Torsional Analysis

THE STRUCTURAL DESIGN OF TALL AND SPECIAL BUILDINGS
Struct. Design Tall Spec. Build. 22, 126143 (2013)

Published online 2 December 2010 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/tal.665
Torsional analysis of multi-storey building structures

under horizontal load
K. A. Zalka*,
Department of Mechanics and Structures, Szent Istvn University, Budapest, Hungary
SUMMARY
New closed-form formulae are presented for the torsional analysis of asymmetrical multi-storey buildings
braced by moment-resisting (and/or braced) frames, (coupled) shear walls and cores. The analysis is based
on an analogy between the bending and torsion of structural systems. A closed-form solution is presented
for the rotation of the building. The torsional behaviour is defined by three distinctive phenomena: warping
torsion, Saint-Venant torsion and the interaction between the two basic modes. Accordingly, the formula
for the maximum rotation of the building consists of three parts: the warping rotation is characterized by
the warping stiffness of the bracing system, St Venant rotation is associated with the St Venant stiffness
of the building and the third part is responsible for the interaction. It is demonstrated that the interaction
between the warping and St Venant modes is always beneficial, as it reduces the rotation of the structure.
It is shown how the proposed formula for torsion can be used for the determination of the maximum deflection of multi-storey asymmetrical building structures. The results of a comprehensive accuracy analysis
demonstrate the validity of the method. A worked example is given to show the ease of use of the procedure. Copyright 2010 John Wiley & Sons, Ltd.
1. INTRODUCTION
The torsional analysis of multi-storey building structures braced by frameworks, (coupled) shear walls
and cores, subjected to lateral load represents a formidable task. Even the deflection analysis of a
single frameworka representative bracing elementleads to a complex problem. The main difficulty is caused by the fact that the framework develops both bending and shear deformations. This
problem was recognized as early as in the 1940s, when L. Chitty (1947) published an excellent paper
on parallel bars interconnected by crossbars. However, her solution was quite lengthy and fairly
complicated for structural engineers to use for design purposes. Several attempts were subsequently
made to produce simpler solutions (Csonka, 1950; Beck, 1956). In parallel with these developments,
considerable attention was also paid to systems of frameworks and shear walls (Rosman, 1960;
Despeyroux, 1972). When several frameworks and (coupled) shear walls are put together to create a
bracing system, the situation becomes even more difficult, even if the arrangement is symmetrical,
mainly because of the interaction among the elements of the bracing system. Simplified models were
also introduced to enable easier treatment of the complex problem, which, occasionally produced
methods of very limited range of application. Khan and Sbarounis (1964), for example, showed that
even for buildings whose primary lateral-resisting system consists of shear walls, the use of pure
flexural model was not appropriate. The combined flexural and shear deformation in frame buildings
was investigated by Blume (1968), who introduced a dimensionless parameter to monitor the sensitivity of the system to bending (or shear) deformation. Miranda (1999) developed an approximate
method to estimate the maximum lateral deformation demands in multi-storey buildings responding
primarily in the fundamental mode when subjected to earthquake ground motions. However, these
(and a number of other) investigations only centred on the two-dimensional (lateral) problem.
* Correspondence to: Karoly A. Zalka, Department of Mechanics and Structures, Szent Istvn University, Budapest,
Thkly t 74, Hungary, H-1146
E-mail: zalkak@btinternet.com
Copyright 2010 John Wiley & Sons, Ltd.
TORSIONAL ANALYSIS OF BUILDINGS
127
Because of the complexity of the torsional behaviour, not many authors deal with the problem.
Considerable efforts have been made regarding the torsional behaviour of individual structural elements (Council on Tall Buildings, 1978; Seaburg and Carter, 2003) but the global torsional behaviour
of whole structural systems is a less cultivated area. There are some excellent publications that offer
relatively simple solution for the global torsional problem (Council on Tall Buildings, 1978; Irwin,
1984; Coull and Wahab, 1993; Hoenderkamp, 1995; Nadjai and Johnson, 1998; Howson and Rafezy,
2002; Schueller, 1990) but they are either still too complicated or of limited applicability, and neither
of them is backed up with a comprehensive accuracy analysis.
To handle this three-dimensional problem in a simple way seems to be hopeless using conventional
tools. However, by relying on an analogy between bending and torsion, a relatively simple solution
can be produced. The aim of this paper is threefold: (a) to establish a new model for the analysis
using this analogy; (b) to produce a simple closed-form solution for the rotation of a building that
clearly shows the contribution of the different stiffness characteristics to the torsional resistance; and
(c) to show how the proposed method can be used for the determination of the maximum deflection
of multi-storey asymmetrical building structures.
Although large frameworks and even whole buildings are now routinely analysed using computer
packages, the proposed method may be useful from several aspects. It helps the structural engineer
to understand the complex three-dimensional behaviour, and thus enables the manipulation of the
stiffnesses and the location of the bracing units in such a way that optimum structural arrangement
is achieved. The proposed method may also prove to be useful at the preliminary design stage when
quick checks are needed with different structural arrangements. Its usefulness cannot be overemphasized for checking the results of a finite element (computer-based) analysis when the input
procedure involves tens of thousands of data and mishandling one datum may have catastrophic
consequences.
The continuum method will be used for the analysis, and it will be assumed that
(1)
(2)
(3)
(4)
The structures are at least four storeys high and the storey heights are identical.
The floor slabs of the building have great in-plane and small out-of-plane stiffness.
The structures are subjected to uniformly distributed lateral load.
The structures develop small deformations and their material is linearly elastic.
2. TORSIONAL BEHAVIOUR AND BASIC CHARACTERISTICS

As with thin-walled bars, multi-storey building structures react to torsion by utilizing their torsional
resistance. As with thin-walled bars, the torsional resistance of multi-storey buildings originates from
two sources. The warping stiffness is associated with the in-plane bending stiffness of the individual
bracing units, which is activated by their moment arm (perpendicular distance) measured from the
shear centre of the bracing system. This phenomenon is best demonstrated by the torsional behaviour
of a single I-column on a fixed base and with a free upper end, whose warping stiffness EI is calculated by multiplying the (in-plane) bending stiffness of its flanges and the square of the perpendicular distance of the flanges from the shear centre of the column (Figure 1(a)):
2
h
tb3 h 2
tb3h 2
EI = EI flange 2 = E
2=E
2
12 4
24
(1)
where E is the modulus of elasticity of the material of the column. Point O marks the shear centre of
the column and z-axis passing through the shear centre is the axis of rotation.
The St Venant torsional stiffness of the bracing system is associated with the in-plane shear stiffness of the bracing units, which is activated by its moment arm (perpendicular distance) measured
from the shear centre of the bracing system. For its demonstration and using the same I-column as
above, assume that the flanges are pierced with big openings of rectangular shape (they are in fact
frames). The St Venant torsional stiffness K is calculated by multiplying the shear stiffness of the

DOI: 10.1002/tal
128
K. A. ZALKA
t
O
a)
b)
Figure 1. Rotation of an I-column on a fixed base. (a) with solid flanges; and (b) with flanges
with openings.
flanges (i.e. the frames) and the square of the perpendicular distance of the flanges from the shear
centre of the column (Figure 1(b)):
2
h
Kh 2
K = 2 K =
2
2
(2)
where K is the shear stiffness of the flanges. It is easy to see that in building structures, the floor slabs
of the building (with their great in-plane stiffness) play the role of the web of the I-column in making
the bracing elements (the flanges) work together.
It is also clear that, apart from the distance of the bracing units from the shear centre of the building, the bending and shear stiffnesses of the individual bracing units play an important role, so these
characteristics are given first. The continuum model is applied where the stiffness characteristics are
considered continuously distributed over the height of the structure.
In the case of a framework (the most characteristic bracing unit) two types of bending stiffness are
considered. The local bending stiffness of the framework is the sum of the bending stiffnesses of the
columns:
n
EI c = E I c,i
(3)
i =1
where Ic,i is the second moment of area of the ith column, n is the number of columns and E is the
modulus of elasticity of the framework.
The global bending stiffness of a framework is defined by
n
EI g = E Ac,i ti2
(4)
i =1
where Ac,i is cross-sectional area of the ith column and ti is the distance of the ith column from the
centroid of the cross-sections.

DOI: 10.1002/tal
129
The shear stiffness of the framework (distributed over the height) is given as
K = K br = K b
Kc
Kc + Kb
(5)
The above shear stiffness basically depends on two terms. Term Kb is the stiffness of the beams
(distributed over the height):
n 1
K b = 2
i =1
6 EI b,i
li h
(6)
where Ib,i is the second moment of area of the ith beam, h is the storey height and li is the bay between
the ith and (i + 1)th columns. Term Kc represents the stiffness of the columns:
Kc =
12 EI c
h2
(7)
where Ic is the second moment of area of the columns as in Equation (3). Finally,
r=
Kc
Kc + Kb
(8)
in Equation (5) is a reduction factor.

To avoid some overrepresentation of the columns of the framework (they appear in the formulae
of both the local bending stiffness and the shear stiffness), the local bending stiffness should be
reduced by factor r (Hegedu s and Kollr, 1999). Accordingly, instead of Equation (3), the following
expression will be used as the local bending stiffness of the framework:
n
EI = rEI c = rE I c,i
(9)
i =1
Shear walls are frequently used for bracing purposes. In terms of the above three stiffness characteristics, only their EI value is of finite magnitude and it corresponds to the local bending stiffness of
a framework. A shear wall can be modelled for the analysis of a system consisting of shear walls and
frameworks as a special framework whose resistance against shear deflection and global bending
deflection is infinitely great. In this way, the analysis can be carried out in a relatively simple way as
all the bracing units are frameworks.
Coupled shear walls can also be considered special frameworks (with some difference), and the
characteristics given above for frameworks above can also be used (with some modification). The
determination of the local and global bending stiffnesses is identical to those of frameworks. Because
of their slightly different behaviour, instead of Equations (5) and (6), the following equation should
be used for the determination of the shear stiffness:
K * = K br* = K b
Kc
K c + K b
(10)
In Equation (10) Kb* is the modified beam stiffness as

n 1
K b =
1
6 EI b,1 ((li + si )2 + (li + si +1 )2 )
EI
li3h 1 + 12 2 b,i
li GAb,i
(11)
where G is the modulus of elasticity of shear of the beams, Ab,i is the cross-sectional area of the ith
beams, si, si+1 are the width of the ith and (i + 1)th wall section, li* is the distance between the ith and

DOI: 10.1002/tal
130
K. A. ZALKA
(i + 1)th wall sections, is a constant depending on the shape of the cross section of the beams ( =
1.2 for rectangular cross sections).
In-filled frameworks and frameworks with cross-bracing can also be part of a bracing system. When
their bending stiffness is calculated, their local bending stiffness is calculated directly as the sum
of the bending stiffnesses of the columns, with r = 1, i.e. EI = EIc. The calculation of the global
bending stiffness EIg is identical to that of the rigid frames, i.e. according to Equation (4). Their shear
stiffness K should be determined according to the different types of bracing. Ready-to-use formulae
are available in structural engineering monographs, e.g. in Stafford Smith and Coull (1991) and Zalka
(2000).
In addition to the stiffnesses of the bracing units, their distance from the shear centre is also needed.
The location of the shear centre is defined as the centre of stiffnesses of the bracing units.
The stiffness of each bracing unit is defined as the reciprocal of the top (in-plane) deflection of the
unit:
Sj =
1
yj (H )
(12)
With the stiffnesses of the units, the calculation of the location of the shear centre is carried out in
the co-ordinate system x y, whose origin lies in the upper left corner of the plan of the building
and whose axes are aligned with the sides of the building (Figure 2):
m
S x
j
xo =
S y
j
, yo =
(13)
where x j and y j are the perpendicular distance of the jth bracing unit from x and y and m is the number
of bracing units. For the calculation of the location of the shear centre, only the in-plane stiffness of
the frameworks is taken into account. Once the location of the shear centre is determined, coordinate
system x y has fulfilled its role and the new coordinate system x y is established with its origin
in the shear centre (Figure 2).
Detailed explanation of the phenomena discussed and the terms used in this section are to be found
in Zalka (2009).
w
x
2 EI2, EIg,2, K2
t2
tm
1
EI1 = EI
EIg,1 = EIg
K1 = K
yo
x
C
j
EIj
EIg,j
Kj
xc
m EIm, EIg,m, Km
xo = t1 = t
tj
xj
Figure 2. Plan arrangement of the bracing system.


DOI: 10.1002/tal
131
3. TORSIONAL ANALYSIS
Knowing the stiffness characteristics of the individual bracing units as well as their perpendicular
distance from the shear centre, it is now possible to carry out the torsional analysis of the bracing
system of the building. The torsional analysis is based on an analogy well known in the stress analysis
of thin-walled structures in bending and torsion (Vlasov, 1961; Kollbrunner and Basler, 1969).
According to the analogy, translations, bending moments and shear forces correspond to rotation,
warping moments and torsional moments, respectively. It follows from the analogy that the results of
the deflection analysis of a system of frameworks, (coupled) shear walls and cores can be used for
the torsional analysis if the characteristic stiffnesses of the deflection analysis are matched with
stiffnesses that characterize the torsional problem. The governing differential equation of the deflection of the bracing system (Zalka, 2009) assumes the form
y 2 y =
w az 2
1
EI 2
(14)
The solution to this differential equation is

y (z) =
w
wEI cosh ( H z ) + H sinh z
H 3 z z 4 wz 2
1
+
2 3
24
2 Ks
cosh H
E (I + I g ) 6
K s
(15)
where
w = wq
(16)
with w being the intensity of the horizontal load, q being an apportioner determining how much the
base unit takes of the total horizontal load, and
2 = a + b, s = 1+
a
b
(17)
a=
aj =
K
EIg
aj
j =2 c j
1+
f
aj
1+
j =2 b j
, b=
K
EI
f
f
aj
j =2 b j
1+
(18)
I j +1
K j +1
I g , j +1
, bj =
, cj =
Ij
Kj
I g, j
Maximum deflection develops at z = H:

ymax = y ( H ) =
wH 4
wH 2
wEI 1 + H sinh H
+
cosh H
8 E ( I + I g ) 2 Ks 2 K 2 s 3
(19)
In the above equations, stiffnesses EI, EIg and K are those of the base unit, and f is the number
of frameworks/coupled shear walls. The base unit is the bracing unit whose K/EI ratio is the greatest.
The bracing units are numbered in such a way that the base unit is always given number 1 and then,
for the sake of simplicity, the subscript 1 is dropped, hence the stiffness characteristics of the base
unit are EI = EI1, EIg = EIg,1 and K = K1 (Figure 2). Details of the derivation and full explanation of
the terms in the above equations are given in Zalka (2009).
All we have to do now is to identify the torsional characteristics that are analogous to the bending
(EI and EIg) and shear (K) stiffnesses with the deflection analysis.
Stiffness EI is the local bending stiffness with the deflection analysis. The corresponding stiffness
with the torsional analysis is the local warping stiffness:

DOI: 10.1002/tal
132
K. A. ZALKA
EI = EIt 2
(20)
where t is the perpendicular distance of the base unit from the shear centre (Figure 2).
Stiffness EIg is the global bending stiffness with the deflection analysis. The corresponding stiffness
with the torsional analysis is the global warping stiffness:
EI g = EI g t 2
(21)
Stiffness K is the shear stiffness with the deflection analysis. The corresponding stiffness with the
torsional analysis is the warping shear stiffness:
K = Kt 2
(22)
With the above analogous characteristics, the governing differential equation of torsion assumes
the form
m az 2
1
EI 2
2 =
(23)
The solution is given by
(z) =
m
mz 2
mEI cosh ( H z ) + H sinh z
H 3z z4
1
+
2 3
E ( I + I g ) 6
K s
cosh H
24
2 K s
(24)
Maximum rotation develops at z = H:
max = ( H ) =
mH 4
mH 2
mEI 1 + H sinh H
+
2 3
1
2
K s
cosh H
8 E ( I + I g ) 2 K s
(25)
Instead of the lateral load on the base unit (w) in Equations (14) and (19), Equations (23)(25)
contain the torsional moment m that the base unit takes of the total torsional moment. Its value is
determined as follows.
Each of the bracing units takes torsional moment according to their torsional stiffness. The torsional
stiffness of the jth unit is defined as
S , j = S j t 2j =
t 2j
yj (H )
(26)
where tj is the perpendicular distance of the jth bracing unit from the shear centre and yj(H) is the
(in-plane) top deflection of the jth unit. Thus, the torsional apportioner related to the base unit assumes
the form
q =
S
m
(27)
,j
where S is the torsional stiffness of the base unit

S =
t2
= t2S
y(H )
(28)
and m is the number of bracing units. The torsional moment the base unit takes is, therefore,
m = mt q = wxc q
(29)
where mt = wxc is the total torsional moment on the bracing system.


DOI: 10.1002/tal
133
Equivalents of coefficients and s in Equations (14), (15) and (19) also have to be established for
use in Equations (23)(25). Careful investigation of Equations (17) and (18) shows that if the torsional
equivalentsstiffness (moment arm)2are substituted for the stiffnesses in Equations (15) and
(19), the moment-arms drop out of the formulae. It follows that the coefficients defined by Equations
(17) and (18) remain unchanged and should be used for the torsional analysis as well.
The evaluation of Equations (24) and (25) using the rotational data of 126 bracing systems ranging
in height from 4 to 80 storeys (cf Validation of the continuum model later on) leads to the following
observations:
(1) The torsional behaviour of the building can be separated into three distinctive parts. The
bending stiffness of the individual bracing units (activated through rotation around the shear
centre) is associated with warping torsionfirst term in Equation (25). The shear stiffness of
the bracing units (activated through rotation around the shear centre) results in pure, St Venanttype torsionsecond term in Equation (25). Because of the different (bending-type and
shear-type) rotation shapes (Figure 1), there is an interaction between the two modes, defined
by the third term in Equation (25). Figure 3 shows the characteristic types of rotation of a
40-storey building braced by frameworks.
(2) The effect of interaction between the warping and St Venant modes is always beneficial as it
reduces the rotation of the structure.
(3) The effect of interaction significantly becomes smaller as the height of the structures increases.
(4) The effect of interaction is roughly constant over the height of the structure (Figure 3(c)).
To conclude the investigation of the torsional behaviour, some special cases will now be considered
as their analysis leads to extremely simple solutions in many practical cases.
Case A: The horizontal elements of the bracing system (including connecting beams in the frameworks and the floor slabs) have negligibly small bending stiffness.
This case is characterized by Kb 0 (for the frameworks). Consequently, the shear stiffness of the
system tends to zero (K 0), which leads to a 0 and b 0 and 0. The governing differential
Equation (23) simplifies to
mt
EI
(30)
and the solutions for the rotation and the top rotation assume the form
z/H
z/H
z/H
z/H
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
a)
b)
-
c)
d)
Figure 3. Typical rotation shapes. (a) warping; (b) St Venant; (c) interaction; and (d) combined.

DOI: 10.1002/tal
134
K. A. ZALKA
mt H 3 z z 4

24
EI 6
(31)
mt H 4
8 EI
(32)
and
max = ( H ) =
where EI is the local warping stiffnesses. This case is identified in Figure 3(a). The use of
Equation (32) should be considered when the shear stiffness of the bracing units are very small
and/or when the bracing system consists of shear walls only. It should be noted that in this case,
mt is the total torsional moment and I represents the sum of the warping stiffnesses of the shear
walls.
Case B: Bracing systems comprising multi-bay, low-rise frameworks tend to develop predominantly St Venant-type rotation and the effect of the warping stiffness becomes insignificant.
This case is characterised by a 0 and b and governing differential Equation (23) cannot
be used directly. However, after some rearrangement, the original derivation leads to
m
K
(33)
where K = Kbt2. This differential equation, together with the boundary conditions (0) = 0 and
(0) = 0, lead to the rotation and the top rotation as
mz 2
2 K
(34)
and
max = ( H ) =
mH 2
2 K
(35)
The characteristic rotation shape is shown in (Figure 3(b)). It is certainly worth considering the use
of Equation (35) when the building is relatively low and the bracing system only consists of (mainly
multi-bay) frameworks.
Case C: The structure is relatively very slender (with great height/width ratio). The structure develops predominantly (global) warping rotation. The second and third terms in Equations (24) and
(25) tend to be by orders of magnitude smaller than the first term and the solutions for the rotation
and the top rotation effectively become
m
H 3z z4

24
E ( I + I g ) 6
(36)
mH 4
8 E ( I + I g )
(37)
and
max = ( H ) =
This case is illustrated in Figure 3(a). It is interesting to note that both Case (A) and Case (C) are
characterized by warping-type rotation.

DOI: 10.1002/tal
135
4. STRUCTURES WITH VARYING STIFFNESS OVER THE HEIGHT

In many practical cases, the stiffness of the bracing system varies over the height (Figure 4(a)). The
structural designer likes to have a quick assessment of such situations as to the significance of the
effect of the variation of stiffness. A lower bound is readily obtained if the calculation is carried out
using the smallest value of stiffness (I1 at the top of the structure; Figure 4(a)). In a similar way, an
upper bound is easily available using the biggest value of stiffness (Io at the bottom of the structure;
Figure 4(a)). The difference between the two results shows if a more accurate calculation is
warranted.
For a more accurate calculation, a decision has to be made about the distribution of the stiffness
over the height (shown by dashed line in Figure 4(b)). The distribution can be given as
I z = I1
z
f
(38)
where m can assume 1, 2, 3 and 4.

Using the above law for the distribution of the stiffness, a reduction factor can be derived examining the behaviour of the column under a compressive force (Dinnik, 1929 and 1932; Timoshenko,
1968). The results of this derivation (values of reduction factor rI) are given in Figure 5 and in Table
1 as a function of I1/Io.
The question remains: which distribution should be used? It depends on the actual situation, but
according to the graphs in Figure 5, as a rule, the distribution of the stiffness has no great effect on
the situation. As the subject of the investigation in this paper is torsion, m = 2 is recommended as the
behaviour of the built-up column (Figure 4(c)) in bending is the closest to that of a bracing system
in torsion.
x
f
I1
I1
H
Iz
Io
Io
z
z
y
a)
b)
c)
Figure 4. Building with stiffness varying over the height.


DOI: 10.1002/tal
136
K. A. ZALKA
rI 1
0.9
0.8
m=1
0.7
m=4
m=2
0.6
0.5
m=3
0.4
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
I1
Io
Figure 5. Values for reduction factor rI.
xc
F
a) Symmetrical arrangement
b) Asymmetrical arrangement
Figure 6. (a) Symmetrical; and (b) asymmetrical arrangement.
Table 1. Reduction factor rI for the variation of stiffness.

I1/Io
rI
m=1
m=2
m=3
m=4
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.657
0.547
0.508
0.487
0.710
0.645
0.622
0.610
0.756
0.715
0.694
0.684
0.797
0.771
0.762
0.758
0.835
0.820
0.813
0.809
0.871
0.862
0.861
0.858
0.905
0.901
0.899
0.898
0.939
0.936
0.935
0.935
0.970
0.969
0.969
0.969
1.0
1.0
1.0
1.0
Once the reduction factor is obtained, all the stiffnesses should be modified according to
I = rI I 0
(39)
The stiffnesses modified in this manner should be used in Equations (20)(22) to obtain the torsional
characteristics that now take into account the variation of stiffness.
5. PRACTICAL APPLICATION
Multi-storey buildings under horizontal load never develop torsion only. When the bracing system of
the building is doubly symmetrical, the shear centre of the bracing system (O) and the centre of the
plan of the building (C) coincide (Figure 6(a)). Under horizontal loadrepresented by its resultant F

DOI: 10.1002/tal
137
vo
A
O
O
M=Fxc
xc
xmax
F
xmax
a)
b)
c)
Figure 7. Displacements. (a) v: maximum displacement; (b) vo: displacement due to force F; (c) v:
displacement due to torsional moment M.
in Figure 6the building develops lateral displacement and no rotation occurs. This case is fully
investigated in Zalka (2009).
When the building is asymmetrical, the shear centre of the bracing system and the centroid of the
plan of the building do not coincide (Figure 6(b)). The external load passing through the centroid
causes two things: lateral displacement in the plane of the load and rotation around the shear centre
(Figure 7(a)). The behaviour of the building is best analysed by transferring the load to the shear
centre. This procedure results in a horizontal load passing through the shear centre and a torsional
moment M = Fxc, where xc is the distance between the shear centre and the centroid (Figure 7). Force
F develops lateral displacements only (vo in Figure 7(b)) and torsional moment M develops rotation
() around the shear centre (Figure 7(c)), which causes additional displacement (v). At any given
location, the total displacement is obtained by adding up the two components:
v = vo + v
(40)
The maximum displacement of the building develops at the top at a corner of the plan of the building (point A in Figure 7) and, making use of the angle of rotation, is obtained from
vmax = v ( H ) = vo + xmax
(41)
where xmax is the distance of the corner point (where maximum deflection occurs) from the shear
centre. The first term (vo) on the right-hand side of Equation (40) can be obtained using Equation (19)
and the angle of rotation is determined by Equation (25).
6. VALIDATION OF THE CONTINUUM MODELACCURACY ANALYSIS

OF 126 SYSTEMS
It is essential for any respectable approximate method to examine its range of validity and accuracy.
The continuum model for the deflection analysis has already been thoroughly investigated and proved
to be a reliable one (Zalka, 2009): a comprehensive accuracy analysis involving 270 test structures
measured the accuracy of Equation (19), the very formulae on which the present proposed method is
based.
As for the proposed method itself, the results of another accuracy analysis is given in this section.
The aim of the accuracy analysis was to validate Equation (25) derived for the maximum rotation.
The results obtained using the approximate formula were compared to the results of the Finite Element
solution. The AXIS VM finite element package (Axis, 2003) was used for the comparison, whose
results were considered exact.
The top rotation of fourteen bracing systems (Figure 8) under uniformly distributed torsional load,
using 11 individual bracing units was calculated. The height of the structures varied between 4 and
80 storeys in eight steps (4, 10, 16, 22, 28, 34, 40, 60 and 80 storeys), creating 126 test cases. The

DOI: 10.1002/tal
138
K. A. ZALKA
W3
O
F5
12
F5
O F5
F5
F5
F5
F5
W3
24 m
a) F5+F5
b) F5+F5+F5
W2
c) F5+F5+W3+W3
F3
W2
O F5
F5
24 m
24 m
W2
12
F5
F5
F7
F3
O
W1
W1
F3
W2
F3
24 m
d) F5+F5+F5+W2+W2
e) F5+F7
f) F3+F3+F3+F3+W1+W1
F6
F6
W1
F7
6
12 m
g) F6+F7+W1+W4
W4
W4
24 m
F7
W1
W4
W4
W4
12 m
h) F6+W1+W4+W4
F7
W1
W1
i) W1+W1+W4+W4
6
F7
F7
F7
O F7
F7
F7
F7
6
412 = 48 m
312 = 36 m
j) F7+F7+F7+F7
k) F7+F7+F7+F7+F7
F6
F10
18
F5 F7
F7
F7
F7 F5
l) F6+F10
W4
56 = 30 m
m) F5+F7+F7+F7+F7+F5
F5
6
24 m
F1
W4
U
F1
24 m
n) F5+W4+W4+F1+F1+U
Figure 8. Bracing systems for the accuracy analysis.

DOI: 10.1002/tal
139
Table 2. Accuracy of the proposed method related to the maximum rotations of the 126 bracing systems
of frameworks and shear walls (Figure 8).
Method
Range of error
(%)
Average absolute
error (%)
Maximum error
(%)
0 to 25
25
Proposed method (Equation (25))
w
3 : W4
1 : F1
yo = t1 = t3 = 6 m
6m
C
4 :U
O4
2 : F5
xc = 1.4
6m
e = 1.714
5 : W4
6m
t5 = t6 = 6 m
6 : F1
6m
6m
xo = t2 = 13.4
h = 4.0
6m
t4 = 4.886
yo
Figure 9. Layout and geometrical characteristics for the worked example.
bays of the one-, two- and three-bay reinforced concrete rigid frames were 6 m and the storey height
was 3 m (F1, F3, F5, F6, F7 and F10 in Figure 8). The rectangular cross sections of the columns (in
metres) were 0.4 0.4 for F1, F3, F6 and F7; 0.4 0.7 for F5 and 0.4 2.0 for F10. The rectangular
cross sections of the beams (in metres) were 0.4 0.4 for F1, F5, F7 and F10; 0.4 1.0 for F3 and
0.4 0.7 for F6. The cross sections of the shear walls were 0.3 12.0 for W1; 0.2 2.0 for W2; 0.2
6.71 for W3 and 0.3 6.0 for W4. The wall sections for the U-core were h = 4.0 and b = 4.0 with
a wall thickness of t = 0.3. The modulus of elasticity for the structures was E = 25 kN/mm2.
The cross sections of the beams, columns and shear wall were chosen in such a way that the structures covered a wide range of stiffnesses. The torsional shapes represented predominant warping-type,
mixed warping-type and St Venant-type, and predominant St Venant-type deformation. The summary
of the results (range of error, average absolute error and maximum error) is given Table 2.
The results summarized in Table 2 demonstrate the performance of the proposed method. It should
be emphasized that the proposed method produced conservative estimates in every test case. The error
range of the method was between 0% and 25%. In the 126 cases, the average difference between the
results of the proposed analytical method and the finite element solution was around 9%.
7. WORKED EXAMPLE
Calculate the maximum deflection of the 28-storey building shown in Figure 9, subjected to a uniformly distributed lateral load of intensity w = 1.0 kN/m2 in direction y. The building is braced by
two one-bay frameworks (F1), one two-bay framework (F5), two shear walls (W4) and a U-core (U).
The FE computer analysis resulted in ymax = 404 mm, and this result is to be checked. The bracing

DOI: 10.1002/tal
140
K. A. ZALKA
Table 3. Stiffness and other characteristics of the bracing units for the worked example.
Bracing
unit
F1
F5
W4
U
K (MN)
r ()
Ic (m4)
EI
(MNm2)
EIg
(MNm2)
b
(1/m2)
y (m)
(Equation (19))
S (1/m)
(Equation (12))
28.444
66.947
0.8
0.941
1
1
0.004267
0.0343
5.4
11.245
85.333
807.29
135 000
281 125
72 000
504 000
0.333
0.008
4.92
1.457
1.106
0.531
0.203
0.686
0.904
1.882
units are numbered as shown in Figure 9. General data for the bracing units are given in the previous
section and the stiffness and other characteristics are summarized in Table 3.
The maximum deflection of the building consists of two parts. The deflection of the shear centre
is calculated first, then the additional deflection due to the rotation around the shear centre is added
(cf. Figure 7).
The deflection of the shear centre comes from a lateral deflection analysis when the horizontal load
is acting through the shear centre and causes deflection only. The horizontal load is resisted by framework F5 and core U. The two perpendicular frameworks (F1) and two shear walls (W1) have insignificant resistance in the direction of the external load and therefore they can safely be neglected.
Framework F5 is chosen as the base unit (as no other framework is available). In using the stiffnesses
of framework F5 and core U (second and last rows in Table 3), the load share and the load on framework F5 are
q=
S2
0.686
=
= 0.2671 w = qw = 0.2671 24 = 6.41 kN m
S2 + S4 0.686 + 1.882
Coefficients a, b, and s are calculated using Equations (17) and (18), bearing in mind that the
number of frameworks is now one:
a=
K
66.947
K
66.947
=
= 0.000133 b =
f =
1 = 0.08293
EI g 504000
EI
807.285
= a + b = 0.2882 s = 1 +
a
0.000133
= 1+
= 1.0016
b
0.08293
With the above coefficients, the top deflection of the shear centre is calculated from Equation (19):
vo =
6.41 84 4
6.41 842
+
6
8 5.04807 10
2 66947 1.00162
6.41 807290 1 + 0.2882 84 sinh ( 24.2)

1 = 0.079 + 0.337 0.027 = 0.389 m
cosh ( 24.2)
669472 1.00163
The location of the shear centre is needed for the rotation analysis. As the bracing system has an
axis of symmetry (x), one coordinate of the shear centre is readily available. The other coordinate is
obtained as the centroid of the stiffnesses, using Equation (13), so the shear centre coordinates are as
follows:
m
S x
i i
xo =
1
m
1.882 18.286
S4 x4
=
= 13.40 m, yo = 6.0 m
S4 + S2 1.882 + 0.686
Knowing the location of the shear centre, the (perpendicular) distances of the bracing units from
the shear centre (tj) and the arm of the load (xc) can be established as given in Figure 9.

DOI: 10.1002/tal
141
All the six bracing units participate in resisting torsion. Core U is the only three-dimensional item,
but because it lies on the symmetry axis, it can only utilize its second moment of area with respect
to axis x. Framework F1 is the base unit (as it has the greatest b = K/EI ratio). Its stiffness characteristics are calculated using Equations (20)(22):
EI = EIt 2 = 85333 62 = 3.072 106 kNm 4
EI g = EI g t 2 = 72 106 62 = 2592 106 kNm 4
K = Kt 2 = 28444 62 = 1.024 106 kNm 2
The torsional moment share on the base unitEquation (27)is calculated using the torsional
stiffnesses of the bracing units
S ,1 = S = t 2 S = 62 0.203 = 7.308 = S ,6 , S ,2 = 13.42 0.686 = 123.2
S ,3 = S ,5 = 62 0.904 = 32.544, S ,4 = 4.8862 1.882 = 44.929
as
q =
7.308
= 0..0295
7.308 + 123.2 + 32.544 + 44.929 + 32.544 + 9.308
,i
With apportioner q and the distance between the centroid and the shear centre xc, the torsional
moment share on the base unit is given by Equation (29):
m = wxc q = 24 1.4 0.0295 = 0.9912 kNm m
Coefficients a, b, and s for the calculation of the rotation are obtained using Equations (17) and
(18), bearing in mind that the number of frameworks is now three (F1, F5 and F1):
I2
I
K
a=
EI g 1 + I 2
I
1+
b=
I g I3 I g
0.03229 2.88
+
+1
1+
I g 2 I I g 3 28.444
0.003413 20.16
= 0.00022
=
K I3 K
72000 1 + 0.03229 28444 + 1
+
K2 I K3
0.003413 66947
K
f
28444
3
=
= 0.16613
.
I
K
I
K
0
03229
28444
EI 1 + 2
85333 1 +
+ 3
+1
I K2 I K3
0.003413 66947
= a + b = 0.16635 = 0.408 s = 1 +
a
0.00022
= 1+
= 1.0013
b
0.16613
With the above coefficients, the top rotation around the shear centre is calculated from Equation
(25):
max =
0.991 84 4
0.991 842
+
8 25.95 106 2 1.024 106 1.00132

0.991 3.072 106 1 + 0.408 84 sinh (34.27)
1 = 0.00238 + 0.00340 0.0001 = 0.00568
669472 1.00132
cosh (34.27)

DOI: 10.1002/tal
142
K. A. ZALKA
The maximum deflection of the building can now be obtained using Equation (41):
vmax = vo + xmax = 0.389 + 0.00568 13400 = 0.454 m
The exact solution using the FE package is vmax = 0.404 m.
8. CONCLUSIONS
An analogy between the bending and torsion of bars makes it possible to carry out a simple torsional
analysis of multi-storey building structures braced by frameworks, (coupled) shear walls and cores.
Application of the analogy leads to a closed form formula for the rotation of the building.
The torsional behaviour is defined by three distinctive phenomenon: warping torsion, St Venant
torsion and the interaction between the two modes. The separation of the basic (warping and St
Venant) modes enables the structural designer to identify the key contributors to the torsional resistance of the building and to achieve optimum structural performance. The interaction between the
warping and St Venant modes is always beneficial as it reduces the rotation of the building. Its effect
may be significant for low-rise structures but becomes rapidly negligible as the height of the structure
increases.
Using the formula for the rotation of the building, the maximum deflection of any asymmetric
multi-storey buildings can be readily determined.
REFERENCES
Axis VM. 2003. Finite Element Program for Structural Analysis. Version 7. Users Manual. Civilax, Inc.: Highlands Ranch,
CO.
Beck H. 1956. Ein neues Berechnungsverfahren fr gegliederte Scheiben, dargestellt am Beispiel des Vierendeltrgers. Der
Bauingenieur 31: 436443.
Blume JA. 1968. Dynamic characteristics of multi-story buildings. Journal of Structural Division, ASCE 94: 337402.
Chitty L. 1947. On the cantilever composed of a number of parallel beams interconnected by cross bars. Philosophical
Magazine, London Ser. 7, 685699.
Coull A, Wahab AFA. 1993. Lateral load distribution in asymmetrical tall building structures. Journal of Structural Engineering, ASCE 119: 10321047.
Council on Tall Buildings. 1978. Planning and Design of Tall Buildings, A Monograph in 5 Volumes. ASCE: New York.
Csonka P. 1950. Procedure for rectangular sway frames (in Hungarian). Mrnki Tovbbkpz Intzet. ptszet, 39.Tudomnyos Knyvkiad: Budapest.
Despeyroux J. 1972. Analyse Statique Et Dynamique Des Contraventments Par Consoles. Annales de lInstitut Technique du
Btiment et des Travaux Publics: No. 290.
Dinnik AN. 1929. Design of columns of varying cross section. Transactions ASME 51.
Dinnik AN. 1932. Design of columns of varying cross section. Transactions ASME 54.
Hegedus I, Kollr LP. 1999. Application of the sandwich theory in the stability analysis of structures. Structural Stability in
Engineering Practice. Kollr L. (ed.). E & FN Spon: London and New York. 187241.
Hoenderkamp HJCD. 1995. Approximate deflection analysis of non-symmetric high-rise structures. Proceedings of the Fifth
World Congress, L.S. Beedle(ed.). Council on Tall Buildings and Urban Habitat, Lehigh University: Bethlehem, PA. May
1419, 1995.
Howson WP, Rafezy B. 2002. Torsional analysis of asymmetric proportional building structures using substitute plane frames.
Proceedings of the 3rd Int. Conf. on Advances in Steel Structures. Hong Kong, Elsevier II: 11771184.
Irwin AW. 1984. Design of shear wall buildings. Construction Industry Research and Information Association, Report 102,
Construction Industry Research and Information Association: London.
Khan FR, Sbarounis JA. 1964. Interaction of shear walls and frames. Journal of Structural Division, ASCE 90(3): 285335.
Kollbrunner CF, Basler K. 1969. Torsion in Structures. Springer-Verlag: Berlin, New York.
Miranda E. 1999. Approximate lateral deformation demands in multistory buildings. Journal of Structural Engineering 125(4):
417425.
Nadjai A, Johnson D. 1998. Torsion in tall buildings by a discrete force method. The Structural Design of Tall and Special
Buildings 7(3): 217231.
Rosman R. 1960. Beitrag zur statischen Berechnung waagerecht belasteter Querwnde bei Hochbauten. Der Bauingenieur 4:
133141.
Schueller W. 1990. The Vertical Building Structure. Van Nostrand Reinhold: New York.
Seaburg PA, Carter CJ. 2003. Torsional analysis of structural steel members (Design Guide 9). American Institute of Steel
Construction: Chicago, Illinois.

DOI: 10.1002/tal
143
Stafford Smith B, Coull A. 1991. Tall Building Structures. Analysis and Design. John Wiley & Sons: New York.
Timoshenko S. 1968. Strength of Materials. Part II: Advanced Theory and Problems, Third edn. D. Van Nostrand Company,
Inc.: Princeton, NJ.
Vlasov VZ. 1961. Tonkostennye Uprugie Sterzhni. Moscow. 2nd edn. Thin-walled elastic beams. Israeli Program for Scientific
Translations: Jerusalem.
Zalka KA. 2000. Global Structural Analysis of Buildings. E & FN Spon: London.
Zalka KA. 2009. A simple method for the deflection analysis of tall wall-frame building structures under horizontal load. The
Structural Design of Tall and Special Buildings 18(3): 291311.

DOI: 10.1002/tal

Torsional Analysis

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Torsional Analysis

Uploaded by

Copyright:

Available Formats

THE STRUCTURAL DESIGN OF TALL AND SPECIAL BUILDINGS

Struct. Design Tall Spec. Build. 22, 126143 (2013)