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

net/publication/258265685

Wind response of cable-stayed masts

Conference Paper · July 2005

CITATION READS

1 192

2 authors:

A. Luigi Materazzi Ilaria Venanzi

Università degli Studi di Perugia Università degli Studi di Perugia
114 PUBLICATIONS   868 CITATIONS    55 PUBLICATIONS   198 CITATIONS   

SEE PROFILE SEE PROFILE

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

Life-cycle cost a View project

Life-cycle cost analysis of tall buildings View project

All content following this page was uploaded by Ilaria Venanzi on 04 June 2014.

The user has requested enhancement of the downloaded file.

 EACWE4 — The Fourth European & African Conference on Wind Engineering
J. Náprstek & C. Fischer (eds); ITAM AS CR, Prague, 11-15 July, 2005, Paper #199

Wind response of cable-stayed masts

2
A.L. Materazzi1, I. Venanzi

ABSTRACT: The standard approach for the equivalent static analysis of line-like structures, based on the
gust factor, is reviewed, with special reference to the case of structures with intermediate elastic supports, as
cable-stayed masts. For this class of structures the first modal shape presents an inversion point and the effect
of the higher modes can not be neglected. A new analysis method, based on the definition of a gust function
variable along the height has been proposed and applied to a case study. The results are compared with those
obtained by a random dynamic analysis and a non-linear step-by-step analysis.

1 INTRODUCTION 2 EQUIVALENT STATIC WIND LOADS

The response of slender structures exposed to 2.1 Vertical cantilever
turbulent wind is characterized by both the Let us consider a prismatic vertical cantilever. Its
background component, mainly depending on the along-wind response is composed of two terms: the
spectral properties of the incoming flow and on the
aerodynamic structural behaviour and by the static response to the mean wind speed V z  and the
resonant component due to the structural modal dynamic response to the fluctuating component of
properties. the velocity v~  z , t  .
Pointing the attention to the case of line-like Almost all the Authors agree that for this class of
structures, the structural response is studied, in the structures only the first mode contributes
common design practice, using proper equivalent significantly to the overall dynamic response.
static loads. In fact when the static scheme is a Following this approach the total response is
cantilever beam the structural response is mainly evaluated by multiplying the static displacements by
governed by the background and the first mode a suitable amplifying factor, the gust factor, which
response only. Taking advantage of this behaviour, takes into account also the effects of the
most of the equivalent static loads supplied by the aerodynamic admittance or the incomplete spatial
International Codes in the field of Wind Engineering coherence of the incoming flow.
are calibrated with reference to this class of As a consequence, the gust factor approach
structures. implicitly assumes that the dynamic response at
When the static scheme is more complicated, as is every point is a simple multiple of the static response
the case of cable-stayed structures, the shape of the to the steady wind, as pointed out by Davenport
first mode departs significantly from that of a (1992).
cantilever beam and the standard formulation of the How this hypothesis departs from the real
gust factor is no longer applicable. Moreover the behaviour may be discussed as follows.
effect of the upper modes may also be relevant. Let us consider a location of the cantilever at the
As a consequence the response under dynamic elevation z above the ground. The 1st mode
loads carried out considering the actual first modal displacement is given by:
shape, along with the upper modes, generally results
more severe than the static response computed using x1  z  \ 1  z sin 2S n1t  (1)
the equivalent static loads.
Both problems were clearly addressed since 1992 where:
by Davenport, who proposed to use a series of static \ 1 z  is the amplitude of the 1st mode displacement,
load patterns, arranged following a patch load which depends on the intensity of the dynamic load
scheme, instead of the gust factor technique. and on the elevation z;
In the present paper the problem is revisited,
n1 is the 1st mode frequency;
discussing firstly the formulation of the gust factor
approach, then a unified solution for evaluating the t is the time.
equivalent static loads acting on any type of vertical The maximum acceleration reached at the same
line-like structures is proposed. location is :
1
Prof. Annibale Luigi Materazzi, DICA, University of Perugia, Italy, e-mail [email protected]
2
Eng. Ilaria Venanzi, PhD Student, DICA, University of Perugia, Italy, e-mail [email protected]

1
 The Fourth European & African Conference on Wind Engineering, Paper #199
x1,max z  4S 2 n12\ 1 z  (2) It is thus clear that the gust factor G can be
adjusted to estimate properly only one aspect of the
which is proportional to the modal displacement structural response: for instance the top displa-
amplitude \ 1 z  . cement, or the maximum bending moment, or the
maximum shear (Zhou et al, 2000; Zhou & Kareem,
Saying P1  z  the mass density per unit length of the 2001; Piccardo & Solari, 2002).
cantilever in the first mode, the static force per unit The cantilever displacements at all elevations,
length which produces the maximum 1st mode along with the associated curvatures, cannot be
acceleration may be evaluated, applying the estimated accurately.
D’Alembert Principle, as: The formulations of G commonly established in
literature (see, among the others, Solari, 1988) are
p1  z  P1  z  x1 ,max z  P1  z  4S 2 n12\ 1  z  (3) calibrated, in most cases, to give the exact value of
the top displacement.
If the first mode effective mass m1 is uniformly A viable solution to overcome this problem is
distributed over the height H of the cantilever, then defining a “gust function” G(z), which depends on
the elevation above the ground.
m1 Multiplying and dividing the last term of eq. 5 by:
P1  z  (4)
H
1
CD  z B z  UV  z 
2
(6)
Further comments on the evaluation of the 2
effective modal masses, which is not a trivial task in
case of wind loads, are presented in the following we obtain:
section 3.
The static load p1 z  has the same shape of the 1
pequiv  z  CD  z B z  UV  z  
2

first eigenvector and is proportional to it. 2

The equivalent static load is then: 1
CD  z B z  UV  z 
2
(7)
 P1 z  4S n1\ 1  z 
2 2 2
1 1
pequiv  z  CD  z B z  UV  z   p1  z 
2
CD  z B z  UV  z 
2
(5)
2 2
The terms on the right hand side of eq. 5 are By defining the gust function G(z) as:
qualitatively represented in fig. 1.
The difference of curvature of the two P1 z 4S 2 n12\ 1 z 
contributions is immediately evident. G1 ( z ) 1  (8)
1
CD  z B z  UV  z 
2

an alternative expression of the equivalent static load

may be defined as:

1
pequiv  z  G1 ( z ) pstat G1 ( z )CD  z B z  UV  z  (9)
2

It leads to accurate results in the estimate of all

the cinematic and static parameters.
2.2 Multimode response of cable stayed masts
The previous discussion can be easily extended to
the case of the multimode response, whose effects
are relevant in the case of vertical structures with
intermediate elastic supports, like the cable-stayed
masts.
Figure 1 Shape of the mean wind load (on the left) and of the The equations from 1 to 5 still stand true and give
static load which gives the first mode response (on the right). the i-th mode contribution to the dynamic response.

2
 A.L. Materazzi, I. Venanzi
Supposing that the modal responses are be defined as:
statistically independent, the equivalent static load
Leq i2
becomes, in the case of multimode response: mi (15)
Meqi
1
p equiv  z  C D z B z  UV  z  
2
2 When the external system excitation is due to
2 (10) wind induced forces, the previous definition is not
N
 ¦ >P i z 4S 2 2
ni \ i z @ exactly applicable and a suitable multiplicative
correction factor k i must be applied to every modal
i 1

In this case the gust function becomes: mass Meqi .

Thus the force pi  z  whose static application
N 2

¦ >P i z 4S 2 2
ni \ i z @ leads to the i-th mode displacement modifies as:
i 1
GN ( z ) 1  (11) pi  z  ki Pi  z  4S 2 ni2\ i  z  (16)
1
C D  z B z  UV  z 2
2
Consequently the Eqs. 9 and 11 become:
where the subscript N in GN(z) points out that the
first N modes have been taken into account.
However Eq. 8, which accounts for the effect of k1 P1  z 4S 2 n12\ 1  z 
G1 ( z ) 1  (17)
the first mode only, may be used as a first 1
C D  z B z  UV  z 2
approximation to evaluate the gust function also in 2
the case of cable stayed masts.
2
3 THE EFFECTIVE MODAL MASSES N

The effective mass of each mode is commonly

¦ >ki P i z 4S 2 2
ni \ i z @
i 1
evaluated in the field of earthquake engineering. In GN ( z ) 1  (18)
1
that case the typical dynamic problem is represented C D  z B z  UV  z 2
2
by a MDOF system acted upon by a support
acceleration x0 t  .
The corresponding equilibrium equation is 4 NUMERICAL EXAMPLES
therefore:
4.1 Programme of the numerical analyses
Mxt   Cx t   Kxt  MLx0 t  (12) In order to highlight the importance of the higher
modes contribution to the structural response in case
where M, C and K are, respectively the mass, the of structures with intermediate elastic supports, the
damping and the stiffness matrices. L is a location gust function G(z) has been evaluated with reference
vector composed by a column of ones and zeroes to a cantilever beam and to a guyed tower.
used to apply the ground acceleration only to the
translational DOF’s.
Calling ĭ the matrix of the system’s
eigenvectors, supposed to be real, we can write:
x ĭq (13)
Substituting the Eq. 13 into the Eq. 12 and pre-
multiplying by ĭ T we obtain the uncoupled system:

t   ĭ T Cĭq t   ĭ T Kĭĭt 
ĭ T Mĭq
(14)
ĭ T MLx0 t 

Saying Leqi ĭ Ti ML and Meqi ĭ Ti Mĭ i the

modal earthquake excitation factors and the modal
Figure 2. Geometry and modal properties of the steel stack.
masses respectively, the effective modal masses may

3
 The Fourth European & African Conference on Wind Engineering, Paper #199
The cantilever beam is a welded steel stack, 100 considering only the first mode and another one
meters high, whose geometry and modal properties considering only the second mode.
are represented in Figure 2 (Solari 1988). In the case of the guyed tower a supplementary
The guyed tower is a 150 meters high time-domain analysis, considering the geometrical
broadcasting antenna. The mast is a steel lattice non-linearity, was performed. For this purpose
tower having a square cross-section in plan (Fig. 3). artificially generated time histories of the wind
It is supported by three levels of stays whose velocity were used.
geometry is represented in Fig. 3. The corresponding Following the Eurocode 1, the sample structures
modal properties are given in Fig. 4. were considered exposed to a wind having the
In Table 1 the significant natural frequencies for following characteristics:
both the analyzed structures are reported. x mean wind speed at 10 m: V10 = 27 m/s;
x site roughness parameters: z0=0.05 and Kr=0.19.
4.2 Frequency domain analyses
For the power spectral density of the null mean
random component of the wind speed the following
expression, due to Solari, was used:
5
n ˜ S v P , n  fL § fL · 3
6.868 v ¨¨1  10.302 v ¸¸ (19)
V v2 z P  © z P  ¹

For the coherence function, the expression

proposed by Davenport (1967) and adopted by the
Eurocode 1 was employed:
Figure 3. Geometric characteristics of the guyed tower. CohP,P',n 
­  2n C 2  z P   z P'2  C 2  xP   xP'2 ½ (20)
° z x °
exp® ¾
°̄ V  z  P   V  z  P ' °¿

with Cx=Cz=11.5.
The cross power spectral density matrix of the
wind load was obtained assuming the quasi-steady
relationship between the mean square pressure
fluctuations and the mean square longitudinal
velocity fluctuations.
The results of the static and the random dynamic
analyses are reported in the Figures 5 and 6. Each
figure contains the displacements under the mean
wind, and the displacements under the random wind
component in the first two modes.
Figure 4. Modal characteristics of the guyed tower.
Following the formulation proposed in the
Table 1. First 3 natural periods of the analyzed structures. previous chapters, the gust function GN(z) may be
written as:
1st natural 2nd natural 3rd natural
period (s) period (s) period (s) N

Stack 1.24 0.29 0.12

GN ( z ) 1  ¦ >g i z @2 (20)
i 1

Guyed tower 1.24 1.07 0.71 where the contribution of the generic mode i is:
mi 2 2
On the selected examples, several types of ki 4S ni \ i  z 
analysis have been carried out: gi ( z ) H (21)
1
CD  z B z  UV  z 
2
x a static analysis under the mean wind;
2
x a random dynamics analysis, carried out

4
 A.L. Materazzi, I. Venanzi
nd 160
2 mode
100
140
Mean wind

80 120
g1(z)
100
60
Z (m)

Z (m)
st 80
1 mode

40 60 g2(z)

40
20
20

0 0
0,0 0,1 0,2 0,3 0,4 0,5 0,6 -3 -2 -1 0 1 2 3 4 5
Displacement (m) Modal contributions to the gust function G(z)

Figure 5. Displacements of the steel stack. Figure 8. Guyed mast: 1st and 2nd mode contributions to G(z).
160 160

140 Mean wind 140

120 120

100 100
st
1 mode
Z (m)

Z (m)

80 80

60 nd 60
2 mode
40 40

20 20

0 0
-0,2 -0,1 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0 1000 2000 3000 4000 5000 6000 7000 8000
Displacement (m)
Load (N/m)
Figure 6. Displacements of the guyed tower. Figure 9. Guyed mast: equivalent static load obtained using
the G2(z) gust function.
120

assumption commonly adopted in most codes to

100 consider the first mode contribution only.
Different is the case of the guyed mast, where the
80
g2(z) g1(z) second mode contribution is relevant and can hardly
be neglected.
Anyway also in the case of the cantilever, the first
Z (m)

60

mode contribution to the gust function GN(z) departs

40 significantly from a constant value.
Using the gust functions G2(z), which consider
20
only the first 2 modes, the equivalent static load was
computed for both the stack and the guyed tower.
0
-1 0 1 2 3 4 5 6
The resulting equivalent static load for the guyed
Modal contributions to the gust function G(z) mast is shown in Figure 9. It represents the static
Figure 7. Steel stack: 1st and 2nd mode contributions to G(z).
load that applied to the structure produces
deflections equal to the sum of the response to the
The contribution of the first two modes has been static mean wind and the dynamic response
computed accordingly to Eq. 21. It is reported in computed considering the 1st and the 2nd modes.
graphical form in Figures 7 and 8 and in Tables 2 The shape of the equivalent load departs
and 3. significantly from the one proposed by Davenport
It is possible to see that in the case of the which used, to simulate the effect of the dynamic
cantilever the second mode contribution g2(z), only behaviour, a set of patch loads, each one being
slightly departs from zero, thus confirming the constant.

5
 The Fourth European & African Conference on Wind Engineering, Paper #199
Table 2. First two modal contributions to the gust function G(z) performed.
for the steel stack. The Newmark’s constant-average-acceleration
Height (m) g1 g2 integration scheme has been used in the step-by step
0 0.000 0.000 analysis. At every time step 't the incremental
5 0.012 -0.006 equilibrium equation:
10 0.046 -0.018
15 0.101 -0.037 M t  't x C t  't x K t  't x t  't
F (22)
20 0.179 -0.062
25 0.284 -0.092 was solved by computing the displacements,
30 0.418 -0.125 velocities and accelerations vectors x , x , x :
35 0.586 -0.158
40 0.794 -0.189
t  't
x t x  t x ˜ 't (23)
45 1.049 -0.213
1
50 1.346 -0.226 t  't
x t x t x ˜ 't  ˜t  't x ˜ 't 2 (24)
55 1.682 -0.226 2
60 2.049 -0.210
65 2.443 -0.179
being F the load vector, M, C and K the mass,
70 2.858 -0.131 damping and stiffness matrices.
75 3.291 -0.068 The geometric nonlinearity was considered in the
80 3.737 0.008 procedure by updating the stiffness matrix.
85 4.189 0.093 The stays pretension was modeled by means of
90 4.644 0.183 equivalent thermal loads applied to the relevant
95 5.098 0.277
members.
100 5.548 0.370
For a generic frame member the stiffness matrix
Table 3. First two modal contributions to the gust function G(z) was obtained as the sum of the linear stiffness matrix
for the guyed mast. and the geometric stiffness matrix as follows:
Height (m) g1 g2 >K @ >K lin @  >iter K geom @ (25)
0.00 0.000 0.000
6.25 -0.001 0.011 in which:
12.50 -0.002 0.029
18.75 -0.006 0.041 ª EA º
« L »
25.00 -0.014 0.036 « »
12 EJ
31.25 -0.026 0.006 « 0 SIMM . »
37.50 -0.044 -0.059 « L3 »
43.75 -0.067 -0.161 « 6 EJ 4 Ej »
« 0 L2 L »
50.00 -0.086 -0.299 >K lin @ « EA EA »
56.25 -0.094 -0.466 « 0 0 »
62.50 -0.082 -0.664 « L L »
« 12 EJ 6 EJ 12 EJ »
68.75 -0.043 -0.897 « 0  3  0 »
75.00 0.027 -1.176 « L L2 L 3

« 0 6 EJ 2 EJ 6 EJ 4 EJ »»
81.25 0.135 -1.525 0  2
87.50 0.282 -1.955 «¬ L2 L L L »¼
93.75 0.464 -2.380
100.00 0.681 -2.681 ª0 º
« 6 »
106.25 0.936 -2.757 «0 SIMM . »
112.50 1.232 -2.559 « 5L »
118.75 1.577 -2.083 «0 1 2 »
125.00
131.25
1.994
2.537
-1.364
-0.448
> iter
K geom @ ( iter 1)
T ˜ ««
0
10
0
15 L
0 0
»
»
137.50 3.258 0.605 « 6 1 6 »
143.75 4.081 1.730 «0   0 »
« 5 L 10 5L »
150.00 4.932 2.876 «0 1 1 1 2 »
 0 
«¬ 10 30 L 10 15 L »¼
4.2 Time domain analyses where: E is the elastic modulus;
Since the guyed towers are highly deformable A is the member’s area;
structures, their response is significantly affected by L is the member’s length;
non-linear effects. For this reason a time domain J is the modulus of inertia;
analysis of the structural response has been T is the member’s axial load.

6
 A.L. Materazzi, I. Venanzi
Within each time step, equilibrium iterations were Zi i  1 2'Z , i=1,…,N;
performed, updating the geometric stiffness matrix Zi ' Zi  GZi , GZi is a small random frequency;
using the value of each member’s axial load T at the
end of the previous step. M is an independent random phase uniformly
The convergence tolerance for the nonlinear distributed between 0 and 2ʌ.
iterations was set to 1% of the maximum axial loads. The time domain analysis was performed for a
The finite element model of the mast is composed duration of 600 s, with a time step of 0.01 s.
of 24 beam elements and 60 rope elements, Following the Rayleigh approach, the damping is
connected in 83 nodes, each one having three assumed to be proportional to a combination of the
DOF’s. mass and the stiffness matrices as follows:
In Figure 10 the mesh used to assemble the finite
element model for the time domain analysis is
>C@ a>M @  b>K @ (27)
represented. The damping ratio [ of the first two modes was
assumed as 1% of the critical value and the
coefficients a and b in the previous equation were
computed considering the first two circular
frequencies Ȧn:

a bZ n
[n  n 1,2 (28)
2Z n 2

In Figure 11 is represented the time history of the

top node maximum displacement is about 1.40 m,
reached after roughly 350 s.

1,5

1,0
Displacement (m)

0,5

Figure 10. Joints numbering of the finite element model used in

the time domain analysis.
0,0
A realization of the uni-dimensional multi-variate
wind velocity process was artificially generated and
-0,5
the corresponding load was applied to the joints of
the structure.
The simulation of the stochastic process of the -1,0
0 100 200 300 400 500 600
wind velocity at the generic node j was performed
Time (s)
using a well established technique (Shinozuka & Jan
Figure 11. Time history of the top node displacement.
1972) as follows:
j N
4.2 Comparison of the results
v j t  ¦¦ H Z  jm i > @
2'Z cos Zi ' t  T jm Zi   M mi (26) The deflections of the guyed tower obtained by the
m 1 i 1
time domain, the frequency domain and the
where: equivalent static analyses are compared in Figure 12.
The time domain response is that corresponding to
§ Im H jm Zi  · the time step when the maximum top node
T jm Zi  tan 1 ¨¨ ¸;
¸
© Re H jm Z 
i ¹
displacement occurred.
As it was expected, the results of the time domain
H Ȧ  is a lower triangular matrix, obtained by the analysis, which takes into account the geometric non
Cholesky decomposition of the target cross-spectral linear effects, are greater that the others.
density matrix S(Ȧ) of the wind speed; The deflections obtained with the frequency
N is the number of intervals along the i-th axis of the domain and the equivalent static analysis turned out
frequency domain; to be very similar.

7
 The Fourth European & African Conference on Wind Engineering, Paper #199
160 6 ACKNOWLEDGEMENTS
140
The Authors gratefully acknowledge the financial
120
Time domain analysis
support of the Italian Ministry of the University and
100 Static response using G2(z)
Scientific Research (MIUR) which funded this study
through the Project VinCES (Vibration in Civil
80 Frequency domain analysis
Engineering Structures).
Z (m)

60

40
7 REFERENCES
Davenport, A. G. (1967), Gust loading factors, Journal of the
20
Structural Division, 93 (3), pp. 11-34.
0 Vellozzi, J., Cohen, E. (1968) Gust response factors, Journal of
the Structural Division, 94, pp. 1295-1313.
-20 Shinozuka M., Jan C. M., (1972), Digital simulation of random
-0,2 0,0 0,2 0,4 0,6 0,8 1,0 1,2 1,4 1,6
processes and its applications. J. Sound Vibration, Vol. 25,
Displacement (m)
pp. 111-128.
Figure 12. Comparison between the displacements obtained Simiu, E. (1973), Gust factors and alongwind pressure
using the TD, the FD and the equivalent static analyses. correlation, Journal of the Structural Division, 99, pp. 773-
783.
The slight difference still present can be ascribed Solari, G. (1988), Equivalent wind spectrum technique: theory
to the fact that the frequency domain response was and applications, Journal of Structural Engineering, Vol.
114 n°6, pp. 1303-1323.
carried out by the superposition of 4 modes, while Solari, G. (1989) Wind response spectrum. Journal of
the gust function accounted for two modal Engineering Mechanics, 115, no. 9, pp. 2057-2073.
contributions only. Davenport, A. G., Sparling, B. F. (1992), Dynamic gust
response factors for guyed towers. J. Wind Eng. Ind.
5 CONCLUSIONS Aerodyn, 41-44, pp. 2237-2248.
Kasperski, M., Niemann, H. J. (1992) The L.R.C. (load
The response of flexible structures to turbulent wind response correlation) – Method. A general method of
is evaluated in most cases amplifying the static estimating unfavourable wind load distributions for linear
response to the mean wind by a suitable constant, the and non-linear structural behavior. J. Wind Eng. Ind.
Aerodyn, 41-44, pp. 1753-1763.
gust factor G. Holmes, J. D. (1994), Alongwind response of lattice towers:
In the case of cable stayed masts, whose first part I - derivation of expression for gust response factors,
modal shape departs significantly from that of a Engineering Structures, 16, pp. 287-292.
cantilever, the method is not applicable and Holmes, J. D. (1996), Alongwind response of lattice towers –
alternative solutions, like the one that uses patch III. Effective load distributions, Engineering Structures, 18,
pp. 489-494.
loads, may be used. Piccardo, G., Solari, G. (1998), Closed form prediction of 3-D
In the present study the problem was reviewed wind excited response of slender structures, J. Wind Eng.
and a unified formulation which leads to the use a Ind. Aerodyn, 74-76, pp. 697-08
gust function G(z), variable with the elevation above Zhou, Y., Gu, M., Xiang, H. (1999) Alongwind static
the ground, was proposed. equivalent wind loads and responses of tall buildings. Part I:
Unfavorable distributions of static equivalent wind loads, J.
Some numerical examples showed that in the case Wind Eng. Ind. Aerodyn, 79 , pp. 135-150.
of cantilever structures, like stacks and chimneys, Zhou, Y., Gu, M., Xiang, H. (1999) Alongwind static
only the first mode contribution may be considered equivalent wind loads and responses of tall buildings. Part
in the analysis. Anyway, the gust function G(z) II: Effects of mode shapes, J. Wind Eng. Ind. Aerodyn, 79,
departs significantly from a constant value and thus pp. 151-158.
Zhou, Y., Kareem, A., Gu, M. (2000) Equivalent static
its variability with the height should be taken into buffeting loads on structures, J. Struct. Eng, 126, pp. 989-
account. 992.
In the case of cable-stayed masts, at least the Zhou, Y., Kareem, A. (2001) Gust loading factor: new model,
second mode contribution to the dynamic response J. Struct. Eng, 127, pp. 168-175.
must be considered and the gust function G(z) may Piccardo, G., Solari, G. (2002), 3-D gust effect factor for
slender vertical structures. Prob. Eng. Mech., 17, pp. 143-
be evaluated accordingly. 155.
The proposed equivalent static formulation, which
uses the SRSS mode superposition rule, leads to
results close to a random dynamic analysis but,
obviously, can not take into account the second order
effects which play a non-negligible role when
flexible structures are analyzed.