You are on page 1of 8

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

net/publication/235738293

Advanced techniques in tall building design

Conference Paper · September 2006

CITATIONS READS

0 508

6 authors, including:

Radu Pascu Mircea Degeratu


Technical University of Civil Engineering of Bucharest Technical University of Civil Engineering of Bucharest
24 PUBLICATIONS   32 CITATIONS    51 PUBLICATIONS   71 CITATIONS   

SEE PROFILE SEE PROFILE

Hasegan Liviu Valer Andrei-Mugur Georgescu


Technical University of Civil Engineering of Bucharest Technical University of Civil Engineering of Bucharest
11 PUBLICATIONS   35 CITATIONS    100 PUBLICATIONS   245 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

SEISMOCODE: Lifelong e-learning platform for active implementation of the new Romanian seismic regulations harmonized with European
standards View project

SEISMOCODE - Lifelong e-learning platform for active implementation of the new Romanian seismic regulations harmonized with European
standards View project

All content following this page was uploaded by Andrei-Mugur Georgescu on 23 May 2014.

The user has requested enhancement of the downloaded file.


Advanced techniques in tall buildings design
P. Ioan, Şt. Beţea, R. Pascu, M. Degeratu, L. Haşegan, A.M. Georgescu

Technical University of Civil Engineering Bucharest

ABSTRACT: This paper presents the approach used for the design of a 106 m tall office building, which
will be the tallest in Bucharest. The response at severe seismic actions implies the development of a favorable
energy dissipation mechanism and this can be checked only by nonlinear dynamic analysis. The behavior at
wind actions is more complex than for ordinary buildings and comprises a dynamic response. Conventional
(code) analysis was considered not enough accurate and wind tunnel tests were performed, on static and elas-
tic models in the wind tunnel facility of the Technical University of Civil Engineering. The 3D dynamic
nonlinear analysis and the wind tunnel tests allowed a safer design of the building structure.

1 INTRODUCTION last level is at +106.30 m. The basement is a rein-


forced concrete structure, 54.09 by 26.10 m. The
Steel structures are frequently used for tall multi- stiffness of the basement at horizontal actions is
story buildings because their advantages compared given by the perimeter wall which is 55 cm thick
to concrete structures: less self weight and greater and the interior walls which are 60 cm thick. The
erection speed. During the last years, their use is superstructure is made of steel. The first 4 stories are
becoming more and more frequent in Romania, after 44.60 m by 26.10 m. The first story is 5.40 m high,
more than 60 years. while the levels 2 to 4 are 4.20 m high. Stories 5 to
In the same time, they arise some problems spe- 23 are 41.95 long by 21.40 m wide and 4.00 m high.
cific to tall buildings: the need to ensure an appro- The last 3 penthouse stories are also 4 m high, but
priate behavior at wind and seismic actions. only 30.90x9.65 m in plane.
The response at severe seismic actions implies the The columns and the braces have hollow rectan-
development of a favorable energy dissipation gular sections made of welded plates while beams
mechanism and this can be checked only by nonlin- are made of europrofiles (HE-A sections for main
ear dynamic analysis. beams and IPE sections for purlins). Column–beam
The behavior at wind actions is more complex and beam-to-beam joints are made with high-
than for ordinary buildings and comprises a dynamic strength bolts grade 10.9 pretensioned at 50% of the
response. Conventional (code) analysis is not torque recommended in C133/82 norm. Pretension-
enough accurate and wind tunnel tests have to be ing is necessary to eliminate the danger of fragile
performed. failure of the high-strength bolts. Joints between
At the Technical University of Civil Engineering column elements are performed on site using butt
(TUCEB), part of the authors developed software, welds on a support plate. Brace joints are made with
based on ANSR computer code (Mondkar & Powell high-strength bolts grade 10.9, and are designed to
1975), for 3D dynamic nonlinear analysis of steel take only axial forces (hinged connections).
structures. At the same university there is also a The horizontal forces (from wind or earthquake)
wind tunnel facility, which can be used for the as- are taken basically by steel shear walls and X braced
sessment of buildings behavior under various wind frames. Additionally, rigid frames on two orthogonal
directions complete this system for the horizontal
conditions.
forces and take all the vertical loads.
The ductility of the structure is ensured at the
global level by the dual system consisting of shear
2 DESCRIPTION OF THE STRUCTURE
walls and braced frames and special moment frames,
and at the local level by the class of the sections for
The building has 4 underground stories and 26 sto-
the zones and bars which can undergo plastic defor-
ries above ground. The mat is at –11.60 m., and the
mations and the grade of the steel.
Figure 1. 3D view and current level plan

Steel shear walls are used for the first time in a standards, as follows : wind loads (STAS 10101/20
high-rise building in Romania. They were chosen – 90) for zone "C" with base pressure gv = 0.55
for their good stiffness and ductility properties kN/m2, soil roughness type III, building class C2 ;
(Astaneh-Asl & Zhao 2000, Astaneh-Asl 2001, Li- snow loads (STAS 10101/21 – 92) – zone "C" with
ang et al. 2004). basic snow load gz = 1.50 kN/m2.
The position of the steel shear walls and of the Seismic actions were taking according to P100-92
braced frames is given in Figure1, respectively code. The structure was also verified at seismic ac-
with plain and dashed thick lines. Their position tions according to the new seismic design code
was established in order to give enough stiffness P100/2005.
on every direction (including torsional stiffness), Thus, according to P100-92 code, seismic loads
but taking also into account the functional and ar- are defined by the following parameters: ks =0.20;
chitectural constraints. Tc = 1.5 s.; importance class II, α = 1,2.
Beam-column joints are rigid and use an end According to P100-2005 code, seismic loads are
plate, a “chair” to take shear (dissociation of the defined by the following parameters: ks =0.24, Tc =
stress state), while the bending moment is taken by 1.6 s.; importance class II, α = 1,2.
high-strength pretensioned bolts. This type of joint
leads to a small number of bolts, stresses in only
one direction of the joint and a reduction of stress
concentration in the maximum stressed part of the
joint. Both principal and secondary beams are pro- 3.2 Dynamic characteristics of the structure
vided with Φ19 studs at 15 cm distance, which en-
sures composite action with the concrete deck. The
deck is a 12 cm grade 35 concrete slab poured on The eigenfrequencies and the modal mass partici-
trapeze sheet type Hösch T40.1-0.88. pation ratios give a global view of the dynamic be-
havior of the structure, which is of first importance
under seismic or wind actions.
3 RESULTS OF THE ELASTIC ANALYSIS The first 9 vibration periods and the corre-
sponding mass participation factors are given in
3.1 Loads
table 1. It can be seen that the first two vibration
periods correspond to translation vibrations and are
As the building is placed in Bucharest, wind and very close, while the 3rd corresponds to a torsional
snow loads were taken according to Romanian vibration.
Table 1 – Vibration periods and mass participation factors
Period UX UY UZ SumUX SumUY SumUZ RX RY RZ
Mode
sec % % % % % % % % %
1 2.29 0 47 0 0 47 0 96 0 0
2 2.02 47 0 0 47 47 0 0 83 2
3 1.36 3 0 0 50 47 0 0 6 41
4 0.64 1 12 0 51 60 0 2 0 0
5 0.62 10 1 0 61 61 0 0 1 0
6 0.43 0 0 0 61 61 0 0 0 11
7 0.35 0 0 0 61 61 0 0 0 0
8 0.33 5 0 0 66 61 0 0 0 0
9 0.32 0 5 0 66 66 0 1 0 0

3.3 Results of the seismic analysis 3.4 Results of the wind analysis
Maximum horizontal displacement at the top story
under the design wind are, in the longitudinal di-
Story drift rection ∆x,max = 44.8 mm = H/2360 and in the
transversal direction ∆y,max = 157.4 mm = H/674.
In both directions the maximum displacement is
0 0.005 0.01 0.015
bellow the accepted limit of H/500, where H is the
ROOF height of the building.
MECH FLOOR 2 A preliminary verification of the sensibility of
MECH FLOOR 1
MAIN ROOF the structure according to the 1990 Canadian Code
ST22 (Bungale 1998) is given by Equation (1):
ST21
20TH W ⋅D 21.40 ⋅ 41.95 1
19TH = = 0.282 < (1)
18TH H 106.30 3
17TH
16TH where H is the height of the building and W and D
15TH the dimensions in plan.
Story

14TH
13TH As the aspect ratio is les the 1/3, the building
12TH should by checked for the dynamic effects of the
11TH
10TH
wind, especially at the maximum accelerations in-
9TH duced by the wind at the last levels of the building.
8TH The check was performed according to the Cana-
7TH
long.
6TH
dian Code (Bungale 1998). The maximum accel-
transv. 5TH eration normal to the wind direction at the last of-
4TH fice level is given by Equation (2):
3TH
2TH  ar 
1TH
a w = n w2 g p WD   = 0.355 m (2)
ρ g β  s2
 B w 

Figure 2. Envelope of story drifts This is 20% higher than the maximum allowed a
value which is 3%g, where g is gravity accelera-
tion.
The maximum story drifts are, for the design
earthquake, in the longitudinal direction
max(∆x/He) = 0.0107 (at 12th story) and in the 4 RESULTS OF THE NONLINEAR DYNAMIC
transversal direction max(∆y /He) = 0.01084 (at ANALYSIS
16th story), which are less then the limit allowed by
the design code P 100-1/2004, which is 0.02 at the The nonlinear dynamic analysis was performed
ULS. with ANSRX computer code, which is based on
At the SLS, the computed values are 0.00535 ANSR I (Mondkar & Powell 1975). This is a finite
and 0.00542, in the longitudinal and the transversal element code for 3D analysis of structures under
direction respectively and the code limit is 0.008. seismic actions represented by ground acceleration
histories.
The finite elements library of the program in- The maximum top displacement recorded in the
cludes several types of linear elements suited for transversal direction during the time-history analy-
the nonlinear analysis of steel structures: beams sis is 0.67 m (Fig. 3), compared to the equivalent
with interaction curves M – N and M – Q, columns inelastic displacement of 0.75 m obtained from the
wit interaction surfaces Mx – My – N and braces linear analysis, with ag = 0.2g.
with post-critical behavior (Ioan et al. 2005).

Top displacement history Drift history


0.8 0.01

0.008 15th story


0.6
23rd story
0.4 0.006 9th story
0.004
0.2
Disp. (m)

0.002

Drift
0
0 5 10 15 20 0
-0.2 0 5 10 15 20
-0.002
-0.4
-0.004
transv.
-0.6 long. -0.006
-0.8 -0.008
Time (sec) Time (sec)

Figure 3. Top displacement history Figure 4. Time-history of the drift - transversal direction

The code is written in FORTRAN 77. The version


used presently at TUCEB is completed with some
new finite elements, like the shear panel element The maximum drift value in the transversal di-
(Ioan et al. 2006), and modified in order to allow rection (Figure 4) is 0.00789, (recorded at the 15th
the use NTABS code (which is developed at level), compared to the value of 0.00903 obtained
TUCEB) as a pre and post processor. from elastic analysis with ag = 0.2g. Drift values
The model described in section 2 was submitted along the height of the structure are of the same
to the N-S component of the accelerogram re- magnitude from the 9th to the 23rd story, and their
corded at INCERC Bucureşti during the Vrancea distribution is similar to the elastic one (see Fig. 2).
1977 earthquake. The main results of this analysis
are presented in the following.

T C I S e is m O y m o d e l p o s t - c r it ic T C I S e is m O y m o d e l p o s t - c r it ic
t im e = 7 .6 0 0 y ie ld e d lin e s = 1 3 9 t im e = 8 .5 5 0 y ie ld e d lin e s = 1 8 5
y ie ld e d q u a d s = 6 y ie ld e d q u a d s = 1 0

E E
D 9 9
8 D 8
C' 7 7
C 5 '6 C' 5 '6
5 C 5
z B 4 4' z B 4 4'
AA' 3 AA' 3
y 1 2 y 1 2
x x

Figure 5. Plastic hinges pattern at the time moments when the base shear force is at maximum (in positive and negative direction)
Both displacements and drifts are inferior to model. The pressure heads were linked, on head
those obtained in the elastic analysis. groups, at a 48 gates scanivalve. The pressures
Plastic hinges develop in beams on the whole stored by the pressure heads and passed by the
height of the structure and there are also some scanivalve unit are successively transmitted to a
shear panels buckling (Fig. 5). differential pressure transducer.
The greatest plastic deformations were recorded
in the “coupling beams” placed between the frame
with shear panels and the bracing in axis 2. How-
ever, the maximum value, recorded at the 12th
floor, is only ϕp,max/ ϕy = 4.93, which is inferior to
the limit value of 7 recommended in the FEMA
Guide (FEMA 1997) for the “Life safety” per-
formance level.

5 BOUNDARY LAYER WIND TUNNEL


MODELLING

Two types of tests were performed: i) rigid model


tests (involving the measurement of mean pressure
distribution on the building); ii) aeroelastic model
tests (involving measurements of accelerations,
amplitudes and frequencies induced by the wind on
the structure).
Both types of tests were conducted in the Aerody-
namics and Wind Engineering Laboratory of the
Technical University of Civil Engineering Bucha-
rest using the boundary layer wind tunnel
(T.A.S.L. 1) facility. The tunnel assures the veloc-
ity profile and turbulent properties of wind to be
modeled correctly. For the purpose of this research
the coefficient of the power law velocity profile
obtained after the calibration of the tunnel was Figure 6. Rigid model of the building 1:100
0.23, value that corresponds to the suburban areas
where the building will be erected. The turbulent
intensity varied between 0.02 and 0.14, values that The measurements performed in the wind tun-
are in good agreement with the natural longitudinal nel on the rigid model of the building (Fig. 6),
turbulence of the wind. gave the local pressure distribution pi on the lateral
surface of the model. Based on these values of the
pressure pi, with the static pressure from the tunnel
5.1 Rigid model tests p∞ and with the airflow velocity U∞, the local pres-
The research followed basically the determina- sure coefficient values Cpi, corresponding to the
tion of the local pressure coefficient distribution Cp pressure heads, were determined for the considered
on the lateral surface of the building model. The straight eight directions of the wind. Integrating
aerodynamic force coefficients Cx and Cy, and the the values Cp on the lateral surface of the model,
aerodynamic momentum coefficients CMx, CMy and the aerodynamic force and momentum coefficients
CMz were also determined, based on the measured values CF and CM were obtained:
local pressure coefficients, and allowed the calcu-
lation of the aerodynamic forces and momentum at Fx
natural scale. Cx = (3)
ρ U ∞2
The rigid model of the building was made from Ac
transparent plastic material, at a scale of 1:100; it 2
was equipped with 114 pressure heads placed on
the lateral surface and was fixed in the experimen- Fy
Cy = (4)
tal vein of the aerodynamic tunnel T.A.S.L.1, on a ρ U ∞2
bed-plate which assures the positioning at various Ac
2
incidence angles between the airflow and the
Mx steel bar, with added discrete masses and with a
CM x = 2
(5) discontinuous lateral side surface, assures also the
ρU ∞
Ac ⋅ l c the similitude of the first 3 eigenfrequencies of the
2 structure and the corresponding mass repartition.
My
CM y = (6)
ρ U ∞2
Ac ⋅ l c
2
Mz
CM z = 2
(7)
ρU ∞
Ac ⋅ l c
2
with Fx and Fy – the force components of the force
and momentum matrix parallel to the Ox and Oy
axes; Mx, Mz şi Mz – the momentum components of
the force and momentum matrix round the Ox, Oy
and Oz axes; ρ - the air density; U∞ - the reference
airflow velocity upstream the model; Ac – the char-
acteristic area of the model; lc – the characteristic
length. As an example, the variation of the coeffi-
cients CMx and CMy with the positioning of the
model is presented in Figure 7.

Figure 8. Aeroelastic model of the building 1:100

Figure 7. The variation of aerodynamics momentum coeffi-


cients CMx, CMy and CMz with the positioning model
Figure 9. Dynamic amplitude variation with the incidence
angle of the wind

6 AEROELASTIC MODEL TESTS

The research followed the determination of the dy-


namic answer of the aeroelastic model of the build-
ing at the wind simulated action on the experimen-
tal vein of the T.A.S.L. 1. Though, we intended to
obtain the acceleration values, the velocity values
and the frequencies corresponding to the dynamic
answer of the aeroelastic model. Figure 10. Acceleration variation with the incidence angle of
The aeroelastic model 1:100 of the building (Fig. the wind
8) is a complete model of the building which mod-
els, at scale, the mass, stiffness and geometry of
the real building. The equivalent structure consists The aeroelastic model of the building is positioned
in a vertical steel bar, with an appropriate section in the experimental vein of the T.A.S.L. 1, being
which was calculated in order to assure the elastic fixed, by the steel bar, at the ground level of the
stiffness required by the similitude condition. That aerodynamic tunnel. For the determination of the
vibrations of the aeroelastic model of the building, per stories, due of the superior vibration modes ef-
a specialized equipment for vibration measurement fect.
was used: a vibration measurement unit and two The wind tunnel tests were performed in the
vibration transducers (acceleration piezoelectric wind tunnel test facility at TUCEB. The dynamic
transducers) placed at the upper side of the build- response of the building was considered within ac-
ing on the steel bar, with the two axes perpendicu- ceptable limits. However, the tests showed athe
lar one by the other allowing the measurement on sensibility of the building in the transverse direc-
both Ox and Oy model axes). tion, where both displacements and accelerations
The tests where done at a speed of the airflow are quite high.
which conducted to a Reynolds number attached to This study shows that for tall and slender multi-
the flow Re = 7.5·105, that assures the existence of story buildings, conventional elastic code design
the sub inertial domain of the longitudinal velocity needs to be supplemented with a deeper insight in
spectrum. Measurements where performed for dif- the dynamic and nonlinear behavior of the struc-
ferent incidence angles θ between the airflow and ture, in order to assure a good safety level.
the model. The main characteristics of dynamic re- The authors developed at the TUCEB the nu-
sponse of the aeroelastic model for x and y direc- merical and experimental tools needed for such
tions were determined: accelerations, frequencies studies.
and dynamic amplitudes. In Figure 9 and 10 are
presented the variations with the wind direction of
the dynamic amplitudes and accelerations along REFERENCES
the x and y directions of the model, at the top of
the building. Astaneh-Asl, A. & Zhao, Q. 2000. Cyclic Tests of Steel Plate
The maximum measured amplitudes are about Shear Walls. Research Report to Sponsor. Department of
Civil and Environmental Engineering., Univ. of Califor-
H/400 which is more then the limit of H/500, but nia, Berkeley.
they correspond to a wind blowing from the SE di- Astaneh-Asl, A. 2001. Seismic Behavior and Design of Steel
rection, which is not likely to be the direction of Shear Walls. SEANOC Seminar, San Francisco.
the strong winds in Bucharest. Bungale, S.T. 1998. Steel, Concrete & Composite Design of
The maximum accelerations are measured in the Tall Buildings. New York: McGraw-Hill.
transversal direction of the building (y direction), Davenport, A.G. 1968. The application of the boundary layer
wind tunnel to the prediction of wind loading. Proc. of In-
when the wind blows from the SE, S or W direc- ternational Research Seminar of Wind Effect on Build-
tion, but they are not greater than 3%g. ings and Structures, Toronto.
The wind tunnel study shows that the building Degeratu, M. 2002. Athmospheric limit layer (in Romanian),
is sensible to the dynamic effect of wind, but the Timişoara: Orizonturi Universitare.
response is within acceptable limits. Federal Emergency Management Agency (FEMA) 1997.
NEHRP Guidelines for the Seismic Rehabilitation of
Buildings (FEMA Publication 273). Washington D.C.
Liang, Q.Q., Uy, B., Write, H., Bradford, M. 2004. Local
7 CONCLUSIONS Buckling of Steel Plates in Double Skin Composite Pan-
els under Biaxial Compression and Shear. Journal of
For the design of a tall multistory steel building in Struct. Engr. 130(3).
Bucharest two special studies where performed: a Iamandi, C., Degeratu, M., Georgescu, A. 1996. A wind tun-
nel study for a TV tower in Bucharest – Romania. Inter-
3D nonlinear dynamic analysis for the seismic ac- national Conference on Urban Engineering in Asian Cit-
tions and wind tunnel tests. ies. Bangkok, Thailand.
The Seismic nonlinear analysis was performed Ioan, P., Betea Şt., Pascu, R., Guscetti, G., Perregaux, N.,
with computer program developed at the TUCEB, Lelli, L., Ponzo, Fr. 2005. Seismic performance assess-
based on ANSR I nonlinear structural analysis ment of steel structures by 3d nonlinear dynamic analy-
sis: case study of an industrial multi-story building in
program. The nonlinear dynamic analysis showed Geneva. Proc. of Eurosteel 2005 Volume C, p. 5.2-59 –
that the response of the building at the most severe 5.2-64, Maastricht, Netherlands.
accelerogram recorded in Bucharest is acceptable: Ioan P., Pascu, R., Beţea, Şt. 2006. Steel shear walls (in Ro-
story drifts are within code limits, plastic rotations manian). Revista construcţiilor, no. 12 & 13.
are not excessive and during the dynamic response Mondkar, D.P. & Powell, G.H. 1975. ANSR – I General
a great number of plastic hinges develop, mainly at Purpose Program for Analysis of Nonlinear Structural
Response. Report No. EERC 75-37. Berkeley, California.
beam ends, allowing for good energy absorption. Sandu, L., Degeratu, M., Haşegan, L., Georgescu, A.,
Steel shear panels entered also in the plastic range, Coşoiu, C.I. 2005. Modelling of wind action on the Bu-
not only at the base of the walls, but also at the up- charest Tower Center International Building (in Roma-
nian). Research report no. 427/2004. Technical Univer-
sity of Civil Engineering, Bucharest.

View publication stats

You might also like