. (10)
The change of aggregate size made it necessary to change
the proportions of sand and rock also. To provide the same
grain size distribution in all the three series, 16.7 percent
TABLE miKBEH AND SIZE OF oPECXKESS X£SI£D«
Specimen Maxlra Batiosize Aggregate Specijnen ffUBbep Tested
size in width toinches raaxlmum Molded Saved
Aggregate
(1)size (d/a)
(2) (^) Ch) (^)
Series I
8«0 in* cubes 10.67 7 9
6*0 in. cubes 8.00 7 6
^•0 in, cubes3A
5.33 6 6
3*0 in. cubes h»00 9 9
h^O in. cubes 10.67 9 9
3*0 in. cubes 8.00 8 9
2*0 in. cubes3/8
5.33 9 9
1*5 in. cubes Jf.OO 6 9
h»0 in. cubes 10.67 k
2.0 by 2.0 by 2.0 in. 3/8 5.33 16 16
1.33 by 1.33 by U.O in. 3.56 6 7
.-1^®^^®*^ ^^^^ corrected to (1/d) a 2.00 accordingto ASIM Designation: C ^2.^^^^)
6
rock and 83*3 percent sand were used for series II and III*
Fresh Contsretc , Ten-cublc-foot and two-cubic-foot
concrete mixers were used for series I and the other two series
respectively. The properties of fresh concrete are shown in
table !!•
TABLE II.—PROPERTIES OF FRESH CORCRSfB
^Je^les Concrete Datch -ise C-luiap Unit .Gijjht
ft.3 inch lb,/ft.3(1) (2) (32 (^)
I 10.0 5.5 1^7.0
n 1.5 0.0 1^2.8
III 1*0 0.3* 1^2.0
^Added a pound of extra water to achieve sooeworkability.
Molding and Curing. Figure 1 shows a typical mold used
in series I and II. Each mold was partitioned into two parts*
Three beams were cast in one part; as many as nine cubes were
cast in the other part*
Test specijaens for series I were molded and cured in
accordance with ASTM Designation i C 192* The beawi mftaken out of the fog room on the sixth day, and three cubes
tMKre sawed fron each beaa. All test specimens were then
capped with a coaunerical capping eoBpound and their dimensions
were recorded. Testing was done on the seventh day*
The sane procedure was adopted for series 11, with two
«36Beptions. Instead of a 5/8 in. tastplng rod. a hand
ylbrator was used for compaction^ and tb« speelfflezis ii»r«
kept Bubwrged in water for curing*
Series III was designed to remoTe some of the variables
Inherent in series I and II* Instead of using different molds
for different sizes, only one mold was used* This mold was
partitioned into various different sizes as shown in Fig* 2*
She entire nold was then fastened securely to a vibrating table
Concrete was vibrated for 20 seconds when t^e mold was half
full Iand far a minute when the mold was full* This laroeedure
was followed to achieve uniform compaction for all speeiB»ns*
The mold was then left undisturbed and fully covered '.fith a
polyethylene sheet for six days, after which the specimens were
moved from the mold* The slab was sawed into different sizes
as shown in Fig* 6* All specimens were left in water until
the next day, when they were capped and tested,
Mttihg^s of Till
Both sawed and molded specimens were tested in accordance
with the applicable provisions of the Method of Test for
Compressive strength of Molded Concrete Cylinders (ASM
Designation: C 39).^^^^
Table III shows the testing machines used for various
SP^imens*
tABLE TES7BIC KA.CHIN£3 USED*
Series
(1)
Size ofspecimentested
9
inches(?)
Machine used
CapacityOf theBachine.
inpoundsih)
t 3 and h Southifark finery 120,000
6 and 8 Iftiiversal HydrauLio 300,000
II 1.5 and 2*0 Eteery Hydraulic 75,000
3 and k Southwark Steer
y
120,000
III 1.33 an*! ^ Southwark Eodvy 120,000
Only one testing naehine was used for series III to
9lifflinate between-^aachine variation* All speclstons vere
tasted in the as-cast position*
The results of these tests have be^ sOBBarizod in
Table IV for ready reference* Detailed results are given
in Tables VI, VII, and VIII in the Appendix.
In Table IV, relative cojapresslve strengths have been
calculated for each series based on the average caajaressive
strength of the molded cubes for which d/a is 10.6? as 100
percent. In all but three cases, values are lover for sawed
specimens than for molded speciisens. Series I and II shcv a
marked increase in strength with decreasing speeisien size
(decreasing d/a)| there is no clear-KJut tread in series III.
O -H <»H -P ©O O O «] P.
•H "H -P
©O
o
cd i
CJ«H
•p © -p
<U -P I
M M +>ctj C © c3^ © G. ©© ^ o> -p ta<i; n cJ
© -P • •
O Ut> +3 q o
j4 -S
I
•H t
© o ^ o
o o >-P a.
u C n c
^+2H S3©o©
£5$ J> OP ©"HCO -d -P
:i u q
©-CJ © XI1> -p -P
PC ©0} © ca c
1^ 05(U -p ©
© P
© c © o• »M 3 a d> -p o ^Oi 0) ft
ON
CO
O iH jG ctJ © ^Moa-PEH©<S>'^'+> 0) ©rd ba-p N\ro]fl5 ftQ«-lO McjiH-d'CXJ 0) ? -P 0} to W'^
I
I m© ©M-P N ctJ d o CM
I
IH(0
©" ®o d N d o
Ho. © <H -H d
©
©CO
O C\3 <M m• • • •
OCSCJvOtN CM00 ONCMCM J-
• • • •vOnOCOUn
ooooCo.\£) ON^3-\r\vOvO
CO cvjoom
• • • •
d©
O© O HOJ-
H<d
•H^ OCOvO H^ • • • •O O H CMHH H r-IHHH
0000CMC^c^cJ\0 OlfNlfNJ- NO NO NO
tS-QrOQNO O roO• • • •Oc»\f\J-H
<
0000• • • •
00 NOJ- ro
o
O
HfH<UI
IMMto
©
©CO
\f\vOlfNU\Cs.lS.ONCJ• • • •
CJ CJJ- J±
C«-HCO OHmo HH H CJ CJ
8 000OCn.00
0000J-COOJ-CJ mcM ON
NO CO fO CM• • • •
NO tNCJtTN
c^3- OnoCJfOH CJ
oCJN
s
• •C*-CM roOH CJ
000ONCOlfN
CN.OO oNO 00 O• • • •
OCOnO^I—
I
CO\
• • • •
J- mcMi-f
0$
•H.d
d©
op.o
H
tMMH11
©
HJ-nOHVnnO• •
J-J-NOI
OJroOHHCJ
J- CJO• •
OnnO O^COCO CO
000J-rOCJO ONOroCMfx-
NO roO>CS.ONO• • •
NO ONf»-CJHr^HH
OCJnO• • «
o ov>OC0CJ>H
J-OCJ
INrONONO mU>• •
CO
<v00 •
• 'H
O O r<"
• • «
CMH
OH
0)
•PHwo«
0]
Series III shows an increase in coefficient of variation
with decreasing specimen size 5 there is no such trend in
series I and II, For series III, the coefficient of
variation is larger for sawed specimens tliah for molded
specimens for each of the three sizes.
The maximum coeffieient of variation is 7. 8^+ percent
for the 3-ln. molded specimens (d/a = 8,00) of series II.
The minimum coefficient of variation is O.76 percent for
the J+'in. molded specimens (d/a = 10.67) of series III*
Test of Significance ,
This test is performed only on the results obtained
for series III. Table V shows the three analyses. The
analyses were carried out seperately because of the unequal
number of test specimens and the differing variances for
the three sizes.
The analyses, therefore, indicate that the differences
in the strengths of the sawed vs. molded specimens was
significant for U—in, size specimens and not significant
for 2-in, sizes specimens. For 1.33-in. size specimen*
the difference In the compressive strengths of molded and
sawed specimens was significant at o< » #05 and not significant
at - .025. ' '•
. ;-
10a
Table V — Analysis of Variance « Tables for Strength
test data.
Source L-egrees Sum of Mean sum "F" Decisionof squares of squares
.
Freedom
^•*in. size
Total 298,000
Treatment 1 253:,000 251,000 32.06 sirnifi^(5.99) cant
w'ithin 6 1+7,000 7,830
2-in. size
Total 5^2 ,000
Treatment 1 30,000 39,000 1.76 Motslgnifi
(^.17) cant
Within 30 512,000 17,067
1.33 -in. size
Total ^90,000
Treatment 1 182,000 182,000 ^.91 signifi(M-.96) cant
Within 10 308,000 30,800
aValues in parenthesis are the "F" -table values.
The sum of squares are computed on the basis of table IX( see appendix).
\ •
DISCUSSION OF RESULIS
11
All iziferencea have been based on the coirpressive
strength test data from series III, Proper control over
wme of the variables affecting compressive strengUi could
not be exercised for the first two series. Laca of uniforiti
compaction, different molds for different sizes, and ^
different testing machines were a few of those
uncontrolled variables. Though both the beams and molded
specimens of series II were submerged in water for curing,
yet the moldeu specimens had six faces In contact with
water while the still un-sawed specimens had only four.
In series Illf however, these variables were eliminated.
The relative corainressive strengths for series III
show lower values for sawed specimens than for molded
specimens, which is In accord with some of the previous
findings. Though the strength of U-in. sawed specimens is
significantly lower tlian that of U-in. molded specimens,
it is not significantly different for 2-in. soecinsens.
Also the strength of 1.33-in* sawed speciraens is significantly
less than that of 1.33-ln. molded specimens atcx:=.05, and
not significantly less ato<=.»025»«
No simple relationship between the ratio of prism width
to maxiaum slse of coarse aggregate and the strength of test
specimens is appare.at from th© results obtained. However,
it is apparent that the compressive strength decreases as
the ratio d/a increased except where d/a = 10.67. Also,
the within-cell variance increases as the size of specimen
decreases* In other words, the data are more hetrogeneous
for the smaller sized specimens. This effect may be attr-
ibuted to the increasing influence of coarse aggregate on
the strength properties of specimens as the edge dimension
of the cube approaches the diameter of the maximum size of
coarse aggregate.
CONCLUSION
From the experimental data obtained from this study,
the following conclusions may be drawn:
1« The ^-in. sawed specimens were significantly weaker
than the ^-in. molded specimens*
2. There was no statistically significant difference
= - between the compressive strengths of molded and
sawed concrete specimens of 2-in size.
3« The strengths of 1.33-in. molded specimens are
.significantly lower than those of 1,33-in. molded
specimens at«x = .05, and the difference in their
strengths is not singificant at o<-.025.
k. The coefficient of variation decreased as the
patio d/a increased.
The conclusions drawn above hold good only ^Aien the
concrete mix design remained fixed with a cement factor of
5.5 sacks per ca. ya.j the water cement ratio was 0.53?
aggregate gradation was fixed; and, molding and fabricating
procedures were held constant.
ACKNOWLEOOMENTS
The author wishes to express his thanks and
appreciation to his Major Instruetori Br. Cecil H* Best^
for his help, guidance, advice, and criticism. An expres
slon of gratitude Is extended to Professor C. E. Scholar,
for permitting an easy access to his personal library.
He Is also deeply Indebted to Dr. Jack B. Blackburn, for
granting an opportunity to undertake this study.
BIBLIOGRAfHY
(1) "Concrete Manufacturing" Concrete Manual. Iftilted
States Department of Interior. Bureau of Reclamation,United States Government Printing Office, Washington, D« C*.7th Ed., 1963, p. 277.
(2) "Factors Influencing Concrete Strength,** by H«Price, Proceeding Aswrlcan Concrete Institute, vol* k7m1951, p. Vl7.
(3) "Static and Fatigue Strength of Hardened Concrete,*'tojr C, E, Kesler and C, P, Sless. Special Technical Publication,AsMrlcan Society for Testing and Materials, No* 169, 1956*
(^) "Experimental Fly Ash Concrete Pavement In Michigan,"by F* E. Logg. Jr*, Presented at Annual Meeting of HighwayResearch Board, January 196^*
(5) "Effect of Sampling and Job Curing Plroeedurcs onCompressive 3trenpth of Concrete," by 'jU K* Wagner, MaterialsResearch and Standards, Vol* 3, No. 8, 1963, pp. 629 to 63^*
(6) "Strength Development of High Cement Content ConcreteCast in I#arce Sections," by R. A* Laplnas, Presented at FallMeeting of American Concrete Institute, Toronto, Ontario,Ilovember 1963 •
(7) "Investigation of Compressive Strength of MoldedCylinders and Drilled Cores of Concrete," by Bryant Matherand 0. Tynes, Journal of the /imerican Concrete Institute,January I96I, pp* 767-778*
(8) "Research Highlight" by R* D. Gaynor, Pi-esentedat 3^th Convention of National Ready Mixed ConcreteAssociation, i'ebruary 196^*
(9) "Design and Control of Concrete Mixture," Bulletin,Portland Cement Association, Chicago, Illinois, 10th Ed*p* 16*
(10) "Standards Book," American Society for Testing andMaterials, Philadelphia, Pa., 1963*
(11) "The Numbers of Specimens or Test of Concrete andConcrete Aggregates Required for Reasonable Accuracy of theAverage," by R, v. Crum, Report on Significance of Tests ofConcrete and Concrete Aggregates, American Society forTesting and Materials, 1935, PP* 111 to 120.
(12) "Statistical Hethods," by G. -.4^ Snedcor, The Iowa
State Otiiversity Press, Aaes, Iowa, Ed. 1956, p.
(13) "Experimental Statistics," by H, C. Fryer, KansasState University, Manhattan, Kansas, 196^, pp. 221-2^2*
{Ik) "Effects of Aggregate Size on Properties of Concrete,"by S. Walker and D, Bloem, Journal of the ilmerican ConcreteInstitute, September, I960, pp. 283-298.
16
APPENDIX
^ o oO -HOU O o^ -HO.
o > -J-l
« 3
p. © a ^^-^© h _ 3 qt>
n G9 o
to a w
»4 a
a> ta
u o(D d :^
-Si
d rs g ^
i ® "H t)
d ts
(V
00
oCi
NOVO vOlTv
o8S o
O^OvC OvCO OCC CnC?CV v.-y rr> CJ Ci CM C4 t\J
8\OvO OCCOO WOO CJCn. CNvO \CNO ON CM
• •••••••<\0 vO NO NO vO \0NOvOvO
J- Cv«N^ 00 OnO^nO C^I
OnO^CTv OnCTv-CNOnOnO• •••••••
COOOCO 00 «SSNCO
c
oo
o
(04J
p;
c•r
'O
o
I
I
OS
0)
H
J3
I
NO
oINcvt
00o •tN NO
5 £
CM O00 tfN C4 Cn-OvONO <^nO roCJNO^f^C?"
888888888O U\O CO IrvXtMfNO lr>
CO NO jNco t^oaC\i CM C4 04 CM CM Oi CJ CV
^nOOO QnO CJJ-OnOCM rOND O INVnco^ rr
nO NO nO nOnO nO nON& nO
rOi'NOCOtrNCO 0\^lPO CTv OvO C?v <5n T>O O"• ••«••••<
C30 tvC^|>.C^tNtS«) Cn
o^oooaNC^oo'CO C^CO00 CO IN C««XO Cn
H CM rOH CM fOH CME-t N H 2: a » W fQ PQ
NO
o
18
5
8
gO-P
1O
C3
Q•§
O
I
:CO
•a«
HO
OJ C4 C\J CM C4 CJ C\J
• ••••••IfSlTwO\0 ^roro rO «^ r<-»rof
O iH O^fHO• •••••«
\0 IfVvOvO ^vO vO
\fsU\\0 NO vD U\Vr
4
O
c
-J
I
G]
09
o
Cvj
o
KMfNUMrsUMr
88888aONOXnco Ovo•k 4h •« » «
vO CJrovO CJCOO\OncO00 HO(Hr-IHH CMH
• •••••
• • • • •
CTsO OVr* p-OOs«HOO«HO«* • • • «
H CIHOJH CM
C
ncSo>4I
i
CQ
c
•o913HO
CM
NOo •
88888CO JNtvtTvO
IK •> » «
iH I—I fH rH r->H
CO WOOCOO• • • • •
l-i t-i r-i r-4 r-i tH
COVNOstMXiQ
a\Qro^cococ^o o o o <^• • • • •
c
5
8
o
aPi
o
o
I
00CM
H
00
o •
vOfH
COi-»0\±^pCOOOCo-VNiHlr
888888
oooooo
18qooo\o^ OOO roO^CM• « « • •
vDvO >0 VnvOHHHiHH
^X0>0>00* • • • •
• «•••<
HO+3
c>4
I
VM
a>
t3o
H
o •
r* Hso H
Ov O O iH O NOJ-OOINCJiH OHvOvO\0 C*^v0v0>O C>»c^lr\
888888888iHvOtO HtOO H J- OCO IfNrOH iTvlTN fOlr\O\rMr\\0vO ifNifN\0 vo^
^)fJ888S8• •••••
00 O^OvO^OvO^ O^C^CO
OOOOnOnO OOOn• •••romro CJ CJ fO ro Cvl
OOOOOO OOO• •••••
C4 rommnro ror^ro
20
«B
i
M
OS
CO 4
v-« fi
•r
/aH
^-v QCv o
4
H•i
NO
E•r
<^ 1
1
so
tcH
CO»
<^<M
<^H
f
o •
>0 r-i
mu± cas? oc^-;t>.^OO ONO ONj-Tr\± J- INlTN CJ HS)
80000Q00O OOOOOOncJ CNOsO ON H'»r\UN
CO Qoa CJcoOcOnO roVn%0 ^0<>^0^0^0^0 trv
H OJ-^ OV± trvj- \r\ONrO (\1 C4J- CMJ- CM r-i• ••••••••
CO O^Os<yv OnOC?*<jNn
OCO M VnoO O O O OONOvOOOHOr-IO• ••••«•••
OJ CM r<^ rorom ro r' i
CO C\JNOroO00OlfNCO'J\onH o o o c^H €^o• •••*••••M (fn CO CM ro CM
S do
•H V4 -PU O cS
O
o•H c o
•O O'd
esj P 3 • ^
_ 3 5a'
t3 n p o
to
to C ^
^5
1
o
5
«
I
o •
j J-J" rOui- I 'Vi
888888888C«. O \f\ CvJ O Vs CV
gOMOOOOOOCCJ OSO O O 04 r-
• •••••••«\0 vO V>vO>0 vDvO vOvO
•••••••••J- J- ji-^^
88^888^^3••••••••«
MO
i-<COOH HOO 2fO I
H HO fO<\l> rOH C\
888888888f*Vl- 00J- so CO^ \0
OCX) O 0«» C\J QsOvO04 C\JJ- vD OHO trsCs• •••••••rHHi-liHHHHHH
8<\)OHQCX^OOQCOHOOINOOO• •••••••
S 8 O 8.C^8 S^'i^• •••••••a
H c\* \e\>o t^-oo a
22
HO
J-r-f
ON CM OJCO CJ 0^tr\^,
8888888:H r-f iH OJ f'-i'?'
J- vO H H CJ CJtrsOO H2- fO^ J-j-^ roror^
rrjCO r-lCUHlJ- H C\IHC4O00• •••••••
\C\Q HHr-lvOlfNC4QOOOOOO^ONOO
H <M roj- IfNvO tNCO Cr
i
r-4|
^5
05
w
o
r--<
CM
H
Oo •
no
88885.8888>0 C^ 'O 01 VN CM C>J- tS
r-i ts.oco CO 00 c^'ir>
CMOVnOiHOO0\\0 CO tVCO CNOJNJ-• ••••••••
CO caCO COCO00 CD 00CO
oo<^oo^c^c^c o• ••*••••«CM CMm CM CM CM fO rn
C0^^QOQC^OOO<7vo o o cr»o 000• •••«•••CMrorOCMr^C^rr.c j CM
OJ
>
>O
-P
0)
o0)
HP!
0)-pa
bO
II
I
23
I
C
aPC?
c
o
o
CI
O
4)
•ar-C
C4
.5
888888:CJ rj oooir\V\<
OO CO00 t«UX}00 OnHHHHHHHH
«••••••»oi CJ M CJH C4 CJ ca
«^) oooooocS• •••••«•
CJ H CJ CJ W C: CI Cvj
V91.3
iS
OJ-
CJO
H CJOO HJ* CJrOCJlfN
-4"^ fn^ J-^ roul-
OOQOOOOOQOOOOOOVM^O\£) xr\VrsxC trsvO
Wj Hi-IHr-lrHr~tHiHH
g
a
s
t JCOnO OlO Cl^IS.nO CM>wfi^l>.<MNp«.*«•••••••
8H Qv0 IS.VO VNvO Qo occcDa) c^co o• •••»•*•C\JCJCl»-««-JHr4HCI
vO^ 00 O H CNQO Cv-CO 00 0*>.0 O CNO OWHHi-ICJCSHCJClH
H cv m^- tf>NO isoo cv
CI
Vr\ro
cviooeococMn
HHOOr-l^^HHH»-lr-»i-l
CM CU CviiHCjH
CM CVi CM Ci CM CJ
HHOOOO• •••••iHiHHHHH
• •••••
HC\4*nJ"VW>
01
P4
5
r-
c5
i
CO
J-QvOvOtO^vO rots.CMQ O^rom CM CJv*- C^OOHO <>0 ImrvO
888)° JOOQO^ C>\0 roO VfNxO CM Its
O O O O CT^O^O
Cj-COJ-'^ fOO O roD-O OnO OOH HO O*••*•••••CJHCMCMCMCMCMCMCJ
CO c4\Doo \fNH o c^^«.
• •••••»••fi r-tH r4 r-* f-i H
• •«••••••H rt rHHHHHHH
H CMrOJ-^vOCs-COCJs
I 43 I
0«H >I I
4^ c3 O OCO
> w o
9HH
I-P
a om (9 o >i o« h -p o»4 ^* •
CO O ® H
*H O O« p c:
•rt ^ O •jJj.Q
oJ « ©«HH
•p •>
- *
6 mo «
u a
^ «P Q)
c» c
I
IsCO a
NO
oo •o o^ o
QO o oQO<<
x>COCO CO CO• • •oo o o
o o
88 88•« *• •» «fc •> » *»4i
i
O
f Ov£>
• •
^ * *5 Hf-»
v>On CC
8H•»
ro HvC
<VO"
•NO
H
O O• «
H H
J-
00cr.
ric
8oI
©
Id
93o
o <:^ro 00
oo O OC««.C«. f-» ONHOv H CO
CO00 €0 00• • • •OO o o
OsO
OOvOvO
oovs^^iO• •
x:
gM
O
I
iHO
IN-CdOO
f^^, ^ 8• • • •
• • • •
a CO
3
O CMro •O
00
ooooof\J CM Wrororo
o o o o o o• •••••HHHHr-4H
CM CM CJ mron-i
OOOOOOcj3- o\0•t Cb «k •> 4|k a»
H O H C4 CJH
• ••«••
••••••CM CM CM CM (M CM
r-< r-1 rH r-I iH rH• •••••
OS C3N• ••••«
CO\r\vOC»-COtNOs CJ^ Cfs CS CJs• •••••
OHOno
OH
COc^
CO
CM
CCM
8vO cm3o5hh<OhCMHHO Hooooororomror^roro CM fO
HHHHHHHHHOOOO OOOOO«••••••••HHHHHHHHH
O c^c^v^u^CNOvD v%^..^Jco Q J- ^xoirsooHHOO HONOsOOON
CJ ro<^<^rOf^eM CM CM CM
O OOOOOOOOOCM CO^ J-nO^V\OJ-nO
vO H O vO H ^r\00 HCJ C4 CM H CJHH H HHHHHHHHHH
vO QOOOCOvOCOnOO<±^ oco<^oo«Jco<^o^os• ••••••••
• ••••••••CMCMCMCMCMCMCMCvJCM
ro m Cvi ro«*>H c*<O f-lOH HHHHHHHHH• ••••••••^ J- J- -d-^J- J- -d- -d-
CMHHHHHHCJH
80000VO Cs.a-'QVrNpocrooooooCMHHHHHCJHCM
4s
J-
^4
O 01
ONti
oooooo
80 H rj '"•c^OOOOO«••«••r-4 r-4 H H f~{ r-l
pOOCvONQpv
OOQQOOvo cjovoiVNro
•> «!. o. •> at. «k
OlrHHO i-iHHr-lf-IHHH
OJ CM 01 CM (MM
ocovo moo on
• •••••CJ H HHr-JH
o OON• •••••
(H Ol OJCMHH
gg§§|g
oOOJ
HO
OwON
o
f-4HCM
O OOO O• •«••« •
O
INCO
OQOQO-!>COX oo<?^
ro CM CM CMm oi OJ fo ca
27
ooooooooo• ••••••••HHHHHHHHH
-j-O CVIOJoOH«X»>rv.Ov.vsXrvONQJ- roco CJNlfvocooo a»oco c».<^com OJ OJ CM Ol CM CM OJ
oooooooooCMCO^ -vO OXTvO CM CM
HH 04 1-! iH O ri OHHHHHiHHHH
CN.OQlr\co\r\J- ojlrsOOOC^OOHOHCMCMCMHCMCMCMaCM
•••••••••(HOICJOJrHOlOJOJH
vo H Q s^f*^jr o C4mOOOOOvOOOO• ••«•••••HCMCMCMHCMHCMCM
COCOCO
a
8
8
I
«
p >o
CM
§8 s s a ICMJ- CM H ^ Ororo ro rO
OO* •
CJO
o O O OiH H H H• • • •H H
H00C4
{r\lf\\f\ o Oc\: rO O CMcono c^coo^ ONc\jm CM J-J* 3
• • • • • •H H H H
tN NOCJCVl CM H• t • « tmm m m rO
NOrH 1*^ H H H
• « • • • • • • •J-
IS O CJo ro m c\j fa CO• • • • • • • • •HH H HH H H H H
OtrWrsOtN m 00 Hcoca rorOCJ m 04 m m• • • • • • • • •HH H HH H H H H
:s CO MCO CO
H
r^
*^CMH
HH
•
q ar-5
- .*•ri
•
8
CmCO
Xi
oCv
I
1
'.'J
•
<H
XT'. O
Bm•
rorO
•H
/—
\
CM
<^f-l
;8
8
oOSm CO
o8
^ oooO rO*© r-lm C4 roro
• • • •
C4 CIO«
H
CO
• • •
vO H CJ 0\• • • •
O rH OO• • • •
• • • •
00 roC^
• • • •H iHHH
O
COa-C4
so
«H
•
COHH
oooo
H O O O* • • •
o^oo
XACO MJ-r«.>o !nnZ># « • *iHHHH
o^coev.• * • •
OOOO• • * •
• * • •iHHHH
CJ C4 010J• • • •HHHH
03 M
30
Fig. 1
31
Fig. 2
32
ICO
o.-I
.-)
n
oCO
olOoCO
ooCMCO
O
CM
oCMOCM
O
CM
OCM
00
o o oo oCM CMCO CO CO
o o or-l 0^
r-l OCO CO CO
ooCO
o\0
CO
oo00
oCM
o00
CO
-D
0)•HM0)
CO
ew•HMa.
O)
•P
UOcooTS0)
73
o
CO
ot-1-4
CO
OoONCM
OlOoCM
oONoCO
oCOoCO
or-l
-ICO
oCO
oCOCDCNJ
ovOoCO
oIT)
c^CSl
COONCO
oOJ
CM
OvOC7N
CM
OOOCO
OCOCM
OCOCO
oino00
oCO
o00
oCM
CM
\0CM
OoCO
oCO
o00CM
osoCOCO
o
CM
OvOCOCM
ONOONCNJ
oCJN
COCM
O00oCO
73o
oCM00CM
OCOCOCM
oCJN
00CM
ca73
toto
•p
c<u
u4->
CO
>•H(010<1>
^1a6oO73Cro
co•H-Pra-pc(U-rt
CO
Fig. 3
o•H
Fig. 4
34
x:+>
^ cnc
C CD•H UU 4->
D c/)
o<U 1-1
CO (0u
O XQ)
0) .-H
•DOx: T3•P c(U (0
•o >
(0 w-ac
-P Q.CO s
oo
l-l
o
(0
c ccn CD
10 MO) TOQ :r
;s sH OCO fH< M-l
ao
tj-.
U) o(H CD
O Q.4-> COOto cr>
U-, c•H
C -Po y)
•H 0)
-p Ho
f-1 c(0
o
CP -H
c m<u o,l-l 0)
CO a.
I
I
in
•H
j:oq.OB£ uotq.oea:a:o3 i[q.6u8J::|.s
Fig. 5
35
MV) to0)•H(4
cn0)
O UCD
(/)
c <<u
oo
oQ. N
•nl
CO•o
Sro ecn
Xm
c 2CO
oT3 -t->
Q)
TD1—
(
•aO
cM-i oO •H
in
cP 0)
c •H<D Q4-> 1-1
cn to
fH
> -Psi La
Q) -pM to
Q. (0
S (D
oo(-1 Ooq-i o
•Hc +^o 0)•H DS-P(0 0)•H
4->
(0
> 1-O
M-i
O coP •H
C +Jcu o•H cO•H
m(D
oo CO
00 NO CM
•H
Fig. 6
Table EC — Nested Classification
SizeKind ^-in, 2-in. 1.33-in. Total Of
Each Row(1) ^22 (1} QlI ilL^
Molded 3^00 29^+0 32003^00 2750 3\003^-20 2950 32^0
' C 3360 3120 31^0r 3110 3^10i 3020 3090M 3010
3200•
': 3160- 3120
30»+0
3180301030102880 I9lf803010
^13^80
Sawed 3170 306027202950
30^029903110
312033^0
2890 261029903090
271030502QhO
2950308028602890296030802960282030002880
771603.2160 H9oy
Total OfEach 257^0 95^10 37580 158730Colunin
RECOMMENDATIONS FOR FUTURE RESEARCai
37
! Experimental Design*
Four ^-in. cubes, sixteen 2-by 2 by ^-in. ,nine 1,33 by 1,33
by h-'in, prisms and a slab were cast for series III. The slab
was then sawed into various sizes as has been shown in Fig, 6.
Both unequal numbers and different error variances for the
three sizes made a two-way analysis of variance,which would
be desireable, difficult to interpret for the strength test
data.
It would be desireable to increase the number of specimens
for all sizes; however, it may still be desireable to use
unequal numbers to gain the same precision for all comparisons.
The analysis of variance technique would be complicated by
the inhomogeneity of varhnces which equal numbers for each
size would not overcome. Further investigation is needed on
the relationship between variance (or coefficient of variation)
and diameter to validate the quote from Tucker by llather.^"
2, Mix Design
A more comprehensive study in this field will be to
compare the strengths of molded ai^ sawed specimens by varying
the concrete mix design variables such as water-cement ratio,
cement factor, ratio d/a etc.
I ,,---•»
COMPARISON OF C0MF51ESSIVE SIR2HGTHS
OF MOLDED AND SAWED CONCRETE
SFEGII4ENS
by
MOHAN SITALDAS JETHWANI
B, S., Kansas State University, 196M-
AH ABSTRACT OF MASTER'S THESIS
submitted in partial fulfillment of the
requirements for the degree
MASTER OF SCIENCE
Department of Civil Engineering
KANSAS STATE UNIVERSITY
MANHATTAN, KANSAS '
. .
196^-
When standard concrete test cylinders fall to yield the
specified 28-day compressive strength, it is the usual practice
to test cores or prisms taken from the concrete in place to
confirm that results previously obtained were not chance
events.
There is only a limited amount of data on the relationship
between compressive strengths of cores and molded cylinders in
the literature. These data show considerable conflict. Some
show that cores are stronger than molded cylinders; others
show that they are weaker.
So far there is no generally accepted method of correlat-
ing the strengths of cores or sawed prisms and molded speci-
mens. It is, therefore, essential that the relationship
between their strengths be established, so that the results
can be interpreted correctly.
The work reported here was designed to compare the com-
pressive strengths of sawed and molded concrete prisms that
were cast, cured, and tested imder similar conditions.
Opportunity also existed to investigate the relationship
between the ratio of prism width, d, to maximum size of
coarse aggregate, a, and the strength of the test specimen.
The experiment was conducted in three series. For the
first two series, three beams, and as many as nine molded
cubes were cast for every size. Only one mold was used for
series III, and a slab, four ^—in. cubes, sixteen 2 by 2 by
^-in. prisms and nine 1.33 by 1.33 by '^-in. prisms were cast.
2
The beams of series I and II and the slab of series III were
sawed into various sizes of cubes and prisms. All the
specimens were beisted in accordance with A3TM Designation:
C 39. The results for series III were correcued to a ratio
of length to diameter of 2.00.
All the conclusions have been drawn from the results
of series III, as some of the variables inherent in first
two series were eliminated in this series.
The results indicate that the k-in, sawed specimens
are significantly weaker than ^-in« molded specimens; the
strengths of 2-in. specimens do not differ significantly;
and the 1.33-in. sawed specimens have lower strengths at
~c =. .05 and the strengths of 1,33-in. molded and sawed
specimens do not differ significantly at «^ = .025; provided,
that the concrete mix design remained fixed with cement factor
of 5.5 sacks per cu. yd.; the water- cement ratio v/as 0,53;
aggregate gradation was fixed; and, molding and fabricating
procedures were held constant*