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

net/publication/307967899

A continuous–discrete approach for evaluation of natural frequencies and

mode shapes of high-rise buildings

Article in International Journal of Advanced Structural Engineering (IJASE) · September 2016

DOI: 10.1007/s40091-016-0129-6

CITATIONS READS

18 363

4 authors:

Mohsen Malekinejad Reza Rahgozar

Islamic Azad University, Sirjan Branch, Iran Shahid Bahonar University of Kerman
29 PUBLICATIONS 435 CITATIONS 94 PUBLICATIONS 1,783 CITATIONS

SEE PROFILE SEE PROFILE

Ali Malekinejad Peyman Rahgozar

Shahid Bahonar University of Kerman University of Florida
11 PUBLICATIONS 30 CITATIONS 16 PUBLICATIONS 373 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Reza Rahgozar on 15 December 2020.

The user has requested enhancement of the downloaded file.

THE STRUCTURAL DESIGN OF TALL AND SPECIAL BUILDINGS
Struct. Design Tall Spec. Build. 24, 591–606 (2015)
Published online 21 November 2014 in Wiley Online Library (wileyonlinelibrary.com/journal/tal). DOI: 10.1002/tal.1201

Dynamic analysis of combined system of framed tube and shear

walls by Galerkin method using B-spline functions

Reza Rahgozar*, Zahra Mahmoudzadeh, Mohsen Malekinejad and Peyman Rahgozar

Civil Engineering Department, Shahid Bahonar University of Kerman, Kerman, Iran

SUMMARY
In this article, dynamic parameters (natural frequencies and mode shapes) of tall buildings that consist of
framed tube and shear walls are obtained using a simple approximate method. The three-dimensional struc-
ture is replaced by an equivalent cantilever beam, considering both bending and shear deformations. On the
basis of dynamic equilibrium, the governing differential equation of motion is obtained and converted to its
corresponding weak form. B-spline functions are then utilized to approximate the weak form and to obtain
the final matrix form of the problem. Finally, by applying essential boundary conditions, the natural fre-
quencies and corresponding mode shapes are calculated. To demonstrate the accuracy of the proposed
method, numerical examples are solved, and the results are compared with those obtained from SAP2000
computer analysis. The results show that the proposed method is efficient and accurate enough to be used
in preliminary design. Copyright © 2014 John Wiley & Sons, Ltd.

Received 13 April 2014; Revised 1 October 2014; Accepted 17 October 2014

KEY WORDS: tall building; framed tube; shear wall; B-spline; natural frequencies; mode shapes

1. INTRODUCTION

When increasing the height of structures, the effect of lateral loads such as wind and earthquake should
be considered as an important factor in design so that in addition to obtaining sufficient strength, lateral
displacements are also limited to an admissible value. Therefore, choosing an appropriate lateral-
resistant system is one of the main concerns for designers. Today, on the basis of the height of the
structure, material types, architectural limitations, and so on, various systems are suggested to
designers such as rigid frames, braced frames, shear-wall frames, framed tube, braced tube, and shear
wall-frame tubes (Taranath, 1988; Smith and Coull, 1996, Halis and Emer 2007, Paulino, 2010).
Framed tube system is one of the most significant modern developments in high-rise structural forms.
In such systems, close spacing of columns and use of deep beams around the structure create a
continuous system, which acts like a hollow cantilever beam against lateral loads. Despite the overall
efficiency of framed tube system, shear-lag phenomenon attenuates the performance of this system. To
reduce the effects of shear lag and increase stiffness of the structure, combinations of framed tube and
other lateral-resistant systems such as bracing or shear wall have been suggested. The combined
system of framed tube and shear walls has been shown to be suitable for concrete structures (Rahgozar
et al., 2011).
To calculate dynamic parameters of tall buildings such as natural frequencies and mode shapes,
three-dimensional modeling and analysis are necessary, which is difficult and time consuming. In
addition, designers repeatedly need to model, analyze and change design parameters until the
desired model is reached. Hence, it is beneficial to utilize an approximate method, which can

*Correspondence to: Reza Rahgozar, Civil Engineering Department, Shahid Bahonar University of Kerman, Kerman,
Iran.
E-mail: [email protected]

Copyright © 2014 John Wiley & Sons, Ltd.

592 R. RAHGOZAR, Z. MAHMOUDZADEH AND M. MALEKINEJAD

calculate the dynamic parameters of a structure with acceptable accuracy (Malekinejad and
Rahgozar, 2011, 2012a, 2012b).
There are numerous studies that obtain the natural frequencies and mode shapes of framed tube and
shear-walled frame systems separately. In past investigations, Khan and Amin (1973), Coull and Bose
(1975), Kwan (1994), and Malekinejad and Rahgozar (2014) presented a simplified method for the
analysis of framed tube structures in which the rigidly jointed perimeter frame panels are replaced with
equivalent orthotropic plates. Singh and Nagpal (1994) by reducing the number of degrees of freedom
(DOFs), and gradually relaxing the constraints unveiled an efficient two-stage procedure for free
vibration characteristics of framed tube. Kaviani et al. (2008) employed a sandwich beam model to
determine the natural periods of multistory buildings subjected to earthquake. Kamgar and Rahgozar
(2013) obtained natural frequencies of framed tube structures through Timoshenko’s beam model,
which considers the influence of shear and flexural deformation. On the basis of D’Alembert’s
principle, Malekinejad and Rahgozar (2014) presented approximate formulas for dynamic response
of tubular tall building structures. Shear-walled frame systems have also been studied by many re-
searchers over the past decades. Kuang and Ng (2001) investigated asymmetric shear wall structures
in tall buildings and presented an approximate dynamic analysis for coupled flexural–warping tor-
sional vibration analysis. Wang and Wang (2005) presented a simple method for determining the first
two periods of natural vibration for buildings with coupled shear walls. On the basis of the continuum
approach and D’Alembert’s principle, Meftah and Tounsi (2008) presented an analytical method for
three-dimensional frequency analysis of buildings braced by shear walls and thin-walled open-section
structures. Rahgozar et al. (2011) considered axial force effects on the vibrational frequencies of
coupled shear walls. Bozdogan (2011, 2012) studied free vibration analysis of wall-frame structures
and coupled shear walls in which the governing differential equation was solved using Differential
Quadrature method.
Although free vibration analysis of framed tube system and shear-walled frame has been studied
extensively over the past few decades, there have been few research efforts related to determining
vibrational characteristics of the combined system of framed tube and shear-wall system. Therefore,
to fill in the gap, in this study, a simple yet accurate formulation for calculating natural frequencies
and mode shapes for the combined system of framed tube and shear walls is presented. On the basis
of the continuum approach, framed tube and shear walls are replaced by an equivalent cantilever beam
located at the mass center. Dynamic equilibrium concepts are used for obtaining the governing
differential equation of vibration. The differential equation is converted to its corresponding weak
form. Then, it is approximated using B-spline functions thereby constructing the eigenvalue problem
(Hopkins and Wait, 1979; Mittal and Jain, 2011; Gu et al., 2011; Yanan et al., 2011; Mohammadi,
2014; Xiaoyun and Weiyin, 2014). Finally, after applying essential boundary conditions, natural
frequencies and corresponding mode shapes are calculated and compared with those obtained from
finite element (FE) analysis. At the end of the study, numerical examples have been solved, and
comparison of the proposed method with finite element method demonstrates that the proposed
method is efficient and accurate enough to be employed at preliminary design stages of tall buildings.

2. FORMULATION AND SOLUTION

In this section, an approximate mathematical model is presented to obtain natural frequencies and
corresponding mode shapes of tall buildings with combined system of frame tube and shear wall.
The whole structure considered, which consists of frame tube and shear walls, is modeled as a
cantilever beam. On the basis of D’Alembert’s principle, the partial differential equation of motion
for this model can be obtained. The solution of the resulting partial differential equation is considered
as the product of approximation functions and a time harmonic function resulting in an ordinary
differential equation (ODE) as a function of spatial variable. Integration by parts is applied to reduce
the order of ODE and obtaining the final form of the weak form. Finally, by approximating the weak
form using B-spline functions, the resulting linear system of equations (matrix form) is obtained. After
imposing boundary conditions and solving eigenvalue problem, the natural frequencies and corre-
sponding mode shapes are obtained.

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal
 BUILDING ANALYSIS BY GALERKIN METHOD USING B-SPLINE FUNCTIONS 593

2.1. Mathematical formulation

A typical rigid frame subjected to lateral loading adopts essentially a shear deformation shape; hence, it
can be modeled by an equivalent shear beam. However, modeling the framed tube as an equivalent
beam with shear rigidities is not sufficient to capture the true behavior of this system. Thus, in the
analysis, framed tube is replaced by a shear beam located at the center of shear rigidity and a flexural
beam at the center of flexural rigidity. Material and cross-sectional properties of equivalent beams are
obtained by considering framed tube as a hollow box section beam (Kwan, 1994) (Appendix A). In
addition, wall members are lumped and replaced by equivalent flexural and shear cantilever beams
located at the flexure center and shear rigidity of the shear walls at the shear center. On the basis of
symmetry, flexure and shear centers of the framed tube and the shear walls are coincident. Therefore,
a cantilever beam with the following equivalent characteristic can be analyzed to obtain the natural
frequencies and corresponding mode shapes of tall buildings with combined system of framed tube
and shear walls (Malekinejad and Rahgozar, 2014).

EI ¼ ðEI Þw þ ðEI Þf (1a)

GA ¼ ðGAÞw þ ðGAÞf (1b)

m ¼ mw þ mf (1c)

where EI and GA are the flexural and shear rigidities of a cantilever beam, respectively; they are
computed as the sum of rigidities for framed tube and shear walls. m denotes the mass per unit length
of the shear-walled tube structure. Subscripts f and w denote the framed tube and shear wall, respec-
tively. E and G are the flexural and shear moduli of elasticity. A and I are the effective shear area
and moment of inertia, respectively. A typical combined system of framed tube and shear walls as well
as the equivalent cantilever beam are shown in Figure 1.

2.2. Governing equation

As mentioned above, the combined system of framed tube and shear walls is considered to be a doubly
symmetric structure. Thus, in this case, the three vibration components, which are the lateral flexural
vibrations in the y and z directions and the torsional vibration about the shear center, are independent
of each other. Consequently, for evaluating dynamic lateral response, the three-dimensional behavior
of a tall building is converted to a one-dimensional problem.
Consider a distributed system with flexural stiffness EI, shear stiffness GA and mass per unit length
m defined over the closed domain 0 ≤ x ≤ b. The variable x is the spatial position of any point of system
from the base of the structure. b is the total height of the structure with lateral displacement denoted by
u(x,t). The internal moment and shear force can be explained by the displacement function as follows
(Paulino, 2010):

d2 u
M ¼ EI (2a)
dx2

du
V ¼ GA (2b)
dx

Equations. (2a) and (2b) represent the bending and shear behavior of an equivalent cantilever beam,
respectively. The corresponding distributed loads can be calculated using the following equations
(Meirovitch, 1980):

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal
594 R. RAHGOZAR, Z. MAHMOUDZADEH AND M. MALEKINEJAD

Figure 1. Combined system of framed tube and shear walls: a) actual structure, (b) flexural and shear
equivalent beams and (c) equivalent cantilever beam model.

d2 M d4 u
wb ¼ 2
¼ EI 4 (3a)
dx dx

dV d2 u
ws ¼  ¼ GA 2 (3b)
dx dx

in which wb and ws are distributed loads associated with bending and shear behaviors. On the basis of
D’Alembert’s principle (Meirovitch, 1980), considering inertia force as an external loading, the
dynamic equilibrium equation can be written as follows:

wI þ wb þ ws ¼ 0 (4)

where wI is the inertia force per unit length of the equivalent beam that can be calculated using the
following equation:

∂2 u
wI ¼ m (5)
∂t 2

Substituting Equations (3a), (3b) and (5) into Equation (4) yields the following (Heidebrecht and
Stafford Smith, 1973):
∂2 u ∂4 u ∂2 u
m þ EI 4  GA 2 ¼ 0 (6a)
∂t 2 ∂x ∂x
In addition, the boundary conditions for equivalent cantilever beam are as follows:

uð0; t Þ ¼ 0 (6b)

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal
 BUILDING ANALYSIS BY GALERKIN METHOD USING B-SPLINE FUNCTIONS 595

∂uð0; t Þ
¼0 (6c)
∂x


∂2 u
EI 2 ¼0 (6d)
∂x x¼b


∂3 u ∂u
EI 3  GA ¼0 (6e)
∂x ∂x x¼b

where b is beam’s length. Here, Equations (6b), (6c), (6d), (6e) are the boundary conditions that rep-
resent zero displacement and rotation at base, and zero moment and shear force at top of the structure,
respectively. Using separation of variables method and considering harmonic response, transverse
displacement, u(x,t), can be written as follows:
uðx; t Þ ¼ X ðξ Þsinωt (7)
in which ξ = x/b and ω is a natural frequency of the system. Substituting Equation (7) into differential
equation (Equation (6a)) and boundary conditions (Equations (6b), (6c), (6d), (6e)), the boundary
value problem is rewritten in terms of ξ as follows:

EI d4 X GA d2 X
  bmω2 X ¼ 0 (8a)
b3 dξ 4 b dξ 2

X ð0 Þ ¼ 0 (8b)

dX ð0Þ
¼0 (8c)
dξ

2 
d X ð1Þ
¼0 (8d)
dξ 2 ξ¼1


EI d3 X GA dX
 ¼0 (8e)
b3 dξ 3 b dξ ξ¼1

2.2. Solution method

Equation (8a) is a fourth-order ODE, which can be converted to integral form by multiplying a test
function denoted by W in Equation (8a) and integrating on a whole domain, i.e.

EI 1 d4 X GA 1 d2 X 1
∫ W dξ  ∫W dξbmω2 ∫0 WXdξ ¼ 0 (9)
b3 0 dξ 4 b 0 dξ 2
Integrating by parts results to the following:


 1
EI d3 X GAdX dW d2 X EI 1 d2 W d2 X
W 3 3  þ ∫ dξ
b dξ b dξ dξ dξ 2 0 b3 0 dξ 2 dξ 2
(10)
GA 1 dW dX 1
þ ∫0 dξ  mbω2 ∫0 WXdξ ¼ 0
b dξ dξ

On the basis of boundary conditions (Equations (8c), (8d), (8e)), the first expression is equal to zero. The X
and W functions are approximated using B-splines basis functions (Rostami et al., 2012, 2013), i.e.

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal
596 R. RAHGOZAR, Z. MAHMOUDZADEH AND M. MALEKINEJAD

X≈NX (11a)

W≈NW (11b)

in which N is the 1 × n matrix of B-spline basis functions. n is the number of B-spline basis used in the
approximation. X and W are the vector of displacement and test value of control points, respectively.
To satisfy displacement boundary conditions (Equations (8a) and (8b)), it is necessary to remove B-
splines with a non-zero value and slope at the base (ξ = 0) from the N matrix.

(
 )
dN i
N¼ ½N i jN i ðξ ¼ 0Þ ¼ 0 and ¼0 (12)
dξ ξ¼0

Matrix form of the problem is obtained by substituting Equations (11a) and (11b) into Equation (10):

 
K  ω2 M X ¼ 0 (13)

where K and M are the stiffness and mass matrices, respectively, which are defined as follows:

EI 1 ∂2 N T ∂2 N GA 1 ∂N T ∂N
K ¼ Kb þ Ks ¼ ∫  2 dξ þ ∫  dξ (14a)
b 3 0
∂ξ 2
∂ξ b 0 ∂ξ ∂ξ

1
M ¼ mb∫0 N T Ndξ (14b)

In Equation (14a), Kb and Ks are the stiffness matrices associated with bending and shear deformations,
respectively. The
 eigenvalue
 problem defined by Equation (13) has trivial and non-trivial solutions. The
trivial solution X ¼ 0 implies that the structure does not vibrate. The non-trivial solution was obtained
by setting the determinant of the coefficient matrix to zero, i.e. |K  ω2M| = 0, a solution that leads to nat-
ural frequencies and corresponding mode shapes.

3. NUMERICAL EXAMPLES

To demonstrate the accuracy and efficiency of the proposed approximate method, three numerical
examples are presented for free vibration analysis of symmetric buildings with framed tube and
shear walls, lateral load-resisting systems. High-rise reinforced concrete buildings, which are 70-story,
80-story and 90-story high, are analyzed using Galerkin method based on B-spline shape functions
(BSFs) and Lagrange Shape Functions (LSFs). Then, the results are compared with those obtained from
SAP2000 analysis (SAP, 2000). In all examples, basis functions of third order and 40 DOFs are imple-
mented for both BSF and LSF methods. Also, in the analyses using SAP2000 for 70-story, 80-story and
90-story tall buildings, 210, 240 and 270 DOFs are considered.
At first, two high-rise reinforced concrete buildings, which are 70-story and 80-story high, with
two different plans as shown in Figure 2, are considered here. Dimensions of columns and beams
are 0.9 m × 0.9 m in the 70-story building and 1 m × 1 m for the other one. Center-to-center spac-
ing of columns and beams is 3 m. Thickness of shear walls and slab floor are 0.35 m and 0.3 m,
respectively. Mass per unit volume of the material is assumed to be ρ = 2400 kg/m3, and Poisson
ratio is ϑ = 0.25.

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal
 BUILDING ANALYSIS BY GALERKIN METHOD USING B-SPLINE FUNCTIONS 597

Figure 2. Floor plans of the 70-story and 80-story framed tube buildings: (a) rectangular plan and (b)
octagon plan.

The framed tube is considered as a hollow box section (plan (a)) and hollow octagon section (plan
(b)) beam with flexural and shear behavior. Material and cross-sectional properties of the equivalent
beam are given in Table 1. Also, Shear walls are replaced by an equivalent shear-bending cantilever
beam with properties listed in Table 2. Finally, circular frequencies of the structure are calculated
and compared with results from FE analysis employing computer program SAP2000. Comparison
of results between the proposed method and FE method (FEM) are given in Tables 3 and 4.
It can be seen from Tables 3 and 4 that results from the proposed method are in good agreement with
the more detailed FE computations. For instance, the corresponding two first-mode shapes of the 70-
story structure with plan (a) in comparison with those obtained from SAP2000 are plotted in Figure 3.
Also, as shown in Figures (4–7), divergence in respect to increase in DOFs and order is investigated
for the 70-story tall building with rectangular plan. These graphs indicate that the proposed method
is more accurate and covers more quickly than LSF method. These advantages are due to high conti-
nuity of B-Spline functions throughout domain of solution.

Table 1. Properties of framed tube replaced by equivalent beam for the 70-story and 80-story shear-walled
tube buildings.
Number Ef Gf t If (m4) Aef ðm2 Þ mf (kg s2/m2)
of (GPa) (GPa) (m)
stories Plan (a) Plan (b) Plan (a) Plan (b) Plan (a) Plan (b)

70 20 2.03 0.27 13 080 5515 18.9 13.5 641 088 406 172
80 20 2.49 0.33 16 060 6707 23.33 16.5 692 160 442 652

Table 2. Equivalent properties of shear walls for the 70-story and 80-story shear-walled tube buildings.
Number Ew Gw Iw (m4) Aew ðm2 Þ mw
of (GPa) (GPa) (kg s2/
stories Plan (a) Plan (b) Plan (a) Plan (b) m2)
70 20 8 2.16 1.62 4.9 3.675 40 320
80 1.86 1.4 4.66 3.5

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal
598 R. RAHGOZAR, Z. MAHMOUDZADEH AND M. MALEKINEJAD

Table 3. First natural frequency of the 70-story and 80-story shear-walled tube buildings.
Number of stories First natural frequency error (%)
BSF LSF SAP BSF LSF
70 Plan (a) 1.104 0.842 1.097 0.7 23
Plan (b) 1.08 0.89 1.05 2.8 15
80 Plan (a) 0.995 0.786 0.99 0.5 20
Plan (b) 0.973 0.824 0.92 5.7 10

Table 4. Second natural frequency of the 70-story and 80-story shear-walled tube buildings.

Number of stories Second natural frequency error (%)

BSF LSF SAP BSF LSF
70 Plan (a) 4.2 2.53 3.75 12 32
Plan (b) 3.86 2.68 3.76 2.6 28
80 Plan (a) 3.66 2.36 3.45 6 31
Plan (b) 3.38 2.48 3.45 2 28

Figure 3. Comparison of mode shapes 70-story structure with plan (a): (a) first mode and (b) second mode.

Figure 4. Convergence of first natural frequency of 70-story tall building with rectangular plan (basis
functions of order 3).

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal
 BUILDING ANALYSIS BY GALERKIN METHOD USING B-SPLINE FUNCTIONS 599

Figure 5. Convergence second natural frequency of 70-story tall building with rectangular plan
(basis functions of order 3).

Figure 6. Convergence of first natural frequency of 70-story tall building with rectangular plan with
different orders of basis functions.

Figure 7. Convergence of second natural frequency of 70-story tall building with rectangular plan with
different orders of basis functions.

Furthermore, the accuracy of the proposed method is verified through another numerical example, a
90-story concrete reinforced tall building with combined system of framed tube and shear wall
(Figure 8). All beam and columns have same size of 1.1 m × 1.1 m in all stories. The thickness of
the shear wall is 40 cm. Other properties are same as those given in previous examples. The properties
of equivalent beam of framed tube and shear walls are listed in Tables 5 and 6.
Two first natural frequencies of the structure are calculated by both BSF and LSF methods and com-
pared with results obtained from SAP2000 free vibration analysis (Table 7).

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal
600 R. RAHGOZAR, Z. MAHMOUDZADEH AND M. MALEKINEJAD

Figure 8. Floor plan of the 90-story framed tube building with rectangular plan.

Table 5. Properties of framed tube replaced by equivalent beam for the 90-story shear-walled tube
buildings.
Number of stories Ef (Gpa) Gf (Gpa) t (m) If (m4) Aef ðm2 Þ mf (kg s2/m2)
90 2 × 109 2.99 × 108 0.403 19 339 27.6 748 608

Table 6. Equivalent properties of shear walls for the 90-story shear-walled tube buildings.
Number of Stories Ew (GPa) Gw (GPa) Iw (m4) Aew ðm2 Þ mw (kg s2/m2)
90 2 × 109 8 × 108 2.74 7.6 69 120

Table 7. First two natural frequencies of the 90-story shear-walled tube building.

First frequency Second frequency Error (%)

BSF LSF SAP BSF LSF SAP 1st frequency 2nd frequency
BSF LSF BSF LSF
0.969 0.811 0.967 3.402 2.44 3.44 0.2 16 1 29

The results for the 70-story, 80-story and 90-story tall buildings with shear-walled tube that are
listed in Tables 3, 4 and 7 show differences that are acceptable for the preliminary design stages, which
requires less time.

4. CONCLUSION

In this paper, free vibration analyses of doubly symmetric tall buildings consisting of framed tube and
shear walls have been studied. On the basis of the continuum approach and due to symmetry, the
whole structure is considered as an equivalent beam. By applying the Galerkin method and using B-
splines as shape functions, the governing differential equation of motion is solved to determine the

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal
 BUILDING ANALYSIS BY GALERKIN METHOD USING B-SPLINE FUNCTIONS 601

natural frequencies and associated mode shapes. To show the accuracy and simplicity of the proposed
method, numerical examples have been solved. Comparisons show good agreement between the
results obtained from the presented method and SAP2000. As seen from Tabs. 3, 4 and 7, errors are
less than 5.7% and 12% for the first and second modes, which is acceptable. Consequently, the
proposed method is not only simple, accurate, yet reliable enough for quick evaluations during the
preliminary design stages but also comprehensive to be used for tall buildings with various and also
complex structural forms.

APPENDIX A

In this section, equivalent geometric and elastic properties of framed tube and shear walls are obtained.
A framed tube structure as shown in Figure A.1(a) can be analyzed as a continuum, which is composed
of four equivalent orthotropic plate panels (Figure A.1(b)).
The method for evaluating the equivalent properties has been presented by Kwan (1994). To
calculate equivalent properties of the hollow box section cantilever beam, a subassemblage is extracted
from a framed tube by assuming that contraflexure points occur at mid-span of the beams and columns
as shown in Figure A.2.

EQUIVALENT MATERIAL PROPERTIES OF FRAMED TUBE

Under the action of vertical axial forces, load–deformation relationships for both the framed tube and
the equivalent membrane are equal if

Est ¼ E m Ac (A:1)

where E is equivalent elastic modulus of the membrane, t is thickness of the membrane; Em is elastic
modulus of the construction material, and Ac is sectional area of the column. It is normal practice to fix
the value of t such that the area of the membrane is equal to the sectional area of the column
(i.e. st = Ac) and so that the axial stress in the column and that in the membrane are equal. In such a case,

Ac
t¼ (A:2)
s

Figure A.1. Typical plane of framed tube structure: (a) actual plan, (b) equivalent hollow box section.

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal
602 R. RAHGOZAR, Z. MAHMOUDZADEH AND M. MALEKINEJAD

Figure A.2. Sub-assemblage extracted from framed tube.

and

E ¼ Em (A:3)

To obtain equivalent shear modulus for the membrane, consider the case where the frame unit is
subjected to a lateral force Q. Lateral deflection may be computed as the sum of that due to bending
Δb and due to shear Δs. The bending deflection Δb is given by

2
Δb ðh  d b Þ3 h ðs  d c Þ2
¼ þ (A:4)
Q 12E m I c s 12E m I b

where Ib and Ic are the moments of inertia for the beam and column, respectively. h, db and dc are the
height of each story and cross dimensions of the beams and columns, respectively. Also, the shear de-
flection Δs is

2
Δs h  db h s  dc
¼ þ (A:5)
Q Gm Asc s Gm Asb

in which Asb and Asc are effective shear areas of the beam and column, respectively, and Gm is the shear
modulus of the material. Equating the total lateral deflection of the frame unit and shear deflection of
the membrane leads to the following equation:

h
Q ¼ Δb þ Δs (A:6)
Gst

where G is the equivalent shear modulus of the membrane. In Eqs. (A.4), (A.5), (A.6), the value of G is
derived as follows:
,
st
G¼ h
Δb Δs
(A:7)
Q þ Q

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal
 BUILDING ANALYSIS BY GALERKIN METHOD USING B-SPLINE FUNCTIONS 603

EQUIVALENT CROSS-SECTIONAL PROPERTIES OF FRAMED TUBE AND SHEAR WALLS

The equivalent shear area and moment inertia for the framed tube are derived as follows:

5
Aef ¼ Af ; Af ¼ 2ðL  2t Þt (A:8)
6

1 1
If ¼ BL3  ðB  2t ÞðL  2t Þ3 (A:9)
12 12

in which Aef and If are the effective shear area and the moment for inertia of the hollow box section,
respectively. Similarly, the above parameters for shear walls are as follows:

5 X
Aew ¼ Aw ; Aw ¼ Awi (A:10)
6

X
Iw ¼ I wi (A:11)

where Aew and Iw are the effective shear area and the moment of inertia corresponding to the equivalent
beam modeling the shear walls. Awi and Iwi are the cross-sectional area and the moment of inertia for
each shear wall in direction of vibration.

APPENDIX B

A B-spline is a polynomial function defined on a non-decreasing set of numbers called knot vector. A
knot vector consists of start and end knots repeated p times and middle knots repeated maximum p  1
times (Rostami et al., 2012, 2013).

η ¼ f η1 … η1 η2 … ηn1 ηn … ηn g (B:1)

where η1 and ηn are the start and end knots, respectively. η2 … ηn  1 are the middle knots repeated once
in this case. The p-order (p  1 degree) B-splines are formed using the following recursive equation:


1 if ηi ≤η < ηiþ1
N i;1 ðηÞ ¼ (B:2:a)
0 otherwise

η  ηi ηiþj  η
N i;j ðηÞ ¼ N i;j1 ðηÞ þ N iþ1;j1 ðηÞ
ηiþjþ1  ηi ηiþj  ηiþ1
(B:2:b)
j ¼ 2; 3; …; p
i ¼ 1; 2; …; l  j

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal
604 R. RAHGOZAR, Z. MAHMOUDZADEH AND M. MALEKINEJAD

in which Ni,1 and Ni,j are ith B-spline with 1 and p order respectively. l is the length of knot vector.
Third-order B-splines defined on η ¼ f 0 0 0 0:5 1 1 1 g are shown in Figure B1.
B-spline functions are polynomial functions, and the values of these in the denominator of
Equation (B.2.b) are constant values. Therefore, the number of required Gauss points for integra-
tion in the proposed method is exactly same as those required for FEM. Indeed, on the basis of
Cottrell et al. (2009), p + 1 Gauss points are required if B-spline functions of order p are implemented
for integration.

B-SPLINES AND APPROXIMATION

One-dimension parametric space is a knot vector in which all are knots repeated once. B-splines are
defined on parametric space and can be used as basis functions in mapping the parametric space into
the physical or geometric space using the following linear combination:

X
lp
X¼ N i;p C i (B:3)
i¼1

where X and Ci are coordinate vectors of geometric space and the ith control point on geometric space,
respectively. Control points are similar to the nodes in FEM but with little difference. The position of
control points on geometric space depends on the form of geometric space as shown in Figure B2.
Although B-splines can model most of usual geometric spaces exactly, this is not true for solution
space. Assume that u is a function (solution space) defined on the geometric space that consists of a
straight line over the closed domain 0 ≤ ξ ≤ 1 (geometric space) as shown in Figure B3.
It is worthwhile to note that the geometric space can be modeled using only two control points
located at the start and end of the domain. However, because of the complex form of the solution, it
is necessary to use more control points. Each control point has a geometric value (coordinate) and
solution value. Solution space can be defined using the following linear combination:

X
lp
u¼ N i;p ui ¼ Nu (B:4)
i¼1

in which N is the matrix of p-order B-spline functions and u is the vector of solution coordinates
associated with control points.

Figure B.1. Third-order B-splines defined on η ¼ f 0 0 0 0:5 1 1 1 g..

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal
 BUILDING ANALYSIS BY GALERKIN METHOD USING B-SPLINE FUNCTIONS 605

Figure B.2. Two examples of geometric space and position of control points: (a) linear geometric
space and (b) nonlinear geometric space.

Figure B.3. Modeling geometric and solution space using B-splines.

REFERENCES
Bozdogan KB. 2012. Differential quadrature method for free vibration analysis of coupled shear walls. Structural Engineering
and Mechanics an International Journal 41(1): 67–81.
Bozdogan KB. 2011. Free vibration analysis of wall-frame structures by differential quadrature method. International Journal of
Engineering and Applied Sciences, IJEAS 3(4): 7–20.
Coull A, Bose B. 1975. Simplified analysis of frame-tube structures. Journal of Structural Engineering, ASCE 101(11): 2223–2240.
Cottrell JA, Hughes TJR, Buzilevs Y. 2009. Isogeometric Analysis: Toward Integration of CAD and FEA. John Wiley & Sons:
Chichester, United Kingdom.
Gu J, Zhang J, Sheng X, Li G. 2011. B-spline approximation in boundary face method for three-dimensional linear elasticity.
Engineering Analysis with Boundary Elements 35(11): 1159–1167.
Halis Gunel M, Emer Ilgin H. 2007. A proposal for the classification of structural systems of tall buildings. Journal of Building
and Environment, Elsevier, Science Direct 42: 2667–2675.
Heidebrecht AC, Stafford Smith B. 1973. Approximate analysis of tall wall-frame structures. Journal of Structural Division,
ASCE 99(2): 199–221.
Hopkins TR, Wait R. 1979. Some quadrature rules for Galerkin methods using B-spline basis functions. Computer Methods in
Applied Mechanics and Engineering 19(3): 401–416.
Kamgar R, Rahgozar R. 2013. A simple approximate method for free vibration analysis of framed tube structures. The Structural
Design of Tall and Special Buildings 22: 217–234.
Kaviani P, Rahgozar R, Saffari H. 2008. Approximate analysis of tall buildings using sandwich beam models with variable cross-
section. The Structural Design of Tall and Special Buildings 17: 401–418.
Khan FR, Amin NR. 1973. Analysis and design of frame tube structures for tall concrete buildings. Journal of Structural
Engineering, ASCE 51(3): 85–92.
Kuang JS, Ng SC. 2001. Dynamic coupling of asymmetric shear wall structures: an analytical solution. International Journal
Solids and Structures 38: 8723–8733.
Kwan AKH. 1994. Simple method for approximate analysis of framed tube structures. Journal of Structural Engineering, ASCE
120(4): 1221–1239.
Malekinejad M, Rahgozar R. 2011. Free vibration analysis of tall buildings with outrigger–belt truss system. Earthquakes and
Structures 2(1): 89–107.
Malekinejad M, Rahgozar R. 2012a. A simple analytic method for computing the natural frequencies and mode shapes of tall
buildings. Applied Mathematical Modelling Journal, Elsevier 36(8): 3419–3432.
Malekinejad M, Rahgozar R. 2012b. A closed form solution for free vibration analysis of tube-in-tube systems in tall buildings.
International Journal of Engineering 25(2): 107–114.
Malekinejad M, Rahgozar R. 2014. An analytical model for dynamic response analysis of tubular tall buildings. The Structural
Design of Tall and Special Buildings 23: 67–80.
Meftah SA,Tounsi A. 2008. Vibration characteristics of tall buildings braced by shear walls and thin-walled open-section
structures. The Structural Design of Tall and Special Buildings 17(1): 203–216.
Meirovitch L. 1980. Computational Methods in Structural Dynamics. Sijthoff & Noordhoff International Publishers: Alphen
aanden Rijn, The Netherlands.

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal
 606 R. RAHGOZAR, Z. MAHMOUDZADEH AND M. MALEKINEJAD

Mittal RC, Jain RK. 2011. B-splines methods with redefined basis functions for solving fourth order parabolic partial differential
equations. Applied Mathematics and Computation 217(23): 9741–9755.
Mohammadi R. 2014.Sextic B-spline collocation method for solving Euler–Bernoulli beam models. Applied Mathematics and
Computation 241: 151–166.
Paulino MR. 2010.Preliminary Design of Tall Buildings, MSc Thesis, Civil and Environmental Engineering Department,
Worcester Polytechnic Institute, Massachusetts, United States.
Rahgozar R, Malekinejad M, Jahanshahi MR. 2011. Free vibration analysis of coupled shear walls with axial force effects, The
IES Journal Part A: Civil & Structural Engineering 4(4): 224–231.
Rostami S, Shojaee S, Moeinadini A. 2012. A parabolic acceleration time integration method for structural dynamic using
B-Spline functions. Applied Mathematical Modelling, Elsevier 36(11): 5162–5182.
Rostami S, Shojaee S, Saffari H. 2013. An explicit time method for structural dynamics using cubic B-spline polynomial
functions. Scientia Iranica 20(1): 23–33.
SAP 2000. Advanced 14.2.2, Computers and Structures, Berkeley, California, USA.
Singh Y, Nagpal AK. 1994. Two stage consideration procedure for free vibration characteristics of framed-tube buildings. The
Structural Design of Tall and Special Buildings, John Wiley & Sons 3(1): 37–49.
Smith S, Coull A. 1996.Tall Building Structures, McGraw Hill Book Company, New York.
Taranath BS. 1988.Structural Analysis and Design of Tall Buildings. McGraw-Hill: New York.
Wang Q, Wang LY. 2005. Estimating periods of vibration of buildings with coupled shear walls. Journal of Structural Engineer-
ing, ASCE 131(12): 1931–1935.
Xiaoyun Y, Weiyin M. 2014. Mapped B-spline basis functions for shape design and isogeometric analysis over an arbitrary
parameterization. Computer Methods in Applied Mechanics and Engineering 269: 87–107.
Yanan L, Liang S, Yinghua L, Zhangzhi C. 2011.Multi-scale B-spline method for 2-D elastic problems. Applied Mathematical
Modelling 35(8): 3685–3697.

Copyright © 2014 John Wiley & Sons, Ltd. Struct. Design Tall Spec. Build. 24, 591–606 (2015)
DOI: 10.1002/tal

View publication stats