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

net/publication/316051145

METHODS TO DESIGN A GUIDED TOWER CONCERNING DYNAMIC BEHAVIOR

Conference Paper · January 2014

DOI: 10.7712/120113.4527.C1740

CITATIONS READS

0 1,093

2 authors:

Cecília Mendonça Rui Barros

SENER Ingeniería y Sistemas University of Porto
4 PUBLICATIONS   0 CITATIONS    189 PUBLICATIONS   415 CITATIONS   

SEE PROFILE SEE PROFILE

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

Intercity Railway Toluca - Mexico View project

Traffic vibrations View project

All content following this page was uploaded by Rui Barros on 05 May 2017.

The user has requested enhancement of the downloaded file.

 COMPDYN 2013
4th ECCOMAS Thematic Conference on
Computational Methods in Structural Dynamics and Earthquake Engineering
M. Papadrakakis, V. Papadopoulos, V. Plevris (eds.)
Kos Island, Greece, 12–14 June 2013

METHODS TO DESIGN A GUIDED TOWER CONCERNING

DYNAMIC BEHAVIOUR
Cecília Mendonça1 and Rui Carneiro de Barros2
1
Faculdade de Engenharia da Universidade do Porto
Rua Dr. Roberto Frias, s/n 4200-465 Porto, Portugal
e-mail: [email protected]
2
Faculdade de Engenharia da Universidade do Porto
Rua Dr. Roberto Frias, s/n 4200-465 Porto, Portugal
e-mail: [email protected]

Keywords: Guyed tower, design, dynamic action, P-Δ effect, geometric non-linearity.

Abstract. The work presented in this paper pretends to structure the main concepts associat-
ed with the design of a guyed tower. In this context, are presented the theoretical foundations
and the essential steps for the various methods of analysis.

This topic is therefore of major interest for Metalogalva (Trofa, Portugal), with which the au-
thors collaborate and interact, since this industry complex of metallic constructions is also
involved in the design and construction of tall telecommunication towers and pole structures.

Two dynamic actions are essential for the guyed tower design: The gust effect of wind and the
action of earthquakes.

Therefore this paper presents various considerations and topics involved in the modeling of a
185 m height guyed tower, that enable to characterize their structural response. Since this is
a slender structure it is was necessary to take into account the effects of second-order P-Δ
analysis and also the nonlinear behavior of cables.

332
 Cecília Mendonça and Rui Carneiro de Barros

1 INTRODUCTION
Guyed towers are comprised of a very slender mast supported laterally by inclined and in
tension cables, which are anchored in concrete foundations. This setting appears as an eco-
nomical option to self-supporting towers. However, its behavior is generally nonlinear and
therefore is complicated to study and describe. The simplifications made and the approximate
models used in their design over the years, often unjustified, led to the collapse of many struc-
tures.

For the study of the dynamic behavior of the guyed tower it is really important to under-
stand how it reacts to the wind and to an earthquake. Figure 1 shows typical normalized spec-
tral densities of wind and earthquake actions, where average frequencies of concern for wind
and earthquake actions are also emphasized. As described, these two dynamic actions excite
quite differently any type of tall (and low) structures.

Figure 1: Dynamic excitation frequencies of structures by wind and earthquake. HOLMES [1]

The natural frequencies of truss structures are between 0,5 a 3 Hz. When the masts are very
flexible it is essential to study the dynamic behaviour in the first mode response due wind ac-
tion, as well as the contribution of the second degree P-Δ effect related to structural instability.
The resonant response of slender structures becomes important when the natural frequency
of the structure is below 1 Hz as Carril [2]. Between the two dynamic actions considered for
this study on guyed towers, the wind is generally the controlling design action in this kind of
structures.

2 DYNAMIC ACTIONS

2.1 Wind
The wind speed varies with height and its variation in time varies with the type of ground
where it is deployed. The wind action is represented by a simplified set of pressures or forces
whose effects are equivalent to the effects of turbulent wind. Ferreira et al.[3]
The generalized forces of wind on the masts can be characterized by quasi-static and dy-
namic components. Both forces and the displacements associated depend primarily on the
fundamental mode and frequency and its damping.

333
 Cecília Mendonça and Rui Carneiro de Barros

The quasi-static behavior and the time-varying behavior of these pole masts structures oc-
curs along-the-wind (that is, in the direction of propagation of the wind) and is due to the ad-
dition of a constant wind pressure with a non-permanent gust pressure. The purely dynamic
vibratory behavior of the pole-masts occurs in the transverse direction of propagation of wind
(across-the-wind) and is due to the aerodynamic phenomenon of vortex shedding at the criti-
cal wind speed.

2.2. Standards
The reference standards used in the study of the behaviour of lattice towers (fixed or guyed)
subjected to wind are the British Standard BS 8100 [4] American Standard ASCE [5] the
German standard DIN 4131 [6] and European standard EN1991-1-4 [7].
The purely dynamic behaviour of masts occurs mainly in the transverse direction of the
wind. Based on Annex E of the standard [7] and Barros[8]. One may check the elements of
the structure to vortex shedding and ovalling.
The force applied to the model was obtained following the steps described in [7].

The characteristic values of the wind actions calculated in EN 1991-1-4 present an annual
probability of being exceeded (2%) equivalent to an average recurrence period of 50 years.
The purely dynamic behavior of masts occurs mainly in the transverse direction of the wind
(across-the-wind).
The mast structure under analysis and design verifications is a 185 m height tall guyed
tower, with structure and cables layout detailed in Fig. 2, located in Lisbon area of soil type D
Mendonça, C.[9]

Fig. 2 – Layout (elevation and plan) of the guyed tower 185 m height

Based on Annex E of EN 1991-1-4, equation (1) and on details in [7], it is possible to eval-
uate the forces acting on a specific guyed tower model (Fig. 3) and access the mast structural
response to vortex shedding as well as ovalization, Zar and Chu[10].

334
 Cecília Mendonça and Rui Carneiro de Barros

Fig. 3 – Force of the wind o the 185 m guyed tower mast (kN/m)

2.2 Earthquakes

2.2.1. Spectral Analyses

According to Chopra [11] when analyzing the response of a general structure to seismic ac-
tion, the equation of motion of the multi-degree-of-freedom structure is:
m utotal + c u + k u =
0 (1)
Where m is the mass, c is damping, u is the displacement and k is the stiffness of the struc-
ture. The total displacement utotal = u is equal to the displacement of soil u (t), plus the rela-
t
g
tive displacement of the structure u(t).

(2)
The seismic movement only causes a dynamic response because the inertia forces depend
on the total displacement of the structure (in fact they are expressed in terms of the total ac-
celeration), while the elastic forces and damping depend only on the relative motion.

(3)
From equation (3), it may be inferred that the structure is acted upon by the seismic force
that is defined as F (t ) = m ug (t ) .

Fig. 4 – Acceleration response spectra for ξ =0.05 (Lisbon area, soil type D)

335
 Cecília Mendonça and Rui Carneiro de Barros

To design the structure to seismic action two norms were used: the Portuguese regulation
RSA [12] and Eurocode 8.[13,14]

2.2.2 Time domain analysis for a given earthquake

The use of a time domain analysis is interesting since it allows the input of scaled ground
accelerations based upon time-histories of several earthquakes (recorded or synthetically gen-
erated). Since earthquakes occur with different frequencies, it is possible that a given earth-
quake has the ability to excite a given structure much more than others.
When using the information of past recorded earthquakes of certain frequency content,
these accelerations can be applied at the structure ground-site after normalizing the infor-
mation of these earthquakes by the PGA (Peak Ground Acceleration) at the site; the later is
associated with a certain recurrence period and statistical distribution of earthquakes.
The time histories and associated data of earthquakes used in the time domain analyses of
the guyed tower were taken from the site of the Pacific Earthquake Engineering Research
Center , PEERC [15]. The present design study used data from four earthquakes: Çaldiran
(Turkey, 1976), Loma Prieta (USA, 1989), Kobe (Japan, 1995) and Chi-Chi (Taiwan, 1999).
It is also noticeable that the frequency band of these earthquakes is mainly 1 Hz to 10 Hz (Fig.
6).

Fig. 5 – Band of frequencies of the four earthquakes

3 MODELING
As a first approach to the design of a guyed tower, Gantes [16] suggest a simplified
equivalent beam model with springs to simulate the cables that constitute an interesting initial
assessment. It should be used to frame the initial structural characteristics of the guyed tower
so that it is possible to model it in any finite element program.

3.1 Requirements
The guyed towers designed for telecommunication have to meet service requirements of
the telecommunication devices themselves, since a slight misalignment of the satellites may
result in loss of signal, which may lead to poor quality of service for thousands of customers.

336
 Cecília Mendonça and Rui Carneiro de Barros

Section 3.8.2 of the American standard TIA 222 [17,18] specifies a maximum horizontal
displacement of 3% of the height of the guyed tower structure; wherefore for lattice structures
the limiting value of the horizontal displacement is only 1.5% of the tower height.
TIA 222 also specifies a maximum value for the rotation of the antennas of 4º 00' 00'',
which is also the limit imposed by Telebrás for VHF antennas. When it comes to broadcasting
on UHF antennas Telebrás is more restrictive and imposes a maximum rotation of 1º 40' 00".

3.2 Mast
The analysis of more traditional tower masts proposes the modeling as a simple truss struc-
ture. As the links are not rigid, the structure appears more flexible than it actually is. To par-
tially solve this problem, it resorts to the use of dummy bars. These prevent the occurrence of
undesired degrees of freedom leading to the occurrence of mechanisms. The use of these bars,
with very little axial stiffness, allows the structure to be stable nevertheless flexible and there-
fore enables analyses of the tower under design study using some software based upon the
finite element method. Many manufacturers still rely on full-scale tests to verify that the re-
sults are as expected using a simpler truss model for design of the guyed tower.
Oliveira et al. [19] proposes a less conservative analysis method which combines three-
dimensional framed members with horizontal and diagonal lattice members, so it is not neces-
sary to use dummy bars. It constitutes a better and more real approximation of the structural
behavior.

3.3 Cables
The cables were modeled by existing cable element in SAP 2000 v15 [20]. The program
models the cable as a catenary to represent the elastic behavior of a cable subjected to its own
weight. Its behavior is nonlinear and takes into account the P-Δ effects as large displacements
and large deformations are accounted for.
A cable without tension is not stable and has not an unique position, so all cables should be
loaded. The Canadian standard CSA S37-01 [21] requires that the values of the initial tension
in the cables should be between 8% and 15% of the final cable capacity.
The environmental temperature and the applied loads (namely due to wind and earthquake)
can change the cable length. The effect of these changes is similar to changes of length of the
undeformed cable with the exception that there is no change in self-weight.
An alternative model that can be programmed for the dynamic study of the cables was giv-
en by Desai and Punde [22] which obtained values very close to the analytical values availa-
ble, and is very quick to apply. Also Bertero [23] and Naguib and El-Saad [24] state that one
can not disregard the initial deflection, or the pressure exerted by the wind on the cable itself,
which otherwise would reach different values in the order of 10% to 15%.
Menin [25] also used the initial tension as 10% of the ultimate stress, designing the guyed
tower from such hypothesis. As expected, and also as it can be seen by the results obtained by
Naguib and El-Saad [24], the higher the value of the initial tension employed in the cables,
the smaller the tower displacements would generally be. Naguib and El-Saad [24] program
allows initial tensions up to 40% of the ultimate strength, which is beyond the control parame-
ters adopted in this work and taken from the Canadian standard CSA S37-01 [21]. In this de-
sign study it was used an initial tension close to 10% of the ultimate stress. But if it would be
needed an expeditious manner, without recourse to a FEM software or specific program, the
one proposed by Bertero [23] is advised.

337
 Cecília Mendonça and Rui Carneiro de Barros

3.4 Load Cases Combinations

The combinations used for the verification of the stress in the guyed tower structure were:

(4)
Where Q1 is a variable action based on wind overload, Q2 is the variable action based on
temperature variation and CA is the accidental combination.

4 ULTIMATE LIMIT STATES VERIFICATIONS

For a structure the ultimate limit states that need to be considered are: loss of static equilib-
rium, internal failure of the structure, excessive deformation of the ground and fatigue. In a
steel structure as the guyed tower the limit state referring to internal failure involves the re-
sistance of cross sections, the resistance of the structure to instability.
The recommended values in EC3 for the partial safety factors are the following:

The resistance of cross sections depends on the class of the section. The verification made
using the elastic resistance is valid for all classes of sections, although for class 4 only the ef-
fective area is taken into account. For classes 1 and 2 the use of the plastic resistance is advis-
able, so that the solution achieved is economical.

4.1 Axial Force

Replacing the axial force N of expression (5) for the maximal axial force N máx obtained
from the load combinations one gets the minimal area needed Amin to support those actions.
(5)

4.2 Bending Moment

Replacing the bending moment M of expression (6) for the maximal bending moment M máx
obtained from the load combinations, one gets W pl,min so that the section verifies for those ac-
tions.
(6)

4.3 Bending Moment and Axial Force interaction

(7)

4.4 Bending Moment and Shear Force interaction

(8)

4.5 Asymmetrical bending

(9)

338
 Cecília Mendonça and Rui Carneiro de Barros

4.6 Buckling effect

Guyed Towers are a very slender structure so it’s very important to avoid the loss of stiff-
ness in the structure when it is subjected to high compressive stresses. This effect, known as
tower global buckling leads to sudden failure of the structure by progressive and excessive
lateral deformations.

Guyed towers are very slender structures, so it is very important to avoid the loss of stiff-
ness in the structure when it is subjected to high compressive stresses. This effect, known as
(tower global) buckling, leads to sudden failure of the structure The evaluation of all the Euler
critical loads of each individual structural member (member buckling) is not a good approach
to access the buckling capacity of the tower, when trying to describe what happens with more
complex structures instead than the behavior a single column. The critical load of the guyed
tower – Ncr – can be determined by a series of tests, simulating the degradation of the stiff-
ness of the tower mast.
It is known that the general equilibrium equation of a structure analyzed by the displace-
ment method is F= F0 + KT D where KT is the total stiffness (elastic stiffness, geometric
stiffness contribution – positive for tensile members, negative for compressed members – and
decrease in stiffness due to material non-linearity). In the vicinity of a previous state of stable
equilibrium, the incremental equilibrium insures that:
∆F =KT ∆D ⇒ KT =∆F / ∆D (5)
Generalizing this tangent total stiffness concept to the tower structure, and labeling Fx as a
disturbing horizontal load applied on the top of the tower mast and dx as the tower resulting
horizontal displacement in the same point, then the ratio Fx/dx is an index or a measure of the
transversal stiffness of the tower mast for each axial compressive load N applied to the mast.
A total of 15 individual computational load tests were performed in the tower model, with
different vertical and horizontal forces, and the corresponding top lateral displacements were
evaluated by tower structural analyses. A linear regression on the computational results ob-
tained permits to determine computationally when the stiffness of the structure would vanish;
in fact:
Fx b
≈ aλ +b =0 → λ=−
dx a (6)
where λ is the buckling load factor insuring null total stiffness of the tower at the onset of
elastic instability.

5 RESULTS

5.1 Natural vibration frequencies and vibrations modes

The analyses of the natural vibration frequencies and mode shapes are very important for un-
derstanding the dynamic behavior of the structure, as well as for evaluating the effective mod-
al masses and the percentage of these needed to access the response with desired accuracy.
Even though the first mode of the 185 m height guyed tower under analysis has a very low
frequency (0.625 Hz), it has almost no mass associated with that torsion mode. So it is ex-
pected that the next two natural frequencies of the guyed tower (2nd and 3rd, equal by sym-

339
 Cecília Mendonça and Rui Carneiro de Barros

metry) of 1st (and 2nd) longitudinal bending mode along 2 perpendicular directions, would
have significant effective modal masses vibrating close to 1 Hz natural frequencies.
Table 1 indicates the values and shapes of the natural frequencies (Hz) and vibration mode
shapes of the considered design case study of the guyed tower.

Table 1 – Vibration frequencies and mode shapes

mode 1 – Torsion - f=0.625 Hz mode 2 - f=1.011 Hz mode 3 - Perpendicular to 2 - f=1.011 Hz

mode 4 - f=1.358 Hz mode 5 – Perpendicular to 4 - f=1.358 Hz mode 6 – Torsion - f=1.498 Hz

mode 7 - f=1.983 Hz mode 8 – Perpendicular to 7 - f=1.983 Hz mode 9 – Torsion - f=2.379 Hz

340
 Cecília Mendonça and Rui Carneiro de Barros

mode 10 - f=2.840 Hz mode 11 – Perpendicular to 10 - f=2.840 Hz mode 12 – Torsion - f=3.261 Hz

5.2 Load Cases Combinations

Fig. 6 – Load case combinations

One can see from this analysis that the most onerous load case combination is that of acci-
dental earthquake load using spectrum given by Eurocode 8 for earthquake type 1; followed
by the case of earthquake type 2. The load case S1 appears in 3rd place.
However when individually examined (without load combinations) the wind load would un-
doubtedly be more demanding for the structure than the earthquake load alone.

5.3 Overall Global Buckling

The 15 load cases applied and the displacements resulting from such load disturbances are
detailed in Table 5. As it can be seen, the stiffness index F x /d x is getting closer and closer to
zero. It can never hit zero, since the analysis program would return an error for having be-
come unstable.
From a linear regression on the computational data (Fig. 8) the following equation is obtained:
(7)

341
 Cecília Mendonça and Rui Carneiro de Barros

For vanishing stiffness F x /d x → 0, from which λ cr = 75.618 / 12.789 = 5.912738 .

Multiplying the initial axial force (for axial load factor λ=1) by λ cr , the value of N cr would
be:
=N cr 5.912738 × 600 kN = 3547.643 kN (8)

Table 5 – Determination of N cr of the guyed tower

1 Column 3 Columns
Load
Fx' P' Fx P dx F x /d x λ
Case
- kN kN kN kN M kN/m -
1 4 200 12 600 0,1952 61,48 1,00
2 2 200 6 600 0,0945 63,49 1,33
3 4 250 12 750 0,2139 56,10 1,67
4 4 270 12 810 0,2223 53,98 1,80
5 4 300 6 900 0,2361 25,41 2,00
6 2 300 6 900 0,1145 52,40 2,00
7 4 350 12 1050 0,263 45,63 2,33
8 4 400 12 1200 0,1903 63,06 2,67
9 2 400 6 1200 0,1441 41,64 2,67
10 4 500 12 1500 0,3883 30,90 3,33
11 4 600 12 1800 0,5434 22,08 4,00
12 4 700 12 2100 0,8306 14,45 4,67
13 4 800 12 2400 1,5091 7,95 5,33
14 2 800 6 2400 0,9853 6,09 5,33
15 4 820 12 2460 1,8657 6,43 5,47

Fig.7 – Determination of N cr through linear regression of stiffness indexes from a numerical simulation

342
 Cecília Mendonça and Rui Carneiro de Barros

6 CONCLUSIONS
Some relevant concepts and steps to design tall guyed towers were addressed, particular-
ly those related to wind and seismic actions. The wind pressures were evaluated accord-
ing to Eurocode 1 (EN 1991-1-4). The seismic actions were evaluated using RSA and
also by Eurocode 8 (EN 1998-1-1 , EN 1998-6); for the design study case of the guyed
tower mast, the seismic actions evaluated by EC8 are more severe than those evaluated
by RSA, or even those evaluated by a time domain analysis of four historic earthquake
records scaled by the PGA at the site (Lisbon area). The natural vibration frequencies and
mode shapes were obtained using SAP 2000 v15. The ultimate limit states verifications
were address. A general methodology for calculating the tower mast buckling load, also
applicable to general complex structures, is applied.

ACKNOWLEDGMENTS
This work was co-participated by funds from the project “VHSSPOLES-Very High Strength
Steel Poles” (Faculty of Engineering of the University of Porto, reference 21518) sponsored
by the European Fund for Regional Development (FEDER) through COMPETE (Operational
Program Competitiveness Factors - POFC). The Authors acknowledge the financial support
and the opportunity to contribute to the development of the transmission towers testing site of
Metalogalva (Trofa, Portugal).

REFERENCES
[1] Holmes, J.D. – Wind Loading of Structures. Spon Press, Taylor & Francis Group, 2001.
[2] Carril JR., C. Análise Numérica e experimental do efeito dinâmico do vento em torres
metálicas treliçadas para telecomunicaçõess. Dissertação de Doutoramento, Escola
Politécnica da Universidade de São Paulo, 2000.
[3] Ferreira, N.A., Barros, R.C., Delgado, R. – Comparisons of a tall building wind re-
sponse with and without a TMD. 3rd International Conf on Computational Methods in
Structural Dynamics and Earthquake Engng (CompDyn 2011: # 654), Corfu, 2011.
[4] BS 8100-3 – Lattice towers and masts. Code of practice for strength assessment of
members of lattice towers and masts. British Standards Institute, London, 1999.
[5] ASCE Standard – Guide for the Dynamic Response of Lattice Towers. American
Society of Civil Engineers, New York, 2000.
[6] DIN V 4131 – Steel Radio Towers and Masts. German Standard, 2008.
[7] EN 1991-1-4 – Ações em estruturas: Ações gerais – Ações do vento. LNEC, Lisboa,
2010.
[8] Barros, R. – Dimensionamento estrutural de mastros metálicos. Revista Internacional de
Métodos Numéricos para Cálculo y Diseño en Inginiería, pp: 351-365, Spain, 2002.
[9] Mendonça, C.M.A. – Dimensionamento de torres espiadas de grande altura. Dissertação
de Mestrado Integrado em Engenharia Civil, Faculdade de Engenharia da Universidade
do Porto (FEUP), Porto, 2012.

343
 Cecília Mendonça and Rui Carneiro de Barros

[10] Zar, M. e Chu, S. – Chimneys. In: E.H. Gaylord and C. N. Gaylord, Structural Engi-
neering Handbook. McGraw-Hill Book Company, New York, 1979.
[11] Chopra, A. – Dynamics of Structures: Theory and applications to Earthquake Engineer-
ing. Pearson Education Inc., New Jersey, 2007.
[12] RSA – Regulamento de Segurança e Ações para Estruturas de Edifícios e Pontes, Diário
da Républica, 1ª Série nº 125, p. 1991, INCM, Lisboa, 1983.
[13] EN 1998-1-1 – Projecto de estruturas para resistência aos sismos – Parte 1: Regras
gerais, acções sismicas e regras para edifícios. Eurocódigo 8, Norma Europeia, Bruxelas,
2010.
[14] EN 1998-6 – Design provisions for earthquake resistance of structures – Part 6: Towers,
masts and chimneys. Eurocode 8, European Norm, Brussels, 2003.
[15] PEERC - Pacific Earthquake Engineering Research Center. Ground motion database,
http://peer.berkeley.edu/peer_ground_motion_database , 2012.
[16] Gantes, C., Khoury, R., Connor, J.J., Pouangare, C. – Modeling Loading and Prelimi-
nary Design Considerations for Tall Guyed Towers. Computers and Structures, Vol. 49,
No 5, pp: 797-805, 1993.
[17] TIA 222-F – Structural Standards for Steel Antenna Towers and Antenna Supporting
Structures. Telecommunications Industry Association, American National Standard,
Arlington VA, 1996.
[18] TIA 222-G – Structural Standard for Antenna: Supporting Structures and Antennas.
Telecommunications Industry Association, American National Standard, Arlington VA,
2006.
[19] Oliveira, M.I.R., Silva, J.G.S., Vellasco, P.C.G.S., Andrade, S.A.L., Lima, L.R.O. –
Structural Analysis of Guyed Steel Telecommunication Towers for Radio Antenna,
Journal of the Brazilian Society of Mechanical Sciences and Engineering, Vol XXIX, nº
2, pp: 185-195, Rio de Janeiro, 2007.
[20] SAP 2000 v15 – Integrated Software for Structural Analysis & Design. Computers &
Structures Inc. (CSI),
http://www.comp-engineering.com/products/SAP2000/sap2000.html Berkeley CA,
2012.
[21] CSA S37-01 – Antennas, towers and antenna-supporting structures. CSA Group, 2011.
[22] Desai, Y.M., Punde, S. – Simple model for dynamic analysis of cable supported struc-
tures. Engineering Structures, 23, pp: 271-279. Elsevier Science, Mumbai, 2001.
[23] Bertero, V – Amplification of stress and displacement in guyed towers. Journal of the
Structural Division, ASCE, Vol 85 nº ST5, pp: 89-101, 1959.
[24] Naguib, M., El-Saad, A. – Effect of Initial Sag and Tension in Cables on Outcome Re-
sponses in Cable Structures. 2006
[25] Menin, R. – Análise Estática e Dinâmica de torres metálicas estaiadas. Dissertação de
Mestrado, Faculdade de Tecnologia – Universidade de Brasília, 2002.

344

View publication stats