Materials and Structures/Matériaux et Constructions, Vol.

34, January-February 2001, pp 3-6

RILEM TC 162-TDF: TEST AND DESIGN METHODS FOR STEEL FIBRE

RILEM TECHNICAL COMMITTEES

REINFORCED CONCRETE

Uni-axial tension test for steel fibre reinforced concrete

Recommendations

The text presented hereunder is a draft for general consideration. Comments should be sent to the TC Chairlady: Prof. Dr. ir. Lucie Vandewalle, K.U.
Leuven, Departement Burgerlijke Bouwkunde, de Croylaan 2, 3001 Heverlee, Belgium. Fax: +32 16 321976; e-mail:
lucie.vandewalle@bwk.kuleuven.ac.be, by 30 June 2001.

TC MEMBERSHIP: Chairlady: L. Vandewalle, Belgium; Secretary: D. Nemegeer, Belgium; Members: L. Balazs, Hungary;
B. Barr, UK; P. Bartos, UK; N. Banthia, Canada; A. Brandt, Poland; M. Criswell, USA; F. Denarié, Suisse, M. Di Prisco, Italy;
H. Falkner, Germany; R. Gettu, Spain; V. Gopalaratnam, USA; P. Groth, Sweden, V. Haüsler, Germany; F. Katsaragafis, Greece;
A. Kooiman, the Netherlands; K. Kovler, Israel; J. Lehtonen, Finland; B. Massicotte, Canada; S. Mindess, Canada; H. Reinhardt,
Gernany; P. Rossi, France; S. Schaerlaekens, Belgium; B. Schnütgen, Germany; S. Shah, USA; Å. Skarendahl, Sweden; H. Stang,
Denmark; P. Stroeven, the Netherlands; R. Swamy, UK; P. Tatnall, USA; M. Teutsch, Germany; J. Walraven, the Netherlands;
A. Wubs, the Netherlands.

1. SCOPE and corresponding displacement shall be recorded digi-

tally. The following data interpretation shall be based on
The test described in the following is used to deter- at least 6 specimens per test series.
mine the so-called stress-crack opening relationship,
σw(w), for Steel Fibre Reinforced Concrete (SFRC). The
test method is applicable for steel f ibre reinforced 3. SPECIMEN SIZE AND SHAPE
cements and steel fibre reinforced concretes that exhibit
strain softening behaviour. The method can also be used The standard test specimen is cylindrical with a cir-
for other fibre reinforced concretes which exhibit strain cular cross-section with nominal diameter 150 mm. The
softening behaviour as well as plain concrete (which is specimen is notched with a circumferential notch with a
assumed always to exhibit strain softening behaviour). depth of 15 mm +/- 1 mm and a width of 2-5 mm. The
The test is not intended for determination of the ten- notch shall be cut with a diamond saw. Mould induced
sile strength. It is recommended that the tensile strength notches are acceptable in situations where cutting is
is determined independently, e.g. indirectly from the made difficult, e.g when testing at early ages. The nomi-
compressive strength using empirical relations derived nal length of the specimen shall be equal to 150 mm, see
from plain concrete. Fig. 1. Only fibre concretes with a maximum aggregate
size of 32 mm and a maximum fibre length of 60 mm
can be used.
2. BASICS
The test shall be conducted as a displacement con-
trolled tensile test on a notched specimen. The specimen
can be either cast or cored from an existing structure or
structural element. The test shall be conducted in a
closed loop mechanical testing machine using the aver-
age measurement of the displacement over the notch as
the feed back signal or in another test setup which
ensures stable post cracking response. Both ends of the
testing specimen shall be clamped to the testing machine
and fixed with respect to rotation in such a way that
rotation is sufficiently eliminated. At the same time care Fig. 1 – The geometry of the standard test specimen for uni-axial
shall be taken not to introduce pre-stressing of the speci- tensile testing. Planes A, B and C are parallel. Plane B is a sym-
men. This can, for example, be achieved by providing metry plane. The total nominal length of the prepared specimen
is 150 mm and the nominal diameter of the specimen is 150 mm.
special fixtures which allow the test specimen to be A central, circumferential notch with a depth of 15 mm +/- 1 mm
glued directly into the test machine or by other means is cut with a diamond saw. The resulting, maximum width of the
providing similar performance. During testing the load notch is 2-5 mm.

1359-5997/01 © RILEM 3
Materials and Structures/Matériaux et Constructions, Vol. 34, January-February 2001

5. TEST SETUP

5.1 Basics
An example of a viable test setup is shown in Fig. 2.
The specimen is glued to metal plates that are connected
to the testing machine using rigid, bolted connections.
The bolts can be pre-stressed to make the connection as
stiff as possible.

5.2 Instrumentation
The displacement across the notch shall be measured
using no less than 3 displacement transducers arranged at
equal distances along the perimeter of the test specimen.
The effective gauge length lg shall not be larger than
40 mm. The resolution of the displacement transducers
shall be 1 µm or less and the accuracy 1% or better.
The load shall be measured using a load cell connected
in series with the test specimen as shown in Fig. 2. The
accuracy of the load cell shall be 1% of the maximum load
observed during the experiment, or better.
Fig. 2 – An example of the principles of test setup used for uni-
axial tension testing.
6. TESTING PROCEDURE
For special applications where the material characteris-
tics are not well represented by those of the standard speci- 6.1 Basics
men it is recommended to use a more appropriate notched
specimen. The ability of the test specimen to represent the The testing procedure contains the following steps,
material in a given application should be based on consid- assuming that the specimen is fixed to metal plates in the
erations of fibre distribution, fibre orientation and the testing machine using adhesives or a similar approach
relation between structural and test specimen dimensions. which eliminates pre-stressing due to clamping, see Fig. 3.
1. Balance the load cell of the test setup.
2. Glue the specimen against the first metal plate and fix
4. SPECIMEN PREPARATION and pre-stress the bolts of the plate in the testing
machine after the glue is hardened.
In general, care must be taken to obtain a fibre orienta- 3. Fix and pre-stress the bolts of the second metal plate in
tion in the test specimen which is representative for the the testing machine, apply glue on the surface and bring
orientation in the structural application. The specimens the specimen in contact with the second metal plate.
used can be either cast or cored from an existing structure 4. Mount and subsequently balance the displacement trans-
or structural element. The complete procedure for speci- ducers after the second glued connection has hardened.
men preparation, including fabrication, curing and/or stor- 5. Run the test under displacement control at the
ing, shall always be recorded.
In the case of cast specimens the proce-
dure for filling of the mould and compaction
of the concrete shall always be decided on
beforehand. Care must be taken not to
introduce an interface at the fracture plane
when filling the mould. The method of
compaction used must be the same as in the
practical application. The specimens shall be
stored under appropriate conditions until
sufficient strength has been achieved to
allow de-moulding.
Whenever possible, cored specimens shall
be extracted perpendicular to the expected
crack plane in the structure or structural ele- Fig. 3 – Schematic representation of the testing procedure assuming that the spec-
ment to be examined. imen is fixed to metal plates in the testing machine using adhesives.

4
TC 162-TDF

required rate using the average signal of the displacement

transducers as control signal.
6. Record the load and the measured displacement by
the individual transducers as functions of time.

6.2 Test control and data acquisition

The required displacement rates are: 5 µm/min up to
a displacement of 0.1 mm and 100 µm/min until the
completion of the test when a displacement of 2 mm is
reached.
Throughout the test, data shall be logged with a fre-
quency higher than 0.5 Hz.

6.3 General requirements

Fig. 4 – Calculation of crack opening w from averaged, measured
displacement over the notch.
The axial stiffness of the testing machine and the
grips shall be sufficient to ensure stable testing without
any snap-back in the averaged displacement transducer _
signal. It is not possible to express this requirement mean stress-crack opening diagram σ w(w) shall be
quantitatively, since suff icient stiffness is typically reported. Furthermore, a characteristic diagram σw,k(w)
achieved through a combination of mechanical stiffness, can be calculated according to the following procedure:
servo valve performance and performance of the elec- Calculate from each test the dissipated energy
tronic control system. between two crack openings wi and wm where wi is the
The bending stiffness of the machine and the grips crack opening corresponding to the situation where the
shall be sufficient to prevent significant rotation of the crack has crossed the failure plane, while wm may be
crack surfaces during testing. This requirement is ful- taken equal to the ultimate crack opening, depending on
filled if, at the end of the test, the maximum difference the type of application:
between the individual transducer signals is less than
( )
wm
10% of the mean displacement. WF = ∫w i
σ w w dw (4)

Typically, wm = 2 mm, while wi is determined as the

7. INTERPRETATION OF RESULTS smallest value of w where:
δ j ≥ 2 10 −4 lg for all j (5)
The stress-crack opening relationship is obtained
from the raw data of each test in the following way: (see also Fig. 5).
Stress σw is calculated from the load P simply by From a statistical analysis the characteristic value of
dividing with the cross sectional area at the notch An:

σw = P (1)
An
The crack opening w is calculated from the average
signal of the displacement transducers by subtracting the
average displacement at peak stress, neglecting the elastic
unloading. Thus, denoting the displacement readings of
the individual
_ n transducers δj, j = 1,2,...,n, the average
signal, δ , is calculated from:
n
∑
δ = 1 δj
n j =1
(2)
_
Denoting the average displacement at peak stress δ p,
the crack opening w is calculated from:
w = δ − δp (3)
(see also Fig. 4).
The stress-crack opening relationship σw(w) is obtained Fig. 5 – Schematic representation of the determination of wi
from corresponding values of σw and w with w > 0. based on a real data set. For the sake of simplicity only two trans-
From a series of tests of at least 6 specimens, the ducer signals are considered.

5
Materials and Structures/Matériaux et Constructions, Vol. 34, January-February 2001

–
the dissipated energy WF,k, as well as the mean value WF, 8. TEST REPORT
can be calculated. A characteristic stress-crack opening
diagram can then_ be obtained from the mean diagram The test report on a test series shall contain the fol-
based on all tests, σw(w), according to: lowing items:
• Geometry of all specimens.
( ) ( ) WF ,k
W
σ w ,k w = σ w w (6) • Description of method of placing and compaction
F of SFRC in the mould (if cast specimens are used).
The choice of fractile and confidence level for the • Description of coring procedure (if cored speci-
calculation of the characteristic value WF,k shall be made mens are used) and location and orientation of the core
in correspondence with national or international safety relative to the structure or structural element.
standards. _ • Description of curing and conditioning prior to
Furthermore, a mean value for wi, w i, is def ined testing.
according to: • In one diagram all the observed σw(w) – curves
t
from w = 0 to w = wm. _
wi = 1 ∑ ( )q
w (7) _ diagram the calculated σw(w) and the σw,k(w)
• In one
t q=1 i from w =_w i to w = wm along with information about the
value of w i and wm.
where (wi )q is wi for test q and t is number of tests. • Any other information typically required by the
relevant national standards.
Furthermore, it is recommended that the following
information is recorded:
• For each specimen observations of uniformity of
fibre distribution, preferably in the form of photographic
registration of the fracture surface.
• Type of testing equipment.