Professional Documents
Culture Documents
S h a h I
REFERENCE: Suaris, W. and Shah, S. P., "Inertial Effects in the In- and the test specimen is commonly measured by cementing strain
strumented Impact Testing of Cementltlous Composites," Cement, gauges to the tup. Assuming that the load measured by the tup is
Concrete, and Aggregates, CCAGDP, Vol. 3, No. 2, Winter 1981, pp. the same as the bending load experienced by the test beam, one
77-83.
can deduce the load versus deflection behavior of the test beam
ABSTRACT: The effects of specimen inertia on the instrumented im- and its dynamic fracture toughness from the observed load-time
pact testing of cementitious composites are discussed and a method for trace. The possibility of such calculations makes the instrumented
reducing such effects is proposed. A two degree of freedom model impact test a useful tool in evaluating the dynamic response of
capable of predicting the amplitude and period of inertial loading materials.
oscillations on the tup-load versus time-trace is presented. Results of
Many investigators have recognized that during an initial period
tests conducted on asbestos cement and fiber-reinforced concrete are
presented and are compared with those predicted using the analytical of time the load measured by the tup and that resisted by the beam
model. It is concluded that the model developed is satisfactory in pre- are not the same as a result of specimen inertia effects [1-8], which
dicting inertial loading effects. manifest as oscillations on the signal obtained from the instru-
mented tup. Server et al [7] have recommended that reliable load
KEYWORDS: cements, impact tests, strain rate, concretes, fiber- measurement be made after three periods of oscillations, after
reinforced concretes, dynamic loading, impulsive loading, fracture
toughness, modulus of rupture, energy, strains, debonding which time the effects of inertia can be ignored. They proposed the
following formula to predict the period of oscillations:
In recent years there has been a rapid increase in the use of the
r = 3.36 ( W / S 0) (EBCs) 1/2 (1)
instrumented impact test for evaluation of the dynamic response of
a wide range of materials. During an instrumented impact test where
(Fig. 1) the compressive force interaction between the striker (tup)
r = period of inertial oscillations,
W ----specimen width,
B = specimen thickness,
S O = sonic velocity of the material,
Cs = specimen compliance, and
E = Young's modulus of elasticity of the test material.
results of which validate the model, and the analytical model are
described in this paper. (6)
Analytical Model where g is the gravitational acceleration and v0 the impact velocity
of the hammer. To simplify the analysis, if it is assumed that the
A rigorous analysis of the impact test can be conducted by ana-
mass of the hammer is large when compared to the mass of the
lyzing the stress wave propagation within the specimen, tup, and
specimen, the solutions for xl and x 2 can be written as
anvils. As the first wave from the impact travels through the beam,
more of the body becomes involved in the motion and reflections
(0/ -I- fl)l/2 Ot'y
and dispersions will occur at all faces. However, when these tran-
sient phenomena have continued for some time, the motion will be-
X1(t) -- X/~fll/23/1/2vosin(wlt + 01) + x/2(c~ + fl)s/2
come periodic and simple solutions can be obtained for the ensuing
steady state of motion. X vOsin(~2t + q)2) (7)
The proposed two degree of freedom model that represents a OL
drop weight impact system is shown schematically in Fig. 2. The x2(t) -- "~(ot + f l ) l / Z f l l / 2 . y 1 / 2 v0sin(wlt + ~1)
falling hammer is represented by a lumped mass m h and the effec- o/
tive stiffness between the tup and the beam-specimen by a spring X/2(cx + t3) 3/2 v O s i n ( ° ° 2 t + 02) (8)
of stiffness k e. The beam is represented by an equivalent lumped
mass m s and an equivalent spring of stiffness k s . It is assumed that where w1 and w2 are the two natural frequencies given by
the anvils are infinitely stiff compared with the stiffness of the
beam-specimen. o~1 = (X/2 f l l / 2 v l / 2 ) / ( C ~ + f l ) I / 2 (9)
Assuming that the beam vibrates in its first mode, expressions
for m s and k s can be written as W2 = X/2 (Ot+fl) 1/2 (10)
k s = 7r4EI/213 (3)
g (~+fl)l/2 ]
where (11)
~1 = - - t a n - I x/~v ° 71/2B1/2
p = mass per unit length of the specimen material,
E = Young's modulus of the specimen material,
1 = moment of inertia of the beam cross section, and
02 = - - t a n -1 ,/~v ° (Ot+fl) 1/2' (12)
l = support span of the specimen.
The equations of motion for the modelled system can now be writ- ot = k e / 2 m s
ten as
fl = k s / 2 m s
(oh mOs) I~121+ (_k;e --ke ~12
l I:l ,4, .)1 = k e / 2 m h
where x I and x2 are the displacements of the masses, mh and m s The load measured by the tup is given by:
from their static equilibrium positions, and 21 and 22 the corre-
IXl°l
sponding accelerations. PT(t) = ke[Xl(t ) --x2(t)] + mhg (13)
The initial conditions for the motion are given by
The bending load carried by the beam can be expressed as:
= ks/2m h
Copyright by ASTM Int'l (all rights reserved); Mon Jun 17 18:15:57 EDT 2019
Downloaded/printed by
Univ de los Andes (Univ de los Andes) pursuant to License Agreement. No further reproductions authorized.
SUARIS AND SHAH ON CEMENTITIOUS COMPOSITES 79
12.0 and
8.0
{]ENDING L O A D - / - - ~ / / Note that the most significant term in Eq 15 is the first term. Thus
z
the energy dissipated by the hammer can be closely approximated
a
< 4.0
o as [1]:
A E o = E a [1 (Ea/Eo)] (20)
0
where
-2.0
o~2 0:4 o.'6 0.7 E a = v 0 ftoPT(t) dt (21)
TIME (m s)
(a)
and
8.0
Eo : 1/2 mhvo 2 (22)
LOADBENDIN
TUP-
G LO ~ - - - - ~ ~ ~ Figure 4 shows the plots of A E 0, EME, and ESD versus time for
6.0
the k e values of 87.5 M N / m (6.0 × 106 lbf/ft) and 5.84 M N / m
z (0.4 X 106 lbf/ft). It can be seen that for a low value of k e (that is,
v using a rubber pad) the energy consumed by bending ESD is
o 4.0
substantially smaller than the energy dissipated by the h a m m e r
A E 0. Thus when a padding material is introduced it becomes
2.0 much more important to consider the energy stored by the speci-
men tup interaction EME while calculating the specimen bending
energy.
I i
0.5 1.0 1.5 1.75
TIME ( m s) Experimental Program
(b)
A series of impact tests were carried out to evaluate the validity
FIG. 3--Analytical tup load and beam bending load curves for: (a) k e = of the above model in predicting the amplitude and period of iner-
8Z5 MN/m and (b) k c : 5.84 MN/m. tial load oscillations of the tup load versus time record. Tests were
3.5
The tup-load trace is seen to indicate the oscillatory behavior 3.0
which has been observed generally. The bending load taken by the
beam is seen to oscillate with the same frequency but about 180 ° z
out of phase with the tup load as was experimentally observed by
Venzi et al [5]. Figure 3b shows a tup-load time trace evaluated for ~ 2.0
a lesser value of k e (5.84 M N / m [0.4 × 106 lbf/ft]). As will be
shown later, this value corresponds to the case when a rubber pad 1,0
is used. Although the oscillatory behavior is still present, the
magnitude of the oscillations is found to have decreased con- / ," / \- ,.,
siderably and on" inflection points are observed instead of the
0.2 0,4 0,6 0.7
sharp peaks seen in the case with higher k e. Also significant is the TIME (m s)
smaller deviation between the tup load and the beam bending (o)
load. 7.O
For any time t after impact, expressions for the energy dissipated
6.0
by the hammer A E 0, the strain energy consumed by bending of the
beam ESD, the kinetic energy of the beam E/, and the elastic
z
energy absorbed by the machine EME resulting from the interac- 4.0
tion between the tup and the specimen can be written as:
i
uJ
A E 0 = Y2mh(v02--~l 2) + m h g x I + [(mhglke) + ( m h g l k s ) ] m h g 2.0
(15)
Copyright by ASTM Int'l (all rights reserved); Mon Jun 17 18:15:57 EDT 2019
Downloaded/printed by
Univ de los Andes (Univ de los Andes) pursuant to License Agreement. No further reproductions authorized.
80 CEMENT, CONCRETE, AND AGGREGATES
between the tup and the specimen was also investigated by con- i,i
ducting tests with two different rubber pads. A total of 108 speci-
mens were tested. The measured modulus of rupture of 2.5- by TIME: 0,2 ms/div
5-by 25-cm specimens tested at a stress rate of 0.14 MPa/s (20
psi/s) was 22.4 MPa (3250 psi).
Tests were also c o n d u c t e d on fiber-reinforced concrete beam
specimens with dimensions of 5 by 10 by 30 and 5 by 10 by 50 cm. .z ze
The mortar mixing proportions by weight for the specimens were ZtD
(No. 4) sieve and ASTM Type III portland cement (per ASTM
Specification for Portland Cement [C 150]) were used. The average
uniaxial compressive strength of the mortar matrix measured on W
Copyright by ASTM Int'l (all rights reserved); Mon Jun 17 18:15:57 EDT 2019
Downloaded/printed by
Univ de los Andes (Univ de los Andes) pursuant to License Agreement. No further reproductions authorized.
S U A R I S A N D S H A H ON C E M E N T I T I O U S C O M P O S I T E S 81
20,0 15.0
O NO PAD 0
15.0 zlO.O
/ Ke=BT:6 MN/M
a
A S
,o.o
/ ~ " (I.OXl0' Ib/ft)
d. 5.0
QH
F-
5.0
oJ 0,2 O.3
TIME (m s)
(o)
I00 200 300
9.0
IMPACT VELOCITY (era/s)
?
13-
Pr(~\\
point for the specimen tested with a r u b b e r p a d (Fig. 7b). T h e poor / Pz(OBSERVE[))
~- 3.0
agreement for the specimen without any r u b b e r p a d may be a re-
sult of the actual value k e being different from t h a t assumed in the
analysis. The actual value of k e is influenced by the roughness of
the b e a m contact zone a n d possible local crushing, which may vary o12 0:4 0.6
TIME (rn s)
from one specimen to another. (b)
From the modulus of rupture ( M O R ) values shown in Table 1, it
is evident that the values of M O R for specimens tested without any FIG. 7--Comparison of analytical attd experimental tup load versus
r u b b e r pads are considerably higher t h a n those with the r u b b e r time curves for: (a) test with no rubber p a d used and (b) using Pad B.
pads, especially at the highest impact velocity tested. T h e M O R
values of specimens tested without any p a d d i n g material at the im-
pact velocity of 294 c m / s are four times the static M O R value a n d solely to the strain rate effects b u t is largely a consequence of tup-
this increase is substantially higher t h a n t h a t observed by other specimen interactions. Note t h a t the average impact strength of
researchers for cementitious composites [12-14]. For example, an the specimens tested with r u b b e r pads is a b o u t 1.7 times the static
increase in tensile strength of a b o u t 1.5 to 2.5 times the static strength of 22.4 M P a (3250 psi).
strength has been observed from "split H o p k i n s o n b a r tests" on Some of the results of tests c o n d u c t e d on concrete a n d fiber-
concrete specimens [12]. T h u s the unusually high value of M O R reinforced concrete specimens are given in T a b l e 2. Both the ex-
for specimens tested without any r u b b e r pads c a n n o t b e a t t r i b u t e d perimental a n d the predicted values of the inertial load PI a n d the
Copyright by ASTM Int'l (all rights reserved); Mon Jun 17 18:15:57 EDT 2019
Downloaded/printed by
Univ de los Andes (Univ de los Andes) pursuant to License Agreement. No further reproductions authorized.
82 CEMENT, CONCRETE, AND AGGREGATES
Impact Effective
Dimensions Mix Velocity Stiffness PI, kN ti, ms
B by D by L, Proportions, v, k e,
cm Cement : Sand : Water: Fibers cm/s MN/m Analytical Experimental Analytical Experimental
time tl at which PI occurs are shown. The density of concrete used 7.5 m m corresponds approximately to 15 times the deflection at
for the calculations was 2240 k g / m 3 (140 l b / f t 3) and a value of the peak load. It is seen t h a t the energy absorbed by the b e a m s
42.7 GPa (6.2 × 106 psi) was used for its Young's modulus, ob- under impact loading is about 30% higher t h a n that absorbed un-
tained from resonance tests on cantilever beams of fiber-reinforced der static loading. Since for steel fiber-reinforced concrete the
concrete (2.5 by 2.5 by 10 cm). primary source of energy absorption is during the fiber pulling-out
The ke values for the specimens tested with the rubber pads were process, the above results indicated that d e b o n d i n g processes of
taken as the stiffness values of the corresponding rubber pads, ob- fibers are not very rate sensitive.
tained as described earlier. A k e value of 175.2 M N / m (12.0 X 106 Figure 8 shows a typical load versus time and the strain versus
lbf/ft) was found to fit well with the observed results for the
specimens tested without any r u b b e r pad. A good agreement be-
tween the predicted and observed PI values can be seen in Table 2.
The experimentally observed period of inertial load oscillations
15,0
was, however, slightly greater t h a n the predicted values.
The values of M O R of fiber-reinforced concrete specimens are
shown in Table 3. The M O R values were calculated from the mea-
I0.0
sured peak load with the assumption that the material behaves z
elastically. The stress rate was calculated from the observed peak
load and the corresponding time. It can be seen that the dynamic
5.0
modulus of rupture is about twice t h a t of the specimen tested at
the slow static rate. This increase in the value of M O R caused by
impact loading is comparable to that observed by other in-
vestigators [12-14]. ,'.o z'.o ~:o 4'.0 .5.0
The energy absorbed by the b e a m ESD was calculated from the TIME (m s)
(o)
load-time trace A E 0 and subtracting the energy absorbed by the
rubber pad EME and the kinetic energy of the b e a m E 1.
These values when the central deflection of the b e a m was
7.5 mm (0.30 in.) are shown in Table 3. Note that the deflection of
-I00
~ ~ LJ / . . . . c.~UaTED
TABLE 3--Fiber-reinforced concrete beams under static and
x -200
1 /c ooo
impact loading."
_z
Bending Energy in N.m at n.-
Copyright by ASTM Int'l (all rights reserved); Mon Jun 17 18:15:57 EDT 2019
Downloaded/printed by
Univ de los Andes (Univ de los Andes) pursuant to License Agreement. No further reproductions authorized.
SUARIS AND SHAH ON CEMENTITIOUS COMPOSITES 83
Copyright by ASTM Int'l (all rights reserved); Mon Jun 17 18:15:57 EDT 2019
Downloaded/printed by
Univ de los Andes (Univ de los Andes) pursuant to License Agreement. No further reproductions authorized.