See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/273370255

Flexural Behavior of Concrete-Filled Circular Steel Tubes under High-Strain

Rate Impact Loading

Article in Journal of Structural Engineering · March 2012

DOI: 10.1061/(ASCE)ST.1943-541X.0000464

CITATIONS READS
80 421

3 authors:

Yaohua Deng C. Y. Tuan

Iowa State University University of Nebraska at Lincoln
20 PUBLICATIONS 334 CITATIONS 64 PUBLICATIONS 1,354 CITATIONS

SEE PROFILE SEE PROFILE

Yan Xiao
Zhejiang University
341 PUBLICATIONS 9,528 CITATIONS

SEE PROFILE

All content following this page was uploaded by Yaohua Deng on 28 March 2016.

The user has requested enhancement of the downloaded file.

 Technical Note

Flexural Behavior of Concrete-Filled Circular Steel Tubes

under High-Strain Rate Impact Loading
Y. Deng, S.M.ASCE1; C. Y. Tuan, Ph.D., P.E., F.ASCE2; and Y. Xiao, Ph.D., P.E., M.ASCE3

Abstract: Nine simply supported circular steel concrete-filled tubes (CFTs), two circular steel posttensioned concrete-filled tubes (PTCFTs),
and one circular steel fiber–reinforced concrete-filled tube (FRCFT) have been tested in an instrumented drop-weight impact facility. The
Downloaded from ascelibrary.org by Iowa State University on 04/21/13. Copyright ASCE. For personal use only; all rights reserved.

weight and the height of the drop-weight were varied to cause failure in some test specimens. The failure modes and local damages in those
specimens have been investigated extensively. Failure in the steel tubes was commonly tensile facture or rupture along the circumference.
Concrete core in the impact area commonly crushed under compression and cracked under tension. The use of prestressing strands and steel
fibers significantly restrained the concrete tension cracks in the PTCFT and FRCFT specimens, respectively. The experimental results
are analyzed in the context of principles of energy and momentum conservation. DOI: 10.1061/(ASCE)ST.1943-541X.0000464.
© 2012 American Society of Civil Engineers.
CE Database subject headings: Dynamic response; Concrete; Tubes; Steel; Posttensioning; Strain rates; Impact tests; Fiber-reinforced
materials; Flexural strength.
Author keywords: Dynamic response; Concrete; Tubes; Steel; Posttensioning; Strain rates; Impact tests; Fiber-reinforced concrete.

Introduction The adequacy of the procedure for predicting the flexural impact
resistance capacity of RC members was validated by a series of
Many researchers have studied the behavior of reinforced concrete physical experiments. Saatci and Vecchio (2009a) investigated
(RC) beams under impact by drop-weight tests (Bentur et al. 1986; the effect of shear capacity on the impact behavior of RC beams.
Kishi et al. 2001, 2002; Fujikake et al. 2009; Silva et al. 2009; All the specimens had identical longitudinal reinforcement, but
Saatci and Vecchio 2009a, b). Bentur et al. (1986) demonstrated with varying shear reinforcement ratio. Saatci and Vecchio
that the inertial effect should be separated from the total impact (2009b) also developed a disturbed stress field model on the basis
load measured at drop hammer, because only a small portion of of a smeared rotating crack approach to modeling shear behavior
the total load caused beam bending. Kishi et al. (2001, 2002) in- under impact conditions.
vestigated the shear and the flexural failure modes of RC beams The behavior of fiber-reinforced concrete (FRC) beams under
under impact loading. They concluded that the impact resistance impact loading has been investigated by Banthia (1987), Wang et al.
may be more rationally estimated from the maximum reaction force (1996), and Bindiganavile (2003). Banthia (1987) used fibers, par-
rather than from the maximum impact force and that both the shear ticularly the steel fibers, to significantly increase the ductility and
and the bending capacities of RC beams under impact were much the impact resistance of FRC beams. Wang et al. (1996) subjected
higher than the corresponding static capacities. Fujikake et al. small beams with various volumes of steel and polypropylene fi-
(2009) examined the overall flexural failure of RC beams under bers under drop-weight impact loading. It was reported that steel
impact by taking into account the influence of drop height and fibers increased the fracture energy much more than polypropylene
the effect of longitudinal steel reinforcement. A two–degree-of- fibers did. Although breakage of steel fibers was the primary failure
freedom mass-spring-damper model was developed, and the pre- mechanism when volumetric fraction was less than 0.5%, fiber
dicted results agreed well with the experimental results. Silva et al. pullout was the primary mechanism at more than 0.75% volumetric
(2009) presented the basis of a procedure using Monte Carlo sim- fraction. Bindiganavile (2003) evaluated the performance of four
ulations in conjunction with a displacement-based design method. types of fibers in FRC beams under impact testing.
1
The impact behavior of steel tubes filled with foam and concrete
Ph.D. Candidate, Univ. of Nebraska–Lincoln, Durham School of has recently been studied by drop-weight tests (Mantena and Mann
Architectural Engineering and Construction, Construction Engineering
2003; Wang et al. 2007; Bambach et al. 2008; Remennikov et al.
and Management, Peter Kiewit Institute, 1110 S. 67th St., Omaha, NE
68182-0571 (corresponding author). E-mail: [email protected] 2011) and by blast tests (Fujikura et al. 2008). Mantena and Mann
2
Visiting Professor, Hunan Univ., College of Civil Engineering, Chang- (2003) investigated the dynamic response of three high-density pol-
sha, 410082, China; Professor, Univ. of Nebraska–Lincoln, Civil Engineer- ymeric structural foams used as filler inside a circular steel tube
ing, Peter Kiewit Institute, 1110 S. 67th St., Omaha, NE 68182-0178. with an instrumented impact tester. The results indicate the one
3
Dean, College of Civil Engineering, Hunan Univ., Changsha, 410082, with higher foam density exhibited larger energy-absorbing capac-
China; Professor, Univ. of Southern California, Dept. of Civil and Envir- ity. Wang et al. (2007) conducted experimental and numerical sim-
onmental Engineering, Los Angeles, CA 90089. ulations to study the dynamic response of simply supported beams
Note. This manuscript was submitted on October 22, 2009; approved on
of steel concrete-filled tubes (CFTs) with identical outer-diameter
June 15, 2011; published online on September 16, 2011. Discussion period
open until August 1, 2012; separate discussions must be submitted for in- but different tube thicknesses under lateral impact loading. The re-
dividual papers. This technical note is part of the Journal of Structural sults indicate that diameter-to-thickness ratio has significant influ-
Engineering, Vol. 138, No. 3, March 1, 2012. ©ASCE, ISSN 0733- ences on the energy-absorbing capacity. Bambach et al. (2008)
9445/2012/3-449–456/$25.00. subjected square hollow-section and concrete-filled steel square

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / MARCH 2012 / 449

J. Struct. Eng. 2012.138:449-456.

 hollow-section beams to low-velocity, large-mass, transverse the circular tube satisfies Eq. (1). The selected section had a section
impacts at the beam midspan. Three section sizes were tested as slenderness of 80.5 and is deemed adequate for this study.
fully clamped beams. Remennikov et al. (2011) compared the The test specimens were made from seamless steel tubes. After
behavior of square hollow sections filled with rigid polyurethane covering one end of the steel tubes with wood plates, the steel tubes
foam (RPF) and with concrete undergoing transverse impact load- were erected vertically, and fresh concrete was poured into the steel
ing. The CFTs had higher impact resistance and energy-absorbing tubes from the open top. Small air vents were drilled along the
capacity than did the steel tubes filled with RPF. Fujikura et al. length of the steel tubes to ensure complete fill of concrete. For
(2008) investigated the dynamic response of a multi-column the PTCFT beams, 0.5-in. (13-mm) seven-wire, low-relaxation,
pier-bent with CFT round columns under blast loading. Simplified prestressing strands were used for post-tensioning. Spacers were
blast analysis was performed and compared with the experimental placed along a posttensioning strand at equal spacing, as shown
results. Fujikura et al. (2008) reported that the prototype bridge in Fig. 1(a), to ensure that the strand was placed at the center along
CFT columns can be designed to provide both satisfactory seismic the steel tube. Because the thickness of the spacers was small, as
performance and adequate blast resistance. shown in Fig. 1(b), the spacers would not prevent concrete from
This paper presents an experimental investigation on drop- filling the tubes. Two steel end plates were fitted inside each steel
Downloaded from ascelibrary.org by Iowa State University on 04/21/13. Copyright ASCE. For personal use only; all rights reserved.

weight impact at midspan of simply supported CFT beams under tube such that uniform compressive prestress could be applied on
various velocities. The posttensioning and fiber-reinforcing effects the concrete infill. To minimize prestress loss, posttensioning force
on the impact resistance of the CFT beams are also evaluated. was applied shortly before the impact test.

Materials
Experimental Program
Three coupons were taken from each group of the steel tube spec-
Nine circular steel CFTs, two posttensioned CFTs (PTCFTs), and imens for tensile tests to determine the stress-strain relationship.
one steel fiber–reinforced CFT (FRCFT) have been tested in a Self-consolidating concrete containing silicon powder, water
drop-weight impact facility at Hunan University, Changsha, China. reducer, and expansive admixture was used to cast the specimens.
The drop-weight consists of poises on top and a tup at the bottom. Corrugated steel fibers of 0.8-mm diameter and 35-mm long,
The weight is controlled by the number of poises and is adjustable amounted to approximately 1.5% by volume, were mixed into
from 125 kg to a maximum of 825 kg, with fourteen 50-kg steel the concrete for the FRCFT specimen. The tensile strength of
plates. The drop-weight is clamped by the clamp check at top, run- the steel fibers is 590 MPa. For each batch of the mixed concrete,
ning freely along two rails, and the impact velocity can be calcu- three 150 × 150 × 150-mm cubes and three 70 × 150-mm cylin-
lated from the potential energy of a free-falling weight. A control ders were cast and cured under the same condition with the spec-
system is used to raise, lower, or release the drop-weight. The maxi- imens. The nominal dimensions of the test specimens and the
mum drop height is 16 m. materials properties are Shown in Table 1.

Fabrication of Test Specimens Test Setup and Procedures

A slenderness limit of λey ¼ 125, is proposed by Bradford et al. As shown in Fig. 2, the test beams were simply supported and im-
(2002), whereas the section slenderness λs is given by pacted at the midspan by a drop hammer. The drop height was var-


 ied from 3.022 to 14.388 m, and the drop-weight was varied from
D fy 525 to 675 kg. Initially, the specimens were impacted with a 3.15-m
λs ¼ ≤ λey ¼ 125 ð1Þ
t 250 clear span by a cylinder-shaped tup. Those specimens sustained
large flexural deformations without rupture. The clear span of
Bradford et al. (2002) stated that a circular steel section would the test specimens was subsequently reduced to 1.88m and im-
reach the yield strength postbuckling, if the section slenderness of pacted by a wedge-shaped tup in an attempt to induce tensile

Fig. 1. Strands and spacers: (a) inserting the strand into the steel tube; (b) dimension of the spacers

450 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / MARCH 2012

J. Struct. Eng. 2012.138:449-456.

 Table 1. Specimen Properties
Identification D × t (mm × mm) D=t ratio λs f y (MPa) f u (MPa) f 0c (MPa) f 0t (MPa) L (mm) Lo (mm)
CFTa1 219 × 3:7 59.2 80.5 340 438 58 6.60 3,400 3,150
CFTa2 219 × 3:7 59.2 80.5 340 438 58 6.60 3,400 3,150
CFTb1 219 × 3:7 59.2 80.5 340 438 58 6.60 3,400 1,880
CFTb2 219 × 3:7 59.2 80.5 340 438 58 6.60 3,400 1,880
CFTb3 219 × 3:7 59.2 80.5 340 438 58 6.60 3,400 1,880
CFTb4 219 × 3:7 59.2 80.5 340 438 58 6.60 3,400 1,880
CFTb5 219 × 3:7 59.2 80.5 340 438 58 6.60 3,400 1,880
CFTb6 219 × 3:7 59.2 80.5 340 438 58 6.60 3,400 1,880
CFTc1 219 × 3:7 59.2 80.5 340 438 45.5 5.53 3,400 1,880
PTCFT1 219 × 3:7 59.2 80.5 340 438 58 6.60 3,400 1,880
Downloaded from ascelibrary.org by Iowa State University on 04/21/13. Copyright ASCE. For personal use only; all rights reserved.

PTCFT2 219 × 3:7 59.2 80.5 340 438 58 6.60 3,400 1,880
FRCFT 219 × 3:7 59.2 80.5 340 438 45.5 7.82 3,400 1,880
Note: D = tube outer diameter; t = tube thickness; λs = section slenderness f y = yield strength of steel; f u = ultimate strength of steel; f 0c = concrete strength at
test date; f 0t = concrete split tensile strength at test date; L = length of specimens; and Lo = clear span.

Fig. 2. Schematic test set up

rupture at midspan in the steel tubes. Kishi et al. (2001, 2002) im- Experimental Results
pacted RC beams using different tup shapes and reported that the
impact behavior of the RC members was not significantly affected Failure Modes
by the tup shape.
Strain gauges were mounted at midheight and bottom in the ax- Four damage modes—no crack, crack at the bottom, crack up to
ial direction at midspan of each specimen. Strain gauges were also lower half of the circumference, and rupture—were observed at
the midspan of the beam specimens, as shown in Table 2. There
mounted in both axial and hoop directions on the top and the bot-
was significant local damage in the CFTs produced by the drop
tom at a section 0.23 m from the support to capture stress waves
hammer. The high-speed video revealed that, for the beams not in-
propagating from the midspan to the support without the interfer- stantly ruptured, the midspan moved up toward the original plane
ence of reflected stress waves from the support. Reaction forces and vibrated until the energy was damped out. The hammer
were measured by two 2,000-kN load cells at the supports. The bounced up a small distance relative to the beam and was briefly
deflection at midspan was recorded by a displacement sensor, detached from the beam. When a CFT beam failed in rupture, a
which was spring loaded under the specimen such that it would crack initially formed at the bottom of the steel tube and quickly
not be detached from the specimen during testing. The data- propagated upward through the beam section, and the beam sub-
acquisition system consisted of dynamic strain gauge amplifiers sequently ruptured into two halves. Slices of concrete core fell out
and Tektronix 4034 oscilloscopes. The amplifiers also served as of the crack from the steel tube, indicating flexural failure at the
low-pass filters at 100 kHz. The voltage signals were displayed beam midspan. For instance, CFTb4 failed in rupture under a
and digitized at a sampling frequency of 1 MHz by the oscillo- 675-kg weight dropped from 8.5-m high; snapshots from high-
scopes and stored on a computer. A high-speed digital video cam- speed video are shown in Fig. 3.
era was used to document the dynamic response of beams during
the impact of drop-weight. The images were recorded at 1,000 Concrete Damage Patterns inside Steel Tubes
frames per second with a resolution of 1;280 × 512 pixels. After The steel tubes were cut along the length with a blow torch to reveal
each impact testing, the permanent deflection at midspan of the test the damage patterns of the concrete core. Fig. 4(a) shows that the
specimen was measured with a ruler. cracks at midspan of CFTb2 propagated from the bottom to the top

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / MARCH 2012 / 451

J. Struct. Eng. 2012.138:449-456.

 Downloaded from ascelibrary.org by Iowa State University on 04/21/13. Copyright ASCE. For personal use only; all rights reserved.

Table 2. Summary of Drop-Weight Impact Test Data

Identification M (kg) H (m) Ei (kJ) W if (kJ) Ea (kJ) δm (mm) δ f (mm) F m (kN) Ru (kN) K m (kN-s) I f (kN-s) I r (kN-s) Damage at mid span
CFTa1 525 3.022 15.56 — 8.71 93 69 1,100 450 4.04 — — No crack
CFTa2 625 10.170 62.35 — 39.13 337 310 1,718 807 8.83 8.71 9.26 No crack
CFTb1 675 6.000 39.73 — 25.34 135 120 910 1,590 7.32 — — No crack
CFTb2 675 6.016 39.84 29.49 26.44 141 125 923 1,614 7.33 7.36 8.13 No crack
CFTb3 675 6.056 40.10 30.10 28.55 142 135 953 1,385 7.36 7.50 7.82 Crack at bottom
CFTb4 675 8.492 56.23 34.80 — — — 996 1,627 8.71 — — Rupture
CFTb5 675 10.046 66.52 — — — — 1,132 1,756 9.47 — — Rupture
CFTb6 675 14.388 95.27 52.50 — — — 1,231 1,409 11.34 — — Rupture
CFTc1 675 5.533 36.64 25.56 30.70 152 148 876 1,560 7.03 7.11 7.71 Crack half circumference
PTCFT1 675 6.046 40.04 29.40 23.82 122 105 969 1,402 7.35 6.88 7.97 No crack
PTCFT2 675 8.450 55.95 42.26 32.89 193 145 1,092 1,592 8.69 8.80 8.77 Crack half circumference
FRCFT 675 5.961 39.47 34.26 26.27 141 126 766 1,243 7.30 7.53 7.92 No crack
Note: M = drop-weight; H = drop height; E i = initial kinetic energy of drop-weight; W if = work done by impact force; Ea = energy absorbed by plastic hinge; δ m = maximum dynamic deflection; δ f = residual
deflection; F m = maximum impact force; Ru = maximum reaction force; K m = initial momentum of drop-weight; I f = impulse of impact force; and I r = impulse of reaction force.

452 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / MARCH 2012

Table 2.

J. Struct. Eng. 2012.138:449-456.

Impact Forces and Reaction Forces

Concrete-Filled Tubes with 1.88-m Span

Fig. 3. Failure sequence of concrete-filled tube b4 during impact

cracks in tension zone and provided additional loading capacity.

readings from the two supports. The maximum impact forces F m

bottom to top, as shown in Fig. 4(d). Therefore, both steel fibers

Three regular CFTs were subjected to the impact of a 675-kg

son, CFTb4 ruptured when the crack propagated completely from

and the maximum reaction forces Ru from all the tests are shown in
mer. The total reaction force was obtained by adding the load cell
strain gauges mounted symmetrically on the rigid base of the ham-
and tension strands served to significantly restrain the concrete
crete of PTCFT2 crushed, indicating that PTCFT2 did not reach its
height of the concrete core in FRCFT and PTCFT2, as shown in
of the section without rupture. The cracks did not go beyond half

The impact force was obtained by averaging readings from the four

tom of the steel tube in CFTb3, whereas CFTb1 and CFTb2 had
weight dropped from 6 m at midspan. A crack initiated at the bot-
ultimate strength until the concrete crushed at the top. For compari-
Figs. 4(b) and 4(c), respectively. Fig. 4(c) shows that the top con-
Downloaded from ascelibrary.org by Iowa State University on 04/21/13. Copyright ASCE. For personal use only; all rights reserved.

Fig. 4. Concrete damage at midspan after removing steel tubes: (a) CFTa2; (b) FRCFT; (c) PTCFT2; (d) CFTb4

large deformation without fracture. As an example, the time histor- the concrete core. The PTCFTs were subjected to the impact of a
ies of the impact and reaction forces from the CFTb2 test are shown 675-kg weight dropped from 6 and 8.5 m at midspan. Similar to the
in Fig. 5. The impact force reached its peak value on contact with CFT impact tests, a plateau was present in the impact force time
the specimen and decreased as a result of beam deflection. If the history of the PTCFT1 test (6-m drop height). A crack formed
beam moved downward faster than the hammer, the hammer would at the bottom of the midspan section of PTCFT2 (8.5-m drop
detach briefly. There would be a secondary peak when the hammer height) and propagated upward during impact. A plateau was
re-established contact with the specimen. There is a plateau in the not present in the impact force of PTCFT2 because the crack
impact force time history as the hammer and the specimen moved formed up to half of the steel tube circumference.
downward together, as shown in Fig. 5. However, there was no pla-
teau if the specimens failed in rupture. The time lag of the reaction Comparison of Concrete-Filled Tubes with
force relative to the impact force would represent the time required Posttensioned Concrete-Filled Tubes with 1.88-m Span
for a shear wave to travel from midspan to the support because
The dynamic deflection time histories at the midspan of CFTs and
shear waves travel faster than bending waves.
PTCFTs are shown in Fig. 6. Under the same 6-m drop height,
Three additional regular CFTs were subjected to the impact of a
PTCFT1 had approximately 16% less maximum dynamic deflec-
675-kg weight dropped from 8.5 , 10 and 14.4 m at midspan. All
tion than CFTb2. The final deflections were measured by a ruler
three specimens ruptured into two halves.
when the beams were at rest. The maximum dynamic deflection
Posttensioned Concrete-Filled Tubes with 1.88-m Span δ m and the residual deflection δ f are shown in Table 2.
From the results in Table 2, the impact resulting from a 675 -kg
Two CFTs were posttensioned to 166 kN (approximately 75% of weight dropped from 6 m cracked the bottom of regular CFTb2
the ultimate strength) using one prestressing strand in the center of

Fig. 6. Midspan deflections of concrete-filled tube and posttensioned

Fig. 5. Impact and reaction forces from concrete-filled tube b2 test concrete-filled tube specimens

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / MARCH 2012 / 453

J. Struct. Eng. 2012.138:449-456.

 with 125-mm permanent deflection at midspan, whereas the same
weight dropped from 8.5 m completely ruptured CFTb4. In com-
parison with the CFTs, the impact resulting from a 675-kg weight
dropped from 6 m produced only 105-mm permanent deflection in
PTCFT1 without cracking, whereas the same weight dropped from
8.5 m caused 145-mm permanent deflection in PTCFT2 with no
rupture. Therefore, PTCFTs had superior impact resistance to
the regular CFTs.

Comparison of Concrete-Filled Tube c1 and Fiber-

Reinforced Concrete-Filled Tube with 1.88-m Span
The concrete compressive strength used for both CFTc1 and FRCFT
specimens was 45.5 MPa. The FRCFT specimen showed no cracks
Downloaded from ascelibrary.org by Iowa State University on 04/21/13. Copyright ASCE. For personal use only; all rights reserved.

from a 6-m drop height, whereas the CFTc1 had crack propagated up
to the lower half of the steel tube circumference from a 5.5-m drop
height. Results showed that CFTc1 and FRCFT sustained 148- and
126-mm permanent deflections at midspan, respectively. These re-
sults indicated that the use of steel fibers effectively restrained the Fig. 8. Strain time histories on a section 0.23 m from the support
tensile cracks in the concrete core and hence increased the impact
resistance. Therefore, FRCFT would have better impact resistance section 0.23 m from the support at 10.5 ms. At that time instant, the
than regular CFTs if steel fibers are mixed in the concrete with ad- axial strain was tensile on the top of the section and compressive on
equate volumetric fraction (i.e., approximately 1.5%). the bottom. This was caused by the inertia effect that there was a
convex curvature (or negative moment) in the beam at the shear stress
Concrete-Filled Tubes with 3.15-m Span
wave front, whereas the reaction forces at the supports were zero.
The structural response of the two CFTs with 3.15-m span was sig- Similar to the impact force and reaction force time histories shown
nificantly different from those of the CFTs with 1.88-m span under in Fig. 7, there was also a plateau in the strain time histories near the
drop-weight impact. Both the impact force and reaction force from support. Within this plateau, the “quasi static” impact force caused
the CFTa2 test had a significant plateau, and the peak value of the concave curvature (or positive moment) in the beam. The hoop stress
impact force was much larger than that of the reaction force, as in the steel tube was tensile at the top and compressive at the bottom,
shown in Fig. 7. The forces within the plateaus were nearly con- opposite to the axial stress at the respective locations. The impact
stant, indicating a “quasi static” loading condition. The structural force diminished at approximately 100 ms, and the beam started
response of the beam is predominantly flexural rather than shear to vibrate freely afterward, as evidenced by the sinusoidal waves
within the plateau. in the strain time histories.

Strains on Steel Tubes

Two strain gauges were mounted on each specimen at midheight and
Analysis of Experimental Data
the bottom of the midspan section in the axial direction. The average During an impact event, the momentum is conserved, and energy of
strain rate at the midspan was calculated to be approximately 3:3 s
1 the system may be dissipated because of friction and damping. The
for specimen CFTb2. Four other strain gauges were mounted on the experimental results are analyzed in the context of energy and mo-
top and the bottom at a section 0.23 m from the support in both axial mentum to validate the drop-weight testing.
and hoop directions (Fig. 2). These four gauges were expected to
capture the stress waves from midspan to the support without re- External Work by Impact Force
flected stress wave from the support. The strain data obtained from
CFTa2 are shown in Fig. 8, in which the stress wave arrived at the The product of impact force and midspan displacement of the beam
equals the work done to the beam by the drop-weight W if . For in-
stance, the relationship between the impact force and midspan de-
flection of CFTb2 is shown in Fig. 9. The external work W if is
simply the area under the curve; the value is given in Table 2.
The area under the sloped portion at the end of the curve represents
the elastic strain energy restored, and the permanent deflection of
CFTb2 was approximately 125 mm. The external work is a good
estimate of the impact energy necessary to cause a specimen to
fracture or rupture. As shown in Table 2, under the same impact
condition, the PTCFTs and the FRCFT required higher W if to cause
fracture than the regular CFTs did.

Kinetic Energy of Drop-Weight

The initial kinetic energy of the drop-weight on impact may be cal-
culated from its potential energy:
E i ¼ MgH ð2Þ

in which M = mass of the hammer; g = gravitational acceleration;

Fig. 7. concrete-filled tube a2 with 3.15-m span under 625 kg at 10 m
and H = drop height of the hammer. These initial kinetic energy

454 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / MARCH 2012

J. Struct. Eng. 2012.138:449-456.

 Summary

Nine simply supported circular steel CFTs, two circular steel

PTCFTs, and one circular steel FRCFT have been tested in an in-
strumented drop-weight impact facility. Failure in the steel tubes
was commonly tensile facture or rupture along the circumference.
Concrete core in the impact area commonly crushed under com-
pression and cracked under tension. However, the use of prestress-
ing strands and steel fibers significantly restrained tension cracks in
the concrete, and the PTCFTs, and steel FRCFT had superior im-
pact resistance to the regular CFTs.
The accuracy of the experimental data was validated against en-
ergy and momentum principles. It is recommended that future tests
should consider different boundary conditions, impact loading,
Downloaded from ascelibrary.org by Iowa State University on 04/21/13. Copyright ASCE. For personal use only; all rights reserved.

diameter-to-thickness ratios of the steel tube, steel yield strength,

and concrete compressive strength.

Fig. 9. Impact force versus mid span deflection from concrete-filled

tube b2 test Acknowledgments

The research was conducted under the collaboration between

values E i are compared with the external work done by the impact the University of Nebraska–Lincoln and the Hunan University of
forces W if in Table 2. It is evident that the external work done by China. Tests described in this paper were conducted at the Ministry
the impact force is always less than the initial kinetic energy of the of Education Key Laboratory of Building Safety and Efficiency,
drop-weight owing to energy loss from impact. the Center for Integrated Protection Research of Engineering Struc-
tures (CIPRES) at the Hunan University. The authors wish to thank
Plastic Strain Energy in Concrete-Filled Tube Boshen Chen for his timely assistance in conducting the impact
Specimens tests. Partial support for this project was provided by the visiting
professorship at the Hunan University. The financial support from
A plastic hinge was formed at the midspan of the specimens that the China Scholarship Council is also gratefully acknowledged.
sustained the impact loading without rupture. The strain energy ab-
sorbed by the plastic hinge of a CFT beam Ea can be calculated as
E a ¼ 2M p θ ð3Þ References
Bambach, M. R., Jama, H., Zhao, X. L., and Grzebieta, R. H. (2008).
in which θ = rotation at the supports; and M p = dynamic plastic
“Hollow and concrete filled steel hollow sections under transverse im-
moment capacity. pact loads.” Eng. Struct., 30(10), 2859–2870.
2δf Banthia, N. P. (1987). “Impact resistance of concrete.” Ph.D. thesis, Univ.
θ ¼ tan
1 ð Þ ð4Þ of British Columbia, British Columbia, Canada.
Lo Bentur, A., Mindess, S., and Banthia, N. (1986). “The behavior of con-
crete under impact loading: Experimental procedures and method of
in which δ f = the residual deflection at midspan; and Lo = clear span analysis.” Mater. Constr., 19(5), 371–378.
of the specimen. Bindiganavile, V. (2003). “Dynamic fracture toughness of fiber-reinforced
The external work done by the impact force W if is theoretically concrete.” Ph.D. thesis, Univ. of British Columbia, British Columbia,
equal to the strain energy absorbed by the plastic hinge Ea plus the Canada.
strain energy recovered from elastic rebound of a specimen. As Bradford, M. A., Loh, H. Y., and Uy, B. (2002). “Slenderness limits for
shown in Table 2, the PTCFT and FRCFT specimens recovered filled circular steel tubes.” J. Constr. Steel Res., 58(2), 243–252.
more elastic strain energy than the regular CFTs did. Fujikake, K., Li, B., and Soeun, S. (2009). “Impact response of reinforced
concrete beam and its analytical evaluation.” J. Struct. Eng., 135(8),
Impulse and Momentum 938–950.
Fujikura, S., Bruneau, M., and Lopez-Garcia, D. (2008). “Experimental
The initial momentum of the drop-weight on impact K m can be investigation of multihazard resistant bridge piers having concrete-filled
expressed as steel tube under blast loading.” J. Bridge Eng., 13(6), 586–594.
pffiffiffiffiffiffiffiffiffi Kishi, N., Mikami, H., Matsuoka, K. G., and Ando, T. (2002). “Impact
K m ¼ M 2gH ð5Þ behavior of shear-failure-type RC beams without shear rebar.” Int. J.
Impact Eng., 27(9), 955–968.
The velocity of the drop-weight must be zero when the maxi- Kishi, N., Nakano, O., Matsuoka, K. G., and Ando, T. (2001). “Experimen-
mum dynamic deflection of a CFT beam δ m was reached. The con- tal study on ultimate strength of flexural-failure-type RC beams under
servation of momentum requires that the impulse of the impact impact loading.” Trans. 16th Int. Conf. on Structural Mechanics in
force up to this time instant I f be equal to the initial momentum Reactor Technology (SMiRT), International Association for Structural
of the drop-weight. The data given in Table 2 validate the accuracy Mechanics in Reactor Technology, Raleigh, NC.
Mantena, P. R., and Mann, R. (2003). “Impact and dynamic response of
of the impact tests conducted. Further more, if the drop-weight and
high-density structural foams used as filler inside circular steel tube.”
the CFT beam are considered as a single system, the external force Compos. Struct., 61(4), 291–302.
acting on the system will be the support reaction force. When the Remennikov, A. M., Kong, S. Y., and Uy, B. (2011). “Response of
velocity of the system reached zero, the impulse of the reaction foam- and concrete-filled square steel tubes under low-velocity impact
force up to this time instant I r should also be equal to K m . As shown loading.” J. Perform. Constr. Facil., 10.1061/(ASCE)CF.1943-
in Table 2, I r compares well with K m within 10% difference. 5509.0000175.

JOURNAL OF STRUCTURAL ENGINEERING © ASCE / MARCH 2012 / 455

J. Struct. Eng. 2012.138:449-456.

 Saatci, S., and Vecchio, F. J. (2009a). “Effects of shear mechanisms “Performance evaluation of flexure impact resistance capacity of rein-
on impact behavior of reinforced concrete beams.” ACI Struct. J., forced concrete members.” ACI Struct. J., 106(5), 726–736.
106(1), 78–86. Wang, R., Li, Z., Ren, G., and Zhang, S. (2007). “Experimental study and
Saatci, S., and Vecchio, F. J. (2009b). “Nonlinear finite element modeling numerical simulation of the dynamic response of concrete filled steel
of reinforced concrete structures under impact loads.” ACI Struct. J., tubes under lateral impact load.” Chin. Civ. Eng. J., 40(10), 34–40.
106(5), 717–725. Wang, N., Mindess, S., and Ko, K. (1996). “Fiber reinforced concrete
Silva, P. F., Mesia, W. D., Marzougui, D., and Badie, S. S. (2009). beams under impact loading.” Cem. Concr. Res., 26(3), 363–376.
Downloaded from ascelibrary.org by Iowa State University on 04/21/13. Copyright ASCE. For personal use only; all rights reserved.

456 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / MARCH 2012

View publication stats J. Struct. Eng. 2012.138:449-456.