Research Article

Impacts of residual stress and shear deformation on 2D steel frames

using fiber plastic hinge element: nonlinear behavior and strength
Phu‑Cuong Nguyen1 · T. D. Tran1

Received: 5 October 2020 / Accepted: 5 May 2021

© The Author(s) 2021  OPEN

Abstract
In this study, impacts of residual stress and shear deformation are investigated on 2D steel frames using a new fiber plastic
hinge method. Geometry and material nonlinearities, residual stress, shear deformation, imperfections are considered
simultaneously in the nonlinear analysis. The proposed method is efficient in computational efforts since the one-element
modeling is used for the nonlinear analysis by employing stability functions for capturing the P-small delta phenomenon.
The P-large delta phenomenon is considered using the geometric stiffness matrix. Plastic hinges are assumed to be formed
at two ends of members. Cross-sections at two ends of members are divided into many fibers. The ECCS residual stress
pattern is assigned directly through fibers as initial stress conditions. A finite element program is coded using the Fortran
programming language for predicting the nonlinear behavior and ultimate strength of planar steel frames. The behavior
and load-carrying capacity of steel frames are predicted precisely and efficiently using the nonlinear inelastic analysis. A case
study of a large-scale planar steel frame is investigated for the frame’s behavior and strength under the effects of residual
stresses and shear deformation. Through numerical examples, we recommend that both residual stress and shear deforma-
tion should be considered in the advanced direct analysis and design procedures for steel-framed structures.

Article highlights

• A nonlinear 2D beam column element is developed suc- • Both residual stress and shear deformation should
cessfully using onlyone element per member for modeling. be considered in the engineering design of steel
• Geometric nonlinearities, material plasticity, residual stress, frames.
and sheardeformation are investigated simultaneously.

Keywords Residual stress · Shear deformation · Steel frames · Nonlinear analysis · Fiber plastic hinge · Geometric
nonlinearity

1 Introduction such as fastly construction, economical solution, sustain-

ability, durability, eco-friendly product, etc. For hot-rolling
Steel frames are used widely in commercial buildings and steel, residual stresses are born in manufacturing proce-
pre-engineering buildings because of their significant dures. Many researchers [1–6] employed the ECCS residual
advantages compared with reinforced concrete structures stress pattern [7], as shown in Fig. 2, for analyzing steel

* Phu‑Cuong Nguyen, [email protected]; [email protected] | 1Advanced Structural Engineering Laboratory, Faculty of Civil
Engineering, Ho Chi Minh City Open University, Ho Chi Minh City, Vietnam.

SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w

Vol.:(0123456789)
Research Article SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w

frames using W-sections. These plastic hinge method stud- efforts, this study employed stability functions [28] to
ies [8–13] considered indirectly residual stresses by using consider the P-small delta phenomenon since only one-
the CRC tangent modulus concept. Some researchers element modeling can precisely predict this effect. An
[1–3, 5, 6, 14] took into account residual stresses directly incremental equilibrium equation for 2D frame elements
through meshing fibers on cross-sections. is formulated as follows:
Many studies developed a lot of nonlinear inelastic
analysis methods for steel frames. They used plastic hinge ⎧ ΔP ⎫ ⎡ A∕I 0 0 ⎤⎧ Δ𝛿 ⎫
⎪ ⎪ EI ⎢ ⎪ ⎪
methods or distributed plasticity methods for predict- ⎨ ΔMI ⎬ = L ⎢ 0 S1 S2 ⎥⎥⎨ Δ𝜃I ⎬ (1)
ing material inelasticity. For geometry nonlinearity, they ⎪ ΔMJ ⎪ ⎣ 0 S2 S1 ⎦⎪ Δ𝜃 ⎪
⎩ ⎭ ⎩ J⎭
employed Hermite interpolation functions [1, 14–17],
high-order interpolation functions [18], stability functions where S1 and S2 are stability functions found in the book
[2, 9, 12, 13, 19, 20], or corotational approach [21–24], in of Chen và Lui [28].
which stability functions is more efficient than other
approaches since they use only one-element modeling for
capturing second-order effects precisely. So this study will 2.2 Fiber plastic hinges
employ stability functions for developing a new method
for analyzing the nonlinear inelastic behavior of steel In this study, two fiber plastic hinges are monitored at two
frames. Recently, da Silva et al. [4] consider shear defor- ends of 2D frame elements. Cross-sections of these hinges
mation using Timoshenko’s beam theory in the nonlinear have meshed into many fibers, as illustrated in Fig. 1. This
inelastic analysis of steel frames. Moreover, no researchers method is more effective than traditional plastic hinge
are discussing in detail the effects of residual stresses and methods since it can directly consider residual stresses as
shear deformation on the nonlinear inelastic behavior and initial conditions, whereas traditional plastic hinge meth-
strength of steel frames. That is the reason why we do this ods consider indirectly residual stresses through proposed
study. equations or formulations of internal forces. An incremen-
This study presents a new method named the fiber tal equilibrium equation for 2D frame elements combining
plastic hinge method for analyzing the nonlinear inelastic
behavior of 2D steel frames, considering both the effects
of residual stress and shear deformation. With the primary
purpose of this study, we investigate the impacts of shear
deformation and residual stress on the nonlinear inelastic
behavior and load-carrying capacity of 2D steel frames in
detail. Next sections, we present the proposed analysis
method, numerical examples, discussion, and withdraw
important conclusions for advanced direct analysis and
design procedures of steel frames.

2 Nonlinear frame element

2.1 P‑small delta phenomenon

P-small delta phenomenon is the impact of axial force on

bending moments (e.i. bending moments are increased
by the axial force), leading to instability of beam-column
elements. P-small delta phenomenon can be considered
by using the geometric stiffness matrix combining with
meshing elements into many short elements, or it can be
considered by using high-order displacement-interpola-
tion functions combining with the flexibility method and
monitoring some integration points along the member
length [25–27], or it can be considered by using corota-
tional formulations, but this method is relatively compli-
cated. To reduce computer resources and computational Fig. 1  Discretion of cross-sections

Vol:.(1234567890)
SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w Research Article

both stability functions and fiber plastic hinges is formu- The incremental sectional deformation vector is esti-
lated as follows: mated by the flexibility matrix and the sectional force vec-
tor as
⎧ ΔP ⎫ ⎡ EA∕L 0 0 ⎤⎧ Δ𝛿 ⎫
⎪ ⎪ ⎪ ⎪ −1
⎡ ∑EA ∑
n n
⎨ ΔMI ⎬ = ⎢⎢ 0 Kii Kij ⎥⎥⎨ Δ𝜃I ⎬ (2) � � − Ei Ai yi ⎤ � �
⎪ ΔMJ ⎪ ⎣ 0 Kij Kjj ⎦⎪ Δ𝜃J ⎪ Δ𝜀 ⎢ i=1 i i ⎥ ΔN
⎩ ⎭ ⎩ ⎭ =⎢ n i=1
� �⎥ (9)
Δ𝜒 ∑ ∑n
ΔM
⎢ − Ei Ai yi Ei Ai yi2 + Ii ⎥
⎣ i=1 i=1 ⎦
S22 EI
Kii = 𝜂I (S1 − (1 − 𝜂J )) (3) where 𝜀 and 𝜒 are the sectional axial strain and the sec-
S1 L tional curvature.
The incremental sectional fiber strain vector is defined
EI by the linear geometric matrix and the incremental sec-
Kij = 𝜂I 𝜂J S2 (4)
L tional deformation vector:

S22 ⎧ Δe1 ⎫ ⎡ 1 −y1 ⎤� �

EI ⎪ Δe2 ⎪ ⎢ 1 −y2 ⎥ Δ𝜀
Kjj = 𝜂J (S1 − (1 − 𝜂I )) (5)
S1 L ⎨ ... ⎬ = ⎢ ... ... ⎥⎥ Δ𝜒 (10)
⎪ ⎪ ⎢
⎩ Δen ⎭ ⎣ 1 −yn ⎦
where 𝜂I và 𝜂J are factors considering the plasticity of fiber
hinges at two ends of frame elements, these factors are The stress–strain relationship of steel is assumed to
value from 0 to 1, the value of 0 is fully elastic, the value of be elastic perfectly plastic. Sectional internal forces are
1 is fully plastic, proposed as: estimated as follows:
∑n � 2 �
i=1 EtIi Ai yi + Ii ⎧ ∑𝜎A ⎫
m
𝜂I = (6) � �
EI N ⎪ i i ⎪
= ⎨ i=1 ⎬ (11)
M ∑
m
∑n � 2 � ⎪ − 𝜎i Ai yi ⎪
⎩ i=1 ⎭
i=1 EtJi Ai yi + Ii
𝜂J = (7)
EI
where n is the number of meshed fibers at two ends of
frame elements; EtIi and EtJi are current modulus of ith fiber 2.4 Residual stresses
at ends I and J, if the strain of fiber is larger than plastic
strain, current modulus of i fiber will be assigned to be By meshing the cross-section into many fibers, as illus-
equal to 0, e.i. contributed stiffness of fiber is equal to 0; trated in Fig. 1, assuming that residual stresses [7] are
Ai is the area of ith fiber; Ii is inertia moment of ith fiber
around its centroid; I is inertia moment of frame cross-
section; yi is the coordinate of the center of ith fiber as
illustrated in Fig. 1.

2.3 Fiber behavior

Fiber behavior needs to be defined for estimating the state

of fiber plastic hinges. Based on the force interpolation
function matrix, we can calculate the incremental sectional
force vector at two ends as follows:

� � � �⎧ ΔP ⎫
ΔN 1 0 0 ⎪ ⎪
ΔM
=
0 x∕L − 1 x∕L ⎨
ΔMI ⎬ (8)
⎪ ΔMJ ⎪
⎩ ⎭

where x = 0 is considering at I end, and x = L is considering

at L end.
Fig. 2  Assumed residual stresses [7]

Vol.:(0123456789)
Research Article SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w

assigned directly to fibers in the beginning, as plotted −Kii Kjj + Kij2 + Kij As GL
in Fig. 2. Dij = (18)
Kii + Kjj + 2Kij + As GL
2.5 Shear deformation
Kii Kjj − Kij2 + Kjj As GL
The factors of bending moments in the element stiffness Djj = (19)
Kii + Kjj + 2Kij + As GL
matrix are developed for calculating the additional flex-
ural shear effects in a frame member. The flexural flexibil-
ity matrix can be obtained by inverting the flexural stiff- 2.6 P‑large delta phenomenon
ness matrix, and the incremental equilibrium equation of
moments and slopes is written as The tangent stiffness matrix of the element, including the
P-large delta phenomenon by using the geometry stiffness
� � ⎡
Kjj −Kij
⎤� �
Δ𝜃IM Kii Kjj −Kij2 Kii Kjj −Kij2
⎥ ΔMI matrix, is summed as follows:
=⎢ (12)
Δ𝜃JM ⎢ −Kij Kii ⎥ ΔMJ [ ] [ ]
⎣ Kii Kjj −Kij2 Kii Kjj −Kij2 ⎦ [K ]T = [T ]T3×6 Ke 3×3 [T ]3×6 + Kg 6×6 (20)

where Kii, Kij, and Kjj are the factors of stiffness matrix in a where the transformation matrix [T ]3×6 for the frame ele-
2D frame element. θIM and θJM are the slopes of the neutral ment is calculated as
axis under bending moments. The flexibility matrix corre-
sponding to flexural shear deformation can be written as ⎡ −1 0 0 1 0 0 ⎤
[T ]3×6 = ⎢ 0 1∕L 1 0 −1∕L 0 ⎥ (21)
{ } [ 1 1 ]{ } ⎢ ⎥
Δ𝜃IS GAs L GAs L ΔMI ⎣ 0 1∕L 0 0 −1∕L 1 ⎦
Δ𝜃JS
= 1 1
ΔMJ (13)
GAs L GAs L and the geometry stiffness matrix is established as
where G is the shear modulus, As is the area subjected to ⎡ 0 MI +MJ M +M
0 0 − I L2 J 0⎤
shear, L is the length of a frame element. The total rotation ⎢ MI +MJ L2
P MI +MJ ⎥
⎢ L 0 − L2 − PL 0⎥
at the I and J ends is summed by Eqs. (12) and (13) as 2 L
� � ⎢ 0 0 0 0 0 0⎥
{ } { } { } Kg 6×6 =⎢ M +M MI +MJ ⎥ (22)
Δ𝜃I Δ𝜃IM Δ𝜃IS ⎢ 0 − I L2 J 0 0 0⎥
= + (14) L2
Δ𝜃J Δ𝜃JM Δ𝜃JS ⎢ − MI +M J
− PL
M +M
0 I L2 J P
0⎥
⎢ L2 L ⎥
⎣ 0 0 0 0 0 0⎦
The force–displacement equation considering shear
deformation is obtained by inverting the flexibility matrix
{ } [ ]{ } 3 Nonlinear algorithm and analysis program
ΔMI Dii Dij Δ𝜃I
ΔMJ
=
Dij Djj Δ𝜃J (15)
The general displacement control method [29] is used for
The incremental equilibrium equation can be written for developing the nonlinear static solution procedure. This
2D frame element considering shear deformation as method owns numerical stability and can solve the prob-
lems with many snap-back and snap-through points. The
⎧ ΔP ⎫ ⎡ E A∕L 0 0 ⎤⎧ Δ𝛿 ⎫ incremental equilibrium equation for solving nonlinear static
⎪ ⎪ t ⎪ ⎪ problems of 2D steel frames is written as follows:
⎨ ΔMI ⎬ = ⎢⎢ 0 Dii Dij ⎥⎥⎨ Δ𝜃I ⎬ (16)
⎪ ΔMJ ⎪ ⎣ 0 Dij Djj ⎦⎪ Δ𝜃J ⎪ [ ]{ } { } { }
⎩ ⎭ ⎩ ⎭ i
Kj−1 ΔDji = 𝜆ij P̂ + Rj−1i
(23)
where
The detail formulation can be found in the original article
Kii Kjj − Kij2 + Kii As GL of Yang and Shieh [29]
Dii = (17) Based on the proposed formulation, an analysis program
Kii + Kjj + 2Kij + As GL
is coded using the Fortran programming language. The pro-
posed program can accurately predict the nonlinear behav-
ior and ultimate strength of 2D steel frames considering

Vol:.(1234567890)
SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w Research Article

START: Input data

1. Form incremental load vector

2. Form tangent stiffness matrix

"Solve equilibrium equations"

Find incremental displacements

j=j+1
1. Find displacements at two ends of members
2. Find sectional force vectors

GDC iterative
1. Calculate strain and stress of fibers

algorithm
2. Calculate charateristic of sections

1. Find vector of total external force

and internal force
Next increment, i = i + 1
2. Form unbalanced force vector

No
Convergence?
Fig. 4  Portal frame
Yes

Yes study is using a fiber plastic hinge method with one-ele-

Next load step? ment modeling.
Figure 5 shows the result comparison of this study,
No Vogel [14], and ABAQUS [30]. It can be seen that the pre-
FINISH: Output data
dicted result is almost identical to the study of Vogel. It
is noted that Vogel’s result (red solid line) and this study
(w/o shear deformation, blue dash line) do not consider
Fig. 3  Flowchart of the proposed program the effects of shear deformation and use the residual stress
pattern [7]. In the case of considering both residual stress
geometric nonlinearity, plasticity, residual stress, shear and shear deformation, ABAQUS’s result [30] (red dot line
deformation using beam-column line elements. Figure 3 with green squares) and the proposed program (blue solid
illustrates the flowchart of the nonlinear inelastic analysis line) are identical in the range of load coefficient from 0.0
procedure of the proposed program. to 6.0 since this load range is in the elastic regime of steel
material. When the load coefficient increases by more than
6.0, the frame behavior is change due to plasticity mainly.
4 Numerical examples In this case, the limit load coefficients are 1.030 of ABAQUS
and 1.007 of the proposed program. Less than 2.23% error
4.1 Portal frame is obtained comparing with ABAQUS’s result. Without the
effect of shear deformation, the limit load coefficient is
Vogel [14] invented a portal steel frame as plotted in Fig. 4. obtained by Vogel’s method to be 1.022 while the limit
This frame is used as a benchmark problem for the nonlin- load coefficient is obtained by this study to be 1.014. Less
ear inelastic analysis of steel frames. Initial imperfections than 0.78% error is obtained comparing with Vogel’s result.
are assumed for columns through the initial out-of-plump- It can be seen that this fiber plastic hinge method can cap-
ness of 𝜓 = 1∕400. Elastic modulus is E = 205000 MPa. The ture the nonlinear behavior and ultimate strength of 2D
yield stress of steel is 𝜎y = 235 MPa. Vogel [14] developed steel frames as the sophisticated plastic zone method but
a plastic zone method for analyzing this frame, while this more effective in the computational effort because the
ABAQUS modeling of Kim and Lee [30] uses 8952 S4R5

Vol.:(0123456789)
Research Article SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w

Fig. 5  Load–deflection rela-

tionship at A of portal frame 1.0
0.9
0.8
0.7

Load coefficient
0.6
0.5
0.4
0.3 This study
0.2 ABAQUS
This study w/o She Def
0.1
Vogel
0.0
0 2 4 6 8 10 12 14 16 18 20 22 24 26
Horizontal deflection at A (mm)

Fig. 6  Load–deflection rela-

tionship at A of portal frame
1.0
with different effects
0.9
0.8
0.7
Load coefficient

0.6
This study
0.5
This study w/o She Def
0.4
This study w/o Res Str
0.3
This study w/o (Res Str + She Def)
0.2 Vogel
0.1 ABAQUS
0.0
0 2 4 6 8 10 12 14 16 18 20 22 24 26
Horizontal deflection at A (mm)

shell elements in the nonlinear analysis while the pro- both shear deformation and residual stress; Case 2 con-
posed program uses only one beam-column line element siders only shear deformation; Case 3 considers only
per member. Moreover, Kim and Lee [30] developed a residual stress; Case 4 does not consider both shear
complicated Fortran subroutine for assigning initial resid- deformation and residual stress. Table 1 shows differ-
ual stresses for shell elements. ences in limit load coefficient and deflection at node
Figures 6, 7, 8, and 9 show load–deflection relation- A generated by this study with four mentioned-above
ships at node A for four different cases: Case 1 considers cases. From Figs. 6, 7 and 9, it can be seen that residual

Vol:.(1234567890)
SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w Research Article

Fig. 7  Load-vertical deflec- 1.0

tion relationship at A of portal
frame 0.9
0.8
0.7

Load coefficient
0.6
0.5
This study 0.4
This study w/o She Def 0.3
This study w/o Res Str 0.2
This study w/o (Res Str + She Def) 0.1
0.0
-14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
Vertical deflection at A (mm)

Fig. 8  Load-vertical deflec- 1.05

tion relationship at A of portal
frame

1.00

Load coefficient
0.95

0.90
This study
This study w/o She Def
0.85
This study w/o Res Str
This study w/o (Res Str + She Def)
0.80
-14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4
Vertical deflection at A (mm)

stresses do not affect global structural stiffness in the stresses are much smaller than those of case 1 and 2
beginning, but shear deformation has significant influ- with the effect of residual stresses. The post-collapse
ences during the incremental iterative solution proce- load curves of horizontal deflection at node A for four
dure. Residual stresses definitely affect the behavior cases are matched together, as plotted in Fig. 6. In addi-
and strength of this frame at the load coefficient of 0.6 tion, from Figs. 8 and 9, the post-collapse load curves of
because, from this time, some fibers start yielding. From vertical deflection and rotation at node A for cases 1 and
Fig. 6 and Table 1, horizontal deflection at node A of 3 are overlapped, and those for cases 2 and 4 are also
the frame in case 3 and 4 without considering residual overlapped. As shown in Figs. 7 and 8, shear deformation

Vol.:(0123456789)
Research Article SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w

Fig. 9  Load-rotation relation- 1.0

ship at A of portal frame
0.9
0.8
0.7

Load coefficient
0.6
0.5
0.4 This study
0.3 This study w/o She Def
0.2 This study w/o Res Str
0.1 This study w/o (Res Str + She Def)
0.0
0.0000 0.0004 0.0008 0.0012 0.0016 0.0020
Rotation at A (rad)

Table 1  Deflection at limit load at point A for portal frame

Case Load coefficient Error (%) Horizontal Error (%) Vertical deflec- Error (%) Rotation (rad) Error (%)
deflection (mm) tion (mm)

Case 1 1.007 – 19.92 – −6.18 – 0.00116 –

Case 2 1.014 0.70 17.93 −9.97 −6.22 0.59 0.00098 −15.63
Case 3 1.022 1.49 11.77 −40.91 −5.08 −17.84 0.00101 −12.94
Case 4 1.027 1.99 10.97 −44.95 −5.16 −16.58 0.00086 −26.05

Case 1 considers both shear deformation and residual stress; Case 2 considers only residual stress; Case 3 considers only shear deformation;
Case 4 does not consider both shear deformation and residual stress

does not affect the vertical deflection at node A or axial in modeling, while this study used a fiber plastic hinge
shortening of members. method using one-element modeling. The limit load
coefficient is obtained by Vogel’s approach to be 1.111.
4.2 Six‑story frame The limit load coefficient is achieved by this study to be
1.116, without the effect of shear deformation. Less than
The six-story frame, as shown in Fig. 10, was also chosen 0.45% error is obtained comparing with Vogel’s result.
by Vogel [14] as one of the benchmark problems for the It can be seen that this fiber plastic hinge method can
nonlinear inelastic analysis of steel frames. Initial imperfec- capture the nonlinear behavior and ultimate strength of
tions of columns are assumed to be out-of-plumpness of 2D steel frames as the sophisticated plastic zone method
𝜓 = 1∕450. Elastic modulus is E = 20500 MPa. Yield stress but more effective in computational effort.
is 𝜎y = 235 MPa. Figures 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, and 22 show
Figure 11 shows the comparison of the load-horizon- load–deflection relationships at node A and B for four dif-
tal deflection relationship predicted by this study and ferent cases: Case 1 considers both shear deformation and
Vogel [14]. Those without the effects of shear deforma- residual stress; Case 2 considers only shear deformation;
tion are in agreement well. It is noted that Vogel used a Case 3 considers only residual stress; Case 4 does not con-
plastic zone method with a lot of beam-column elements sider both shear deformation and residual stress. Tables 2

Vol:.(1234567890)
SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w Research Article

Fig. 10  Six-story frame

Fig. 11  Load-horizontal 1.1

deflection relationship at A of
six-story frame 1.0
0.9
0.8
Load coefficient

0.7
0.6
0.5
0.4
0.3
0.2 This study w/o She Def
0.1 Vogel
0.0
0 30 60 90 120 150 180 210 240 270 300 330
Horizontal deflection at A (mm)

Vol.:(0123456789)
Research Article SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w

Fig. 12  Load-horizontal 1.1

deflection relationship at A of
six-story framewith different 1.0
effects
0.9
0.8

Load coefficient
0.7
0.6
0.5 This study
0.4 This study w/o She Def
0.3 This study w/o Res Str
0.2 This study w/o (Res Str + She Def)
0.1 Vogel
0.0
0 30 60 90 120 150 180 210 240 270 300 330
Horizontal deflection at A (mm)

Fig. 13  Load-horizontal
deflection relationship at A of 1.10
six-story frame with different 1.05
effects
1.00
0.95
Load coefficient

0.90
0.85
0.80
This study
0.75
0.70 This study w/o She Def
0.65 This study w/o Res Str
0.60 This study w/o (Res Str + She Def)
0.55 Vogel
0.50
30 60 90 120 150 180 210 240 270
Horizontal deflection at A (mm)

and 3 show differences of limit load coefficient and deflec- of 0.6 because, from this time, some fibers start yielding.
tion at nodes A and B generated by this study with four From Figs. 12 to 22, Tables 2 and 3, deflection at nodes A
mentioned-above cases. From Figs. 12 and 18, it can be and B of the frame in four different cases are not much
seen that residual stresses do not affect global structural different such as the behavior of the portal frame in the
stiffness in the beginning, but shear deformation has one example. The rotation at nodes A and B in four cases
significant influences during the incremental iterative is not much different. As shown in Figs. 14, 15, 19 and 20,
solution procedure. Residual stresses definitely affect the shear deformation does not affect on the vertical deflec-
behavior and strength of this frame at the load coefficient tion at node A and B or axial shortening of members, while

Vol:.(1234567890)
SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w Research Article

Fig. 14  Load-vertical deflec- 1.1

tion relationship at A of six-
story frame 1.0
0.9
0.8

Load coefficient
0.7
0.6
0.5
This study
0.4
This study w/o She Def 0.3
This study w/o Res Str 0.2
This study w/o (Res Str + She Def) 0.1
0.0
-24 -22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0
Vertical deflection at A (mm)

Fig. 15  Load-vertical deflec-

tion relationship at A of six- 1.10
story frame
1.05

Load coefficient
1.00

0.95

This study 0.90

This study w/o She Def 0.85
This study w/o Res Str
0.80
This study w/o (Res Str + She Def)
0.75
-20 -18 -16 -14 -12 -10 -8 -6 -4
Vertical deflection at A (mm)

residual stresses make the influences on the behavior and 5 Case study
axial shortening of columns in load coefficient range from
0.7 to 1.1. There are differences between the portal frame In the last two examples, we proved the proposed meth-
and the six-story frame in the behavior curves because the od’s accuracy and reliability in predicting the second-
portal frame is collapsed by yielding along the length of order inelastic analysis of steel frames. In this case study,
columns, mainly as shown in Fig. 23. In contrast, the six- we analyze the impacts of residual stresses and shear
story frame is collapsed when both beams and columns deformation for a large-scale steel frame. A five-bay
are yielding, as shown in Fig. 24. nine-story steel frame with geometry, cross-sections, and

Vol.:(0123456789)
Research Article SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w

Fig. 16  Load-rotation relation- 1.1

ship at A of six-story frame
1.0
0.9
0.8

Load coefficient
0.7
0.6
0.5
This study
0.4
This study w/o She Def 0.3
This study w/o Res Str 0.2
This study w/o (Res Str + She Def) 0.1
0.0
-0.024 -0.021 -0.018 -0.015 -0.012 -0.009 -0.006 -0.003 0.000
Rotation at A (rad)

Fig. 17  Load-rotation relation-

ship at A of six-story frame 1.10
1.05
1.00

Load coefficient
0.95
0.90
This study
0.85
This study w/o She Def
0.80
This study w/o Res Str
0.75
This study w/o (Res Str + She Def)
0.70
-0.018 -0.016 -0.014 -0.012 -0.010 -0.008 -0.006 -0.004
Rotation at A (rad)

loadings is studied, as shown in Fig. 25. Elastic modulus HEB240, HEB220. Cross-sections of beams from one-to-
is E = 20500 MPa. Yield stress is 𝜎y = 235 MPa. The beam- three stories, from four-to-six stories, from seven-to-nine
span length of the frame is 5.0 m. The story height of the stories are IPE400, IPE360, IPE330. Applied loadings at
frame is 3.5 m. Cross-sections of exterior columns from exterior column tops are F1 = 120 kN . Applied loadings
one-to-three stories, from four-to-six stories, from seven- at interior column tops are F2 = 240 kN . Wind loadings of
to-nine stories are HEB240, HEB220, HEB200. Cross-sec- F3 = 12 kN and F4 = 24 kN are put at positions, as shown in
tions of interior columns from one-to-three stories, from Fig. 25. We use one element per member for modeling all
four-to-six stories, from seven-to-nine stories are HEB260, beam-columns for the frame. All cross-sections are divided

Vol:.(1234567890)
SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w Research Article

Fig. 18  Load-horizontal 1.1

deflection relationship at B of
six-story frame 1.0
0.9
0.8

Load coefficient
0.7
0.6
0.5 This study
0.4
This study w/o She Def
0.3
0.2 This study w/o Res Str
0.1 This study w/o (Res Str + She Def)
0.0
0 5 10 15 20 25 30 35 40 45 50 55 60 65
Horizontal deflection at B (mm)

Fig. 19  Load-vertical deflec- 1.1

tion relationship at B of six-
story frame 1.0
0.9
0.8

Load coefficient
0.7
0.6
0.5
This study
0.4
This study w/o She Def 0.3
This study w/o Res Str 0.2
This study w/o (Res Str + She Def) 0.1
0.0
-5.5 -5.0 -4.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0
Vertical deflection at B (mm)

into sixty-six fibers for accuracy and considering residual deformation and residual stresses and without these
stresses by inputting initial stress values for each fiber. effects. Based on Figs. 26 and 29, it can be concluded that
Figures 26, 27, 28, 29, 30, and 31 show comparisons shear deformation significantly affects the structure’s stiff-
of the load–deflection relationships at A and B of the ness. While residual stresses impact substantially on the
frame under nonlinear static analysis considering shear structure’s behavior and stiffness as fiber plastic hinges

Vol.:(0123456789)
Research Article SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w

Fig. 20  Load-vertical deflec-

tion relationship at B of six- 1.10
story frame
1.05
1.00

Load coefficient
0.95
0.90
This study
0.85
This study w/o She Def
0.80
This study w/o Res Str
0.75
This study w/o (Res Str + She Def)
0.70
-5.4 -5.1 -4.8 -4.5 -4.2 -3.9 -3.6 -3.3 -3.0 -2.7 -2.4
Vertical deflection at B (mm)

Fig. 21  Load-rotation relation- 1.1

ship at B of six-story frame
1.0
0.9
0.8
Load coefficient

0.7
0.6
0.5
This study
0.4
0.3 This study w/o She Def
0.2 This study w/o Res Str
0.1 This study w/o (Res Str + She Def)
0.0
0.000 0.002 0.004 0.006 0.008 0.010 0.012
Rotation at B (rad)

Vol:.(1234567890)
SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w Research Article

Fig. 22  Load-rotation relation-

ship at B of six-story frame 1.110

1.080

Load coefficient
1.050

1.020

0.990 This study

0.960 This study w/o She Def

This study w/o Res Str
0.930
This study w/o (Res Str + She Def)
0.900
0.000 0.002 0.004 0.006 0.008 0.010 0.012
Rotation at B (rad)

Table 2  Deflection at limit load at point A for six-story frame

Case Load coefficient Error (%) Horizontal Error (%) Vertical deflec- Error (%) Rotation (rad) Error (%)
deflection (mm) tion (mm)

Case 1 1.113 – 306.82 – −13.87 – −0.01241 –

Case 2 1.116 0.30 303.08 −1.22 −14.28 2.96 −0.01352 8.90
Case 3 1.116 0.29 317.50 3.48 −15.00 8.14 −0.01444 16.35
Case 4 1.119 0.58 303.61 −1.05 −15.01 8.25 −0.01471 18.49

Case 1 considers both shear deformation and residual stress; Case 2 considers only residual stress; Case 3 considers only shear deformation;
Case 4 does not consider both shear deformation and residual stress

Table 3  Deflection at limit load at point B for six-story frame

Case Load coefficient Error (%) Horizontal Error (%) Vertical deflec- Error (%) Rotation (rad) Error (%)
deflection (mm) tion (mm)

Case 1 1.113 – 52.17 – −5.19 – 0.00365 –

Case 2 1.116 0.30 48.03 −7.93 −5.19 −0.12 0.00338 −7.24
Case 3 1.116 0.29 51.89 −0.52 −5.11 −1.54 0.00429 17.66
Case 4 1.119 0.58 47.74 −8.48 −5.12 −1.30 0.00389 6.82

Case 1 considers both shear deformation and residual stress; Case 2 considers only residual stress; Case 3 considers only shear deformation;
Case 4 does not consider both shear deformation and residual stress

Vol.:(0123456789)
Research Article SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w

start yielding. The load-carrying capacity of the frame is

no much change as residual stresses and shear deforma-
tion are considered or not.
Tables 4 and 5 list deflection values at the limit load of
the frame for nodes A and B in four cases of considering or
not considering shear deformation and residual stresses.
It can be seen that in the case 1 of considering both shear
deformation and residual stresses, the horizontal deflec-
tion of the frame is biggest. This recommends that struc-
tural designers check the buildings’ service conditions
with the effects of shear deformation and residual stresses
simultaneously.

Fig. 23  Plasticity of portal frame at limit load [1]

6 Conclusion

A fiber plastic hinge method for nonlinear inelastic

analysis of 2D steel frames considering both shear defor-
mation and residual stress is developed successfully.
Geometric nonlinearity is considered by using stability
functions and the geometric stiffness matrix. Material
nonlinearity is simulated by a proposed fiber plastic
hinge model. Residual stress is directly considered by
dividing several small fibers on the cross-sections. The
proposed procedure can predict precisely and effec-
tively the behavior and load-carrying capacity of 2D
steel frames under static loadings by using advanced
nonlinear analysis as complicated plastic zone methods
or commercial general finite element software (ABAQUS,
ANSYS, etc.) done. Local buckling, lateral-torsional buck-
ling, panel zone, etc., are not considered in this study.
In the next work, the effects of shear deformation and
residual stress on the nonlinear inelastic dynamic analy-
sis of steel frames will be investigated and evaluated.
Some original conclusions are withdrawn from this
study:

• Residual stresses definitely affect the behavior and

strength of steel-framed structures when some moni-
toring fibers start yielding.
• Shear deformation significantly affects the global and
local structural stiffness during the analysis procedure.
• The influence of shear deformation on transverse
deflection is considerable in the nonlinear analysis of
steel frames.

Fig. 24  Plasticity of six-story frame at limit load [1]

Vol:.(1234567890)
SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w Research Article

F1 = 120 kN F2 = 240 kN F2 = 240 kN F2 = 240 kN F2 = 240 kN F1 = 120 kN

F3 = 12 kN A
IPE330 IPE330 IPE330 IPE330 IPE330

HEB220

HEB220

HEB220

HEB220

HEB200
HEB200
F1 = 100 kN F2 F2 F2 F2 F1
F4 = 24 kN
IPE330 IPE330 IPE330 IPE330 IPE330

HEB220

HEB220

HEB220

HEB220

HEB200
HEB200

F1 F2 F2 F2 F2 F1
F4 = 24 kN
IPE330 HEB220 IPE330 IPE330 IPE330 IPE330

HEB220

HEB220

HEB220

HEB200
HEB200

F1 F2 F2 F2 F2 F1
F4
IPE360 IPE360 IPE360 IPE360 IPE360
HEB240

HEB240

HEB240

HEB240
HEB220

HEB220
F1 F2 F2 F2 F2 F1
F4

9 x 3.5 m = 31.5 m
IPE360 IPE360 IPE360 IPE360 IPE360
HEB240

HEB240

HEB240

HEB240
HEB220

HEB220
F1 F2 F2 F2 F2 F1
F4
IPE360 IPE360 IPE360 IPE360 IPE360
HEB240

HEB240

HEB240

HEB240
HEB220

HEB220
F1 F2 F2 F2 F2 F1
F4
IPE400 IPE400 IPE400 IPE400 B IPE400
HEB260

HEB260

HEB260

HEB260
HEB240

HEB240
F1 F2 F2 F2 F2 F1
F4
IPE400 IPE400 IPE400 IPE400 IPE400
HEB260

HEB260

HEB260
HEB260
HEB240

HEB240
F1 = 120 kN F2 = 240 kN F2 = 240 kN F2 = 240 kN F2 = 240 kN F1
F4 = 24 kN
IPE400 IPE400 IPE400 IPE400 IPE400
HEB260

HEB260

HEB260

HEB260
HEB240

HEB240
5.0 m 5.0 m 5.0 m 5.0 m 5.0 m

Fig. 25  Five-bay nine-story frame

Vol.:(0123456789)
Research Article SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w

Fig. 26  Load–deflection 1.1

relationship at A of nine-story
frame with different effects 1.0
0.9
0.8

Load coefficient
0.7
0.6
0.5 This study
0.4
This study w/o She Def
0.3
0.2 This study w/o Res Str
0.1 This study w/o (Res Str + She Def)
0.0
0 12 24 36 48 60 72 84 96 108 120
Horizontal deflection at A (mm)

Fig. 27  Load-vertical deflec- 1.1

tion relationship at A of nine-
story frame 1.0
0.9
0.8

Load coefficient
0.7
0.6
0.5
This study
0.4
This study w/o She Def 0.3
This study w/o Res Str 0.2
This study w/o (Res Str + She Def) 0.1
0.0
-24 -22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0
Vertical deflection at A (mm)

Vol:.(1234567890)
SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w Research Article

Fig. 28  Load-rotation relation- 1.1

ship at A of nine-story frame
1.0
0.9
0.8

Load coefficient
0.7
0.6
0.5
This study
0.4
This study w/o She Def 0.3
This study w/o Res Str 0.2
This study w/o (Res Str + She Def) 0.1
0.0
-0.0010 -0.0008 -0.0006 -0.0004 -0.0002 0.0000
Rotation at A (rad)

Fig. 29  Load–deflection 1.1

relationship at B of nine-story
frame 1.0
0.9
0.8
Load coefficient

0.7
0.6
0.5 This study
0.4
This study w/o She Def
0.3
0.2 This study w/o Res Str
0.1 This study w/o (Res Str + She Def)
0.0
0 8 16 24 32 40 48 56 64 72
Horizontal deflection at B (mm)

Vol.:(0123456789)
Research Article SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w

Fig. 30  Load-vertical deflec- 1.1

tion relationship at B of nine-
story frame 1.0
0.9
0.8

Load coefficient
0.7
0.6
0.5
This study
0.4
This study w/o She Def 0.3
This study w/o Res Str 0.2
This study w/o (Res Str + She Def) 0.1
0.0
-18 -16 -14 -12 -10 -8 -6 -4 -2 0
Vertical deflection at B (mm)

Fig. 31  Load-rotation relation- 1.1

ship at B of nine-story frame
1.0
0.9
0.8
Load coefficient

0.7
0.6
0.5
This study
0.4
0.3 This study w/o She Def
0.2 This study w/o Res Str
0.1 This study w/o (Res Str + She Def)
0.0
0.0000 0.0002 0.0004 0.0006 0.0008 0.0010
Rotation at B (rad)

Vol:.(1234567890)
SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w Research Article

Table 4  Deflection at limit load at point A for nine-story frame

Case Load coefficient Error (%) Horizontal Error (%) Vertical Error (%) Rotation(rad) Error (%)
deflection(mm) deflection(mm)

Case 1 1.053 – 86.48 – −12.90 – −0.00090 –

Case 2 1.061 0.75 79.30 −8.30 −13.14 1.88 −0.00092 2.43
Case 3 1.060 0.66 81.62 −5.62 −12.66 −1.84 −0.00081 −9.82
Case 4 1.067 1.32 74.62 −13.71 −12.81 −0.68 −0.00084 −6.48

Case 1 considers both shear deformation and residual stress; Case 2 considers only residual stress; Case 3 considers only shear deformation;
Case 4 does not consider both shear deformation and residual stress

Table 5  Deflection at limit load at point B for nine-story frame

Case Load coefficient Error (%) Horizontal Error (%) Vertical Error (%) Rotation(rad) Error (%)
deflection(mm) deflection(mm)

Case 1 1.053 – 47.63 – −11.83 – 0.00077 –

Case 2 1.061 0.75 43.38 −8.92 −12.13 2.53 0.00068 −11.64
Case 3 1.060 0.66 43.56 −8.54 −11.07 −6.43 0.00089 15.65
Case 4 1.067 1.32 39.52 −17.02 −11.31 −4.40 0.00080 3.96

Case 1 considers both shear deformation and residual stress; Case 2 considers only residual stress; Case 3 considers only shear deformation;
Case 4 does not consider both shear deformation and residual stress

• The proposed formulation for the effects of shear use, you will need to obtain permission directly from the copyright
deformation does not affect the axial shortening of holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​
org/​licen​ses/​by/4.​0/.
frame members.
• Both residual stresses and shear deformation should
be considered in the advanced direct analysis and
design procedures for steel frames.
References
• In engineering design, designers should check the
1. Nguyen P-C, Kim S-E (2016) Advanced analysis for planar steel
effects of residual stresses and shear deformation on frames with semi-rigid connections using plastic-zone method.
service conditions of buildings Steel Compos Struct 21(5):1121–1144
2. Nguyen P-C, Kim S-E (2014) An advanced analysis method for
three-dimensional steel frames with semi-rigid connections.
Finite Elem Anal Des 80:23–32
Acknowledgements The authors gratefully acknowledge the financial
3. Chiorean CG (2013) A computer method for nonlinear inelastic
support granted by the Scientific Research Fund of the Ministry of
analysis of 3D composite steel–concrete frame structures. Eng
Education and Training (MOET), Vietnam. (No. B2019–MBS–01). The
Struct 57:125–152
authors would also like to thank colleagues at Ho Chi Minh City Open
4. da Silva RGL, Lavall ACC, Costa RS, Viana HF (2018) Formula-
University for supporting this work.
tion for second-order inelastic analysis of steel frames including
shear deformation effect. J Constr Steel Res 151:216–227
Declarations`
5. Zubydan AH (2013) Inelastic large deflection analysis of space
steel frames including H-shaped cross sectional members. Eng
Conflict of interest The authors declare that they have no conflict of Struct 48:155–165
interest. 6. Jiang X-M, Chen H, Liew JYR (2002) Spread-of-plasticity
analysis of three-dimensional steel frames. J Constr Steel Res
Open Access This article is licensed under a Creative Commons Attri- 58(2):193–212
bution 4.0 International License, which permits use, sharing, adap- 7. European Convention for Constructional Steelwork (1984)
tation, distribution and reproduction in any medium or format, as Ultimate limit state calculation of sway frames with rigid
long as you give appropriate credit to the original author(s) and the joints. Technical Committee 8, Structural Stability. Technical
source, provide a link to the Creative Commons licence, and indicate Working Group 8.2, System publication no. 33. ECCS General
if changes were made. The images or other third party material in this Secretariat
article are included in the article’s Creative Commons licence, unless 8. Ziemian RD, McGuire W (2002) Modified tangent modulus
indicated otherwise in a credit line to the material. If material is not approach, a contribution to plastic hinge analysis. J Struct
included in the article’s Creative Commons licence and your intended Eng 128(10):1301–1307
use is not permitted by statutory regulation or exceeds the permitted 9. Kim S-E, Kim Y, Choi S-H (2001) Nonlinear analysis of 3-D steel
frames. Thin-Walled Struct 39(6):445–461

Vol.:(0123456789)
Research Article SN Applied Sciences (2021) 3:686 | https://doi.org/10.1007/s42452-021-04638-w

10. Kim S-E, Chen W-F (1996) Practical advanced analysis for braced 22. Teh LH, Clarke MJ (1998) Co-rotational and Lagrangian formula-
steel frame design. J Struct Eng 122(11):1266–1274 tions for elastic three-dimensional beam finite elements. J Con-
11. Kim S-E, Chen W-F (1996) Practical advanced analysis for str Steel Res 48(2–3):123–144
unbraced steel frame design. J Struct Eng 122(11):1259–1265 23. Li ZX (2006) A co-rotational formulation for 3D beam ele-
12. King WS, White DW, Chen WF (1992) Second-order inelastic anal- ment using vectorial rotational variables. Comput Mech
ysis methods for steel-frame design. J Struct Eng 118(2):408–428 39(3):309–322
13. Nguyen P-C, Kim S-E (2012) Second-order plastic-hinge analysis 24. Battini J-M, Pacoste C (2002) Plastic instability of beam struc-
of space semi-rigid steel frames. Thin-Walled Struct 60:98–104 tures using co-rotational elements. Comput Methods Appl
14. Vogel U (1985) Calibrating frames. Berlin, Germany, Stahlbau, Mech Eng 191(51–52):5811–5831
pp 295–301 25. Saritas A, Koseoglu A (2015) Distributed inelasticity planar frame
15. Foley CM (2001) Advanced analysis of steel frames using parallel element with localized semi-rigid connections for nonlinear
processing and vectorization. Comput Aided Civ Infrastruct Eng analysis of steel structures. Int J Mech Sci 96–97:216–231
16(5):305–325 26. Du Z-L, Liu Y-P, Chan S-L (2017) A second-order flexibility-based
16. Liew JYR, Chen H, Shanmugam NE, Chen WF (2000) Improved beam-column element with member imperfection. Eng Struct
nonlinear plastic hinge analysis of space frame structures. Eng 143:410–426
Struct 22(10):1324–1338 27. Du Z-L, Ding Z-X, Liu Y-P, Chan S-L (2019) Advanced flexibility-
17. Teh LH, Clarke MJ (1999) Plastic-zone analysis of 3D steel frames based beam-column element allowing for shear deforma-
using beam elements. J Struct Eng 125(11):1328–1337 tion and initial imperfection for direct analysis. Eng Struct
18. Chan SL, Chui PPT (2000) Nonlinear static and cyclic analysis of 199:109586
steel frames with semi-rigid connections. Elsevier, Amsterdam 28. Chen WF, Lui EM (1987) Structural stability: theory and imple-
19. Nguyen P-C, Kim S-E (2017) Investigating effects of various base mentation. Elsevier, New York
restraints on the nonlinear inelastic static and seismic responses 29. Yang Y-B, Shieh M-S (1990) Solution method for nonlinear prob-
of steel frames. Int J Non-Linear Mech 89:151–167 lems with multiple critical points. AIAA J 28(12):2110–2116
20. Nguyen P-C, Kim S-E (2014) Nonlinear inelastic time-history 30. Kim S-E, Lee D-H (2002) Second-order distributed plasticity
analysis of three-dimensional semi-rigid steel frames. J Constr analysis of space steel frames. Eng Struct 24(6):735–744
Steel Res 101:192–206
21. Crisfield MA (1990) A consistent co-rotational formulation for Publisher’s Note Springer Nature remains neutral with regard to
non-linear, three-dimensional, beam-elements. Comput Meth- jurisdictional claims in published maps and institutional affiliations.
ods Appl Mech Eng 81(2):131–150

Vol:.(1234567890)