13th World Conference on Earthquake Engineering

Vancouver, B.C., Canada

August 1-6, 2004
Paper No. 1465

ANALYSIS OF CRITERION FOR TORSIONAL IRREGULARITY

OF SEISMIC STRUCTURES

Nina ZHENG1 Zhihong YANG2 Cheng SHI1 Zhongren Chang1

SUMMARY

The torsion effects caused by irregularity of plan layout of seismic structures have been emphasized for
seismic design in some codes. When and how to consider torsion effects is depended on the criterion and
relative regulations for torsional irregularity. There are some differences between the criterions from one
code to another. In this paper, the differences between the codes of China, USA and Europe are studied.
Through analyzing series of structures with eccentricity in one and two directions, the corelation between
torsion effects and the criterion adopted by the different codes is obtained. According to the analysis
results, it can be indicated that torsion effects has no dependency relation with the criterion adopted by the
codes mentioned above and some regulations in the code are not reasonable.

INTRODUCTION

With the development of the theory of seismic design, the viewpoint is approved popularly that the torsion
effects resulting from the irregular layout of the structure should be considered. The criterion and some
relative regulations for torsional irregularity have been put forward in Code for Seismic Design of
Buildings of China (called GB50011-2001 for short)[1], United Building Code (version1997)(called
UBC97 for short) of USA [2], while the criterion for regularity of structures have been regulated in
Structural codes of Europe [3] (called EC8 for short). The rationality and practicability of these criterion
and relative regulations in these codes will be investigated in this paper.
In these three codes, for structures with torsional irregularities, the torsion effects should be considered in
two aspects, (1) The analysis models of structures should be spatial; (2) The modal analysis considering
torsion coupling should be adopted for seismic design. For the structures with extreme irregularities in
layouts of mass and stiffness, the torsion effects under seismic action in two directions should be
considered simultaneously. That is to say that the criterion for torsional irregularity of the seismic
structures can be further taken as the one for the irregular structures considering seismic action in two
directions simultaneously, thus the criterion for torsional irregularity becomes a keystone for the selection
of different design methods in the seismic design of the structures.

1
College of Civil Engineering, Chongqing University, Chongqing, 400045, China
2
Committee of Construction of Nan'an District, Chongqing, 400060, China
 COMPARISON AND ANALYSIS OF PROVISIONS ON TORSION IRREGULARITY
IN GB50011-2001, UBC97 AND EC8

In GB50011-2001, UBC97 and EC8, the criterion for torsional irregularity of the structures is
approximately in the same form as shown in Figure 1. Torsional irregularity should be considered when
maximum story drift (or inter-story drift) at one end of the structure transverse to an axis ( δ 2 ) is more than
1.2 times the average of the story drifts (or inter-story drifts) at the two ends of the structure .The criterion
is expressed in Equation (1).

Figure 1. Graphic of torsional irregularity in GB50011-2001 and UBC97

 δ1 + δ 2 
δ2   > 1 .2 (1)
 2 
δ +δ2 
θ = δ2  1  (2)
 2 
Where δ 1 and δ 2 are the story drifts (or inter-story drifts) at the two ends of the structure respectively. θ is
the parameter of the criterion.
Some differences between the terms of torsional irregularity are given as following,
1) δ 1 and δ 2 can be story drifts or inter-story drifts in GB50011-2001, and the maximum θ from the two
results of story drifts and inter-story drifts should be adopted, while these can only be inter-story drifts in
UBC97 and story drifts in EC8;
2) θ is calculated considering not only actual eccentricity, but also accidental eccentricity of ±5 percent of
the length of the structure in UBC97 and EC8, while in GB50011-2001 only actual eccentricity needs to
be considered; and
3) In the code of GB50011-2001, the maximum of θ is 1.5 while there is no similar provision in code of
UBC97 and EC8.
When the criterion are used in practice, some problems should be pointed out,
1) It is not convenient for engineers to determine θ at the beginning of the design. θ must be calculated
through the global analysis of structure;
2) The relative eccentricity is usually considered as the important factor influencing torsion effects, the
relationship between θ and relative eccentricity should be studied;
3) Story drifts (inter-story drifts) of different floors may be different. So θ may be different between
different floors. It is not clear that θ of which floor should be adopted in GB50011-2001, UBC97 and
EC8; and
4) If θ is a proper criterion for torisonal irregularity, whether it can further be taken as the criterion for
considering seismic action in two directions simultaneously should be studied.
Therefore, there are some unclearness and inconvenient in the criterion of torsional irregularity. The above
problems have been investigated by example analysis by Zheng [4] and the contents are extracted here.

EXAMPLE STRUCTURES ANALYSIS

Although the criterion for torsional irregularity looks simple, but the application of it has some confusions
and it is necessary to investigate the details of it and make it easy to use. For the factors influencing
torsion effects, many scholars approve the relative eccentricity is an important factor. The relationship
between θ and relative eccentricity and the relationship between torsional effects and θ are respectively
investigated by examples analysis of structures with eccentricity in one and two directions. Based on these
analyses, the applicability of the criterion will be evaluated, the superiority of story drift and inter-story
drift to calculate θ will be evaluated and the value of θ for different stories will be compared.

Method for analysis

The example structures are analyzed through modal α γ
 Tg 
analysis by SAP2000. The response spectrum is adopted η2α max α =   η2α max
T 
given by GB50011-2001 as shown in figure 2, here
α max is 0.16, Tg is 0.45s, η1 is 0.02, η 2 is 1 and γ is
0.45α max α = [η2 0.2γ − η1 (T − 5Tg )]α max
0.09. The spatial model and the postulate that the floor
is rigid are adopted. Three degree of freedom (two for T(s)
lateral translation and one for torsion rotate) for every
0 0.1 Tg 5Tg 6.0
story are considered and the first 15 mode shapes are
combined by the complete quadratic combination Figure 2. Seismic effect coefficient curve
method (CQC) for torsion rotate and lateral translation
coupling model.

Design of example structures

Two type of structures with torsional irregularity are designed, one with eccentricity in one direction and
the other with eccentricity in two directions. The relative eccentricity along X direction of structures with
eccentricities in two directions keeps invariable when the relative eccentricity along Y direction varies
from 0.0 to a value. To investigate the influence of plan size to the torsional effects, three kinds of plan
sketch that the ratios of long size and short size are respectively 1:1, 3:1 and 5:1 (for short called
correspondingly as Str.11, Str.31 and Str.51) are designed. All elevation layouts are regular. The structures
are 5 stories and the height of the ground floor is 4.6m and the above is 3.6m each. The cross section is
500×500mm for columns and 300×700mm for beams. The center of rigidness (CS) of floor is located on
the center of floor and the center of mass (CM) is variable with the variation of mass distribution, so the
total mass and total lateral resisting stiffness keep invariable. Here only the plan layout of Str.31 is shown
in figure 3, where ex ， ey are the eccentricity along x and y direction.
 C
Y
CM

ey
CS X
B
ex

1 2 3 4 5 6 7

Figure 3. Plan layout of Str.31

For describing clearly, defining such parameters as,

exr = ex / r (3)

er = ex 2 + ey 2 / r (4)

r= (a 2
)
+ b 2 12 (5)
Where exr is relative eccentricity for structures with eccentricity in X direction while er in two
directions, r is the radius of torsion rotate of the floor. a , b are the length of structure along X and Y
direction. The above information of all example structures are shown in table 1.

Table 1. Information of structures for analysis

Style of Structures with Structures with eccentricity in one

Number×span
structure eccentricity in X direction ~ two directions

Xdirection Ydirection exr Numbers 0.00～ exr ～ er Number

Str. 11 6×6 6×6 0.00~0.34 8 0.00~0.23~0.41 6+7

Str. 31 6×6 2×6 0.00~0.56 8 0.00~0.37~0.46 6+8

Str. 51 10×5.4 2×5.4 0.00~0.45 11 0.00~0.31~0.34 8+9

Results analysis

Relationship between θ and relative eccentricity

θ y ( θ x ) and θ yd ( θ xd ) are respectively calculated by story drifts and inter-story drifts under earthquake
action in Y(X) direction. The relationships between θ and e are shown in figure 4. (1), (3) and (5) of
figure 4 correspond to the structures with eccentricity in one direction, (2), (4) and (6) of figure 4
correspond to the structures with eccentricity in two directions. Based on figure 4, the following can be
obtained,
1) θ x and θ y are almost same between different stories. θ yd and θ xd are different from one story to
another, the θ of the top story is bigger than that of the other stories, but these differences are not
obvious. The differences between θ y and θ yd , θ x and θ xd are little, which differences is biggest in the top
story. Therefore, the difference of story drifts and inter-story drifts between GB50011-2001, UBC97 and
EC8 can be neglected for design.
2) For regular structure, θ x , θ xd , θ y and θ yd are all equal 1, as shown in (1), (3) and (5) of figure 4. When
exr varies from 0 to a small value, θ y ( θ yd ) varies dramatically from 1 to a value which exceeds the limit
value of torsional irregularity 1.2. For example, when θ y is 1.2, the corresponding exr of Str.11, 31 and
51 are all less than 0.05. Thus according to the criterion of torsional irregularity, torsion effects should be
considered for the structure with a very little eccentricity which less than 0.05. That is to say any structure
may be torsional irregular if accidental eccentricity of 5 percent the length of the structure is considered
according to UBC97 and EC8.

1F 3F 5F 1F 3F 5F
1.4 1.4
1.3 1.3
θyd
θy

1.2 1.2
1.1 1.1
exr exr
1 1
0 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4

(1) Str.11 with eccentricity in one direction

1F 3F 5F 1F 3F 5F
1.4 1.4
1.3 1.3
θyd
θy

1.2 1.2
1.1 1.1
er er
1 1
0 0.1 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3 0.4 0.5

1F 3F 5F 1F 3F 5F
1.15 1.15
1.1 1.1
θx

θxd

1.05 1.05
1 1
er er
0.95 0.95
0 0.1 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3 0.4 0.5

(2) Str.11 with eccentricity in one and two directions

Figure 4. Variations of θ with exr (er )
 1F 3F 5F 1F 3F 5F
1.4 1.4

1.3 1.3

θyd
θy

1.2 1.2

1.1 1.1
exr exr
1 1
0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6

(3) Str.31 with eccentricity in one direction

1F 3F 5F 1F 3F 5F
1.4 1.4
1.3 1.3

θyd
θy

1.2 1.2
1.1 1.1
er er
1 1
0 0.1 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3 0.4 0.5

1F 3F 5F 1F 3F 5F
1.1 1.1

1.05 1.05
θx

θxd

1 1
er er
0.95 0.95
0 0.1 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3 0.4 0.5

(4) Str.31 with eccentricity in one and two directions

1F 3F 5F 1F 3F 5F
1.4 1.4
1.3 1.3
θyd
θy

1.2 1.2
1.1 1.1
exr exr
1 1
0 0.1 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3 0.4 0.5

(5) Str.51 with eccentricity in one direction

Figure 4. Variations of θ with exr (er ) (continue)
 1F 3F 5F 1F 3F 5F
1.4 1.4
1.3 1.3

θyd
θy

1.2 1.2
1.1 1.1
er er
1 1
0 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4

1F 3F 5F 1F 3F 5F
1.15 1.15
1.1 1.1

θxd
θx

1.05 1.05
1 1
er er
0.95 0.95
0 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4

(6) Str.51 with eccentricity in one and two directions

Figure 4. Variations of θ with exr (er ) (continue)
3) After the quick increasing, θ y of Str.11, 31 and 51 decrease with the increasing of exr . It is
contradictory to the results that Wang [5] pointed out lager relative eccentricity, lager torsion effects. The
boundaries of declining points are not same for the three categories of structures, which may attribute to
the different sizes of the structures. The decrease ratio of θ y is largest for Str.51. θ y may be less than 1.2
for the structure with bigger eccentricity. According to the criterion, these structures are irregular and the
torsion effects may be ignored. But this is obviously impossible.
4) As shown in (2), (4) and (6) of figure 4, with ey increasing but ex is constant, θ y of Str.11 increases
firstly and then decreases even less than the one of structure with only ex along X direction. When θ y of
Str.31 and 51 increases to some extent and the values of them are more than those of structures with ex .
This is to say that one er may correspond to several θ y with the variations of the size and eccentricity of
the structures. The θ y of the structures with eccentricities along two directions has not good relationship
with er .
5) As (2), (4) and (6) in Fig. 4 shown, with eccentricity ey increasing but ex keeping constant, θ x
increases gradually. Contrasting with the phenomenon that big θ y corresponds to little exr of structures
with eccentricity in X direction, values of θ x here are all less than 1.1. The factors influencing θ x need be
investigated furthermore.
According to above analysis, the criterion for torsional irregularity has not close relations with relative
eccentricity exr and er . The parameter θ is not proper to represent the torsion effects of the structures
and defined as the criterion of torsional irregularity.

The relationship between λVX and θ y

Defining λVX as following
λVX = Vxy Vxx (6)
Where Vxy is the shear force in X direction under earthquake action in Y direction for structures with
eccentricity in X direction. Vxx is the shear force in X direction under earthquake action in X direction.
λVX denotes the torsion effects of structures with eccentricity in X direction.
For seismic design, Vxy should not be ignored when λVX increase to a prescribed value and the total shear
in X direction should be combined by Vxy and Vxx no matter the structure has eccentricity in one directions
or two directions, otherwise the design is not safety enough. Therefore λVX can be considered as the link
of torsional irregularity and considering earthquake action in two directions simultaneously.
Correlations of λVX and θ y of example structures with eccentricity in X direction are shown in Figure5.

1F 3F 5F 1F 3F 5F
0.5 0.5
0.4 0.4
0.3 0.3
λvx

λvx
0.2 0.2
0.1 0.1
θy θy
0 0
1 1.1 1.2 1.3 1 1.1 1.2 1.3 1.4

(1) Str.11 (2) Str.31

1F 3F 5F
0.3

0.2
λvx

0.1
θy
0
1 1.1 1.2 1.3 1.4

(3) Str.51
Figure 5. Variations of λVX and θ
It can be seen that with the increasing of θ y , λVX of different structures all decrease but the degrees of
decrease are not same. When θ y are in 1.3~1.4, the same θ y may corresponds to several λVX for Str.31
and 51. And λVX increases with the increasing of story for Str.31 and 51. In a word, the correlation of
torsion effects and the criterion for torsional irregularity is not definite. This criterion for torsional
irregularity is not proper to decide considering earthquake action in two directions simultaneously or not.

CONCLUSIONS

In this paper, the criterion and relative regulations for torsional irregularity in GB50011-2001, UBC97 and
EC8 are compared and analyzed from the theoretical and practical aspects. Through designing and
analyzing the series of example structures with eccentricity in one direction and two directions, a
elementary conclusion can be drawn that the corelationships between torsion effects and θ are not
definite, although the criterion about θ is adopted by GB50011-2001, UBC97 and EC8 for torsional
irregularity. This criterion is not proper to decide considering earthquake action in two directions
simultaneously or not.

REFERENCES

1. GB50011-2001, “Code for Seismic Design of Buildings.” (in Chinese) Beijing: China Architecture
& Building Press, 2001.
2. UBC97, “Uniform building code.” International Conference of Building Officials (ICBO). Vol.2.
Structural Engineering Design Provisions, USA, 1997.
3. Cheng Shaoge. “Structural Europe code (EC8).” (in Chinese) Institute of earthquake resistant of
china academy of building research, 1997
4. Zheng Nina. “Research on the scope of structures considering seismic action in two directions
simultaneously”. (in Chinese) Thesis for master degree, Chongqing University, Chongqing, China,
2003.
5. Wang Yaowei, Huang Zongmin. “ Analysis on factors influencing non-linear earthquake response
of structures with eccentricity”. (in Chinese) Journal of Chongqing Jianzhu University. 2001, 23(6):
114-120.