Lateral Impact Response of Rubberized-Fibrous Concrete-Filled Steel Tubular Columns: Experiment and Numerical Study

Abstract

This paper presents an experimental and numerical study on the lateral impact behavior of rubberized-fibrous concrete-filled steel tubular (CFST) columns. Four types of concrete were utilized in the experimental program in the infilled columns: normal concrete (NC), rubberized concrete (RuC), steel fiber concrete (SFC), and hybrid RuC-SFC. Twelve specimens were tested using drop-weight impact with fixed-sliding boundary conditions. Three different transverse impact energies were produced by applying two masses of the hammers dropped from two different heights. A high-speed camera was implemented to measure the mid-span deflection against time. A 3-D finite element model was presented and verified against the tested specimens and some other experimental work from the literature. Load-displacement curves, the impact force time history, impact energy absorption, and failure modes of the CFST columns under the lateral impact were fully analyzed. The present results showed that at, certain impact energies, the steel tubular suffered only from the plastic deformation, beyond which it started cracking depending on the type of filled concrete. The steel tubular filled with hybrid RuC-SFC showed the highest resistance to crack formation, followed by that filled with SFC, while those filled with NC showed the lowest resistance to crack formation. There is an agreement between the numerical and the experimental results.

1. Introduction

In the last few years, the engineering community has been greatly interested in increasing the use of concrete-filled steel tubular (CFST) columns in major construction projects such as bridge piers, super high-rise buildings, large-scale structures, and subway stations worldwide [1]. This is due to their excellent bearing capacity, good resistance to blasting, impact, seismic loading, higher ductility, better toughness, higher stiffness, higher durability, and higher fire resistance. Core concrete can prevent or delay steel tube buckling. Meanwhile, the tube provides excellent confinement for core concrete, thus delaying concrete cracking and enhancing the plastic deformation capability of core concrete. During the life cycle, CFST columns may suffer serious harm or failure due to various transverse impacts from the vehicle and naval vessel collisions that encourage exploring the impact behavior under transverse impact loading. The existence of concrete in the CFST columns helps diminish both transverse failure displacement and the absorbed energy compared to hollow sections under large mass and low-velocity transverse impact loading. Xiuchen et al. [2] summarized analytical, experimental, and numerical studies of reinforced concrete (RC) structures subjected to vehicle collision. They discussed the effects of reinforcement materials and structural factors on impact resistance, impact force response, and failure modes under three different impact modes.

The existence of concrete in the CFST columns helped diminish both transverse failure displacement and the absorbed energy compared to hollow sections under large mass and low-velocity transverse impact loading [3]. The CFST columns showed a high resistance to impact at temperatures up to 400 °C [4]. Many researchers studied experimentally and theoretically the behavior of circular CFST columns under transverse impact loads [5,6,7,8,9]. They considered the influence of tube length, material properties, tube confinement, and projectile configurations on the collision behavior of the CFST columns. The resistance of steel tubes, mild steel, or stainless steel to local buckling increased due to the concrete infilling with higher collision resistance [10]. The effect of axial load application on the collision resistance, impact energy, lateral deflection, collision duration, and residual displacement was examined on the CFST column behavior [8,11,12]. Merwad et al. [13] summarized the transverse impact process on hollow and concrete-filled steel tubular (CFST) members and their dynamic performance under impact loading. They provided detailed discussions on the stages of the impact process, dissipated energy, impact energy, failure modes, and life-cycle performance of concrete-filled steel tubular (CFST) members subjected to transverse impact loading. Ahmadi et al. [14] explored the seismic failure probability and weakness appraisal of nine-story composite structures made up of rectangular concrete-filled steel tube (RCFT) columns and steel beams. According to the reliability analysis, a nine-story building is extremely sensitive to seismic excitation.

Various materials other than normal concrete were used to fill the core of the CFST sections under transverse collision loading to improve their performance and gain additional merits. Wang et al. [15] used ultra-lightweight cement composite (ULCC)-filled pipe in composite pipe tubes under lateral collision loads. This section showed an increase in the capacity of energy absorption. Recycled aggregate showed similar collision and deformation resistance capabilities to normal CFST columns [16,17,18]. CFST members showed a higher collision resistance under low-velocity transverse collision loading [12,19]. Ibrahim et al. [20] explored numerically and experimentally the impact performance of composite RC beams with the I-steel and pultruded I-GFRP beams. They stated that the impact performance of the composite specimens without GFRP I-beams was improved considerably by the concrete compressive strength.

Many researchers experimentally studied the influence of fine aggregate replacement by fine tire aggregate with different replacement ratios on the mechanical properties of concrete [21,22,23,24,25,26,27,28,29,30,31,32]. They concluded that, with the increase of rubber in the concrete, the deformation capacity, toughness, dynamic energy capacity, fatigue life, anti-crack performance, performance under the same stress level, seismic performance, and impact resistance increased, whereas the compressive and tensile strength of concrete decrease. El-Sisi et al. [33] found that concrete panels with rubber replacements for coarse aggregates have worse blast resistance in far fields and reduced overall static resistance compared to normal concrete. One of the most effective replacement materials for increasing the ductility of composite columns is rubberized concrete-filled steel tubular (RuCFST) columns when a percentage of fine aggregate is replaced by crumb rubber [34,35,36], which is considered a green structural material. However, no research has been conducted on the impact behavior. The maximum recommended ratio of rubber replacement is about 15% of fine aggregate to maintain the good mechanical properties of concrete.

Adding steel fibers to the concrete mix with a volume fraction between 0.5% and 2% increased the concrete compressive strength, flexural strength, ductility, modulus of rupture, and splitting tensile strength of fiber-reinforced concrete, but it had little effect on compressive strength [37,38,39,40]. Sahloddin et al. [41] experimentally investigated the flexural performance of built-up composite beams and high-performance SCC composite slabs. They demonstrated that hollow section spacing and fiber content had the largest effect on the composite beams’ load capacity and energy absorption. In some circumstances, experiments are expensive and consume a lot of time and effort, in addition to the difficulty of measuring some phenomena. Therefore, numerical simulation is a suitable alternative to overcome this dilemma. Guo and his colleagues [38,39,42,43] used the numerical manifold method and neural network to analyze the plate/shell problems such as bending, vibration, and buckling.

In the present work, a mixture of rubberized-fibrous concrete was used as the core of circular CFST columns. An experimental and numerical study on the impact behavior of these composite columns was conducted to investigate their dynamic response against transverse impact loading. The influence of the mass and height of the dropped hummer, steel ratio, and constraining (confining factor) were studied. The effects of the crumb rubber replacement ratio and the addition of steel-fiber to concrete on the impact performance of circular CFST members were investigated. The model’s accuracy was estimated by comparison with the present model and the model established by Wang et al. [11].

2. Experimental Work

The main purpose of this research was to study the behavior of the CFST columns infilled with the green structural material (rubberized concrete) under lateral impact load, which contributes to saving the environment from the undesired hazards of waste tires. It is known that the existence of rubber as a concrete component badly affects some of its mechanical properties. On the other hand, adding steel fiber to the concrete mix enhances it. Thus, a comparative study was held on the behavior of concrete-filled steel tubular columns filled with these mixes under lateral impact loading.

2.1. Specimen Description and Material Properties

A total of twelve circular concrete-filled steel tubular specimens were prepared for the experimental program. The steel tubes used in the tests were cut from a cold-formed hollow circular steel section with an average wall thickness (t_s) of 1.2 mm. All the specimens had an outside sectional diameter (D) of 89 mm and a length of 1500 mm. The manufacturer provided the mechanical properties of steel. The average yield strength, tensile strength, modulus of elasticity, and Poisson’s steel tube ratio were 235 MPa, 360 MPa, 2.01 × 10⁵ MPa, and 0.3, respectively. The specimens were chosen because of the following:

• To satisfy the condition that max. (d/t) ≤ 90 ε² (according to Eurocode 4 [44]) to avoid local buckling when the specimens are axially loaded in an extension to this research.
  where d: tube diameter
  t: tube thickness
  $\epsilon =\sqrt{\raisebox{1ex}{$235$}\!\left/ \!\raisebox{-1ex}{$f_y$}\right.}$ for the steel tube material
• These were the most suitable dimensions for our lab capabilities from those available on the market.

Four distinct concrete mixes were used for the infilled core. The first mix was normal concrete. The second concrete mix was rubberized concrete obtained by replacing 15% of fine aggregate with crumb rubber, since the rubber content should not exceed 20% of the total aggregate volume because of the negative effect on the concrete strength [45,46].The third mix had steel fiber with a 1.0% volume fraction since this is the optimum percentage of steel fiber to be added to the concrete mix by volume fraction based on the compressive and tensile strengths [47]. The fourth mix had rubberized concrete and steel fiber with the same percentages as the previous concrete mixes. The mix design of all types was the same, except for replacing 15% of the fine aggregate with crumb rubber in specimens with rubberized concrete. For each of the four groups, four 150 mm × 150 mm × 150 mm cubes and three 300 mm × 150 mm cylinders were prepared while casting the test specimens to obtain the concrete material properties. The compressive strengths of the normal concrete, rubberized concrete, steel fiber concrete, and mixed rubberized-steel fiber concrete were 46.8 MPa, 36.6 MPa, 39.6 MPa, and 39.1 MPa, respectively. The tensile strengths of the concrete mixes were 3.42 MPa, 2.46 MPa, 4.28, and 4.08 MPa, respectively.

A 5 mm thick steel plate was welded at one end of each steel tube, as shown in Figure 1 for the specimen setup. This was very helpful during concrete casting. Before pouring the concrete, the inner surface of the tubes was cleaned. The concrete was poured into the steel tubes in five layers to ensure good compaction of the concrete inside the tubes. This was performed using a steel rod and a vibrator. The specimens, cubes, and cylinders were cured for 28 days before testing. The following labeling was used to define each specimen: “NC, RUC, SFC, or MC” denotes the type of filled concrete; “M1 or M2” indicates the mass of the drop hammer, where M1 = 120 kg and M2 = 150 kg. “H1 or H2” indicates the height of the drop hammer from the specimen, where H1 = 1.5 m and H2 = 1.95 m. It is worth noting that the masses and the heights of the hammers were selected based on the capability of the test machine.

The following labeling was used to define each specimen: “NC, RUC, SFC, or MC” denotes the type of filled concrete, where “NC” denotes normal concrete, “RUC” denotes rubberized concrete, “SFC” denotes steel-fiber concrete, and “MC” denotes mixed rubberized-steel fiber concrete. “M1 or M2” indicates the mass of the drop hammer, where M1 = 120 kg and M2 = 150 kg. “H1 or H2” indicates the height of the drop hammer above the specimen, where H1 = 1.5 m and H2 = 1.95 m. For example, RUC-M1-H2 denotes that the infilling concrete was rubberized concrete, tested under a hummer of 120 kg and a height of 1.95 m.

The important parameters that differed in the testing in this research included:

• Impact energy ($W$): 1765.8 J to 2869.4 J

The impact energy ($W$) is defined as:

$$W=\frac{1}{2}m{V}_0^2=mgH$$

(1)

where ($m$) is the mass of the drop hammer, ($H$) is impact height, ($g$) is the gravity acceleration, and ($V_0$) is the initiation impact velocity, as shown in

Table 1. Data and results of tested specimens.

Group	Specimen Label	D × ts (mm × mm)	$\mathit{\xi }$	r %	Vf %	m	$\mathit{H}$	W (J)	Impact Energy Increase	$\mathbf{\Delta }$ (mm)
Group I (M1-H1)	NC-M1-H1	89 × 1.2	0.28	0	0	120	1.5	1765.8	1.00	56.78
	RUC-M1-H1	89 × 1.2	0.36	15	0	120	1.5	1765.8	1.00	63.74
	SFC-M1-H1	89 × 1.2	0.33	0	1	120	1.5	1765.8	1.00	66.02
	MC-M1-H1	89 × 1.2	0.34	15	1	120	1.5	1765.8	1.00	62.92
Group II (M1-H2)	NC-M1-H2	89 × 1.2	0.28	0	0	120	1.95	2295.5	1.30	77.17
	RUC-M1-H2	89 × 1.2	0.36	15	0	120	1.95	2295.5	1.30	74.46
	SFC-M1-H2	89 × 1.2	0.33	0	1	120	1.95	2295.5	1.30	72.77
	MC-M1-H2	89 × 1.2	0.34	15	1	120	1.95	2295.5	1.30	73.5
Group III (M2-H2)	NC-M2-H2	89 × 1.2	0.28	0	0	150	1.95	2869.4	1.62	95.54
	RUC-M2-H2	89 × 1.2	0.36	15	0	150	1.95	2869.4	1.62	91.14
	SFC-M2-H2	89 × 1.2	0.33	0	1	150	1.95	2869.4	1.62	83.31
	MC-M2-H2	89 × 1.2	0.34	15	1	150	1.95	2869.4	1.62	98.32

.

• Constraining factor ($\xi$): 0.28 to 0.34

The constraining factor ($\xi$) described the “composite action” of CFST and was proposed by Han et al. [9] as follows:

$$\xi =\frac{A_{\mathrm{s}}{f}_{\mathrm{y}}}{A_{\mathrm{c}}{f}_{\mathrm{ck}}}=\alpha \frac{f_{\mathrm{y}}}{f_{\mathrm{ck}}}$$

(2)

where ($A_{\mathrm{s}}$) is the steel tube’s cross-sectional area, ($A_{\mathrm{c}}$) is the core concrete’s cross-sectional area, $\alpha$ = ($A_{\mathrm{s}}$/$A_{\mathrm{c}}$) is the steel ratio of CFST members, ($f_{\mathrm{ck}}$) is the characteristic concrete strength, which is taken as 0.67 of the cube strength of concrete for normal strength concrete, and ($f_{\mathrm{y}}$) is the yield strength of steel.

• Rubber replacement ratio (r): 0 and 15%.

The rubber content is 0 and 15% as a replacement of sand volume.

• Fiber volume fraction ($V_{\mathrm{f}}$): 0 and 1%. The fiber volume fraction is defined as the ratio of the volume of fibers present to the total volume of the concrete.

2.2. Test Setup

The tests were executed using a drop weight impact machine shown in Figure 2. The columns were positioned horizontally with fixed-sliding boundary conditions. The specimens were placed so that the effective length between supports was 1200 mm. The columns were exposed to impact via a drop hammer with a 30 mm × 80 mm solid rigid flat indenter. The hammer was made up of various steel blocks given two different weights by attaching a variable number of steel blocks to the projectile. The weight was guided down the steel rails during the test by greased rollers on its sides to ensure a vertical impact on the specimen’s mid-span. In each test, the drop hammer was raised to the designed height through a mechanical winch and then released to apply the impact load to the specimen. The different initial heights and masses of the drop hammer, as detailed in Table 1, provided a variety of impact energies and velocities. A schematic view of the rest of the setup is shown in Figure 3. During the test, a high-speed video camera at a speed of 2000 frames per second was set to capture the specimen destruction process. A white dot was drawn at the center line of each specimen to track the displacement from the time of impact to the maximum reached position after impact.

2.3. Test Experimental Results and Discussion

The response of the tested CFST specimens with different concrete mixes under lateral impact loading was monitored. The main observed signs were the overall flexure of the specimens and local buckling of the steel tubes. In addition, the mid-span deflection of the tested columns was recorded. A plastic failure mechanism with plastic hinges at mid-span and the support positions were noticed. Between these positions, the columns remained straight, as shown in Figure 4.

Under lateral impact loading, the response of the CFST specimens was dominated by a mix of local buckling and global flexural deformation. The modes of failure of the specimens indicate the interaction effect between the inner concrete and the outer steel tube. Due to the inner concrete’s support, less local buckling was observed in the compression zone at the top of the mid-span section and the bottom of the fixed and sliding-ended sections. However, apparent plastic hinges occurred in the impact area and near the fixed ends, accompanied by the dissipation of a significant amount of the impact energy. Moreover, the inner concrete failure mechanism exhibited ductile behavior due to the confinement effect of the outer steel tube. The tensile fracture of the steel tube resulted in the tension side of the mid-span section in increasing mid-span deflection, such as what happened in (NC-M2-H2), (RUC-M2-H2), (SFC-M2-H2), (MC-M2-H2), (NC-M1-H2), and (RUC-M1-H2). It was noticeable that, in the case of high impact energy, fracture occurred in the specimens. Therefore, less mid-span displacement occurred in the (SFC-M2-H2) column, then the (NC-M2-H2) column, then the (RUC-M2-H2) column, and then the (MC-M2-H2) column.

Specimens tested under low-impact energy (Group I) showed no fracture, as shown in Figure 4a. Specimen (NC-M1-H1) showed less mid-span displacement due to the brittle behavior of the normal concrete, which controlled the flexible performance of the steel tube. On the other hand, it was observed that specimens with a greater constraining factor (in the same group) performed more ductile under impact load, as can be seen in Figure 5a. Generally, in low energy tests (M1-H1), all four test specimens behaved similarly. There was deformation in the CFST with no rupture in the steel tubes. With the increasing energy, local buckling was noticed near the supports in all four specimens. The rupture was observed in the CFST columns with normal and rubberized concrete, as shown in Figure 4b. Adding steel fiber to the concrete mix enhanced the impact behavior and helped prevent rupture in the outer steel tube. This may be due to the occurrence of fiber bridging, which resists the crack growth in the concrete and subsequently dissipates the energy and delays the crack initiation in the steel tubular. This observation agrees with the previous work found in the literature, where they concluded that adding steel fiber improved the impact resistance of concrete [41]. Deng et al. [7] concluded that using steel fibers effectively restrained the tensile cracks in the concrete core and increased the impact resistance. Therefore, the fiber-reinforced concrete-filled tube would have better impact resistance than regular concrete-filled tubes. At a high energy level, there was a rupture in the steel tubes of the four specimens at the impact position, in addition to local buckling near the support. This may be attributed to the strain of the critical point on the steel tubular reaching its ultimate tensile strain and maybe beyond it, i.e., the commencement of the crack initiation in the steel or sudden failure.

On the other hand, Han et al. [9] concluded that the failure modes of the CFST specimens clearly showed the interaction effect between the outer steel tube and the inner concrete due to the support of the inner concrete. Less local buckling was observed, while obvious plastic hinges occurred and dissipated a large amount of the impact energy. The failure mode of the inner concrete also showed ductile characteristics due to the outer steel tube’s confinement effect, which also increased the specimen’s energy dissipation capacity.

The time history of mid-span deflection curves of specimens is illustrated in Figure 5. All curves generally exhibited similar trends. The deflection increased immediately after impact and gradually declined as the specimen releases its elastic strain energy. The elastic deformation returned after the impact was over, leaving the specimen with residual deflection. The mid-point displacement was traced with the help of the marker point in Figure 6 during the test procedure. The figure captured the displacement from the initial contact between the drop-weight and the specimen, remarkably increasing with time until the marker point reached a stable position (maximum deflection). In most cases, indentation was noticed at the contact area on the specimens. Cracks were noticed on the opposite side at this position in all specimens of group III with the highest impact energy and two specimens of group II. This showed that using the green structural material (rubberized concrete) did not reduce the efficiency of the composite column. On the other hand, using steel fiber in the concrete mix helped prevent these cracks in this group, as shown in Figure 4.

2.4. Finite Element Analysis (FEA)

2.4.1. Model Description

A finite element analysis (FEA) model was developed to simulate the impact behavior of CFST with different concrete mixes under lateral low-velocity impact. The model was established with the explicit dynamic nonlinear finite element code AUTDYN. The FEA model consisted of five components: the concrete core of CFST, steel tube, drop hammer, end plates, and blocks representing supports, as illustrated in Figure 7. The FEA model considered the major effects of the impact process on the column, including strain rate effects of the steel tube and concrete, the interaction between the concrete and steel, and the concrete element erosion criterion after impact. To simulate all the model’s components, a three-dimensional eight-node solid element SOLID164 was used. A mesh size of 15 mm was used. Only the drop hammer’s indenter was modeled to save computation time. The indenter mass was set to be the same as the drop hammer.

It is well known that the finite element mesh affects the results, and using a very fine mesh greatly increases the required computational time. In finite element models exposed to dynamic loading, the mesh dependency is even more sensitive than static loading. A mesh convergence study was carried out to investigate its impact on the deflection curves and the simulated impact force and to find the most accurate model with the shortest solution time. Different mesh sizes (25 mm, 20 mm, 15 mm, and 10 mm) were adopted to obtain accurate results and reasonable run-times. The overall behavior of the models was compared to the experimental results. An example of these comparisons is shown in Figure 8, revealing that using a mish size of 15 mm gave the desired accuracy within a reasonable time.

When concrete and steel are subjected to impact loading, the behavior of concrete and steel materials is significantly different from those during quasi-static load. In the dynamic finite element software AUTODYN, the RHT constitutive material model (Riedel–Hiermaier–Thoma) has been adopted since 2000 [48]. This plasticity model with damage softening uses a triaxial strength description that can predict the shock wave regime. It was found reliable in predicting major dynamic failure phenomena of reinforced concrete structures and the behavior of concrete material under impact loading. The RHT concrete constitutive model considers three ultimate surfaces, namely elastic limit surface, failure surface, and residual strength surface, as shown in Figure 9.

The steel tube was modeled with the Johnson-cook constitutive material model to capture the high strain rate response of the steel tube under impact loading. This model considers three components: strain rate, strain hardening, and thermal softening. The FEA model took into account the key influences of the impact process, such as the strain rate effect of the concrete and steel tube and the contact behavior between all model components. The body interaction contact was used to prevent penetration in the normal direction. Full-bond contact between the concrete core and steel tube was considered in the tangential direction. Two blocks were utilized to represent the supports in the two ends, fixed support at one end and sliding support at the other. The contact between the steel tube and the block representing fixed support was fully bonded in the tangential direction. The contact between the steel tube and the block representing sliding support was frictionless. There was contact between the indenter of the drop hammer and the steel tube. The body interaction contact was used between the indenter of the drop hammer and the steel tube to prevent penetration in the normal direction.

The boundary conditions in the model were the same as those in the experimental test. The two blocks representing the supports were restricted from rotation or translation in all directions to achieve such boundaries. The right surface of the end plate on the side of the fixed support was restricted from rotation or translation in all directions, and the left surface of the end plate on the side of the sliding support could move freely in the tangential direction. Finally, the indenter of the drop hammer was restricted from rotation or translation in all directions except the vertical direction (loading direction).

Furthermore, the contact between the steel tube and the sliding support block was frictionless. The convergence study was performed to identify an appropriate contact between the concrete core and steel tube. Then, the contact between the concrete core and steel tube was bonded in the tangential direction. The body interaction contact was used to prevent penetration in the normal direction. The body interaction contact was applied in this model to prevent friction in the normal direction. The rotation and translation of the two blocks representing the supports were limited in all directions. The drop hammer was only allowed to rotate and translate in the vertical direction. The surface of the end plate at the sliding support could move freely in the axial direction. However, the surface of the end plate at the fixed support was prevented from translation or rotation in all directions. The drop hammer was placed close to the specimen and given an initial velocity (V₀). For the sake of the accuracy of the proposed model, the material properties of each component were taken as reported in each experimental specimen, as mentioned previously.

2.4.2. Verification of the FEA Model

The proposed finite element model was validated using the current experimental tests and other tests performed earlier by Han et al. [9]. For comparison, dimensions, material properties, and boundary conditions were simulated exactly as the tested specimens. Han et al. specimens were circular concrete-filled steel tubular columns (DZF25-114×3.5) with a 1500 mm length. The outside diameter of the tubes was 114 mm, the thickness was 3.5 mm, and the boundary conditions utilized at the ends of the specimen were fixed-sliding. The reported cube strength of the concrete was 48.7 N/mm². The steel tube’s yield stress (f_y) and modulus of elasticity (Es) were 298 N/mm² and 2.01 × 105 N/mm², respectively. All the available data results from Han et al. [9] were compared to their finite element counterparts, as shown in Figure 10. The proposed model managed to show the predicted time history of the mid-span deflection, the velocity of the drop hammer, the time history of the impact force of the drop hammer, and the mid-span velocity of the specimen. There was good agreement between the FE and experimental results performed by Han et al. [9].

2.5. Finite Element Results and Discussion

The time history of mid-span deflection curves of some of the tested specimens compared to the counterpart finite element results are illustrated in Figure 11 with acceptable accuracy. The curves generally exhibited similar trends. The mid-span deflection normally grows throughout the peak stage even if the impact force decreases or fluctuates and continues to increase for most of the plateau stage. It demonstrates that the specimen continued to move downwards even if the contact between the specimen and thehammer was lessened or the two objects separated. The mid-span deflection reached its peak value during the impact force’s plateau stage and gradually declined as the specimen releases its elastic strain energy. After the impact was over, the elastic deformation returns, leaving a residual deflection in the specimen. The mid-span deflection in the numerical model was slightly less than in the experimental study, which means that the sample was softer than the model. This might be due to several reasons, such as the initial simplification found in the steel beam and the misevaluation of the bond strength between the concrete and steel tube.

As noticed in Figure 12, the FEA model properly predicted the specimen’s flexural deformation, as well as the local buckling of the steel tube. Due to the small thickness of the tube, all samples experienced local buckling at the parts subjected to compression stress with different levels. The variation of local buckling from sample to sample might be due to the differences in the initial imperfection and the bond between the concrete and steel. For samples (c), (d), and (h), the observed local buckling was slightly lower than in the other samples; the same observation was found in the model.

To identify the phases of the impact process, the impact force time-history curves were obtained from the simulation as illustrated in Figure 13. The whole curves could be generally divided into three stages: (1) the peak stage (the impact force sharply increases to its peak value at a short time, which related to the Specimen’s local deformation), (2) the plateau stage(the impact force retains a steady value for a long time, and most of the impact energy is dissipated, which related to the advancement of the global plastic deformation of the specimen), and (3) the descending stage (the impact force rapidly decreases to zero). For each group, a similar value of the impact force was produced. When the impact energy was 1765.8 J (Group I), the impact force was about 250 kN. When the impact energy increased by about 30% (Group II), the average produced impact force reached 300 kN with an increase of only 20%. For an increase in the impact energy of about 62% (Group III), the average impact force exceeded twice that of the first group. It was noticed that the plateau value and duration of the impact force, as well as the mid-span deflection, increased as the drop hammer mass increased.

Table 1 and

Table 2. Results of experimental tests.

Specimen Label	D × ts (mm × mm)	$\mathit{\xi }$	m	$\mathit{H}$	W (J)	$\mathit{t}$ (ms)	${\mathit{F}}_{\mathbf{stab}}$ (kN)	${\mathit{F}}_{\mathbf{max}}$ (kN)	$\mathbf{\Delta }$ (mm)	E (J)
NC-M1-H1	89 × 1.2	0.28	120	1.5	1765.8	28.7	29.14	272	53.06	1700.52
RUC-M1-H1	89 × 1.2	0.36	120	1.5	1765.8	30.2	29.06	247	57.23	1703.64
SFC-M1-H1	89 × 1.2	0.33	120	1.5	1765.8	28.7	29.28	262.9	55.07	1726.2
MC-M1-H1	89 × 1.2	0.34	120	1.5	1765.8	28.9	29.08	261.9	54.32	1701.29
NC-M1-H2	89 × 1.2	0.28	120	1.95	2295.5	31.1	32.11	302.8	68.38	2231.26
RUC-M1-H2	89 × 1.2	0.36	120	1.95	2295.5	31.9	31.62	309.1	73.22	2241.56
SFC-M1-H2	89 × 1.2	0.33	120	1.95	2295.5	30.9	31.21	295.1	68.46	2242.8
MC-M1-H2	89 × 1.2	0.34	120	1.95	2295.5	31.6	30.6	291.7	69.62	2244.8
NC-M2-H2	89 × 1.2	0.28	150	1.95	2869.4	36	35.52	300.6	84.31	2809.42
RUC-M2-H2	89 × 1.2	0.36	150	1.95	2869.4	36.2	35.89	576.5	78.74	2760.87
SFC-M2-H2	89 × 1.2	0.33	150	1.95	2869.4	34.4	34.29	576.5	81.21	2841.18
MC-M2-H2	89 × 1.2	0.34	150	1.95	2869.4	35.2	34.01	576.5	78.14	2759.05

illustrate the influence of the height of the drop hammer. The peak value, plateau value, and duration of the impact force, as well as the mid-span deflection, all increased as the height of the drop hammer increased. This is because the strain rates of the materials are greater with the higher height of the drop hammer.

This study found that the steel tube suffered only from plastic deformation at low impact energy (Group I). When the impact energy increased by 30%, as in Group II, it started cracking depending on the type of the infilled concrete, e.g., specimens NC-M1-H1 and RuC-M1-H1. The steel tubular filled with hybrid RuC-SFC had the highest resistance to fracture formation, followed by SFC. This may be due to these concrete mixes’ increased stiffness and ductility.

It is known from previous studies that the area under the impact force-displacement curve reflects the external work performed by the impact force, which is nearly equal to the initial impact energy. In contrast, a portion of the energy is lost from impact. The impact energy equals the strain energy absorbed by the plastic hinge (absorbed energy) plus the strain energy recovered from a specimen’s elastic rebound (recovered energy).

Table 3. Rubberized concrete properties.

r (%)	0	10	20	30	40	50
Compressive strength (MPa)	64.4	50.7	40	29.4	24.1	21.6
Tensile strength (MPa)	5.8	4.9	3.5	3.2	3	2.4
Unit weight (kg/m3)	2541	2531	2487	2451	2423	2365
Modulus of elasticity (GPa)	52.9	50.4	44.2	37	34.5	25.7

illustrates that the constraining factor ($\xi$) had a significant effect on the curves. The peak and plateau values of the impact force grew as the steel ratio increased, while the duration dropped drastically. The mid-span deflection also decreased significantly. This indicates that when the constraining factor increased, so did the member’s impact resistances. It was observed that specimens with a higher constraining factor exhibited very ductile behavior under the lateral impact, whereas specimens with a lower constraining factor exhibited brittle behavior.

3. The Parametric Study

Using the verified finite element model, it was interesting to study the effect of increasing the rubber content in the infilled concrete mid. It is known that the recommended rubber replacement ratio in concrete members is about 15%. However, increasing this ratio in the CFST composite columns highlighted some advantages of disposing waste tires to save the environment. Rubberized concrete is well known for its ductility and increased amount of absorbed energy [23]. Five different replacement ratios were used, as illustrated in

Table 4. Results of FE models with different rubber contents.

Specimen Label	D × ts (mm × mm)	$\mathit{\xi }$	m	$\mathit{H}$	W (J)	r (%)	${\mathit{f}}_{\mathbf{cu}}$	${\mathit{f}}_{\mathbf{t}}$	${\mathit{t}}_{}$ (ms)	${\mathit{F}}_{\mathbf{stab}}$ (kN)	${\mathit{F}}_{\mathbf{max}}$ (kN)	Decrease in Fmax	$\mathbf{\Delta }$ (mm)
C1	114 × 2	0.51	250	6	14700	0	64.4	5.8	31.2	109	539.2	1	138.84
C2	114 × 2	0.65	250	6	14700	10	50.7	4.9	31.6	108	493.9	0.915	143.22
C3	114 × 2	0.82	250	6	14700	20	40	3.5	31.45	106.5	463.9	0.860	141.71
C4	114 × 2	1.12	250	6	14700	30	29.4	3.2	31.32	104.5	432	0.801	140.07
C5	114 × 2	1.37	250	6	14700	40	24.1	3	31.17	101.7	402.3	0.746	141.31
C6	114 × 2	1.52	250	6	14700	50	21.6	2.4	31.88	100.9	387.2	0.718	144.36

. The average compressive strength, unit weight, tensile strength, and modulus of elasticity of rubberized concrete are shown in Table 4, according to Youssf et al. [24]. The analysis was performed on CFST columns with the dimensions and data of the applied impact energy detailed in Table 4. It was noticed that increasing the rubber content increased the column ductility and consequently decreased the maximum recorded impact force by about 8.5%, 14%, 19.5%, 25.4%, and 28.2%, with rubber replacement ratios of 10%, 20%, 30%, 40%, and 50%, respectively. However, the column strength must be checked against axial load since the concrete compressive strength was noticeably affected.

The effects of concrete compressive strength and yield strength of the surrounding steel tubes were studied. The concrete compressive strength was increased from 20 MPa to 60 MPa. This indicated an increase in the impact force by 28.2%. In contrast, it reduced the maximum displacement by 8%. The effect of yield strength of steel tubular was studied using values of 235 MPa, 355 MPa, and 450 MPa. The impact force was increased by 36.5%,while the maximum displacement was decreased by 32.3%, as shown in Figure 14. It is clear that the change of the outer steel tube yield strength had a much stronger influence on the impact response than concrete strength. This may be due to the increased plastic moment capacity of the column section.

The effect of increasing the outer tube thickness was studied as well. The steel tube thickness was analyzed using 2 mm and 3 mm thicknesses. The material properties of all materials were kept constant for the sake of comparison. It was noticed that increasing the steel tube thickness decreased the maximum deflection from 141.71 mm to 103.19 mm, a reduction of about 27.2%, while the impact force was increased from 463.9 kN to492.1 kN, an increase of about 6.1%. These changes are due to the change in the moment of inertia of the tube section, which leads to a stuffer column and higher flexural capacity.

4. Conclusions

This paper presented a numerical and experimental study on the lateral impact behavior of the rubberized-fibrous concrete-filled steel tubular (CFST) columns and their dynamic response. The following conclusions could be drawn within the scope of the current research:

• Under lateral impact loading, the response of the CFST specimens wasdominated by a mix of local bucklingand global flexural deformation, as well as tensile fracture of the steel tube, resulting in the tension side of the mid-span section increasing mid-span deflection.
• At certain impact energies, the steel tubular columns suffered only from plastic deformation, beyond which it started cracking depending on the type of filled concrete.
• Rubberized concrete had close lateral impact resistance to normal concrete in the CFST columns.
• Adding steel fiber to the concrete mix enhanced the impact behavior and helped delay/preclude cracks in the outer steel tube.
• The steel tubular filled with hybrid RuC-SFC had the highest resistance to fracture formation.
• The constraining factor (ξ) had a significant effect on the curves. The peak and plateau values of the impact force grew as the steel ratio increased, while the duration dropped drastically. The mid-span deflection also decreased significantly. This indicates that when the constraining factor increases, so do the member’s impact resistance.
• Increasing rubber content raised the ability of the CFST columns to absorb impact energy, which was noticed as the reduction in the resulting impact force. Using 50% rubber replacement decreased the impact force by 28.2% compared to normal concrete.