Determination of the First-Crack

Strength and Flexural Toughness of

Steel Fiber-Reinforced Concrete
S. Mindess, L. Chen, and D.R. Morgan
Department of Civil Engineering, University of British Columbia, Vancouver, British Columbia,
Canada, and HBT AGRA Limited, Burnaby, British Columbia, Canada

The first-crack strength and flexural toughness of steel fiber-rein- load-deflection curves of FRC in terms of some param-
forced concrete specimens containing different volumes of steel fibers eter that could provide a useful comparison of the util-
(0 to 1.5%) were determined using the procedures outlined in ity of different fibers or fiber contents, and for specifi-
ASTM C1018, and the ASTM toughness indices were calculated. It cations and quality control purposes. A number of the
was found that the value of the first-crack strength depended upon proposed methods have been described in detail by
the sensitivity of the load versus deflection measurement system, Kasperkiewicz and Skarendahl [1]. Basically, they can
and upon how "first crack" was defined. In addition, the shapes of be divided into two types: (1) Absolute value descrip-
the load versus deflection curves were strongly dependent on wheth- tions involve the total energy involved in fracturing a
er the deflection measuring system included or excluded the extra- specimen of arbitrary size, or in loading it to a partic-
neous deformations (due to machine deformations and deformations ular end-point deflection. Such tests include those pro-
at the specimen supports). It was concluded that the toughness posed by RILEM [2] for thin sheet specimens up to 25
indices depended both upon how the first-crack deflection was de- m m thick, and by the Japanese Society of Civil Engi-
termined, and upon how deflections were measured. This indicates neers [3]. (2) Relative value descriptions normalize the
that the ASTM toughness indices are strongly influenced by the energy absorbed up to a specified deflection by the
particular test procedures and methods of analysis used. ADVANCED energy absorbed up to approximately the elastic limit
CEMENT BASED MATERIALS 1994, 1,201-208 of the material, or the energy absorbed in fracturing
KEY WORDS: Fiber-reinforced concrete, First-crack strength, the unreinforced matrix. Such tests include those in
Steel fibers, Test methods, Toughness index ASTM C1018 [4] and by the American Concrete Insti-
tute [5]. These methods are compared schematically in
Figures 1 and 2.
In North America, by far the most common method
) lain concrete is a brittle material, with low ten- of characterizing FRC is that proposed in ASTM C1018
sile strength and strain capacities. To improve [4]. Toughness index values are used for both specifi-
these mechanical properties, there has been a cation and quality control purposes, as well as for as-
steady increase in the use of fiber-reinforced concrete sessing the usefulness of different types and addition
(FRC) over the past 25 years. Discrete, discontinuous rates of fibers. Hence, the remainder of this discussion
fibers are not normally added to increase the strength will focus primarily on this test method.
of concrete, though small increases in strength may All of the above methods depend upon an accurate
result from their use. Rather, fibers are added to con- measurement of the area under the load versus deflec-
trol the cracking of FRC, and to alter the postcracking tion (P-8) curves. It is by now well known [1,6-8] that
behavior of the composite, that is, its strain capacity the flexural toughness and the point of first crack de-
and its energy absorption capacity, or toughness. It is pend upon exactly how they are determined experi-
thus particularly important to be able to measure and mentally. The factors to be considered include test con-
interpret the load-deflection curves in the postcracking figuration, type of testing machine, specimen geome-
zone, in order to provide an indication of the quality of try, and the way in which specimen deflections are
the FRC. determined. These factors affect primarily the initial
Various attempts have been made to quantify the portions of the P-8 curves, approximately out to the
first peak in the P-8 curve. Thus, absolute value de-
scriptions of flexural toughness [2,3] are relatively in-
Address correspondenceto: Sidney Mindess, Department of Civil Engineering,
University of British Columbia, 2324 Main Mall, Vancouver, British Columbia,
sensitive to errors in determining the initial part of the
V6T 1Z4 Canada. P-8 curve. However, relative value descriptions of flex-
© ElsevierScience Inc. Received September 14, 1993
ISSN 1065-7355/94/$7.00 Accepted December 22, 1993
202 s. Mindess et al. Advn Cem Bas Mat
] 994; 1:201-208

is = Area0ABCI 12o= Area0ABEG

Area 0AJ Area 0AJ
'~ B = _ Area 0ABDH P~ = First Crack Load
/ /"~ C u 0 - ~
p LAJ ~ ~ = First Crack Deflection
"o
O
,..I
(a)

X
O
LL 0.4N o
I I I I I

'J '1 'H 'G F~

0 6 38 5.58 10.56 Zero
Span Load
Midpoint Deflection 10 Midpoint Deflection Capacity

A B
(b)
'1o _ Area 0ABF
"O t~
m O
O ,..J
,_1 Toughness:
TjscE = A r e a 0 A E F 0 I ~
x
X Toughnes: Factor: ._ 2 ]
_¢
_.e LL
IJ.
i
]F = Midpoint Deflection Load
Capacity
Wtb=L/150
Midpoint Deflection FIGURE 2. Relative value descriptions of the toughness of
FRC: (a) ASTM C1018 [4] toughness indices; (b) ACI Com-
FIGURE 1. Absolute value descriptions of the toughness of mittee 544 [5] recommendation.
FRC: (a) RILEM [2], in which the area is measured either to
the point at which the load drops to 40% of the peak load, or
to a deflection of span/10, whichever comes first; (b) JSCE [3], were cast in brass molds, and were then moist cured
in which the area is measured out to a deflection of span/150.
until the time of testing, at an age of about 190 days.
Seven groups of specimens were cast, with fiber con-
ural toughness [4,5], and in particular ASTM C1018 [4], tents (by volume) of 0, 0.5, 0.75, 1.0, and 1.5%; at the
which are strongly influenced by the initial portion of 0.75% fiber content, three different w:c ratios were
the P-8 curve, may be very sensitive to h o w the P-8 used, 0.35, 0.50, and 0.55. For these mixes, the super-
curve is determined. In the present study, two major plasticizer dosage was varied somewhat in order to
factors in measuring the shape of the P-8 curve, and maintain comparable workabilities. The mix propor-
their effects u p o n the ASTM C1018 toughness indices, tions and certain other parameters for these mixes are
are examined: (1) inclusion or exclusion of the extra- given in Table 1. The fibers used were hooked-end
neous deformations in the flexural test (due to machine fibers 30 mm long and 0.5 mm in diameter. It should be
deformations and deformations at the specimen sup- noted that the mixes shown in Table I are probably not
ports); and (2) definition and location of first crack. optimal FRC mixes, since the sand content is a bit low.
This may account for some of the lower than expected
toughness values, particularly at the high fiber con-
tents.
Experimental Procedures
Specimens ToughnessDetermination
Specimens conforming to the dimensions recom- Three specimens of each type were tested, using a 150
mended in ASTM C1018 [4], 100 x 100 x 350 mm, kN capacity servo-controlled testing machine, oper-
Advn Cem Bas Mat Determination of First-Crack Strength 203
1994;1:201-208

TABLE1. Mix proportions and fresh concrete properties

Fiber Content Mix Proportions* Superplasticizer Slump Unit Weight
(% by Volume) (By Mass) (L/m3) (mm) (kg/m3)
0.00 0.50 : 1.0 : 2.2 : 3.3 1.410 80 2437
0.50 0.50 : 1.0 : 2.2 • 3.3 2.489 120 2503
0.75 0.55 : 1.0 : 2.2 : 3.3 1.199 150 2443
0.75 0.50 : 1.0 : 2.2 : 3.3 3.219 150 2517
0.75 0.35 : 1.0 : 2.2 : 3.3 7.405 30 2537
1.00 0.50 • 1.0 • 2.2 - 3.3 4.315 70 2563
1.50 0.50 : 1.0 : 2.2 : 3.3 4.305 80 2597
*Water: cement: fineaggregate: coarseaggregate.

ated under deflection control. A so-called "Japanese traneous deformations were included, and one in
yoke" [3] was used, so that accurate measurements of which extraneous deformations were excluded.
the mid-span deflection u n d e r third-point loading In the present study, to record the load versus de-
could be carried out, excluding extraneous deforma- flection data directly, an X-Y plotter was used. In ad-
tions due to machine deformations and deformations dition, the output signals from the testing machine and
at the specimen supports. This is shown schematically the LVDTs were acquired by a computer-based data
in Figure 3. In these tests, the net mid-span deflection acquisition system at a sampling rate of 15 points per
was measured from the neutral axis of the specimen to second. This is fast enough so that the entire load ver-
its top surface. In addition, mid-span deflections in- sus deflection curve could be captured accurately, even
cluding the extraneous deformations were measured during the very rapid unloading that often accompa-
simultaneously, for comparison purposes, as shown nies the first major crack. A computer program was
schematically in Figure 4. subsequently used to produce the load versus deflec-
It should be noted that the current edition of ASTM tion curves, and to calculate the various toughness pa-
C1018 [4] appears to be ambiguous about the need to rameters: Is, I1o, and I20. For these calculations, first
exclude extraneous deformations, or at least is open to crack was defined as the point from which a series of
misinterpretation in this regard. Revisions to this stan- 20 consecutive data points (over a total deformation of
dard are, however, proposed which should completely 0.01 mm or more) had a slope at least 5% less than the
remove this ambiguity Uohnston, C.D., personal com- average slope of the curve between 45 and 70% of the
munication]. Nonetheless, even though the impor- peak load.
tance of extraneous deformations has been long under-
stood, it remained possible for some investigators test-
ing to the ASTM C1018 standard [4] to ignore this
problem. For instance, in the study on which ref 6 is
Experimental Results
based [9], the results from the six laboratories involved In Figures 5 to 11, typical load versus deflection curves
had to be divided into two groups, one in which ex- are s h o w n for the seven mixes described in Table 1. For
each mix, two curves are shown, one i n c l u d i n g extra-

Stee~b~VDTho,;i~s~~ Steel bar~ LVDTholdingscrew

~,~"~ F /r~-o -~ Specimen Specimen ~ __Specimen

°

Base of Testing Machine Base of Testing Machine

Side elevation End elevation

Side elevation End elevation
FIGURE 3. Schematic sketch of the test arrangement for flex-
ural toughness measurement, using a Japanese yoke; net FIGURE 4. Schematic sketch of the method used to measure
mid-span deflections exclude extraneous deformations. mid-span deflection, including extraneous deformations.
204 s. M i n d e s s e t al. A d v n C e m Bas M a t
1994;1:201-208

20

... Excluding Extraneous Deformation Including Extraneous

....... Deflection 20 Excluding Ext ...... s Deformation Including Extraneous Deflect,on

.:... ,... , . . . . . . . .
Z
P ..... : ; * i w,c:o ol
First Crack ;/ First Crack
0 . .[_.:... :. . . . . : .._: .... :. . . . . . . . . . : . .
--J 10 - - / W/C = 0.55
__ 1 : :
'/ , '-// . . . . . . . Fibre Content - 0.75 °/c
-- --:I-
X / : x
: / : .......
u--, . 5 . -I:- _.... " ..:... ~ -. : . . - : : . ,-r ..._tK...: . 15 = 4.9 15 4.7
1' 5
I1o = 9 . 6 Ilo 8.5
t . . . . . . . . . : .... : . . . . . . . ~ : " 12o = 1 9 . 0 12o = 13.7
/ / ;/' ;: . . . . . !-1 . . . . . . . . . . . . . . . .
0 " I ; I ; I ; i ; 0 ; I , I J , I ,
0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
Midpoint Deflection (mm) Midpoint Deflection (mm)

F I G U R E 5. Load versus deflection curves for plain concrete, FIGURE 7. Load versus deflection curves for 0.75% fiber con-
with and without extraneous deformations. tent, w/c = 0.55 concrete, with and without extraneous de-
formations.
n e o u s deformations, a n d one excluding e x t r a n e o u s de-
formations. Typically, the deflections at a p p a r e n t first
crack for the curves including e x t r a n e o u s deformation w h e r e d = mid-span deflection; p = load at first crack
were a b o u t 4 to 10 times as great as those for the curves (defined as described above); f = span length; E =
excluding the e x t r a n e o u s d e f o r m a t i o n s ( d e p e n d i n g on m o d u l u s of elasticity (assumed to be 35,000 MPa); ~ =
w h e r e first crack was defined). In Table 2, the first- Poisson's ratio (assumed to be 0.2); I = m o m e n t of
crack deflection (defined as described above), Wlst, a n d inertia; and d = b e a m depth.
the first-crack load, fist, are given, as well as the ulti- It m a y be seen from Table 2 that the theoretical first-
mate load, fult, a n d the t o u g h n e s s indices Is, I10, and crack deflections are in g o o d a g r e e m e n t with the first-
I20. (The effect of the choice of location of first crack is crack deflections d e t e r m i n e d experimentally, w h e n ex-
described below.) All are average values for the three traneous deformations are excluded. In general, the
specimens tested. The p a r a m e t e r s listed are given b o t h theoretical values are slightly lower than the m e a s u r e d
including a n d excluding the e x t r a n e o u s deformations. values. It should, h o w e v e r , be n o t e d that the influence
For comparison, theoretically calculated first-crack de- of the steel fibers on the value of E was not taken into
flections are also given. This calculation is based on the account.
formula In Figure 12, the initial portions of the load ver-
sus deflection curves (excluding e x t r a n e o u s deforma-
23 p£3 1 + 216d2(a(~3+ I~)] tions) s h o w n in Figures 5 to 11 are given at a greatly
1296 EI x 115 e x p a n d e d scale, m a g n i f i e d b y a factor of a l m o s t
=
50. It m a y be seen that these curves are highly nonlin-
flexural c o m p o n e n t shear c o m p o n e n t ear.
(1)

20 30

t Excluding Extraneous Deformation Including Extraneous Deflection - Excluding Extraneous Deformation Including Extr_aneo=us Deflection I
/
25
: : ! : : :[ ...... ~ 1
. . . . . . Fibre Coment = 0 7 5 %

l--~X':--- ,'~¢~itst.Crack~,,~.,,
................
- - : [ .... ~
....'[ I
"~ First Crack j ~ ' "'":" ..... ---- : : : I
~1 15 F=rst Crack _/. - First Crack . . . . . . . . . . . . .

-- ' s/: : ' ' : : ' ' 1 /: . . . . . . .

:3 .... ,._
'/ t_. __:__.[-
' } ............. ~ ..................
" .......... /]--i ' ...... i' ......
| x~lo ---,- : I -, .... i :
x 7 : s = 5.3 15 = 4.0 : : :/ : . 15 = 5.0 15 = 5.0
n- 5 ---;q ...... :-|110 = 10.5 I10 = 7.1 /~ ..... : ] E -"/Z . . . . :' 11o = 1 0 . 2 11o = 8.9
~/:-i . . . . . ! - - I 120 = 19.0 120 = 12.4 l..i .... i-- t 5 --~"~: ...... :' ~l 120 = 2 0 . 6 12o = 14.1 1-iii " :
7/.._: ...... .... ..... ....
0 ; I ; I ; I ; I ; / o ; i ; I ; i ; ~ ;
0 0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0,8
Midpoint Deflection (ram) Midpoint Deflection (mm}

FIGURE 6. Load versus deflection curves for 0.5% fiber con- FIGURE 8. Load versus deflection curves for 0.75% fiber con-
tent, w/c = 0.5 concrete, with and without extraneous de- tent, w/c = 0.5 concrete, with and without extraneous de-
formations. formations.
A d v n Cem 8as M a t Determination of First-Crack Strength 205
1994;1:201-208

30
25 "" Excluding Extraneous Deformation Including Extr..a.~o..u.s Deflection ] - - -
25
...: .~ . . . . . ,. . . . . . ~. . . . ~. ~ ~ . . ~ '-. . . . . .: . . . . . . :. . . . . .
~ 2o
~--: . . . . . . :. . . . . . : , , ~ ~ _ _ ! ~ , ,

--I
0 15 First Crack - -~,/',= . . . . . ,- -
i .. _ ~ . . . . . . : . / . . . . '~ . . . . . ', _ Fibre Content = 1.50%]. _
v'
~ lO
ii ..... ;--~-: ...... ; 15 = 5.0 15 = 4.0 . . . . ,6 :
5 :/" : : I~o 9.9 11o = 6.1 5 ---";i'''--"----" I10 : 11.5 ItO = 7.6
...,./.~. . . . . . ::. . . . . . 12o 18.5 12o = NIL 120 = 23.2 120 = NIL
o : t ; i ; I ; o / i i ; I ; I ; I ;
0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8
Midpoint Deflection (mm) Midpoint Deflection (mm)

FIGURE 9. Load versus deflection curves for 0.75% fiber con- FIGURE 11. Load versus deflection curves for 1.5% fiber con-
tent, w/c = 0.35 concrete, with and without extraneous de- tent, w/c = 0.5 concrete, with and without extraneous de-
formations. formations.

Discussion some instances the use of insufficiently stiff testing

machines was probably also a contributing factor.
Measurement o f Deformation
It is clear from Figures 5 to 11 that the load-deflection Definition of First Crack
curves induding the extraneous deformations are sim- It is, at least in principle, easy enough to measure net
ply incorrect characterizations of the material behavior. deflections accurately. However, defining first crack is
The first-crack deflection one would expect (based on a much more difficult problem conceptually. In ASTM
eq 1) is of the order of 0.03 to 0.05 ram. Here, however, C1018 [4], first crack is defined as "the point on the
the first-crack deflections including the extraneous de- load-deflection curve at which the form of the curve
formations (Table 2) are very much greater than this. first becomes nonlinear." Implicit in this definition is
Interestingly, this large error is not fully reflected in the assumption that the FRC is a linearly elastic mate-
the calculated toughness indices (Table 2), but the dif- rial up to first crack. However, it has long been known
ferences are nonetheless significant in practice. As that, when looked at in detail, the cr-e curve for con-
stated earlier, the importance of correctly measuring crete is everywhere nonlinear, though it is often ap-
net deflections has been recognized for some time proximated as being linear up to about 50% of the ul-
[10,11], but commercial practice has lagged far behind timate stress. In fact, even before it is loaded, concrete
this recognition. There is a great deal of anecdotal in- contains microcracks, and some of these begin to grow
formation that even today different testing laboratories as soon as the concrete is loaded. Thus, the phrase
come up with very different values of the ASTM first-crack deflection has no fundamental meaning.
toughness indices, and this may be attributed largely
to the deflection measurement technique, although in
~_ t FC=0,00 FC=O.50 FC=0.75 FC=0,75 FC=0,75 "FC~I.00 FC=I.50 t
25 - - W/C=0.50 W/C=0.50 W/C=0.55 W/C=0.$O W/C-0.35 W/C=0,$0 W/C=0.~O - " "
3O
_ Excluding Extraneous Deformation Including Extraneous Deflection t - : : : : : : : : -4-'-._
..~ . . . . . . . . . . ~,- . . . . . ; . . . . . . . . _-;~ ,.~ . . . . ; ..... ;, . . . . . . : . . . .J . 20 .......... : ..... i ...... ::............
25 ' ~ FirlltCrllck""~ "(~" ~ ' ~ ~- i , .

..... ..... .... ...... .... --j: .... ......

] " First Crack "'~" . . . . ! . . . . . ! ......
'
:. . . . . . : . . . . . ~.. . . . . . :. . . . . .
-J : I : .:.., j , ,,.I ..2:
15 .............. ;/s .............. W/C = 0.50 t ..... :. . . . . .
:: '. ~ /-
~, :; ";: . . . . : I F i b r e C o_ n_ t e n t =. l . 0 0 % ~ , - . . . .
~-, :: . . . . . .

10 ---; .... /-,- . . . . . "--- [ . --: . . . .

u. ..... : l .... ~,. . . . . . . . :-i - --
m - - ~ / ' /. ". . . :'. . . . . . : - - - / ' 15 = 5.2 16 = 4.6 --:' . . . .
5 ..... , /' . . . . . .= . .
': .
/~ . . .
' r/ ' / ' / ' '' /' . . . . .
i" . . . . . . . . . .
'
. .

5 ..,~, . . . . . . : . . . . . . :. . . . Ilo = 11.1 Ilo = 7.9 __: . . . .

~o : ...... "" : .~.' / : '. •l : ' . ~ . Telting M I I c h i rl e: I N S T R O N J
/ " ~: ...... :...... : - . . -[ 20 = 21 7 120 = NIL - .::. . . .
.I
. .

• " : ,
; I ; I , , I i 0 '°° t I,,'" IJ L.," iS l/ " I I I
0
0 0.2 0.4 0.6 0.8 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
Midpoint Deflection (mm) Midpoint Deflection (ram)

FIGURE 10. Load versus deflection curves for 1.0% fiber con- FIGURE 12. The initial portions of the curves (without extra-
tent, w/c = 0.5 concrete, with and without extraneous de- neous deformations) shown in Figures 5 to 11, at a greatly
formations. expanded horizontal scale.
206 s. Mindess et al. Advn Cem Bas Ma~
1994; 1 : 2 0 1 - 2 0 8

TABLE 2. Fracture toughness parameters

Excluding Extraneous Including Extraneous
Deformations Deformations Theoretical
Fiber Content 1st Crack
(% by Volume fl~t f=lt Wl~t W1,t Deflection
and w/c Ratio) (MPa) (MPa) (ram) 15 Ilo I2o (ram) Is Ilo I2o (mm)

(0.0%) 4.4 4.5 0.036 -- -- - - 0.266 - - - - - - 0.030

w/c = 0.50
(0.50%) 4.6 4.6 0.033 5.2 9.2 17.4 0.265 4.0 7.4 13.1 0.032
w/c = 0.50
(0.75%) 5.0 5.6 0.037 5.1 10.4 21.0 0.367 5.0 9.4 15.5 0.035
w/c = 0.55
(0.75%) 5.4 6.2 0.039 5.4 11.1 23.1 0.321 5.0 8.6 13.0 0.037
w/c = 0.50
(0.75%) 6.6 7.0 0.049 5.4 10.4 19.6 0.537 4.1 6.3 * 0.046
w/c = 0.35
(1.0%) 5.8 6.9 0.039 5.3 11.3 22.6 0.462 4.5 7.7 12.5 0.041
w/c = 0.50
(1.5%) 6.1 7.3 0.048 5.4 11.4 22.2 0.570 4.2 6.7 * 0.042
w/c = 0.50
*First crack deflection too great to permit calculation of I2o.

i ...... ! ..... i ...... i...... i...... i ..... i ...... i...... ..... ....

Z First Crack , : : ,I w/c= o.so I

1" ~ ~ ',. . . . . . . . . . . . . . . . . . . . . . . qFilteContent=O,75%}Fibte
Content = 0 , 7 5 % - -
] , ' : : : :1 J,p.n.s. Yoke'

--I

×"~ 15F = 5.0 15B = 6 . 0 15C = 5.4 -'

® 5 1 1 ~ ~ ~ = 9.3 hoc= 8.4 . . . .
E 5 I10F 10.2 IIOB = 1 2 . 6 I10c = 11.2 :
~- ~' . . . . . . ,2OF= 1,.0 ,20B= 16.8 I ,20C= 15.2 .... : 120F = 20.6 120B = 26.1 120C = 22.9

ol ;, I ~ I :, I i i o i i i I : : ~ :
0 0.1 0.2 0.3 0,4 0,5 0 O.t 0.2 0.3 0.4 0.5
Midpoint Deflection (mm) Midpoint Deflection (mm)

FIGURE 13. To u g h n e s s indices calculated for different defi- FIGURE 15. Toughness indices calculated for different defi-
nitions of first crack for 0.50% fiber content, w/c = 0.50 con- nitions of first crack for 0.75% fiber content, w/c = 0.5 con-
crete. crete.

25

c' ----i-- : ...... ! - : ---:-- ~-~ ~c~o.35 1

_~o -Ix-: ..... ~' . . . . . . ~ o - r ~o'~ r .... !

_3'5 ! i i

01 10 O i : i
"- --i i.... : :..... ! !. . . . : i
~1 .... i ISF = 4.9 15s = 5.6 15c = 4.8 i 'to -! 15F = 5.0 15B = 5.9 15C = 4,8 i......
4}
[J i '1OF = 9.6 hOB= 11.2 '10C = 9.5 : X
~D .... ! IIOF = 9.9 IIOB = 12.0 I10c--- 9.6 '--
U.-- 5 --- ' 12OF= 18.5 120B= 24.1 120C= 17.6 : ..... :

..... : ._ -, ..... • --,--- , --.- . .. _ ,. . . . r ...

0 o i : I i ; I ;
0 0.1 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3 0.4 0.5
Midpoint Deflection (mm) Midpoint Deflection (mm)

FIGURE 14. To u g h n e s s indices calculated for different defi- FIGURE 16. Toughness indices calculated for different deft-
nitions of first crack for 0.75% fiber content, w/c = 0.55 con- nitions of first crack for 0.75% fiber content, w/c = 0.35 con-
crete. crete.
Advn Cem Bas Mat Determination of First-Crack Strength 207
1994;1:201-208

25 25
i ' ' i i i i i
Z~
2°_ ~ . . . ". . . . c. . :: ..... ~...... ,...... ,...... o 20
--;~ ............ ! .... ! 12b,. . . . . . . . . .. . .. . .. . . . . . . . ,,_..
i-
. . . . ~..__- .............. • .... • .... :. . . . _

o) 15 ..... '- . . . . . . . . . . . . . . . . . . . . " ..................

_J -
© : I10
10
" ! ) 15F = 5.2 Iss = 5.7 ISC = 5.5-i .... i !i..~i:':i"::": ........ :.:::..!:::.-: (
...::!,,::. I

~ .... ;~ IIOF = 11.1 II05 = 12.5 10C = 107 ....... I ._ ........................ 15

"~ 5 .... i. 12OF----21.7 1208= 25.7 120C=21.0 -: ...... ~; ~ ¢: : ..... = ...... >
•--)-- ii!iiii!iiii!iii£ii
I.- w
' ~ r , _ -=,n ~on, 5 0w / c l _:: : ..
o ', ; i ; i ; i i i I (/)
<~
. . . . '~ . . . .
.
; . . . .
.
~ T _ _ _ _ - :
_ _ . _ - - _ . _ . .
E X T R A N ~. Q U S D E F O.R M A T I O N S E X G L U I ~ D
I " - - ~
:
- - - - ~
:
- - "

0 0.1 0.2 0.3 0.4 0.5 0

0.4 06 08 1 1.2 1,4 1.6
Midpoint Deflection (ram)
Fibre Content (% by Volume)
FIGURE 17. Toughness indices calculated for different defi-
nitions of first crack for 1.0% fiber content, w/c = 0.5 con- FIGURE 19. Effect of fiber content on the toughness indices.
crete.

curve, three points were chosen. The one marked F

This is seen clearly in the curves of Figure 12, which corresponds to the definition of first crack described
are everywhere highly nonlinear. earlier in this article. The points marked B and 'C cor-
According to the current version of ASTM C1018 [4], respond either to the peak load, or to sharp breaks in
the load-deflection curves should be drawn to a scale the P-8 curve. Note that, for clarity, the curves were
such that on the load scale, 25 mm corresponds to a originally drawn to a horizontal scale about twice that
flexural stress of about I MPa, while on the deflection recommended in ASTM C1018 [4]. Any of the marked
scale, 25 m m corresponds to a deflection of about 0.1 points could plausibly be chosen as defining first crack.
mm, or even less. For such scales, as shown in Figure It may be seen that, at the lower fiber contents, the
5, there is indeed an apparent first crack. However, differences in toughness indices are of the order of 10
when magnified, there is no obvious point on these to 20%; at higher fiber contents, they range up to about
curves (Figure 12) at which a first crack occurs. Logi- 35%. In all cases, the differences are higher for I20 than
cally, a definition of first crack that depends on the for 15, suggesting that the differences increase for
scale to which the load-deflection curve is drawn has greater values of the end-point deflection. These dif-
little validity. ferences are not enormous; however, they are great
Now, what difference does this make in practice? enough to cast doubt on the validity of the current
For the six FRC mixes described in Tables 1 and 2, the ASTM C1018 [4] procedures.
toughness indices Is, I10, and I20 were calculated for From the data of Table 2, it may also be seen that the
different definitions of first crack. In Figures 13 to 18, toughness indices are quite insensitive to the fiber vol-
such calculations are shown for fiber contents ranging ume ratio. As shown in Figure 19, for example, the
from 0.5 to 1.5% by volume. The points for which the toughness indices 15 and I10 are almost completely in-
various calculations are made are marked. On each dependent of fiber volume in the range studied here,
and even I20 remains essentially constant in the range
of fiber volumes from 0.75 to 1.5%. This suggests that
the test may not always be useful in distinguishing
2, ..... ! i : i ...... ! ...... i...... i ..... i ............
between different fiber contents, particularly at higher
z~ 2o fiber volume ratios.

Conclusions

1. When calculating ASTM toughness indices, it is

important to ensure that the load versus deflec-
0
0 0.1 0.2 0.3 0.4 0.5 tion curves are determined such that extraneous
M i d p o i n t D e f l e c t i o n (ram) deformations are excluded.
FIGURE 18. Toughness indices calculated for different defi- 2. The ASTM toughness indices are d e p e n d e n t
nitions of first crack for 1.5% fiber content, w/c = 0.5 con- upon the way in which first crack is defined.
crete. 3. If the load-deflection curves are determined ac-
208 S. Mindess et al. Advn Cem Bas Mat
1994;1:201-208

curately, and plotted to a larger scale than that 3. JSCE-SF4. Concrete Library of JSCE (Japan Society of Civil
recommended in ASTM C1018 [4], it becomes Engineers). 1984, 3, 58-61.
clear that there is no unique definition of first- 4. ASTM C1018-89. Standard Test Method for Flexural Tough-
hess and First Crack Strength of Fiber Reinforced Concrete
crack; the curves are everywhere nonlinear. (Using Beam with Third-Point Loading), Book of ASTM Stan-
4. These observations cast in doubt the reliability of dards, Part 04.02, ASTM: Philadelphia, 1991; pp 507-513.
the current ASTM C1018 [4] procedures for cal- 5. ACI Committee 544. ACI Materials Journal 1988, 85, 583--
culating toughness indices. 593.
6. Gopalaratnam, V.S.; Shah, S.P.; Batson, G.B.; Criswell,
M.E.; Ramakrishnan, V.; Wecharatana, M. ACI Materials
Journal 1991, 88, 339-353.
Acknowledgment 7. Johnston, C.D.; and reply by authors. Gopalaratnam,
V.S.; Shas, S.P.; Batson, G.B.; Criswell, M.E.; Ra-
makrishnan, V.; Wecharatana, M. ACI Materials Journal
This work was supported by the Natural Sciences and Engineering 1992, 89, 304--309.
Research Councilof Canada, through its support for the Network of
8. Lim, T.Y.; Paramasivam, P.; Lee, S.L. ACI Structural Jour-
Centres of Excellenceon High-PerformanceConcrete. nal 1987, 84, 524-536.
9. Gopalaratnam, V.S.; Shah, S.P.; Baston, G.B.; Criswell,
M.; Ramakrishnan, V.; Wecharatana, M. In Report, Task
References Group on CMRC/NSF Research/ACI Committee 544; Univer-
1. Kasperkiewicz, J.; Skarendahl, A. Toughness Estimation in sity of Missouri: Columbia, 1989; pp 1-70.
FRC Composites, CBI Report 4:90; Swedish Cement and 10. Kobayashi, K.; Umeyama, K. Report, Department of Build-
Concrete Research Institute: Stockholm, 1990; pp 1-52. ing and Civil Engineering, Institute of Industrial Science; Uni-
2. RILEM Technical Committee 49 TFR. Testing Method for versity of Tokyo: Tokyo, 1980.
Fibre Reinforced Cement-Based Composites, Materiaux et Con- 11. Johnston, C.D. In Steel Fibre Concrete; Elsevier Applied
structions, No. 102; 1984; pp 441-456. Science Publishers: New York, 1986; pp 333-360.