Journal of Constructional Steel Research 211 (2023) 108159

Contents lists available at ScienceDirect

Journal of Constructional Steel Research

journal homepage: www.elsevier.com/locate/jcsr

Impact behavior of high-strength steel beam with circular web openings

Huayan Chen a, Yanzhang Wang a, Chaojiang Fu a, Ting Zhang a, Caisong Luo a, *, Bizhen Wang b
a
College of Civil Engineering, Fujian University of Technology, Fuzhou 350118, China
b
Fujian Zhonglin Engineering Construction Co., Ltd., Nanping 353000, China

A R T I C L E I N F O A B S T R A C T

Keywords: A drop-weight impact testing machine was used to conduct an impact test on 9 specimens in order to examine the
Steel beams with circular web openings impact resistance of high-strength steel beams with circular web openings (HSSBCWO). The experimental results
High-strength steel indicated that the dynamic response of HSSBCWO is significantly affected by the change of impact energy and
Dynamic response
opening height. A coefficient β is proposed to assess the severity of localized deformation in HSSBCWO, and the
Energy absorption
Damage assessment
range of β values for the specimens of test are found to be approximately between 0.5 and 0.8. Within the range
of impact energy designed for the experiment, HSSBCWO primarily absorbed energy through localized plastic
deformation. A finite element model which can predict the dynamic response of HSSBCWO under impact load is
established. The results show that the optimal ratios of opening height to beam height and spacing between
openings to beam height for the HSSBCWO are determined to be 0.50 and 0.93, respectively. Furthermore, a
damage evaluation system for HSSBCWO under impact load was established, utilizing the ratio of mid-span
residual displacement to calculated beam span Dr / L0 as the damage assessment index, with impact energy,
opening height, and spacing between opening as the main control parameters.

1. Introduction During the service life of structures, they may be subjected to acci­
dental loads such as explosions and impacts [7]. These extreme loading
With the continuous advancement of urbanization and the limitation events can cause significant damage to structures, resulting in severe
of land resources, the construction industry is increasingly inclined to­ property loss and even endangering human life safety. Therefore,
wards using high-performance materials and components [1]. The use of numerous scholars have conducted research on the dynamic perfor­
high-strength steel is becoming more widespread due to its excellent mance of high-strength steel materials and steel beams with web
strength-to-weight ratio. By implementing high-strength steel, it is openings (SBWOs).
possible to substantially decrease the dimensions of structural elements, Alabi et al. conducted a study on the tensile performance of two types
leading to a reduction in the amount of steel required for the entire of high-strength steels, S690QL and S960QL, under different strain rates.
structure [2]. Consequently, this reduction aids in diminishing carbon The results indicated that high-strength structural steels with low yield-
emissions, conserving resources, and enhancing economic advantages tensile strength ratio are more sensitive to the influence of strain rate
[3]. High-strength steel is primarily employed as heavy-duty compo­ [8]. Zhang et al. investigated the mechanical properties of Q345 steel
nents in high-rise buildings, large-span roofs, bridges, and offshore under impact loading. The experimental results demonstrated that the
platforms [4]. Conventional steel beams are limited by deflection con­ strain rate has a significant effect on the mechanical behavior of Q345
straints, which prevent the full utilization of their material properties steel under high strain rate loading. Under extreme loading conditions of
[5]. Utilizing the web openings technique not only minimizes steel high temperature and dynamic impact load, the strength of Q345 steel
wastage but also decreases the self-weight of the structure. Furthermore, exhibits a nonlinear relationship with strain rate [9]. Yang et al. con­
the adaptable design of web openings enables the inclusion of diverse ducted quasi-static tensile tests and dynamic tests on S690 high-strength
utilities, thereby enhancing the spatial efficiency of the building [6]. steel. They also modified important parameters of the dynamic consti­
Incorporating high-strength steel materials in the design of HSSBCWOs tutive models and proposed Cowper-Symonds and Johnson-Cook
is an effective strategy to fulfill the high-performance demands of models with higher accuracy for adapting to S690 high-strength steel
building structures. [10]. Tang et al. conducted tensile tests on a novel NOR anchor rod steel

* Corresponding author.
E-mail address: [email protected] (C. Luo).

https://doi.org/10.1016/j.jcsr.2023.108159
Received 8 June 2023; Received in revised form 24 July 2023; Accepted 7 August 2023
Available online 10 August 2023
0143-974X/© 2023 Elsevier Ltd. All rights reserved.
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 1. Details of the specimens.

Table 1
Design parameters of specimens.
Specimen L(mm) H(mm) bf(mm) tw(mm) tf(mm) h(mm) S(mm) e(mm) v(m/s) m(kg) Ei(kJ)

C10–5–180-310 2300 300 150 5.8 7.9 180 310 130 10 530 26.5
C12–5–180-310 2300 300 150 5.8 7.9 180 310 130 12 530 38.16
C14–5–180-310 2300 300 150 5.8 7.9 180 310 130 14 530 51.94
C12–4–180-310 2300 300 150 5.8 7.9 180 310 130 12 430 30.96
C12–6–180-310 2300 300 150 5.8 7.9 180 310 130 12 630 45.36
C12–5–160-310 2300 300 150 5.8 7.9 160 310 130 12 530 38.16
C12–5–200-310 2300 300 150 5.8 7.9 200 310 130 12 530 38.16
C12–5–180-290 2300 300 150 5.8 7.9 180 290 110 12 530 38.16
C12–5–180-330 2300 300 150 5.8 7.9 180 330 150 12 530 38.16

Note: e represents the minimum distance between two openings.

steels and proposed a constitutive model capable of predicting the dy­

namic response of structural steels with strengths below 960 MPa within
a large strain rate range [12].
Al-Husainy et al. conducted a finite element analysis to analyze the
impact performance of steel beams with circular web openings
(SBCWOs). The analysis results indicated that impact energy has a sig­
nificant influence on the impact force and displacement of the SBCWOs,
while the number of openings has a negligible effect on their impact
performance [13]. Al-Rifaie et al. investigated the influence of steel
beams with rectangular web openings with different impact velocities,
Fig. 2. Geometry of test specimen for tensile tests (ts = 5.8 mm, ts = 7.9 mm). opening areas, depths and numbers (SBRWOs) on the impact perfor­
mance of beams through finite element analysis. The results showed
that, under a constant opening area, SBRWOs with narrow openings
Table 2
exhibit better impact performance than those with wide openings. When
Material properties of high-strength steel.
the opening depth exceeds 70% of the total depth of the rectangular web
Position fy (N/ fu(N/ E (N/ Poisson,s Elongation opening steel beam, the impact response remains essentially unchanged
mm2) mm2) mm2) ratio (%)
[14]. Wang et al. conducted drop hammer impact tests and finite
Web 958.31 984.65 198,634 0.293 17 element analysis to study the dynamic response of steel beams with
Flange 770.02 803.63 207,182 0.289 16
hexagonal web opening (SBHWOs). They suggested avoiding subjecting
Note: fy represents the yield strength; fu represents the ultimate tensile strength; the beam at a position close to 1/4 of the beam's calculated length near
E represents the elasticity modulus. the support to impact loads [15]. Luo et al. performed impact tests on
SBHWOs, revealing the influence of opening height and spacing be­
at strain rates ranging from 0.1 s-1 to 1000 s-1. They improved the tween openings on the dynamic behavior of SBHWOs. They also
Johnson-Cook model to suit the novel NOR anchor rod steel and developed a damage assessment curve suitable for SBHWOs [16].
demonstrated its enhanced energy absorption capacity [11]. Yang per­ Most researchers have focused on studying the dynamic performance
formed Split Hopkinson Pressure Bar (SHPB) tests on various structural of high-strength steel at the material level rather than at the component

2
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 3. Experiment apparatus.

Fig. 4. Sensors location.

level. There is a lack of research on the dynamic performance of was 150 mm, web thickness tw was 5.8 mm, and flange thickness tf was
HSSBCWOs under impact loads. In light of this gap, this study aims to 6.9 mm. The opening heights h of the specimens were designed in three
conduct an in-depth investigation into the dynamic performance of variations: 160 mm, 180 mm, and 200 mm. Additionally, the spacing
HSSBCWOs using a combined approach of experimental testing and between openings S were designed in three variations: 290 mm, 310
finite element analysis. mm, and 330 mm [17].
The numbering scheme for the specimens in Table 1 is as follows: “C”
2. Scheme for test represents circular web openings; “10,” “12,” and “14” denote impact
velocities v of 10 m/s, 12 m/s, and 14 m/s, respectively; “4,” “5,” and “6”
2.1. Design of specimens indicate the mass of the drop hammer m as 430 kg, 530 kg, and 630 kg,
respectively; “160,” “180,” and “200” represent the opening heights h of
A total of nine HSSBCWOs specimens were designed and fabricated the beam as 160 mm, 180 mm, and 200 mm, respectively; and “290,”
for the experiment, as shown in Fig. 1. Six web openings were designed “310,” and “330” signify the spacing between openings S in the beam as
on each specimen, labeled as O1 to O6 from left to right. The cross- 290 mm, 310 mm, and 330 mm, respectively.
sectional dimensions of the specimens are presented in Table 1. The
beam length L was 2300 mm, beam height H was 300 mm, beam width bf

3
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

properties of the high-strength steel. The loading rate was controlled at

2 mm/min, and the specimens were loaded until a displacement of 30
mm was reached [19]. The final experimental results were obtained by
averaging the values from three sets of specimens to obtain more ac­
curate material properties. The material performance indicators of the
high-strength steel are presented in Table 2.

2.3. Test equipment and sensors setting

The drop hammer impact tests were conducted using the ultra-
heavy-duty drop hammer testing machine at Fujian University of
Technology, as shown in Fig. 3. This equipment is equipped with the NI
data acquisition system from the United States, which enables the
simultaneous collection of 16 channels of high-speed analog voltage
signals and 16 channels of strain signals, effectively addressing the
measurement challenges associated with measuring the force and dy­
namic displacement of the heavy drop hammer head. The ultra-heavy-
duty drop hammer testing machine comprises a rigid hammer head,
detachable counterweight plates (ranging from 230 kg to 980 kg), a
14.5 m high guide rail, a force acquisition system, hinged supports that
apply boundary constraints, and a rigid pressure beam that ensures the
safe rotation of the specimen upon rebound.
The arrangement of sensors is illustrated in Fig. 4. At the mid-span
and at intervals of S, a displacement transducer LVTD and an acceler­
ometer AC were installed to obtain displacement and acceleration data.
Due to a positional conflict between the displacements transducer and
accelerometers, the acelerometers were positioned near the midline of
the beam. Additionally, five strain gauges SG1 to SG5 were evenly
distributed at equal intervals from top to bottom at the mid-span of the
web. Furthermore, the impact force data were acquired using pressure
sensors installed within the hammer head.

2.4. Experiment process

First, the specimens must be installed on the hinged supports.

Additionally, the stability of the specimens must be ensured by securely
fastening the pressure beams in place. Second, position the sensors ac­
cording to the corresponding locations indicated in Fig. 4, and the
number of weight plates should be based on the experimental re­
quirements in order to achieve the desired testing conditions. Once the
data acquisition system is confirmed to be functioning properly, the test
can be initiated. This process involves raising the hammer to the target
height and releasing it, allowing it to free fall along the guide rails. When
the hammer reaches a distance of 100 mm from the specimen, it passes
through a velocity laser probe, and the signal conditioner receives the
velocity signal of the hammer. Subsequently, the data acquisition system
begins capturing data at a frequency of up to 200 kHz and triggers a
high-speed camera to record images at a rate of 2000 frames per second.
This procedure ensures the acquisition of precise and reliable experi­
mental data, providing support for subsequent analysis and research.

3. Experimental results

3.1. Impact process

Taking the specimen V12-5-180-310 as an example, the process of its

impact can be observed in Fig. 5, and Fig. 6 showing the impact force
and the displacement-time curve at the mid-span. At the instant of 0 ms,
Fig. 5. Impact process. the hammer makes contact with the upper flange surface of the spec­
imen at its maximum velocity. Under their strong interaction, the impact
2.2. Steel properties force rapidly increases and reaches its impact force maximum value Fm
at approximately 0.25 ms. The specimen undergoes downward accel­
Three tensile specimens were taken from the flange and web plates of eration, leading to a reduction in the contact between the hammer and
the high-strength steel, aligned with the rolling direction at 0◦ [18], as the specimen, resulting in a decrease in the impact force. Due to the
shown in Fig. 2. The specimen dimensions are provided in Fig. 2. Quasi- velocity difference, they come into strong contact again, causing a
static uniaxial tensile tests were conducted to determine the material renewed increase in the impact force. This cyclic behavior repeats twice,

4
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 6. Typical time history curves of impact force and mid-span displacement.

and at 1.5 ms, the impact force curve ceases its significant vibration. 3.2. Impact force
Hence, the stage between 0 ms and 1.5 ms is referred to as the impact
force vibration stage Δv. During this stage, the specimen undergoes Fig. 7 presents the impact force-time curves of specimens in the
flexural deformation along the direction of impact due to the impact of experiment. It can be observed that the impact force-time curves for all
the falling hammer. The displacement exhibits nearly linear increase. specimens can be divided into the three stages described in Fig. 6. This is
After approximately 1.5 ms, the specimen enters the energy dissi­ attributed to the appropriate selection of impact energy, which avoids
pation stage, during which the restoring force provided by the specimen scenarios where the impact energy is too low, resulting in minimal or no
itself and he inertial force becomes relatively stable. The speed of the plastic deformation in the specimen, or when the impact energy is too
falling hammer and the motion of the specimen are nearly identical, and high, causing direct specimen failure. In these two cases, the impact
their contact remains relatively stable. In this stage, the impact force energy is not fully absorbed by the specimen, and as a result, the impact
remains essentially constant, lasting for approximately 4.5 ms. Thus, this force curve does not exhibit impact force stable stage Δs and impact force
stage is known as the stable stage of impact force Δs. The average impact descending stage Δd. where the impact energy is sufficiently dissipated.
force during this stage is represented as Fp, the plateau value of impact The characteristic values of the impact force for each specimen are
force. In this stage, a depression is gradually formed on the contact provided in Table 3.
surface between the upper flange and the hammer head, accompanied As shown in Fig. 8 (a), a higher impact energy results in a stronger
by an increase in the lateral bending deformation of the mid-span web contact between the specimen and the hammer, leading to a larger
plate. However, the impact energy during this stage is mainly absorbed impact force maximum value Fm. Although the strain rate increases
by the overall bending deformation of the specimen. with the increase in impact energy, within the range of impact velocities
As the impact energy is gradually absorbed, the motion velocity of considered in this study, the specimen remains at a moderate strain rate
the hammer and specimen decreases, and the rate of increase in mid- level. Moreover, the strain rate effect of high-strength steel is lower
span displacement slows down until it reaches a peak value Dm. The compared to that of ordinary low-carbon steel, resulting in a limited
mid-span displacement starts to decrease at 6 ms, suggesting a shift in increase in strength obtained through the strain rate effect [20].
energy absorption from overall bending deformation to localized Consequently, the impact force plateau value Fp is minimally affected by
deformation. Due to significant local deformation and the gradual the impact energy, as the impact energy ranges from 26.5 kJ to 51.94 kJ,
decrease in impact energy, the contact interaction between the falling resulting in a mere 7.3% increase in the specimen's impact force plateau
hammer and the specimen weakens, resulting in a decrease in the impact value Fp.
force. When the impact energy becomes excessive and localized defor­ At lower impact energies, as shown in Fig. 9 (a), the overall bending
mation can no longer absorb energy effectively, overall bending defor­ deformation of the specimen primarily absorbs the impact energy. The
mation may resume its role in energy absorption, the mid-span contact between the hammer and the specimen remains stable for an
displacement increases again. As the energy is further dissipated, the extended period of time, thereby resulting in the extended duration of
motion velocity of the specimen and hammer decreases to zero. The the stable stage of impact force Δs. The duration of the impact force
specimen and hammer then rebound until they separate, resulting in a descending stage Δd and the overall impact duration are significantly
constant mid-span displacement while the impact force declines to zero influenced by changes in the impact energy. These findings indicate that
at 21.9 ms. This stage is therefore referred to as the impact force the energy absorption in the perforated beam is mainly achieved
descending stage Δd. through localized plastic deformation.
When the impact energy remains constant, increasing the spacing
between the openings S in the HSSBCWO results in a higher flexural

5
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

stiffness, leading to an increase in the impact force maximum value Fm,

as shown in Fig. 8 (b). However, the impact force plateau value is
minimally influenced by the spacing between the openings S. When the
spacing between the openings S increases from 290 mm to 330 mm, the
impact force plateau value remains nearly unchanged. The increased
spacing between the openings S allows for a wider web width in the
HSSBCWO, enhancing its resistance to web buckling deformation [21].
This, in turn, leads to shorter durations of the impact force reduction
stage and the overall impact duration, as illustrated in Fig. 9 (b).
Increasing the height of the openings h in the HSSBCWO results in a
reduction in the remaining cross-sectional area, consequently, a
decrease in the HSSBCWO ‘s stiffness [22]. This leads to a decrease in the
impact force maximum value Fm and a reduction in the HSSBCWO ‘s
bending capacity. As a result, when subjected to an impact load, the
deformation of the HSSBCWO increases, and a greater amount of impact
energy is absorbed through the its deformation. The stress in the affected
region decreases as a result, leading to a decrease in the impact force
plateau value Fp. Specifically, when the opening height h increases from
160 mm to 200 mm, the impact force plateau value Fp decreases by 27%,
as shown in Fig. 8 (c). The duration of the impact force descending stage
and the overall impact duration of the beam increase as its deformation
increases, as illustrated in Fig. 9 (c).
The impact force maximum value Fm of the specimen is influenced by
factors such as the initial stiffness of the specimen, the contact area
between the hammer and the specimen, impact velocity, and impact
mass. Consequently, Fm is relatively unstable, primarily balanced by
inertial forces, with minimal growth in support reaction forces. Instead,
the support reaction forces reach their maximum value during the
impact force stable stage Δs, as depicted in Fig. 10. At this point, the
inertial forces are relatively small and stable, and the specimen reaches a
state of dynamic equilibrium, the impact force becomes more stable. In
summary, the force plateau value Fp is minimally affected by the impact
energy and is more closely related to the inherent anti-impact properties
of the HSSBCWO, which significantly reflect its impact performance.
Reducing the opening height h can improve the impact performance of
the HSSBCWO, whereas increasing the spacing between openings S has
an insignificant effect on its impact performance enhancement.
The impact duration Δt of a HSSBCWO under the action of impact
load is given by Δv + Δs + Δd, where Δv represents the duration of the
impact force vibration stage and Δs represents the duration of the impact
force stable stage. Both Δv and Δs are minimally influenced by the impact
energy and opening size. On the other hand, Δd, the duration of the
impact force descending stage, to some extent, determines the overall
impact duration. This suggests that the HSSBCWO primarily absorbs
impact energy through localized deformation.

3.3. Displacement

The mid-span displacement-time curves of specimens are shown in

Fig. 11. With the exception of specimen C10–5–180-310, the mid-span
displacement-time curves of each specimen continue to increase after
the initial rebound. Specimen C10–5–180-310, subjected to a relatively
lower impact energy, experiences significant overall bending deforma­
tion that absorbs most of the impact energy, resulting in its mid-span
displacement-time curve not exhibiting a second growth phase.
Fig. 12 illustrates the impact of various parameters on the maximum
mid-span displacement Dm of the HSSBCWOs. With an increase in
impact energy, the specimen absorbs more energy, resulting in higher
deformation and conversion of kinetic energy, which leads to larger
displacements. With an increase in impact energy from 26.5 kJ to 51.94
kJ, there is a 96% increase in energy, resulting in a 31.56% increase in
the maximum mid-span displacement Dm of the specimen. This indicates
that the HSSBCWO exhibits favorable impact performance, as shown in
Fig. 7. Time history curves of impact force. Fig. 12 (a).
The relatively small variation in the spacing between openings S in
the experimental design of this study leads to insignificant performance

6
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Table 3
Characteristic values of each specimen.
Specimen Ei (kJ) tmax (ms) Fm (kN) Fp (kN) Δv (ms) Δs (ms) Δd (ms) Δt (ms) β Dm (mm) Eab (kJ) EAR (%)

C10–5–180-310 26.5 0.272 965.95 493.62 1.63 5.42 8.69 15.74 0.55 38.62 23.5502 88.86
C12–5–180-310 38.16 0.25 1039.16 527.11 1.66 4.21 16.02 21.89 0.73 41.84 35.78133 93.76
C14–5–180-310 51.94 0.246 1261.63 551.14 1.60 3.80 25.22 30.62 0.82 50.81 51.10392 98.39
C12–4–180-310 30.96 0.276 1048.03 520.97 1.54 4.06 11.84 17.44 0.68 30.94 28.07154 90.67
C12–6–180-310 45.36 0.27 1118.61 539.58 1.60 3.83 21.84 27.27 0.80 47.33 43.56003 96.03
C12–5–160-310 38.16 0.27 1253.06 653.24 1.54 4.14 13.78 19.46 0.71 39.00 34.96604 91.63
C12–5–200-310 38.16 0.25 972.95 475.65 1.54 4.29 18.01 23.84 0.76 55.59 36.22531 94.93
C12–5–180-290 38.16 0.26 1064.15 526.14 1.63 3.52 17.95 23.10 0.78 41.73 36.24854 94.99
C12–5–180-330 38.16 0.31 1222.90 531.26 1.53 4.29 14.18 20.00 0.71 42.80 34.68227 90.88

Note: tmax represents the Peak time of impact force; Δt represents the duration of the impact; β represents the local deformation coefficient, calculated by Δd /Δt.

improvements with increased slotted spacing, as illustrated in Fig. 12 plastic deformation of the specimen absorbs the majority of the impact
(b). However, reducing the opening height h can mitigate the decline in energy. The values of β for all specimens range from approximately 0.5
beam strength and stiffness, and increase the cross-sectional area of the to 0.8, suggesting that within the range of impact energy in this exper­
HSSBCWO, and decrease its mid-span displacement, as depicted in iment, the primary response of the HSSBCWOs under impact load is
Fig. 12 (c). It is highly effective to reduce the displacement of the localized plastic deformation. The impact energy has a significant in­
HSSBCWO under impact loads by lowering the opening height h. fluence on β, with an evident increase in the β as the impact energy
increases, indicating larger localized plastic deformation of the
3.4. Deformation HSSBCWO. Additionally, increasing the spacing between openings S and
reducing the opening height h both lead to a decrease in β, thereby
3.4.1. Global deformation reducing the localized plastic deformation of the beam.
The HSSBCWO exhibits significant overall bending deformation
under the impact load. Due to the presence of openings in the web, the 3.4.3. Opening deformation
beam's section continuity is compromised, resulting in a decrease in its The stress distribution around the central openings is depicted in
stiffness. Consequently, the beam experiences severe local deformations Fig. 16, where the red color indicates the tensile stress region, and the
at the impact location, as shown in Fig. 13. Fig. 14 illustrates the blue color represents the compressive stress region [26]. In the tensile
deflection curves of the tested specimens after impact. Both an increase stress zone, the edge of the circular opening is stretched, the opening
in impact energy and an increase in beam opening height h contribute to shape is stretched from the arc to almost a straight line. Local buckling of
larger mid-span deflections. However, the effect of increasing the the web results in out-of-plane twisting deformation above the central
spacing between openings S on mid-span deflection is negligible. It is opening. Simultaneously, under significant compressive stress, the left
worth noting that when an impact load is applied to a HSSBCWO, a upper web and the web above the opening extruding, generating shear
strain wave will be generated near the impact location, which will deformation, as shown in Fig. 13.
propagate to both sides. During the propagation process, encountering a The projected shapes of the central openings in each specimen after
medium will result in the formation of reflected waves. The interaction impact are shown in Fig. 17. The deformation of the central openings
between the strain wave and the reflected wave affects each other, and increases with higher impact energy. Increasing the spacing between
an increase in spacing between openings S enlarges the medium, thereby opening and reducing the opening height can reduce the deformation of
slowing down the propagation speed of the strain wave in the the central openings. After the impact on the HSSBCWO, significant
HSSBCWO. Consequently, the strain wave is weakened by the time it deformations occur in the central openings, resulting in a decrease in the
reaches non-impact locations, resulting in altered strain distribution and projected area of the openings. The magnitude of the decrease in the
reduced deflection at those positions, as illustrated in Fig. 14 (b). projected area reflects the severity of opening deformation. To eliminate
the influence of opening size, the severity of opening deformation can be
3.4.2. Local deformation quantified by the opening area reduction rate (ARR). The ARR can be
Under the combined action of compression force and binding force, calculated by the ratio of the total area A2 after deformation of the
both the solid part and the opening part in mid-span of the beam will be central openings to the original total area A1, as shown in Fig. 18. The
subjected to shearing force, and Vierendeel bending will occur, resulting impact energy significantly affects the ARR, which exhibits approxi­
in local buckling of the mid-span web, as shown in Fig. 13. However, this mately exponential growth with increasing impact energy. Increasing
type of local plastic deformation is irregular and difficult to quantify its the spacing between opening and reducing the opening height can
extent of plastic deformation. As mentioned earlier, the local de­ mitigate opening deformation, thereby reducing the ARR.
formations of the HSSBCWOs primarily occur during the impact force
descending stage Δd. A larger proportion of the impact duration Δt 3.5. Energy absorption
occupied by the impact force descending stage Δd indicates a greater
amount of energy absorbed by the localized plastic deformation. When a HSSBCWO is subjected to impact from a drop hammer, a
Therefore, the ratio β of the duration of the impact force descending portion of the impact energy is absorbed and transformed into elastic
stage Δd to the impact duration Δt can reflect the proportion of energy deformation energy and plastic deformation energy. Additionally, some
absorbed by local deformations relative to the total absorbed energy, energy may be converted into forms such as heat and sound energy,
thereby indicating the magnitude of localized plastic deformation. The which are not absorbed by the HSSBCWO but instead transmitted and
values of β for each specimen are presented in Table 3. The influence of dissipated through the surrounding air and objects [23]. The remaining
different parameters on β is illustrated in Fig. 15. For specimen C10-5- portion of the energy is converted into kinetic energy during the
180-310, β is 0.55, indicating that when the impact energy is <26.5 rebound of the hammer. Neglecting the effects of heat energy, sound
kJ, the overall bending deformation of the specimen absorbs most of the energy, and internal deformation energy of the hammer, the change in
impact energy. When the impact energy exceeds 26.5 kJ, the localized kinetic energy of the hammer corresponds to the work done by the

7
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 9. Characteristic value of impact force under each parameter (time).

impact force, as shown in Eq. (1).

1 ( 2 )
m v − v(t)2 = W(t) (1)
2
In the equation, m represents the mass of the drop hammer, v denotes
the initial impact velocity of the drop hammer, v(t) represents the ve­
locity of the drop hammer at time t, and W(t) represents the work done
by the impact force. By using a high-speed camera, the velocity of the
hammer at the end of the impact can be obtained, allowing for the
calculation of the impact energy absorbed by the HSSBCWO. The energy
absorbed and energy absorption rate by each specimen is presented in
Table 3.
Fig. 19 illustrates the variations in energy absorptions (Eabs) and
Fig. 8. Characteristic value of impact force under each parameter (force).
energy absorption rates (EARs) of the HSSBCWOs under different pa­
rameters. As the impact energy increases, the HSSBCWO undergoes
larger overall bending deformation and local plastic deformation,

8
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 10. Time history curve of impact force and reaction force.

resulting in the absorption of more impact energy. The energy absorp­

tion Eab demonstrates a roughly linear growth trend with increasing
impact energy. Notably, at an impact energy of 51.94 kJ, the HSSBCWO
absorbs 98.4% of the impact energy. Since the deformation of the
HSSBCWO primarily occurs during the impact force descending stage
Δd, a significant portion of the energy (55%–82%) is absorbed through
the HSSBCWO ‘s local plastic deformation [24]. Therefore, the evalua­
tion of the impact resistance of the perforated beam should primarily
focus on the impact force descending stage Δd. As previously mentioned,
the dimensions of the openings have a significant impact on the local
deformation of the HSSBCWO. Decreasing the spacing between openings
S and increasing the height of the openings h both contribute to
increased local deformation in the perforated beam, thus enhancing the
absorption of impact energy. The average energy absorption rate for all
specimens is 93.35%, demonstrating the exceptional energy absorption
capacity of the HSSBCWO.

3.6. Acceleration

During the experiment, the acceleration data at various measure­

ment points were collected using accelerometers and plotted as accel­
eration time history curves, as shown in Fig. 20. The positive
acceleration in the direction of gravity was considered [25]. In the
impact force vibration phase Δv, strong contact occurs between the drop
hammer and the specimen, resulting in significant fluctuations in the
acceleration time history curve of the HSSBCWO, followed by gradual
attenuation. When the HSSBCWO is subjected to the impact load, the
impact force rapidly propagates to the mid-span position of the beam.
The mid-span position A1 exhibits the first acceleration response, while
the stress wave is influenced by the interference of the reflected wave,
causing a delay in the acceleration response at A2 and A3. Among the
specimens, the response time at A2 and A3 is the shortest for specimen
C12–5–180-290 due to its smallest spacing between opening. The
average response times at A2 and A3 for all specimens are 0.07 ms and
0.17 ms, respectively, and the acceleration propagates from A1 to A2
and from A2 to A3 at velocities of 4428.57 m/s and 3100 m/s, indicating
that the propagation speed of the stress wave decreases with increasing
propagation distance. As the impact location of the drop hammer is at
the mid-span of the HSSBCWO, the mid-span experiences greater stress
and strain, resulting in larger vibrations and deformations. Therefore,
the acceleration response in this region is stronger, with the maximum
Fig. 11. Time history curves of mid-span displacement.
acceleration occurring at A1, while the accelerations at A2 and A3 are
smaller. The average maximum accelerations at A1, A2, and A3 are
727,333.3 m/s2, 149,222.21 m/s2, and 46,625 m/s2, respectively. The

9
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

maximum accelerations at A2 and A3 are only 20.52% and 6.4% of the

maximum acceleration at A1.
Nine prominent peaks were selected as feature points near the ac­
celeration decay line in Fig. 21. The acceleration values corresponding
to these feature points were extracted from the acceleration time history
curve to construct the acceleration distribution curve at different time
intervals, as shown in Fig. 22. Following the application of the impact
load on the HSSBCWO, the acceleration is primarily concentrated in the
mid-span region, while the acceleration in the non-mid-span regions is
relatively small. There is almost no acceleration distribution between A3
and the supports. As shown in Fig. 24 (a), at the time of 0.269 ms for
specimen C10–5–180-310, the acceleration at A2 is minimal, and there
is nearly no acceleration at A3. Meanwhile, the peak value of the impact
force occurs around 0.25 ms, during which only the mid-span region
experiences acceleration. As a result, the peak impact force is mainly
balanced by the inertia force generated in the mid-span region.

4. Finite element analysis

4.1. FEA model

4.1.1. Established model

To further analyze the impact performance of HSSBCWOs under
impact loads, finite element analysis was conducted using ABAQUS/
Explicit. Fig. 23 presents a representative view of the finite element
model. 8-node linear C3D8R solid elements were employed to ensure
modeling accuracy [26]. The mesh size was set to 20 mm. To better
simulate the dynamic response of the HSSBCWO under impact loads,
mesh refinement was applied in the hammer impact area, with a mesh
size of 10 mm. The web and flange were divided into two mesh layers in
the thickness direction [27]. As the hammer was considered nearly rigid,
it was modeled as a rigid body using the R3D4 rigid element. Constraints
were applied to the hammer to restrict its movement in the vertical di­
rection using a constraint reference point, in accordance with the
experimental conditions. The integrity of the HSSBCWO structure was
ensured by employing tie connections between the web and flange plates
[28]. Hinged supports were used to constrain both ends of the
HSSBCWO in the experiments. To accurately simulate the effect of the
hinged supports, the hinged supports at both ends were coupled to the
restrained reference points and allowed rotation only about the Z-axis
[29]. In terms of contact, a contact model was employed to simulate the
interaction between the hammer and the HSSBCWO [26]. In the normal
direction, there was a hard contact between the hammer and the upper
flange, while in the tangential direction, Coulomb friction was consid­
ered with a friction coefficient of 0.2 [30]. These contact constraints
ensured accurate simulation of the relative motion and dynamic
response between the hammer and the beam. The impact velocity of the
hammer was implemented by defining a predefined field. These detailed
treatments contributed to model accuracy and provided reliable foun­
dational data for subsequent analysis.

4.1.2. Material properties

The high-strength structural steel was utilized in the experiments
conducted in this study, and its fundamental material properties are
presented in Table 1. Yang et al., in order to predict the dynamic stress-
strain response of structural steels with yield strengths ranging from
235Mpa-960Mpa, a continuous dynamic constitutive model is proposed
from static to high strain rate loading conditions (up to 5000 s− 1) [12],
as shown in Eq. (2).
( ) ( )
σ = σs fy , ε ⋅DIFavg fy , ε̇ (2)

Where, σs (fy, ε) is the static model of steel, as shown in Eq. (3).

Fig. 12. Maximum mid-span displacement of each parameter.

10
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 13. Failure mode of HSSBCWO.

⎧
⎪ Eε ε ≤ εy 4.1.3. Model validation
⎪
⎪
⎪
⎪ Fig. 24 compares the experimental and finite element simulation
⎪
⎪ fy εy ≤ ε ≤ εsh
⎨
⎛ ⎞ results of the deformation patterns for specimen C12–5–180-310. The
σ= (3) results demonstrate that the aforementioned model accurately simulates
⎪
⎪ ( )⎜ 2ε* ⎟
⎪
⎪
⎪
⎪ fy + fu − fy ⎝0.4ε* + [ ]0.2 ⎠ εsh < ε < εu the overall bending deformation, flange buckling deformation, openings
⎪ * 5
⎩ 1 + 400(ε ) deformation, and web buckling deformation of the HSSBCWO. The
impact force and mid-span displacement time-history curves for each
Where σ and ε are the stress and strain, respectively, and ε* is shown specimen are shown in Fig. 25 and Fig. 26, respectively. The impact
in Eq. (4). force and displacement time-history curves obtained from the finite
ε − εsh element model show a significant agreement with the experimental re­
ε* = (4) sults. Table 4 presents the average ratios of the impact force platform
εu − εsh
value Fp and the maximum mid-span displacement Dm obtained from the
Where εu and εsh are ultimate strain, strain hardening strain, finite element model to the corresponding experimental values, which
respectively. As shown in Eq. (5) and (6). are 0.99 and 0.94, respectively. The results reveal a significant
( ) concurrence between the finite element simulation and experimental
f
εu = 0.6 1 − y (5) observations.
fu

f 4.2. Ratio of opening height to beam height h/H

εsh = 0.1 y − 0.055 (6)
fu
Using finite element simulation, the influence of opening height on
DIFavg (fy, ε•) in Eq. (2) is the strain rate effect model of structural
the impact performance of the HSSBCWO was further investigated by
steel, as shown in Eq. (7).
varying the parameter h/H. The range of h/H was chosen between 0.40
( )
( ) ε̇ pavg and 0.70, and the impact energy was set at 51.94 kJ. The impact plateau
(7)
1
DIFavg fy , ε̇ = 1 +
Davg value Fp and the maximum mid-span displacement Dm of each param­
eter h/H are shown in Fig. 27. It can be observed that as h/H increases,
Where
the impact force plateau value Fp decreases nearly linearly, when h/H
(
fy
)6 increases from 0.4 to 0.7, the impact force plateau value Fp decreases by
Davg = 1000 (8) 65.17%, as shown in Fig. 27 (a). When h/H is <0.5, the increase in
235
maximum mid-span displacement Dm is not significant. However, when
(
fy
)0.2 h/H exceeds 0.5, the maximum mid-span displacement Dm rapidly in­
pavg = 3 (9) creases, as shown in Fig. 27 (b). Therefore, when designing HSSBCWOs,
235
it is advisable to keep h/H within 0.5. It is worth noting that when h/H is
An appropriate damage criteria model was selected, which takes into <0.5, the maximum mid-span displacement Dm remains almost un­
account the relationship between fracture strain and stress triaxiality changed, and further reducing the opening height would only increase
[31], as shown in Eq. (10) and (11). the self-weight of the perforated beam. Thus, within the range of impact
/ ⎧ energy considered in this experiment (0 kJ < Ei < 51.94 kJ), setting h/H
⎨ 1.13 T σ ≤ − 1/3
εfd εu = 0.04 + 0.86exp( − 0.7T σ ) − 1/3 < T σ ≤ 10/3 (10) to 0.5 is the most reasonable choice for the HSSBCWO.
⎩
0.12 10/3 < T σ
4.3. Ratio of opening height to beam height S/H
/ ⎧ 0.43 T σ ≤ 5/3
⎨
Taking the ratio of spacing between openings S to beam height H S/H
εfs εu = 0.38 + 0.40exp[6.69(T σ − 2) ] 5/3 < T σ ≤ 2 (11)
⎩ as the independent variable, further investigation was conducted to
0.78 2 < Tσ explore the influence of spacing between openings S on the impact
Where, εfd and εfs are initial fracture plastic strain for ductile failure performance of the HSSBCWO. The range of S/H was chosen between
and shear failure, respectively. εu is the ultimate plastic strain of steel; Tσ 0.87 and 1.13, and the impact energy was set at 51.94 kJ. The impact
sigma means stress triaxiality. plateau value Fp and the maximum mid-span displacement Dm of each
parameter S/H are shown in Fig. 28. It can be observed that the variation

11
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 15. β of each parameter.

Fig. 16. Stress distribution.

energy considered in this experiment (0 kJ < Ei < 51.94 kJ), setting S/H
Fig. 14. Deflection curves of the HSSBCWOs. to 0.93 is the most reasonable choice for the HSSBCWO.

in S/H has a minor effect on the impact performance of the HSSBCWO. 5. Damage assessment
When S/H is <0.93, a significant decrease in the impact force plateau
value Fp is observed. However, when S/H is <0.93, the impact force Among various measurable parameters, the displacement at the
plateau value Fp stabilizes with an average value of 570.90 kN, as shown impact location of the HSSBCWO is the most easily quantifiable.
in Fig. 28 (a). Therefore, when designing HSSBCWOs, S/H should be Therefore, this study adopts the residual displacement criterion to assess
>0.93. Considering that increasing the spacing between openings S the degree of damage of the HSSBCWO [32]. The residual displacement
would increase the self-weight of the beam, within the range of impact criterion evaluates the damage level of the component by comparing the

12
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 17. Deformations of the central openings.

Fig. 19. EARs and Eabs of each parameter.

Fig. 18. ARRs of the O3 and O4.

represent different types of damage, where purple, green, red, and black
represent Class I, Class II, Class III, and Class IV damage, respectively.
residual displacement Dr after impact with its calculated span L0, Cyan, blue, and pink indicate the critical conditions between Class I and
expressed as the ratio Dr / L0. This criterion is used to classify the Class II damage, Class II and Class III damage, and Class III and Class IV
damage level into different categories. The corresponding ranges for Dr / damage, respectively.
L0 are as follows: mild damage, categorized as Class I, with 0 < Dr / L0 ≤ To obtain the critical surface equation between Class I and Class II
1/120; moderate damage, categorized as Class II, with 1/120 < Dr / L0 damage, Class II and Class III damage, Class III and Class IV. Three
≤ 1/30; severe damage, categorized as Class III, with 1/30 < Dr / L0 ≤ 1/ critical conditions were fitted separately.
12; and failure and complete collapse, categorized as Class IV, with Dr / The critical surface equation between Class I and Class II damage is
L0 ≤ 1/12 [45]. Different opening sizes can result in varying degrees of shown in Eq. (12).
damage to HSSBCWOs under impact loads, and as mentioned earlier, the
( )4.31 ( )1.12 ( )4.31 ( )1.12
impact energy is the primary controlling factor influencing the damage h S h S
Ei = 11.03 − 24.37 + 2.60 − 8.53 (12)
in HSSBCWOs components. Therefore, by considering the impact energy H H H H
Ei, opening height h, and spacing between openings S as parameters, this
The critical surface equation between Class II and ClassIII damage is
study investigates the effects of varying combinations of impact energy,
shown in Eq. (13).
opening height h, and spacing between openings S on the damage in
perforated beams. ( )2.63
h
( )− 30.86
S
( )2.63 ( )− 30.86
h S
The results of 324 numerical simulations are compiled, and based on Ei = 94.02 − 123.06
H
+ 0.44
H
− 1.68
H H
the corresponding Dr / L0 values, the damage model of the components is
(13)
obtained and presented in Fig. 29. In the figure, different colors

13
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 20. Time history curves of acceleration.

14
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 20. (continued).

variation in the spacing between openings S has a minor influence on the

degree of damage in the HSSBCWO, as the three critical surfaces are
nearly parallel to the y-axis. The range of impact energy for the critical
surface between Class I and Class II damage is approximately
4.98–13.02 kJ. For the critical surface between Class II and Class III
damage, the range of impact energy is approximately 14.00–107.00 kJ.
The critical surface between Class III and Class IV damage has an impact
energy range of approximately 79.8–120.20 kJ. When the impact energy
is relatively low, the opening height h is the dominant factor affecting
the degree of damage, as shown in Fig. 30 (a). As the impact energy
increases, the damage also increases. The critical surface between Class
III and Class IV damage is nearly parallel to a plane perpendicular to the
y-axis, indicating a linear correlation between the impact energy and the
opening height h, as shown in Fig. 30 (c). Therefore, when the impact
energy is high, both the impact energy and the opening height are
dominant factors in determining the degree of damage in the perforated
beam.
The three E-h-S surfaces divide the damage space coordinate system
into four regions. By assessing the opening height h, spacing between
openings S, and impact energy of the HSSBCWO, its position in the
Fig. 21. Characteristic points on time-history curve of acceleration at position
of A1.
damage space coordinate system can be determined, allowing for a
quick assessment of the degree of damage.
The critical surface equation between ClassIII and ClassIV damage is
6. Conclusions
shown in Eq. (14).
( )1.00 ( )6.86 ( )1.00 ( )6.86
h S h S This study investigates the impact performance of HSSBCWOs by
Ei = 269.02 − 268.97 + 2.60 − 31.11 conducting drop hammer impact tests on nine specimens. The effects of
H H H H
(14) impact energy, opening height, and spacing between openings on the
HSSBCWOs ‘performance are analyzed. Numerical simulations are
The critical surface is depicted in Fig. 30. It can be observed that the employed to analyze the influence of different h/H and S/H ratios on the

15
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 22. Acceleration distribution curve.

16
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 22. (continued).

17
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 23. Finite element model.

The HSSBCWOs exhibit good displacement resistance. The impact

energy and opening height h have a significant influence on the
maximum mid-span displacement, with a 31.56% increase observed
when the impact energy increases from 26.5 kJ to 51.94 kJ, and a
42.54% increase observed when the opening height h increases from
160 mm to 200 mm. However, the variation in spacing between open­
ings S does not significantly affect the mid-span displacement of the
HSSBCWOs. While increasing the spacing between openings S does not
reduce the deflection at the impact location, it effectively inhibits the
increase in deflection at non-impact locations. The coefficient β, which
reflects the magnitude of local plastic deformation in the HSSBCWO,
Fig. 24. Comparison of failure modes between test and finite element method. ranges between approximately 0.5 and 0.8 for all specimens, indicating
that local deformation is predominant within the range of impact en­
ergies considered in the experiment. Increasing the spacing between
openings S and decreasing the opening height h both lead to a reduction
impact performance of the HSSBCWOs. Furthermore, a total of 324 in the β coefficient and subsequently decrease the level of local plastic
models are developed to assess the damage levels of the HSSBCWOs. The deformation in the HSSBCWO. The area reduction ratio δ of the openings
following conclusions are drawn: exhibits an exponential growth trend with increasing impact energy, and
During the impact process, the initial stage of impact is primarily increasing the spacing between openings S or reducing the opening
characterized by overall bending deformation, while the subsequent height h results in a decrease in the area reduction ratio δ.
stage of impact is dominated by local deformation. However, when the The HSSBCWOs demonstrate excellent energy absorption capacity,
impact energy becomes excessively high and local deformation cannot with an average energy absorption rate of 93.35%. The propagation
absorb further energy, the contribution of overall bending deformation velocity of stress waves decreases with the increase in propagation
to energy absorption increases again. distance, and the average maximum accelerations at locations A1, A2,
The impact force maximum value Fm increases with the increase in and A3 are 727,333.3 m/s2, 149,222.21 m/s2, and 46,625 m/s2,
impact energy and spacing between openings S, while it decreases with respectively. At the moment of impact force maximum value Fm, ac­
the increase in opening height h. The impact force plateau value Fp is celeration is only distributed near the mid-span region, with other parts
primarily sensitive to the opening height h. When the opening height h is of the beam experiencing negligible acceleration.
increased from 160 mm to 200 mm, the impact force plateau value Fp Based on the finite element analysis results, it is recommended to
decreases by 27%, while the increase in impact energy from 26.5 kJ to maintain the values of h/H and S/H at around 0.5 and 0.93, respectively,
51.94 kJ results in only a 7.3% increase in the impact force plateau value within the range of impact energies tested in this experiment (0 kJ < Ei
Fp. The variation in spacing between openings S from 290 mm to 330 < 51.94 kJ). A damage coordinate system and damage surfaces are
mm hardly affects the impact force plateau value Fp. The impact force established using the impact velocity, opening height, and spacing be­
plateau value Fp decreases with increasing impact energy until it reaches tween openings as the primary control parameters, with the Dr / L0
stability at 45.36 kJ. Increasing the opening height h or reducing the serving as the damage assessment indicator. This approach enables the
opening spacing S leads to an increase in the duration of impact force rapid determination of the damage level of HSSBCWOs.
descending stage Δd.

18
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 25. Comparison between numerical and experimental time-history curves of impact force.

19
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 26. Comparison between numerical and experimental time-history curves of mid-span.

20
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Table 4
Comparison between numerical and experimental conclusion.
Specimen Fp,test/kN Fp,FE/kN Fp,FE/Fp,test Dm,test/mm Dm,FE/mm Dm,FE/Dm,test

C10–5–180-310 493.62 540.18 0.91 35.21 37.52 0.94

C12–5–180-310 527.11 501.07 1.05 40.84 39.50 1.03
C14–5–180-310 551.14 547.75 1.01 52.06 58.19 0.89
C12–4–180-310 520.97 521.95 1.01 32.83 41.28 0.80
C12–6–180-310 539.58 558.21 0.97 47.91 51.51 0.93
C12–5–160-310 653.24 625.62 1.04 39.00 40.62 0.96
C12–5–200-310 475.65 454.58 1.05 55.59 56.41 0.99
C12–5–180-290 526.14 546.61 0.96 41.73 42.62 0.98
C12–5–180-330 531.26 560.71 0.95 42.80 43.79 0.98

Note: Fp,test represents the impact force platform value obtained from the test; Fp,FE/kN represents the impact force platform value obtained from the finite element
model; Dm,test represents the he maximum mid-span displacement obtained from the test; Dm,FE represents the the maximum mid-span displacement obtained from the
finite element model.

Fig. 27. Fp and Dm of each h/H.

Fig. 28. Fp and Dm of each S/H.

21
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Fig. 29. Model of damage.

Nomenclature

L Beam length
L0 Calculated span
H Beam height
bf beam width
tw Web thickness
tf Flange thickness
h Opening heights
S spacing between openings
O1, O2, O3, O4, O5, O6 Names of openings，respectively
v Impact velocity
m Mass of the drop hammer
Ei Impact energy
fy Yield strength
fu Ultimate tensile strength
E Elasticity modulus
ts Thickness of the material specimen
Fm Impact force maximum value
Fp Impact force plateau value
Δv Duration of the impact force vibration stage
Δs Duration of the impact force stable stage
Δd Duration of the impact force descending stage
Δt Duration of impact
Dm Maximum mid-span displacement
β Local deformation coefficient
ARR Opening area reduction rate
v(t) Velocity of the drop hammer at time t
Eab Energy absorption
EAR Energy absorption rate
σs Quasi-static stresses
εu Ultimate strain
εsh Strain hardening strain
Fp,test Impact force platform value obtained from the test
Fp,FE Impact force platform value obtained from the finite element
model
Dm,test Maximum mid-span displacement obtained from the test
Dm,FE Maximum mid-span displacement obtained from the finite
element model
Dr Residual displacement
Fig. 30. Critical surface of damage mode.

22
H. Chen et al. Journal of Constructional Steel Research 211 (2023) 108159

Declaration of Competing Interest [9] W. Zhang, P. Hao, Y. Liu, et al., Determination of the dynamic response of Q345
steel materials by using SHPB, Proc. Eng. 24 (2011) 773–777.
[10] X. Yang, H. Yang, S. Zhang, Rate-dependent constitutive models of S690 high-
The authors strictly comply with ethical standards of research and strength structural steel, Constr. Build. Mater. 198 (2019) 597–607.
publishing paper. This research work was supported by the Natural [11] J. Tang, M. He, Y. Qiao, et al., Tensile behavior of a novel high-strength and high-
Science Foundation of Fujian Province (No.2022J01928， toughness steel at strain rates from 0.1 s− 1 to 1000 s− 1, Constr. Build. Mater. 304
(2021), 124606.
No.2020J01885) and Fujian Provincial Department of Housing and [12] X. Yang, H. Yang, L. Gardner, et al., A continuous dynamic constitutive model for
Urban Rural Development Science and Technology Plan Project(No. normal-and high-strength structural steels, J. Constr. Steel Res. 192 (2022),
2022-K-71, No. 2022-K-305). The authors declared that they have no 107254.
[13] A.S. Al-Husainy, A. Al-Rifaie, W. Ogaidi, Behaviour of steel beams with circular
conflicts of interest to this work. We declare that we do not have any web openings under impact loading, in: Conference Series: Materials Science and
commercial or associative interest that represents a conflict of interest in Engineering 888, IOP Publishing, 2020, p. 012069 (1).
connection with the work submitted. We declare that we have no [14] A. Al-Rifaie, A.S. Al-Husainy, T. Al-Mansoori, et al., Flexural impact resistance of
steel beams with rectangular web openings, Case Stud. Const. Mater. 14 (2021),
financial and personal relationships with other people or organizations e00509.
that can inappropriately influence our work, there is no professional or [15] F. Wang, C. Fu, H. Chen, et al., Effect of impact loading on the dynamic response of
other personal interest of any nature or kind in any product, service and/ steel beams with hexagonal web opening, Thin-Walled Struct. 180 (2022), 109896.
[16] C. Luo, F. Wang, H. Chen, et al., Castellated steel beams under impact load,
or company that could be construed as influencing the position pre­ J. Constr. Steel Res. 196 (2022), 107394.
sented in, or the review of, the manuscript entitled, “Impact behavior of [17] GB 50017-2017 Standard for Design of Steel Structures, China Architecture and
high-strength steel beam with circular web openings”. Building Press, Beijing, 2017 (in Chinese).
[18] GB/T 228.1-2010 Metallic Materials-Tensile Testing at Room Temperature, China
Planning Press, Beijing, 2010 (in Chinese).
Data availability [19] Y.B. Zhang, L.Y. Sun, L.J. Li, et al., Effects of strain rate and high temperature
environment on the mechanical performance of carbon fiber reinforced
Data will be made available on request. thermoplastic composites fabricated by hot press molding, Compos. Part A 134
(2020), 105905.
[20] H. Yang, X. Yang, A.H. Varma, et al., Strain-rate effect and constitutive models for
Acknowledgments Q550 high-strength structural steel, J. Mater. Eng. Perform. 28 (11) (2019),
6626–6637. 20.
[21] A. Mohebkhah, H. Showkati, Bracing requirements for inelastic castellated beams,
This research work was supported by the Natural Science Foundation J. Constr. Steel Res. 61 (10) (2005), 1373–1386. 21.
of Fujian Province (No.2022J01928，No.2020J01885) and Fujian [22] D. Sonck, J. Belis, Lateral-torsional buckling resistance of castellated beams,
Provincial Department of Housing and Urban Rural Development Sci­ J. Struct. Eng. 143 (3) (2017), 04016197. 22.
[23] H. Hao, T.T. Tran, H. Li, et al., On the accuracy, reliability and controllability of
ence and Technology Plan Project(No. 2022-K-71，No. 2022-K-305). impact tests of RC beams, Int. J. Impact Eng. 157 (103979) (2021) 23.
[24] M.T. Nawar, I.T. Arafa, O.M. Elhosseiny, et al., Full static resistance of castellated
References steel beams with hexagonal web openings for blast response predictions, Eng.
Struct. 245 (2021), 112844.
[25] F.F. Pérez, B. Luccioni, J.C. Vivas, et al., High strength fiber reinforced beams
[1] Y.S. Du, D.H. Guo, Z.H. Chen, et al., Experimental study on the flexural behavior of
under impact load, Int. J. Impact Eng. 159 (2022), 104039.
square high-strength concrete-filled steel tube beams with different CFRP wrapping
[26] F.S. Makarem, F. Abed, Nonlinear finite element modeling of dynamic localizations
schemes, Compos. Struct. 304 (2022), 116325.
in high strength steel columns under impact, Int. J. Impact Eng. 52 (2013) 47–61.
[2] H. Al-azzani, W. Wang, Y. Zhang, et al., Experimental study on behavior of
[27] J. Wei, J. Li, C. Wu, et al., Impact resistance of ultra-high performance concrete
restrained high strength Q690 steel beams exposed to fire, J. Constr. Steel Res. 197
strengthened reinforced concrete beams, Int. J. Impact Eng. 158 (2021), 104023.
(2022), 107463.
[28] B. Yang, H. Wang, Y. Yang, et al., Numerical study of rigid steel beam-column
[3] B.H. Jiang, Z.H. Wang, M.J. Wang, et al., Effects of loading mode on mechanical
joints under impact loading, J. Constr. Steel Res. 147 (2018) 62–73.
properties of high strength steel Q690 and their application in coupon test, Constr.
[29] K. Chen, Y. Zhang, K.H. Tan, Behaviour of steel beam-column joints subjected to
Build. Mater. 253 (2020), 118969.
quasi-static and impact loads, J. Constr. Steel Res. 183 (2021), 106721.
[4] F.F. Liao, K.S. Zhang, W. Wang, et al., Fracture analysis of high-strength steel
[30] E. Munoz-Garcia, J.B. Davison, Tyas., An analysis of the response of structural bolts
beam-column connections with initial defects, J. Constr. Steel Res. 194 (2022),
subjected to rapid rates of loading, in: Euro Steel Conference on Steel and
107301.
Composite Structures, Maastricht, The Netherlands, 2005.
[5] S. Demirdjian, Stability of Castellated Beam Webs, McGill University, Department
[31] B. Yang, H. Wang, Y. Yang, S.B. Kang, X.H. Zhou, L. Wang, Numerical study of rigid
of Civil Engineering and Applied Mechanics, 1999.
steel beam–column joints under impact loading, J. Constr. Steel Res. 147 (2018)
[6] M.R. Soltani, A. Bouchaïr, M. Mimoune, Nonlinear FE analysis of the ultimate
62–73.
behavior of steel castellated beams, J. Constr. Steel Res. 70 (2012) 101–114.
[32] T. Krauthammer, Explosion Damage Assessment[R]. First Structural Forensic
[7] A.Z. Zhu, W. Xu, K. Gao, et al., Lateral impact response of rectangular hollow and
Engineering Seminar on Structural Failure Investigations, University of Toronto,
partially concrete-filled steel tubular columns, Thin-Walled Struct. 130 (114–131)
1999, p. 1.
(2018) 7.
[8] A.A. Alabi, P.L. Moore, L.C. Wrobel, et al., Tensile behaviour of S690QL and
S960QL under high strain rate, J. Constr. Steel Res. 150 (2018) 570–580.

23