Journal of Building Engineering 71 (2023) 106474

Contents lists available at ScienceDirect

Journal of Building Engineering

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

Analytical model for progressive collapse of RC frame structures

subjected to blast loadings
SeokJun Ju, Hyo-Gyoung Kwak *
Dept. of Civil and Environmental Engineering, Korea Advanced Institute of Science and Technology, 291 Daehak-ro, Yuseong-gu, Daejeon, 34141,
Republic of Korea

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

Keywords: An improved analytical model that enables tracing of the progressive collapse of RC frame
RC Frame structures under blast loading is introduced in this paper. The moment-curvature relation
Blast analysis uniquely defined in an RC section is used for the simulation of the bending behavior, and then the
Moment-curvature relation constructed nonlinear moment-curvature relation is modified with the consideration of many
Direct shear failure factors such as the effects of bond-slip, axial force and catenary action, which dominantly affect
Progressive collapse
the nonlinear response of RC structures. Upon the definition of the monotonic moment-curvature
envelope, the hysteretic unloading and reloading paths are defined to trace the dynamic bending
behavior of an RC section on the basis of the hysteretic curve of reinforcing steel. In advance, to
describe the direct shear failure, the empirical direct shear stress-slip relation suggested by
Krauthammer and Astarlioglu is considered in the solution procedure and implemented into the
formulation with the use of non-dimensional spring element. Moreover, as was in the moment-
curvature relation, the hysteretic rule in the shear stress-slip relation proposed by Kraut­
hammer and being used in many previous researches is also adopted. Because the numerical
approach based on the moment-curvature relation of RC sections will dramatically reduce the
solution steps, this approach makes it possible to trace the collapse not only of RC members but
also of RC structures that are comprised of many structural members and require enormous so­
lution procedures. After verifying the reliability and accuracy of the introduced numerical model
through correlation studies of the RC members and a simple frame, the collapse analyses of multi-
bay and multi-story RC frames are performed to prove the possibility for the progressive collapse
analysis of entire frame structures only with the use of beam elements.

1. Inroduction
Because explosive incidents and accidental explosions cause serious damage in structural members which usually lead to the
progressive collapse of the entire structure, the assessment of the resisting capacity becomes more important for the structural
members and, furthermore, for the entire structure subjected to the blast loading. Recent rapid increases in the size and height of
structures have also accelerated the need to secure structural resistance to blast loadings [1]. To exactly evaluate the resistance of RC
structures subjected to blast loadings, accordingly, a lot of numerical and experimental studies have been performed [2–4] from the
verification of material properties to the analysis of structural behavior under blast loadings which are accompanied by a high strain
rate deformation.

* Corresponding author.
E-mail address: [email protected] (H.-G. Kwak).

https://doi.org/10.1016/j.jobe.2023.106474
Received 5 September 2022; Received in revised form 10 March 2023; Accepted 2 April 2023
Available online 3 April 2023
2352-7102/© 2023 Elsevier Ltd. All rights reserved.
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

Especially, reinforced concrete (RC) structures subjected to a blast loading present significantly different structural behaviors from
that observed under a quasi-static loading because of the remarkable change in the material properties of concrete under a blast
loading accompanied by a high strain rate deformation [5]. Moreover, an exact evaluation of the structural response in RC members
requires additional consideration of nonlinear effects induced by the bond-slip effect and the direct shear behavior. To trace the
nonlinear response of the RC structural members subjected to blast loadings, rigorous finite element (FE) analyses with the use of
innumerable solid elements have been done in the case of numerical analyses [6,7]. However, modeling an entire RC structure sur­
passing a member level using solid elements will be unsuitable due to a tremendous increase in the number of finite elements and the
accompanying solution steps, and it might be one of the reasons that most of the FE analyses are limited to an RC member level.
Moreover, it means that the numerical simulation for the progressive collapse of RC structures will be costly and time consuming.
Only a limited number of blast analyses have been performed in a structure level which include many beams and columns.
Especially, to minimize these excessive computational costs, simplified analytical approaches have been used such as an alternative
load path (ALP) analysis [8] which is described in the US General Service Administration (GSA) and US Department of Defense (DoD)
guidelines; however, it is also true that these approaches have the fundamental limitation in describing the progressive collapse
because they do not consider the initial damage or the subsequent damage in the adjacent members of the damaged regions [9]. To
remedy this limitation, Shi et al. [9] proposed an improved ALP analysis reflecting the previous damages assessed by the P–I diagram,
but this approach also has not overcome the limitation in describing the progressive collapse of an entire structure because the sig­
nificant coupling phenomenon between the explosion and the accompanying progressive collapse was excluded.
To address these limitations in conducting a progressive collapse analysis of RC frames under blast loading, this paper presents an
improved nonlinear solution procedure based on the simplified moment-curvature (M-Φ) relation of the RC sections. The use of the M-
Φ relation can dramatically reduce the solution steps [10] because (1) an RC member will be modeled with a very limited number of

Fig. 1. Procedure for constructing the M-Φ relation of an RC section.

2
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

beam elements instead of innumerable solid elements, and (2) even in an RC element, the solution steps, in comparison with the
layered section approach which requires the repeated calculation of the neutral axis [11], are remarkably reduced. It means that the
use of beam elements based on the M-Φ relation will eventually facilitate the progressive collapse analysis of entire RC frame structures
which are made up of many beams and columns.
Upon the definition of the monotonic M-Φ curve, while considering the nonlinear effects of bond-slip and axial force, which in­
fluence the structural response, the hysteretic behavior should be supplemented because an RC member may experience unloading
and/or reloading behavior after yielding due to redistribution of the member force in the entire structure [12,13]. Moreover, to exactly
simulate the progressive collapse behavior, the analytical model proposed in this paper considers the axial force change and the
catenary action, which basically accompany the progressive collapse process. These two effects are implemented through modification
of the M-Φ relation, instead of adopting rigorous independent formulations.
Unloading and reloading responses of an RC section are defined on the basis of the hysteretic behavior of reinforcing steel because
the applied moment will be carried very largely by the steel reinforcement placed in an RC section after the first yield excursion [14].
Especially, because the bond-slip is initiated with the occurrence of cracking and enlarged after the reinforcing steel yields which
accompanies the rigid body rotation, the bond-slip effect is taken into consideration by modifying the slope of the M-Φ relation, and
this modified relation is applied within the plastic hinge regions in an RC member, such as the beam-column joints and mid-span where
the yielding of reinforcing steel is concentrated [15]. To consider the direct shear failure, the direct shear stress-slip relation proposed
by Krauthammer and Astarlioglu [16] is based to define the envelope curve, and the non-dimensional spring element is adopted to
implement the shear slip phenomenon [17]. The direct shear failure of structural members can be described by the magnitude of the
relative displacement in the spring element. After establishing the validity of the proposed approach by comparative studies between
numerical simulations and experiments for RC members and structure under blast loadings, the collapse analyses of multi-bay and
multi-story RC frames are performed to demonstrate its efficiency and applicability in tracing the progressive collapse of RC frame
structures composed of many structural members.

2. Construction of the M-Φ relation

2.1. Construction of the monotonic envelope curve
The accuracy of the structural response in an RC frames will depend on how to exactly describe the structural behavior of the
composed RC beams and columns. In advance, because an RC beam or column is formed through a sequential consecutive integration
of RC sections, it can be inferred that the accuracy and exactness of the structural response are directly related to the response of the RC
sections. Moreover, the characteristic of an RC section is represented by the M-Φ relation which is uniquely determined for the given
conditions such as the size of a RC section, the placement of reinforcing steels, and the properties of constituent materials. Thus, the M-
Φ relation has been widely used in nonlinear analysis of RC structures [18,19]. Fig. 1 shows the procedure to determine the M-Φ
relation of an RC section. In defining the stress-strain relations of the concrete, the material model proposed by Kent and Park later
modified by Scott [20] is adopted for the compressive region, and the bilinear relation with elastic region and strain softening region is
used for the tensile concrete. Moreover, bilinear relation with elastic region and strain region is considered for the steel as used in many
previous studies [13,21].
Then, the determined M-Φ relation of the beam or column can be simplified to the tri-linear relation based on the initial cracking in
concrete (Point A) and yielding in reinforcing steel (Point B) as shown in Fig. 2. Beyond the yielding point B, the resisting capacity is
assumed to be maintained up to the ultimate strain in the compressive concrete. Furthermore, an increase of the deformation will cause
the development of cracks, which accompanies the bond-slip along the interface between the reinforcing steel and surrounding
concrete [22]. This bond-slip will be enlarged as the crack widens and rapidly increase with the yielding of the reinforcing steel [23].
Because this bond-slip causes inelastic behavior, which cannot be covered by the conventional M-Φ relation, additional consideration
of the bond-slip effect is required to simulate the structural behavior more exactly. In this regard, the bond-slip effect was reflected in
the numerical analysis by modifying the monotonic M-Φ relation [10]. Based on the assumption that a large portion of the bond-slip

Fig. 2. Tri-linear monotonic M-Φ relation of an RC section [15].

3
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

will be concentrated within the plastic hinge length at critical regions such as the vicinity of the mid-span or both clamped ends, where
the yielding of reinforcing steel occurs, the equivalent bending stiffness EIeq to be used in these critical regions was evaluated (see
Fig. 2) and successfully applied to the various numerical analyses of RC structures [21,24]. More details for the derivation of the
equivalent bending stiffness can be found elsewhere [10].
When an RC structure or member is subjected to blast loadings which is accompanied by the strain rate deformation in addition to
the loading histories of the unloading and reloading phases, the RC sections in each member will also experience the strain rate
dependent deformation and the loading phases. This means that not only the hysteretic behavior but also the strain rate effect must be
implemented in the M-Φ relation of an RC section.
In order to consider the effect of the strain rate in the M-Φ relation of an RC section, the dynamic increase factor (DIF) needs to be
newly defined as a relation of the curvature rate [15] rather than the strain rate which was usual in the entire previous studies [25–27].
Details on the procedure to construct DIF equations can be found elsewhere [15] and the simplified tri-linear monotonic envelope
curve considering the strain rate effect can be constructed by multiplying the DIF values for cracking (Point A) and yielding (Point B)
points as shown in Fig. 2.

2.2. Definition of unloading and reloading paths

Upon the constructed monotonic M-Φ relation in Fig. 2, the unloading and reloading paths are implemented to trace the hysteretic
behaviors in structural members subjected to blast loadings [24]. The basic hysteretic M-Φ relation of an RC section under high strain
rate condition was proposed in a previous paper [15] and can be expressed by the following equation:
(1 − p) • φ∗
M ∗ = p • φ∗ + (1)
(1 + φ∗G )1/G

where φ∗ = (φ − φr )/(φ0 − φr ), M∗ = (M − Mr )/(M0 − Mr ), and the parameter p is the strain hardening ratio. In Fig. 3, (φ0 , M0 ) is the
i i
point where two asymptotes with slope EIA1 and EIA2 meet (see Point E), and (φr , Mr ) is the point where the curvature is reversed (see
Points A and C). These two points are updated at every curvature reversal. Moreover, parameter G will determine the shape of the
unloading and reloading curves. As mentioned in the previous paper [15], the G value is not affected by the strain rate, but is affected

Fig. 3. Hysteretic M-Φ relation of an RC section [15].

4
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

by the section properties such as the constituent materials, the section dimension, and the placement of reinforcing steels. Therefore, it
is impossible to express parameter G in a unified equation and the process of constructing the expression of the parameter G for each RC
section is required. Details related to the determination of the parameter G can be found elsewhere [15]. As an example, the parameter
18.33ξ
G has the form of G(ξ) = 20 − 0.119+ξ for the section of beam RC 1-4 adopted in the numerical analysis in this paper.

2.3. Consideration of the continuously changing axial force

As is well known, the axial force influences the curvature, and hence, the M-Φ relation of the RC section also depends on the
magnitude of the axial force. As shown in Fig. 4(a) which represents the P-M interaction diagram of the RC column, the yield moment
My of an RC section increases proportionally to the axial force P up to the balanced axial force Pb , but a further increase of the axial
force decreases the yield moment. Moreover, the failure curvature and the corresponding ductility considerably decrease by the
presence of the axial force. The excessive axial force larger than Pb will cause even a strain softening behavior, as shown in Fig. 4(b),
because the inelastic deformation of concrete rather than the ductile behavior of steel governs the structural responses in these axial
force levels. Accordingly, the axial force effect must be considered and can be implemented with the monotonic M-Φ relation cor­
responding to the applied axial force.
On the other hand, most of the RC frame structures are designed to the wind load and/or the seismic load, and these horizontally
applied lateral loads cause a relatively small axial load P and large moment M in the columns. The applied load combination of (P,M)
will be located below the balance point in the P-M interaction diagram for the designed section. To implement the axial force effect into
the M-Φ relation, basically a linear interpolation is used upon the definition of the five simplified boundary M-Φ relations (see points A,
B, C, D and E in Fig. 4), on the assumption that an increase or decrease in the moment capacities and the corresponding curvatures is
linearly proportional to the applied axial load P (see Fig. 5). Especially, the M-Φ relations for the axial forces of Pb and the axial forces
larger than Pb are simplified as the linear elastic and the linear elastic-linear softening relations, respectively, shown in Fig. 4(b).
Actually, because the axial forces in the RC columns will continuously be changed, whether big or small, this linear interpolation
makes it possible to consider the continuous change in the axial force. This consideration for the change in the axial force should be
considered to exactly trace the progressive collapse of the entire frame structure, and details for the implementation in the formulation
can be found in Fig. 10 describing the solution procedure.
Especially, the sudden collapse of a column causes a large deformation in the connected beams, and this large deformation will
accompany the development of a tension force in the beams while resisting not to be separated from their original positions. This
tension force is eventually carried only by the embedded tensile and compressive reinforcements and has been described as a catenary
force. The development of this catenary force will be possible when the beam has sufficient ductility to sustain the large deformation
[28,29]. As can be inferred from the force equilibrium at the beam-to-column joint, the progress in the catenary action will increase the
tension force (see Fig. 6(a)). Even though an increase of the axial tensile force will remarkably reduce the bending resistance and
ductility in the RC beam, as shown in Fig. 6(b), nevertheless the beams still retain a small resistance even at the large deformation.
Accordingly, consideration of the catenary action may cause a relatively stiff structural response in the large deformation stage, as
shown in Fig. 6(c), because exclusion of the catenary action assumes a zero stiffness in the large deformation stage. Because the
presence and absence of a structural member at a certain time may affect the progressive collapse of a structure accompanying the
excessive deformation, the catenary action needs to be considered to exactly trace the collapse sequence of a structure. To consider the
catenary action in this paper, accordingly, the evaluation of the axial tensile force in a beam was based not on the original configu­
ration but on the deformed configuration, and then, the modification of the M-Φ relation corresponding to the obtained axial tensile
force was followed as in the case of the axial compressive force in Fig. 5.

3. Implementation of the direct shear stress-slip relation

Different from the bending failure which has been developed when an RC structure is subjected to blast loadings accompanying a
relatively low pressure over a long time that cover most of the blast loadings, direct shear failure may be developed in the case of blast
loadings with a high pressure over a very short time. It means that two quite different failures of the flexural and direct shear failures do

Fig. 4. Axial force effect on the M-Φ relation of an RC section.

5
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

Fig. 5. Linear interpolation of the moments and curvatures.

Fig. 6. Catenary action in the pushdown analysis of a beam-to-column joint [28].

not occur at the same time. Especially, the direct shear failure will proceed before the bending failure if the shear strength of a structure
is insufficient to resist the applied blast loading, and its occurrence is concentrated at the support regions in which the shear force has
the maximum value. Because the shear crack occurs in the perpendicular direction to the member axis due to the sudden downward
motion of the structure, the direct shear stress-slip relation needs to be implemented into the formulation to consider the shear slip
effect.
Many related relations have been proposed through experimentation [16,30,31], and of those relations, the shear stress-slip

6
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

relation proposed by Krauthammer and Astarlioglu based on the impact test [16] is adopted in this paper because it fundamentally
considers the strain rate effect and also has been generally used in the blast analyses of RC structures [32,33]. Especially, the rapid
loading on the shear test specimen will cause an inertia resistance effect together with the change in the crack paths as in the axially
loaded specimen, and these lead to an increase of the shear strength of the concrete [16]. In addition, the shear stress-slip relation is
modified to implement the axial force effect on the shear behavior in the case of the axial compression, and the ACI 318 code [34] is
N
referred to for the strength increase factor (1 + 2000×A g
). Fig. 7 shows the shear stress-slip relation adopted in this paper, and the values
of the variables in Fig. 7 can be found in Table 1 [16]. The maximum slip is limited to the diameter of the shear reinforcing bar, as
mentioned in previous studies [14,16], because either shear failure of the bars or rotation beyond the point at which direct shear is
applicable could occur.
Generally, the shear behavior of the RC structures will be limited within the elastic range even in the case of a blast loading because
the structure is basically designed to lead to the ductile failure. However, because an unexpected rapid large blast loading may cause
severe shear damage in the structural members, not only the inelastic large shear definition but also the definition of the corresponding
hysteretic paths is required to trace the shear slip behavior, especially the possible shear failure effect in the progressive collapse of RC
frame structures. Upon the definition of the monotonic dynamic shear stress-slip envelope, accordingly, implementing the hysteretic
behaviors needs to be followed to describe the sequential shear damages in structural members as was in the M-Φ relation. Unlike the
hysteretic M-Φ relation, however, defining the unloading and reloading paths in the direct shear-slip relation is almost impossible
because of no related experimental data. For this reason, very few mathematical models have been introduced [31,35], and the model
introduced by Krauthammer et al. [31], which has been adopted in several numerical studies, is adopted in this study.
Fig. 8 represents the unloading and reloading paths in the direct shear-slip relation. After a concrete matrix is cracked due to the
direct shear while developing a permanent relative slippage (points ① and ⑤ in Fig. 8), this developed slippage will not be recovered
even after complete unloading (from ① to ①‘, ③ to ③’ and ⑤ to ⑤’, respectively). The unloading stiffness may be slightly smaller than
the initial elastic shear stiffness, but it is assigned to be equal to the initial elastic shear stiffness [35]. Then, the other face of the
concrete matrix starts to resist the reversed shear forces, but the already cracked surface will not reserve the same shear resistance with
that in the uncracked concrete matrix. Especially, the corresponding shear stiffness will be inversely proportional to the magnitude of
the developed residual slippage (from ① to ②, ③ to ④ and ⑤ to ⑥, respectively). This load reversal will continue until a symmetric
point on the other side in the monotonic envelope is reached if the symmetric cyclic loading is applied. Moreover, the hysteretic curve
in Fig. 8 ignores the friction slip resistance which may develop along the aggregate surface on the crack face and the dowel action in the
reinforcement [36], upon the assumption that it will be small enough.

4. Solution procedure
The rest of the procedures from the composition of the element stiffness matrix to the time integration of the dynamic equilibrium
equation are the same as those used in the nonlinear analysis of the RC frame structures. Different from the use of the layer model [21,
37] generally adopted in constructing an element stiffness matrix to take into account the different material properties of reinforcing
steel and concrete across a section, the M-Φ relation of an RC section [19,38], which has also been broadly used in the nonlinear
analyses of RC structures, is used in this paper. Because the application of the layer model to the nonlinear analysis of entire RC
structures consisting of many structural members requires a remarkable increase in the numerical integration and prediction of
material states induced from the evaluation through the entire layers even in an element, the adoption of this model to large structures
with many degrees of freedom has some limitations and thus, the essential reason for the use of the M-Φ relation rather than the layer
model in this paper. Based on the given section properties, the M-Φ relation is uniquely determined by a section analysis, and its slope
represents the bending stiffness EI. It means that the bending stiffness of an RC member can be directly obtained from the M-Φ relation,
while markedly reducing the calculation procedures.

Fig. 7. The shear stress-slip relation.

7
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

Table 1
The corresponding values for the parameters in Fig. 7

Parameter Corresponding value

√̅̅̅̅̅
τm
1.5(11 fc + 0.6ρvt fy )
′

τl ρvt fy ≤ 0.008fy
Kc 180fc
′

Ku 4( − 2000 − 0.75fc )
′

Δ2 τm /Kc
Δ3 Δ2 + 0.012
Δ4 Δ3 + (τl − τm )/Ku
Δmax Bar diameter

Fig. 8. Hysteretic behavior in the direct shear-slip relation [31].

In addition, the direct shear behavior needs to be considered because a perpendicular crack can be developed by the direct shear
near the end of the structural member before forming bending cracks at critical regions when an RC structure is subjected to a blast
loading with high pressure in a very short time. Usually, direct shear failure occurs suddenly in the perpendicular direction to the
member axis if the shear strength of structure is insufficient to resist the applied blast loading and will develop geometric discontinuity.
The reason is that the vertical shear cracks induced from the direct shear failure will reduce the shear capacity and increase the relative
slip along the shear crack interface. Moreover, the occurrence of shear cracks may cause a change in the entire structural response.
Accordingly, the shear failure must be taken into account to exactly evaluate the structural failure, and it has been considered through
the implementation of the direct shear stress-slip relation into the formulation [17]. The non-dimensional spring element [39] is placed
at the beam-column joint shown in Fig. 9. The vertical stiffness of the spring element is determined from the defined direct shear
stress-slip relation in Fig. 7 but the other two degrees of freedom at the node of the spring element (corresponding to the axial
deformation and the rotational deformation) are restrained. The implementation of this spring element is similar to the consideration
of the bond-slip behavior by using the double node in the conventional nonlinear analysis of RC structures [35,40].
In the blast analysis of RC structures, the finite element formulation is based on the Timoshenko beam theory [41], which considers

Fig. 9. Implementation of shear spring elements at the beam-column joint.

8
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

the shear deformation, but the effect of the shear deformation is not taken into consideration in describing the nonlinear behavior of RC
structures. The reason is that the normal bending stresses reach a maximum at the extreme fibers on an RC section, where the
transverse shear stresses are at their lowest value. It means that the interaction between the two stress components is so small that it
can be ignored. As stated earlier, moreover, the bending behavior is described by using M-Φ relation and the direct shear behavior by
using a spring element with the shear stress-slip relation, respectively. Especially, to take into account the variation in the axial force,
which is prominent during the progressive collapse of RC structures, the modified M-Φ relation described in Fig. 5 is updated at the
next time step. Besides, the relative difference in the lateral deformations at the double node is used to define the shear slip and, in
advance, to evaluate the corresponding shear stiffness. In solving the dynamic equilibrium equation, the lumped mass matrix and 5%
Rayleigh damping are used, and the non-iterative central difference method is applied [42]. More details related to the entire solution
procedures can be found in Fig. 10.

5. Experimental verification
To test the validity of the proposed analytical model, two RC beams and one RC column in a member level and a four-bay three-
story RC frame in a structure level were investigated. A RC beam specimen of RC1-4 performed by Jacques et al. [43], one RC column
specimen of SCC-0%-38 performed by Burrell et al. [44], and another beam specimen of S12-3 tested by Liu et al. [45] were adopted,
and a four-bay three-story RC frame tested by Yi et al. [46] was considered. Especially, because the RC frame was designed to observe

Fig. 10. Solution procedure for the blast analysis.

9
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

the force-deformation response at the collapsed column location during the gradual collapse of the RC frame structure under the
gravity load of the upper floor frame, the applicability of the proposed analytical model in the structure level and the influence of
catenary effect on the structural behavior can be examined. Tables 2 and 3 show the material properties and details of the blast
loadings considered in the experiment for each specimen, respectively, and more details about experiments can be found elsewhere
[43–46].
As shown in Fig. 11, a RC beam specimen of RC1-4 and column specimen of SCC-0%-38 were subjected to a uniformly distributed
blast pressure which was controlled through the shock tube, and in the case of column specimen in Fig. 11(b), a hydraulic jack applied
a constant axial force P to the specimen. Because the plastic deformation is expected to be concentrated at the mid-span and at the end
of the beam and the shear failure at the end of the beam as well, a separate consideration of these regions is required in the finite
element modeling to simulate a more exact structural response [47]. The specimens were modeled along the entire span with an
element which has a span length less than the plastic hinge length. Evaluation of the plastic hinge length was based on a simple
equation proposed by Sawyer [48] that gave the following values: Lp = 236 mm for the RC1-4 specimen, Lp = 60 mm at the fixed end
and Lp = 79 mm at the center for the SCC-0%-38 specimen, respectively. Accordingly, the RC1-4 specimen was divided into 10 el­
ements with the same length of L = 198 mm, and the SCC-0%-38 specimen was discretized with 4 elements with L = 60 mm at the
mid-span and the clamped ends and 12 elements with L = 125 mm for the remaining region. In advance, the shear spring elements in
Fig. 8 were implemented at both ends of the specimens.
Fig. 12 presents a comparison between the analytical results and the measured deflections at the mid-span in both member
specimens, and shows very good agreement between both displacement histories. The specimen RC1-4 had an initial displacement by
the previous shock wave, and numerical analysis was performed considering this initial deformation. The numerical analysis, the
hysteretic behavior in an RC section and the bond-slip effect were basically included, of course. These good agreements represent that
the introduced numerical model can effectively be used in tracing the deformation history of the RC members subjected to blast
loadings. Additional comparisons of and/or differences in the numerical results according to the consideration of various influencing
factors can be found in a previous paper [15].
The geometric configuration and experimental setup of beam specimen S12-3 are shown in Fig. 13. Unlike the previous specimens,
the specimen was directly exposed to a TNT explosion. For a numerical analysis using the proposed model, the specimen was dis­
cretized to 40 elements with element length that is 50 mm shorter than the plastic hinge length. Furthermore, blast pressure-time
histories for each element were evaluated by CONWEP [49], which is also implemented in LS-DYNA [50]. Additional numerical
analyses with LS-DYNA were conducted to verify the accuracy and effectiveness of the proposed model. The Karagozian & Case
Concrete model (MAT_072R3) was used for concrete and the failure was defined when the maximum principal strain exceeds 0.1 using
the MAT_ADD_EROSION option. An elasto-plastic model (MAT_003) was used for reinforcing steel. Concrete and reinforcing steel were
constrained by the BEAM_IN_SOLID option with consideration of bond-slip using an empirical relation by Murcia-Delso [51], and blast
loadings were calculated by CONWEP(LOAD_BLAST_ ENHANCED). In order to determine the proper mesh size, additional analyses
with three different FE mesh sizes, Mesh A (10 mm*10 mm*10 mm), Mesh B (20 mm*20 mm*20 mm), and Mesh C (73 mm*100
mm*100 mm), were performed.
Fig. 14(a) presents a comparison among experimental results, numerical results yielded by the proposed model, and numerical
results using LS-DYNA. Both numerical results show very good agreement with the displacement history in the experiment. Fig. 14(b)
shows the obtained numerical results using LS-DYNA according to the FE mesh size, and presents converged accurate predictions in
Mesh A and Mesh B. However, Mesh B requires more than 18,000 solid elements to obtain converged results, which is a significant
amount compared to the 40 elements used in the proposed model. As is well known, numerical analyses using solid elements usually
produce highly mesh-dependent results because of the stiff characteristic of the elements []. In contrast, the proposed model is
basically based on the beam element, which can describe the structural behavior of the beam with only a small number of elements and
the formulation based on the moment-curvature relation of a RC section leads to an additional reduction in the solution steps. The
difference in the number of elements will be enlarged in structures composed of many beams and columns and the proposed model is
expected to be more efficient in the analysis of these structures.
Next, the numerical results were compared at the structure level. Fig. 15 shows the geometric configuration and experimental setup
for an RC frame. As shown in this figure, the gravity load applied to the middle column from the upper levels of the structure was
simulated using a servo-controlled hydraulic actuator, and the middle column on the first story was constructed by stacking two
mechanical jacks and a load cell. A constant vertical load of 109 kN was applied on the top of the middle column by a servo-controlled
hydraulic actuator, and the same load was transmitted to the lower jack through the middle column. Then, the step-by-step unloading
was initiated by lowering the mechanical jacks to artificially simulate the gradual failure of the lower story column. Because this

Table 2
Material properties and details of each specimen.

Specimens Member Depth Width Longitudinal Reinforcement Lateral Reinforcement fc

′
fy
type (mm) (mm) (mm) (mm) (MPa) (MPa)

RC1-4 (Jacques et al.) Beam 300 150 3–10 M/2–10 M/3-10 M Φ6.3@150 38.5 445
SCC-0%-38 (Burrell Column 152 152 2–10 M/2-10 M Φ6.3@38 51.8 483
et al.)
S12-3 (Liu et al.) Beam 300 220 2D12/2D12 D8@200 40 580
Model frame (Yi et al.) Beam 200 200 4D12 D6@150 25 416
Column 200 200 2D12/2D12 D6@150 25 416

10
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

Table 3
Loading information.

Specimens Loading type Pr (kPa) Ir (kPa ms) td (ms) Axial force (kN)

RC1-4 (Jacques et al.) Blast 59.8 366.2 14.5 –

SCC-0%-38 (Burrell et al.) Blast 40.9 401.4 24.4 294
S12-3 (Liu et al.) Blast (3 kg TNT, 0.65 m) 3.81x104 2443.1 0.39 –
Model frame (Yi et al.) Static loading – – – 109

Fig. 11. Geometric configuration for specimens RC1-4 and SCC-0%-38.

Fig. 12. Displacement history of specimens RC1-4 and SCC-0%-38 [15].

Fig. 13. Geometric configuration for specimen S12-3.

11
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

Fig. 14. Displacement history of specimen S12-3.

Fig. 15. A four-bay three-story RC frame.

gradual unloading will not accompany the strain rate deformation, attention is given to the deformation history of the middle column
whose deformation had to be affected directly and indirectly by the nonlinear behavior of the adjacent members.
In modeling the structure with a finite number of elements, the element length is determined by considering the plastic hinge length
determined by a Sawyer’s simple equation [48] which is generally used in the nonlinear analysis of RC structures [10]. Because the
evaluated plastic hinge lengths were Lp = 108 mm for the columns on the first floor, Lp = 90 mm for the columns on the other floors

Fig. 16. Load-displacement relation at the lower middle column.

12
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

and Lp = 150 mm for the beams, respectively, the two elements at both ends of the beam were modeled with an element length of L =
150 mm, and an additional eight elements were used to discretize the rest of each beam. In advance, the same rule was also applied in
discretizing the columns, and the shear spring elements were added at both ends of all members to describe the shear failure.
Fig. 16 presents a comparison of the obtained numerical result with the experimental data for the downward displacement at the
lower middle column. As shown in Fig. 16, regardless of the shear slip effect, the numerical results show good agreement with the
experimental data from the initial elastic behavior up to the yielding of the reinforcing steel in the adjacent beams, even at the post-
yielding behavior. These results also show that the shear slip effect does not affect the structural behavior in this example case, as was
expected, because this structure was subjected to gradual unloading which does not have the accompanying sudden shear failure.
Moreover, the sufficiently large shear span ratio of the beams in this structure does not enable the occurrence of shear failure.
Because the numerical results depend on the number of finite elements, the convergence test is usually performed before the
numerical analyses. As shown in Fig. 17, the use of more than ten elements, together with the consideration of bond-slip effect in the
two elements at both ends of the beam, is required to obtain the converged numerical results similar to the experimental data, whereas
no consideration of the bond-slip effect will give a slight conservative numerical result. Consequently, the obtained results describe the
fact that the application of the numerical approach introduced in this paper can be extended to the RC structure level.
To perform additional verification of the proposed model in a structural level, blast analyses of a single story two bay simple frame,
as in Fig. 18, were conducted using the proposed model and LS-DYNA. Table 4 shows the assumed material properties of the con­
stituent materials, the size of sections and information for the blast loading. In both numerical analyses, the pressure-time history
induced by TNT is evaluated by the air-blast pressure model based on the CONWEP model [52]. Each member is divided into 20
elements considering the plastic hinge length in the proposed model case. In the numerical analysis using LS-DYNA, the elasto-plastic
model is used for the reinforcing steel, and the Karagozian & Case Concrete (K&C) model, which is one of the generally used plastic
based concrete models, is used for concrete, and the structure is discretized with 20 mm cubed solid elements, which was determined
through a convergence test in the analysis of specimen S12-3. All nodes on the floor are set to a fixed condition.
Fig. 19 shows the displacement time history at the mid-span of column C1. The responses obtained from the two numerical analyses
show very good agreement up to the maximum deformation despite the differences in element type and concrete model used. Even
though there is a slight difference in the post-peak behavior, the hysteretic behavior predicted by the proposed model is still similar to
the residual response obtained by LS-DYNA. Fig. 20 also visualizes the structural behaviors obtained by the proposed model and LS-
DYNA at two representative times (marked in Fig. 19). As shown in Fig. 20(a), the first plastic hinge occurs at the fixed end of column
C1, which is the closest column to the explosive. Sequentially, the plastic deformation proceeds at the mid-span of column C1 and then
at the end of beam B1. Overall structural responses including the order and location of the plastic hinges are very similar to the results
obtained by LS-DYNA, and this comparison shows the reliability of the proposed model at a structural level.

6. Progressive collapse of RC frames

To simulate the progressive collapse of RC frames subjected to blast loadings, a five story five bay RC frame, shown in Fig. 21, was
considered. Two types of TNT explosions, shown in Fig. 21, (Blast 1 and Blast 2) were considered, and Table 5 shows the related
information for the members and loadings. The two TNT explosions are identical except for the location of the explosions. Especially,
the section dimensions of the beams and columns were determined by referring to the pre-existing drawings for RC buildings, and a
concrete cover thickness of 5 cm was assumed. The blast pressure-time history induced by TNT blast was calculated based on the
CONWEP model [52,53], and Fig. 22 shows the obtained pressure-time history to be applied to the mid-span of the interior column C3,
as an example. 20 elements with the consideration of plastic hinge length is used to discretize each structural member.
Fig. 23 shows the progressive collapse of the structure subjected to Blast 1 in Fig. 21 when all four influencing factors of the bond-
slip effect, the axial force effect, the catenary action and the shear-slip effect mentioned in this paper are considered, where the
deformed configuration of the structure is represented with a tenfold amplification. A structural member was removed when the
structural response presents a larger value than any one of three limiting values of the ultimate curvature in the moment-curvature

Fig. 17. Convergence of the numerical results in an RC frame.

13
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

Fig. 18. Configuration of an RC frame under TNT blast.

Table 4
Information on the members and blast loading for a one story two bay RC frame.

Member type Section dimension (width*height) Reinforcement Stirrup fc (MPa)

′
fy (MPa)

Beam 220 mm*300 mm 2D12/2D12 D8@80 40 458

Columns 300 mm*300 mm 2D12/2D12 D8@80 40 458

TNT weight Position

10 kg − 1.5 m*1.5 m

Fig. 19. Displacement time history at the mid-span of column C1.

relation, the ultimate shear slip in the shear stress-slip relation, and the resisting capacity for axial force. As shown in this figure, the
structural collapse is initiated by the bending failure of the nearest beam from the TNT explosion, and then the sequential failure is
propagated through the compressive failure of columns because of the small remaining bending resistance due to the application of
large axial force. The outermost columns are finally collapsed by the bending failure due to the excessive deformation of the structure,
as shown in Fig. 23(d), while developing a pancake collapse.
Moreover, Fig. 24 visualizes the deformed configurations of the RC frame at a representative time of t = 70.0 ms according to the
exclusion of one of the four influencing factors and shows the following: (1) the bond-slip effect does not significantly affect the
progressive collapse of RC structures which is accompanied by a large deformation due to the yielding of the reinforcements in the RC
members around the already lost members because the bond-slip usually affects the structural response at the cracked state before the
yielding of the reinforcement (see Fig. 24(a)); (2) the catenary action also does not dominantly affect the progressive collapse of RC
structures because a slight decrease of the compressive force in the columns on the first floor due to a limited increase of the tensile
force in the adjacent beams does not change the collapse behavior in the structural level. Moreover, the suddenness of the collapse and
the development of very small lateral displacement make it difficult to enlarge the catenary action effect in the collapse behavior (see
Fig. 24(b)); (3) the direct shear may affect the very early failure of the structural members but does not dominantly affect the posterior
collapse behavior as well because all the structural members in this example structure have relatively large span-to-depth ratios which
induce the bending failure in the remaining structural members even when considering the shear slip effect. Nevertheless, it is also true
that the difference in the progressive collapse of the structure depends on the relative resistances of the structural member to the
bending and shear (see Fig. 24(c)); (4) but if the axial collapse of the columns is not allowed due to the adoption of the concept of the
strong column-weak beam design practice, then the structural collapse may be governed by the simultaneous failure of the connecting
beams while producing a completely different collapse behavior (see Fig. 24(d)).
Especially, the difference in the collapse behavior in Fig. 23(d) and 24(d) based on the axial collapse of the columns can also be

14
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

Fig. 20. Comparison of structural behaviors by the proposed model and LS-DYNA.

15
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

Fig. 21. Configuration of a five story five bay RC frame.

Table 5
Related information on the five story five bay RC frame.

Member type Section dimension (width*height) Reinforcement Stirrup fc (MPa)

′
fy (MPa)

Beam 300 mm*800 mm 8D19/8D19 D8@200 28 400

Column 400 mm*400 mm 3D22/3D22 D8@200 28 400

Loading Type Information

Static loading (1.4D+1.7L) Dead load 3.0 kN/m2 (12 cm thick RC slab with 2% reinforcement)
Live load 2.5 kN/m2 (Ordinary office standard in KBC 2019)
Blast loading TNT 2000 lbs (very high level threat in ASCE)

Fig. 22. Pressure-time history at the mid-span of the members adjacent to the explosions.

examined with the variation in the member forces in the adjacent members. As shown in Fig. 25 which represents the variation in the
member forces in the adjacent column and beam due to the TNT explosion, considering the axial collapse of the columns produces a
sharp decrease of the end moment in beam B2 simultaneously with the collapse of column C2, and this decrease of the bending moment
in the beams enables the sustenance of the beams to reach structural collapse at t = 70.0 ms (see Fig. 23(d)). On the other hand,
ignoring the axial collapse of the columns does not enable the failure of the columns to reach the bending failure of the beams due to an
excessive increase of the unequal vertical displacement between the C2 and C3 column lines, and it may cause a different collapse
behavior from reality (see Fig. 24(d)).
Asymmetric collapse of the same structure was also analyzed with the application of the Blast 2 loading in Fig. 21. As shown in
Fig. 26, the structural collapse is initiated by the bending failure of the nearest beam from the TNT explosion, and the collapse of
adjacent columns follows, as with the previous symmetric Blast 1 loading. The posterior collapse behavior is also very similar to that of
the symmetric loading case, and no significant differences are found between both the loading cases for the example structure
considered. However, because the structural collapse is dominantly affected by various factors such as the relative stiffness of the

16
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

Fig. 23. Progressive collapse of the RC frame for the symmetric Blast 1 loading.

Fig. 24. Structural collapse of the RC frame for the Blast 1 loading.

17
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

Fig. 25. Variation in the member force with and without the axial failure.

beams and columns, composition of the members in the structural system and the place of the blast loading, the possibility to reach a
different conclusion still remains. As shown in Fig. 27, which represents the resisting capacity of structural members against flexural
and direct shear failures using the pressure-impulse (P–I) diagram. The P–I diagrams represent four divided regions for beams but only
flexural failure for columns. Therefore, if a beam is subjected to blast loading that has blast pressure and impulse corresponding to
Region A in Fig. 27, then direct shear failure will be developed, and absence of consideration of the direct shear effect may cause

18
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

Fig. 26. Progressive collapse of the RC frame for asymmetric Blast2 loading.

Fig. 27. P–I diagrams for structural members and applied blast loadings.

significant errors in an evaluation of the structural behavior.

From the progressive collapse analyses of an RC frame structure, it can be concluded that (1) the influencing factors of the bond-slip
effect, the catenary action and the shear-slip effect may dominantly affect the collapse of an RC beam or column but do not affect the
progressive collapse of an RC structure composed of many beams and columns, whereas, (2) because the axial force change and, by
extension, the compressive failure of column will change the collapse mode of the structure, the axial force effect must be taken into
consideration.

7. Conclusion
An improved analytical model based on the M-Φ relation was proposed for the progressive collapse analysis of RC frame under blast
loadings. To take the strain rate effect into account, the dynamic increase factor (DIF) was newly defined in terms of curvature rate.
Upon the determination of the monotonic envelope of the M-Φ relation, the relation to describe the hysteretic response of the RC
section was defined with reference to the hysteretic stress-strain curve of reinforcing steel. For the precise evaluation of the progressive
collapse in RC frame structures, many influencing factors were examined such as the effect of bond-slip, the axial force and the
catenary action that are known to have a major impact on the collapse behavior in a member level. In addition, this paper demonstrates
that the inevitable change in axial force and the direct shear failure can also be considered even in the analytical model based on the M-
Φ relation. After verifying the exactness of the introduced model through a comparative study between numerical and experimental
results for RC structural members, the applicability and efficiency of the proposed numerical model even in a structural level were also

19
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

demonstrated by comparing the numerical results with those obtained from LS-DYNA. Finally, additional analysis was performed on
the full-scale RC frame to describe the progressive collapse behavior of the entire structure. The numerical results obtained in this
paper give the following conclusions: (1) the proposed analytical model based on the M-Φ relation can effectively be used in the
progressive collapse analysis of RC frame structures subjected to blast loadings while significantly reducing the calculation steps; (2)
unlike in RC beams and columns, the influencing factors of the bond-slip effect, the catenary action and the shear-slip effect do not
seem to dominantly affect the progressive collapse of RC structures composed of many beams and columns; whereas, (3) the change in
the axial force according to the redistribution of the member forces must be implemented because it affects the collapse sequence and
final collapse configuration of the RC frame structure, and (4) the proposed analytical model enables the simulation of the progressive
collapse of the entire RC frame structures by using a limited number of beam elements. In advance, the use of this model does not
require in-depth prior knowledge for concrete models, which dominantly affect the progressive collapse of structures due to blast
loading.

Author statement
SeokJun Ju: Methodology, Software, Validation, Formal analysis, Visualization, Writing-Original Draft. Hyo-Gyoung Kwak:
Conceptualization, Supervision, Writing-Review & Editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Data availability

No data was used for the research described in the article.

Acknowledgement
This research was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT)
(No. 2023R1A2C2005101).

References
[1] T. Chrysanidis, D. Mousama, E. Tzatzo, N. Alamanis, D. Zachos, Study of the effect of a seismic zone to the construction cost of a five-story reinforced concrete
building, Sustain. Times 14 (2022), https://doi.org/10.3390/su141610076.
[2] Y. Shi, Y. Hu, L. Chen, Z.-X. Li, H. Xiang, Experimental investigation into the close-in blast performance of RC columns with axial loading, Eng. Struct. 268
(2022), 114688, https://doi.org/10.1016/J.ENGSTRUCT.2022.114688.
[3] Y. Liu, Yan J. bo, Huang F. lei, Behavior of reinforced concrete beams and columns subjected to blast loading, Def Technol 14 (2018) 550–559, https://doi.org/
10.1016/j.dt.2018.07.026.
[4] C. Zhao, Q. Wang, X. Lu, X. Huang, Y.L. Mo, Blast resistance of small-scale RCS in experimental test and numerical analysis, Eng. Struct. 199 (2019), https://doi.
org/10.1016/J.ENGSTRUCT.2019.109610.
[5] C.A. Ross, J.W. Tedesco, S.T. Kuennen, Effects of strain rate on concrete strength, ACI Mater. J. (1995), https://doi.org/10.1111/j.1365-2389.2010.01314.x.
[6] B. Yan, F. Liu, D.Y. Song, Z.G. Jiang, Numerical study on damage mechanism of RC beams under close-in blast loading, Eng. Fail. Anal. 51 (2015) 9–19, https://
doi.org/10.1016/J.ENGFAILANAL.2015.02.007.
[7] M.J. Lee, H.G. Kwak, Numerical simulations of blast responses for SFRC slabs using an orthotropic model, Eng. Struct. 238 (2021), 112150, https://doi.org/
10.1016/J.ENGSTRUCT.2021.112150.
[8] H. Helmy, H. Salem, S. Mourad, Progressive collapse assessment of framed reinforced concrete structures according to UFC guidelines for alternative path
method, Eng. Struct. 42 (2012) 127–141, https://doi.org/10.1016/J.ENGSTRUCT.2012.03.058.
[9] Y. Shi Yanchao, Z.X. Li, H. Hao, A new method for progressive collapse analysis of RC frames under blast loading, Eng. Struct. 32 (2010), https://doi.org/
10.1016/j.engstruct.2010.02.017.
[10] H.G. Kwak, S.P. Kim, Simplified monotonic moment-curvature relation considering fixed-end rotation and axial force effect, Eng. Struct. 32 (2010) 69–79,
https://doi.org/10.1016/j.engstruct.2009.08.017.
[11] H.G. Kwak, J.K. Kim, Implementation of bond-slip effect in analyses of RC frames under cyclic loads using layered section method, Eng. Struct. 28 (2006)
1715–1727, https://doi.org/10.1016/j.engstruct.2006.03.003.
[12] B.R. Ellingwood, Earthquake risk assessment of building structures, Reliab. Eng. Syst. Saf. 74 (2001) 251–262, https://doi.org/10.1016/S0951-8320(01)00105-
3.
[13] H.G. Kwak, S.P. Kim, J.E. Kim, Nonlinear dynamic analysis of RC frames using cyclic moment-curvature relation, Struct. Eng. Mech. 17 (2004) 357–378,
https://doi.org/10.12989/sem.2004.17.3_4.357.
[14] R.L. Park, R. Park, T. Paulay, Reinforced Concrete Structures, John Wiley & Sons., 1974.
[15] S.J. Ju, H.G. Kwak, Moment-curvature approach for blast analysis of RC frames with multitudinous members, J. Build. Eng. 42 (2021), 102463, https://doi.org/
10.1016/j.jobe.2021.102463.
[16] T. Krauthammer, S. Astarlioglu, Direct shear resistance models for simulating buried RC roof slabs under airblast-induced ground shock, Eng. Struct. 140 (2017)
308–316, https://doi.org/10.1016/j.engstruct.2017.02.056.
[17] S.J. Ju, H.G. Kwak, A FE model to evaluate the resisting capacity of RC beams and columns under blast loading based on P-I diagram, Int. J. Impact Eng. 161
(2022), 104113, https://doi.org/10.1016/J.IJIMPENG.2021.104113.
[18] G.K. Park, H.G. Kwak, F.C. Filippou, Blast analysis of RC beams based on moment-curvature relationship considering fixed-end rotation, J. Struct. Eng. 143
(2017) 1–11, https://doi.org/10.1061/(ASCE)ST.1943-541X.0001837.
[19] H.Y. Grisaro, A.N. Dancygier, M. Asce, Dynamic response of RC elements subjected to combined loading of blast and fragments, J. Struct. Eng. (2021) 147,
https://doi.org/10.1061/(ASCE)ST.1943-541X.0002866.
[20] B.D. Scott, M.J.N. Priestley Rp, Stress-strain behavior of concrete confined by overlapping hoops at low and high strain rates, J. Proc. 79 (1982), https://doi.org/
10.14359/10875.

20
S. Ju and H.-G. Kwak Journal of Building Engineering 71 (2023) 106474

[21] M.J. Lee, H.G. Kwak, Blast and impact analyses of RC beams considering bond-slip effect and loading history of constituent materials, Int. J. Concr. Struct.
Mater. 12 (2018), https://doi.org/10.1186/s40069-018-0244-9.
[22] T. Chrysanidis, Experimental and numerical research on cracking characteristics of medium-reinforced prisms under variable uniaxial degrees of elongation,
Eng. Fail. Anal. 145 (2023), 107014, https://doi.org/10.1016/j.engfailanal.2022.107014.
[23] W. Gerstle, A.R. Ingraffea, P. Gergely, Fracture mechanics of bond in reinforced concrete, Proc. U.S. Natl. Congr. Appl. Mech. (1982), https://doi.org/10.1061/
(asce)0733-9445(1984)110:4(871).
[24] G.K. Park, H.G. Kwak, F.C. Filippou, Hysteretic moment-curvature relations for the analysis of RC flexural members subjected to blast loading, Comput. Concr.
27 (2021) 537–548, https://doi.org/10.12989/CAC.2021.27.6.537.
[25] L.J. Malvar, J.E. Crawford, Dynamic Increase Factors for Concrete, 1998.
[26] L.J. Malvar, J.E. Crawford, Dynamic increase factors for steel reinforcing bars, Proc Twenty-Eighth DoD Explos Saf Semin (1998).
[27] M. Saatcioglu, A. Lloyd, E. Jacques, A. Braimah, G. Doudak, Focused Research for Development of a CSA Standard on Design and Assessment of Buildings
Subjected to Blast Loads. Ottawa, Canada, 2011.
[28] H. Choi, J. Kim, Progressive collapse-resisting capacity of RC beam-column sub-assemblage, Mag. Concr. Res. 63 (2011) 297–310, https://doi.org/10.1680/
macr.9.00170.
[29] L. Guo, S. Gao, F. Fu, Y. Wang, Experimental study and numerical analysis of progressive collapse resistance of composite frames, J. Constr. Steel Res. 89 (2013)
236–251, https://doi.org/10.1016/J.JCSR.2013.07.006.
[30] R. French, E. Maher, M. Smith, M. Stone, J. Kim, T. Krauthammer, Direct shear behavior in concrete materials, Int. J. Impact Eng. 108 (2017) 89–100, https://
doi.org/10.1016/j.ijimpeng.2017.03.027.
[31] T. Krauthammer, N. Bazeos, T.J. Holmquist, Modified SDOF analysis of RC box-type structures, J. Struct. Eng. (1986), https://doi.org/10.1007/s12603-016-
0817-8.
[32] G.K. Park, H.G. Kwak, F.C. Filippou, Evaluation of nonlinear behavior and resisting capacity of reinforced concrete columns subjected to blast loads, Eng. Fail.
Anal. 93 (2018) 268–288, https://doi.org/10.1016/j.engfailanal.2018.07.024.
[33] R. Yu, D. Zhang, L. Chen, H. Yan, Non-dimensional pressure–impulse diagrams for blast-loaded reinforced concrete beam columns referred to different failure
modes, Adv. Struct. Eng. 21 (2018), https://doi.org/10.1177/1369433218768085.
[34] ACI Committee 318, Building Code Requirements for Structural Concrete and Commentary, vol. 28, 2008, pp. 0–15.
[35] A. D’Ambrisi, F.C. Filippou, Modeling of cyclic shear behavior in RC members, J. Struct. Eng. 125 (1999) 1143–1150, https://doi.org/10.1061/(asce)0733-9445
(1999)125:10(1143).
[36] B. Martín-Pérez, S.J. Pantazopoulou, Effect of bond, aggregate interlock and dowel action on the shear strength degradation of reinforced concrete, Eng. Struct.
23 (2001) 214–227, https://doi.org/10.1016/S0141-0296(00)00004-3.
[37] Y. Fan, L. Chen, Q. Fang, T. Wang, Blast resistance of externally prestressed RC Beam: a theoretical approach, Eng. Struct. 179 (2019) 211–224, https://doi.org/
10.1016/J.ENGSTRUCT.2018.10.079.
[38] J. Dragos, C. Wu, Interaction between direct shear and flexural responses for blast loaded one-way reinforced concrete slabs using a finite element model, Eng.
Struct. 72 (2014) 193–202, https://doi.org/10.1016/j.engstruct.2014.04.043.
[39] C. Wu, H. Sheikh, A finite element modelling to investigate the mitigation of blast effects on reinforced concrete panel using foam cladding, Int. J. Impact Eng.
55 (2013) 24–33, https://doi.org/10.1016/j.ijimpeng.2012.11.006.
[40] M.R. Salari, E. Spacone, Finite element formulations of one-dimensional elements with bond-slip, Eng. Struct. 23 (2001) 815–826, https://doi.org/10.1016/
S0141-0296(00)00094-8.
[41] C.M. Wang, Timoshenko beam-bending solutions in terms of euler-Bernoulli solutions, J. Eng. Mech. 121 (1995), https://doi.org/10.1061/(asce)0733-9399
(1995)121:6(763).
[42] A.K. Chopra, Dynamics of Sturctures, Prentice Hall, 1995.
[43] E. Jacques, A. Lloyd, P. Imbeau, D. Palermo, J. Quek, GFRP-retrofitted reinforced concrete columns subjected to simulated blast loading, J. Struct. Eng. (2015),
https://doi.org/10.1061/(ASCE)ST.1943-541X.0001251.
[44] R.P. Burrell, H. Aoude, M. Saatcioglu, Response of SFRC columns under blast loads, J. Struct. Eng. 141 (2015), https://doi.org/10.1061/(ASCE)ST.1943-
541X.0001186.
[45] S. Liu, Y. Zhou, J. Zhou, B. Zhang, F. Jin, Q. Zheng, et al., Blast responses of concrete beams reinforced with GFRP bars: experimental research and equivalent
static analysis, Compos. Struct. 226 (2019), 111271, https://doi.org/10.1016/j.compstruct.2019.111271.
[46] W.J. Yi, Q.F. He, Y. Xiao, S.K. Kunnath, Experimental study on progressive collapse-resistant behavior of reinforced concrete frame structures, ACI Struct. J. 105
(2008), https://doi.org/10.14359/19857.
[47] S. Bae, O. Bayrak, Plastic hinge length of reinforced concrete columns, ACI Struct. J. (2008), https://doi.org/10.14359/19788.
[48] A. Herbert, J. Sawyer, Design of concrete frames for two failure stages, Spec. Publ. 12 (1965), https://doi.org/10.14359/16726.
[49] D. Hyde, ConWep, Conventional Weapons Effects Program, US Army Eng Waterw Exp Station USA, 1991.
[50] LSTC, LS-DYNA R8.0 Keyword User’s Manual - Volume I, 2015.
[51] J. Murcia-Delso, P. Benson Shing, Bond-slip model for detailed finite-element analysis of reinforced concrete structures, J. Struct. Eng. 141 (2015) 1–10, https://
doi.org/10.1061/(asce)st.1943-541x.0001070.
[52] G. Randers-pehrson, K.A. Bannister, Airblast Loading Model for DYNA2D and DYNA3D, 1997.
[53] G. Le Blanc, M. Adoum, V. Lapoujade, External Blast Load on Structures – Empirical Approach, 5th Eur LS-DYNA Users Conf, 2005.

21