Simulation of the dynamic behavior of a

bi-cable ropeway with modal bases

PhD Gabriel Hurel ([email protected]),
Ing. Jerome Laborde and Pr. Louis Jézéquel
Ecole Centrale de Lyon, LTDS,
36 avenue Guy de Collongue, 69134 Ecully, France
Published on July 5, 2018
doi:10.1007/978-3-319-74700-2 5

Abstract
This article presents the building of a dynamic model of a bi-cable ropeway. Thanks
to the assumption of quasi-static advance of the vehicles, the calculation is performed
step by step. At each step, a transient linearized dynamic solution is calculated around
the quasi-static equilibrium using modal bases. The ropeway system is substructured
considering the different elements: track rope, hauling rope and vehicles represented by
pendulums. The results of time integration give the accelerations felt by the passengers
according to the load case.

Keywords ropeway, dynamic, modal basis, substructuring, quasi-static

1 Introduction
1.1 Problematic
In order to improve mobility in cities, new types of public transport are investigated. In
particular, ropeway is highlighted because it presents several benefits. First, the cost of
building is relatively low compared to subway or tramway. It only needs to build stations
and some pylons. The space used by the infrastructure on the floor is very low: unlike bus
and tramway, the ropeway is not in the way of streets. Moreover, the speed of displacement
is quite high and the throughput is similar to that of a tramway. Finally, this is one of
the most secured type of transport. The rate of injuries and deaths per passengers is the
lowest of all types of transportation. On the other side, ropeways have drawbacks. The
regulation can be an obstacle to the proximity between the cabins and habitation buildings.
In fact, the ropeway can create some discomfort for the neighborhood. The noise and the
obstruction of the view can degrade the comfort of residents. It might also have some
problems for the comfort of the passengers. The comfort criteria are thermic, acoustic and
vibratory. Within the I2TC project, the objective of this article is to provide a tool able to
solve dynamic and vibratory issues of an urban ropeway. In fact, the dynamic requirements
for public transports are higher than touristic installations. For this, a dynamic modelling
of a ropeway line is performed. Thus we aim to understand the dynamic behavior of the
line and be able to predict the acceleration felt by passengers. The sensitive parameters can
be found and an optimization can be performed.

1
1.2 The 3S cable-car
One of the technologies considered for urban transportation using ropes is the 3S cable car.
As its name indicates, the number of cables is three. Two of them are track ropes which
sustain the vehicles. They are fixed and the cabins are rolling on them like a train on a
railroad. The third cable is the hauling rope. Its role is to pull the vehicles along the line.
This cable is moving at the same speed as the vehicles, since they are clamped on it. The
track ropes are embedded in the first and the last station. Their tension is tuned few times
in the year in order to compensate the thermal expansion. In service the tension is varying
mainly due to the load of vehicles. The tension of the hauling rope is kept constant at the
first station by a hydraulic cylinder. The tension at the last station is imposed by the engine
to haul the vehicles. The vehicles are composed of three main parts: the cabin, where the
passengers are, is hanged to a carriage by a hanger. The hanger is fixed to the carriage by a
damped revolute joint. The carriage is fixed to the hauling rope by a clamp and it is rolling
on both track ropes.

1.3 Bibliography
The theory of vibration of a cable was mainly developed by Irvine [1] in the seventies.
He developed analytical models able to calculate the shapes and the frequencies of the
first eigenmodes of a cable with a small sag. Sack [2], Simpson [3] and Triantafyllou [4]
studied the vibrations of a cable in translation. Sergev and Iwan [5] were ones of the first
to study a cable with masses. They represent a taut cable by a succession of segment.
Rosenthal [6] used an iterative method to find the modes of a cable with sag and masses.
Cheng and Perkins [7] calculated the eigenmodes when the position of the mass is varying
from one end of the cable to the other. Al-Qassab [8] computed these eigenmodes with a
Galerkin method. The first model of a ropeway is certainly that of Portier [9]. The vehicles
are represented by pendulums and the cable by springs. The results are compared with
experiments. Brownjohn [10] built a model with a small sag cable to modelize the ropeway
of Singapore. Sofi [11] represented the cabins with damped oscillators.

2 Model building
2.1 Presentation
The proposed model shown in figure 1 corresponds to a span between two pylons. Since
the model is planar, only the half of the system is considered: one track rope, a half of the
hauling rope and half of the vehicles. Both pylons are assumed to be infinitely rigid. The
cables are assumed to be elastic, with a mass. Because of low angles of bending, the bending
rigidity is neglected.
Each vehicle is modelled by a pendulum as shown in figure 1. At the top, the carriage is
represented by a point mass. It is in sliding contact with the track rope and clamped on the
hauling rope. The hanger is attached to the carriage by a damped cylindrical joint. It has
a rotational inertia and a longitudinal stiffness. Another point mass represents the cabin.
In this model, dynamic effects that could occur between cables and the atmosphere are not
considered.

2.2 Assumptions
Considering the studied system, two strong assumptions are made. First, the speed of the
vehicles C is low enough to consider the motion of the vehicles have no dynamic effect on
the system. In other words, the cabins are slow enough to consider their advance as quasi-
static. The system can be represented by a succession of static equilibriums. To validate
this assumption, the results of a simple model of a mass moving on a taut cable is compared

2
Figure 1: Model of the 3S cable car (red line: track rope, green line: hauling rope, circle:
carriage, square: cabin, blue line: hanger)

with the quasi-static solution. Figure 2 shows the state of the cable when the mass is at the
middle of the cable for three values of speed: a low speed similar to the speed of the cable
car vehicles speed, a high speed similar to the speed of high speed trains and a transitional
speed. The solid line represents the quasi-static solution. In every cases, the inertial forces
tend to separate the cable from its static shape but in the high speed case these forces are
higher so the error on the displacement cannot be neglected.

-0.005

-0.01
y/L

-0.015
V=10 m/s
V=60 m/s
-0.02
V=100 m/s
Quasi-static
-0.025
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
x/L

Figure 2: Comparison of the shape of the cable for three different speeds of the mass and
the quasi static solution

At low speed (10 m/s), figure 3 shows that the average error E on the displacement is
below 1%.
1 T 1 L uR − u
Z Z
E= dxdt (1)
T 0 L 0 uR
Where T is the period required by the mass to travel on the cable, L is the length of the
cable, uR is the reference displacement of the cable and u is the calculated displacement.
A correction of the quasi static solution can be made by adding the inertial forces of the
mass and the cable calculated with the succession of equilibrium states. The results with
this correction shown in red line on figure 3 are better, in particular at speeds between 10
and 60 m s−1 .
Around these static equilibriums, we assume the dynamic displacement is small enough
to linearize the behavior equations. This important assumption allows us to project the
problem on the eigenmodes of the structure.

3
 10
Average error E (%)

1
Quasi-static solution
Quasi-static + inertial forces

1 10 100
C (m/s)

Figure 3: Average error with the quasi static assumption for different speeds of the mass

2.3 Computation strategy

Thanks to the first assumption, the computation is performed step by step. The advance
of the vehicles is discretized. Each step corresponds to a position of the whole of vehicles
along the cables. At each step a non-linear static equilibrium will be computed. Under
the assumption of small dynamic displacement, linear eigenmodes of each cable part are
calculated, taking into account the position of the vehicles and parameters of the static
equilibrium: shape and tension of the cables. The eigenmodes of the entire track rope
constitute a modal basis on which the behavior equations are projected. A part of the
hauling rope is the portion between two vehicles or between a vehicle and a pylon. The
extremity of each part of the hauling rope are not always stationary but may be linked to
the vehicles. Static response modes have to be added to the natural modes. The behavior of
the pendulums embodying the vehicles are considered as non-linear to keep a good coupling
with cables. After assembling all parts: reduced track rope, reduced hauling rope parts and
vehicles, a time integration is performed at each step. The initial conditions of a step are
provided by the final state of the previous step. The global behavior of the system is given
by the sum of static and dynamic results.

3 Static equilibrium
3.1 Static equation of a cable
The force on the elementary portion ds at the position (x, y) are the weight and the tension
on each side:
µds~g + T~ (x) + T~ (x + dx) = ~0 (2)

4
Where µ is the linear mass, ~g the gravitational acceleration and T (x) the tension of the
cable at the global coordinate x. The vector equation 2 is projected on x and y axes:

dTx (x)
=0 (3)
dx
dTy (x) ds
= µg (4)
dx dx
The linear mass µ is not constant because of the elastic strain:
µ0
µ= (5)
T
1+
ES
Where E and S are the Young modulus and the area of the section of the cable respectively.
In the ropeways case, the tension T is negligible compared to the product ES. The order of
magnitude of the ratio is about 5 × 10−3 . So we consider µ as a constant. The equations 3
and 4 give us the differential equation:
s 2
d2 y(x)

1 dy(x)
= 1 + (6)
dx2 τ dx

Tx
The constant τ is homogeneous to a length and equal to . The solution of the equation 6
µg
corresponds to a catenary curve:
x 
y(x) = τ cosh + K1 + K2 (7)
τ
The constants K1 and K2 are calculated with the boundary conditions for each part of the
cable. The expression of the tension of the cable is given by:
x 
T (x) = Tx cosh + K1 (8)
τ

3.2 Equilibrium of a vehicle

The forces exerted on a vehicle with a mass M at the abscissa xM are the gravitation, the
tension of the hauling rope TH on the left side and on the right side and the reaction of the
track rope N :
M~g + T~H
L
(x) + T~H
R ~ = ~0
(x) + N (9)
The contact between the vehicle and the track rope is frictionless, hence the force N ~ is
orthogonal to the track rope. However the angle of the cable on the ϕL is not the same as
~ is chosen as the average of the
the angle on the right side ϕR . The angle ϕM of the force N
angles of the track rope:
ϕL + ϕR
ϕM = (10)
2
The equilibrium of the point of the track rope in contact with the vehicle is given by:

T~L (x) + T~T (x) = N

~ (11)

The projection on the direction orthogonal to N ~ shows that the tension of the track rope TT
is equal on the left and on the right of the vehicle.

5
3.3 Equilibrium of the span
In the same way as the linear mass, we consider the length of the track rope Lc is not varying
because of its elasticity, hence it is constant:
s
Z B Z XB 2 n+1
X Z Xi d2 yc (x)

dyc (x)
Lc = ds = 1+ dx = τci dx
A xA dx i=1 xi−1
dx2
n+1
X Z Xi  
x
= cosh + K1ci (12)
i=1 xi−1
τci

The number n is the number of vehicles on the span. Consequently there are n + 1 parts on
each cables in the span. The equations of the equilibrium of the vehicles 9 and 11 and the
equation 12 are implemented in a nonlinear solver to determine the constants K1i and K2i
of each cable part.

3.4 Results

0.12

0.1

0.08
y/L

0.06

0.04 Vehicle
Track rope
0.02
Hauling rope
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
x/L

Figure 4: Shape of a span with three vehicles.

Figure 4 shows the shape of the cables of a span with three vehicles in a certain position.
Because τci > τhi , the sag of the hauling rope is more important than the sag of the track
rope. The tension of both cables are shown in figure 5 for the same position of the vehicles.
The tension of the track rope is continuous whereas the tension of the hauling rope has
discontinuities at each vehicle.

4 Dynamic response
The objective is to predict the dynamic response of the system while the vehicles are moving
forward. Hence, we need to perform transient simulations. First, this kind of simulation is
computationally expensive. The less number of degrees of freedom there is, the faster the
computations are. Secondly we are interested by the low frequency dynamic phenomena
of the system. Hence we choose to make a modal reduction of the cables. This method
allows to take a small amount of generalized degrees of freedom while the low frequency
phenomena are well described. However, at each step, the system is changing. The modal
bases have to be recalculated. The computation of modes has to be fast. The structure of a
cable car is characterized by elongated parts with very small interfaces represented by points
in the model. The substructuring method appears as a good solution to our problem. The

6
 Track rope
T T /T T0 3.125

3.12

3.115

3.11

Hauling rope
1.15

1.1
T H//T H0

1.05

1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
x/L

Figure 5: Tension of the ropes

computation of the eigenmodes of each part of the cable separately is not time-expensive
and the representation is relatively good.

4.1 Track rope

The dynamic equations of an inclined cable with a mass density are given by Irvine [1]:
 2    
∂ u(t, s) ∂ ∂u(t, s) dU (s)
µ − g sin α = T (s, t) + (13)
∂t2 ∂s ∂s ds
 2    
∂ v(t, s) ∂ ∂v(t, s) dV (s)
µ + g cos α = T (s, t) + (14)
∂t2 ∂s ∂s ds

At the curvilinear abscissa s, the static position of the point of the cable is described by
the coordinates U (s) and V (s) in the frame of reference (~eu , ~ev ). The coordinates u(s, t)
and v(s, t) are the dynamic displacements of the point at the curvilinear abscissa s at the
time t. The tension T can be split up in a static and a dynamic contribution:

T (s, t) = TS (s) + TD (s, t) (15)

The equations 13 and 14 describe the dynamic and static behavior. The static terms can be
subtracted:
∂2u
  
∂ ∂u(t, s) ∂u(t, s) dU (s)
µ 2 = TS (s) + TD (s, t) + (16)
∂t ∂s ∂s ∂s ds
∂2v
  
∂ ∂u(t, s) ∂v(t, s) dV (s)
µ 2 = TS (s) + TD (s, t) + (17)
∂t ∂s ∂s ∂s ds

Previously, we assumed the global length of the cable is constant because the tension was
neglected compared to the stiffness. However, in order to express the dynamic tension, we
consider the local strain of the cable ε(s, t) with the Green-Lagrange deformation:

TD (s, t) = ESε(s, t)
"  2  2 !#
∂u(s, t) dU (s) ∂v(s, t) dV (s) 1 ∂u(s, t) ∂v(s, t)
= ES + + + (18)
∂s ds ∂s ds 2 ∂s ∂s

7
The expression of dynamic tension 15 is replaced in the equations 13 and 14. The equations
are linearized neglecting the terms in u and v with an order greater or equal to two:
2 ! 2
∂2u dU d2 U ∂u
  
dTS dU ∂ u
µ 2 = + 2ES + TS + ES
∂t ds ds ds2 ∂s ds ∂s2
(19)
dU d2 V dV d2 U ∂v dU dV ∂ 2 v
 
+ ES + + ES
ds ds2 ds ds2 ∂s ds ds ∂s2
2 ! 2
∂2v dV d2 V ∂v
  
dTS dV ∂ v
µ 2 = + 2ES 2
+ TS + ES
∂t ds ds ds ∂s ds ∂s2
(20)
dU d2 V dV d2 U ∂u dU dV ∂ 2 u
 
+ ES + + ES
ds ds2 ds ds2 ∂s ds ds ∂s2

Because the track rope is only in frictionless contact with the vehicles, the modes of the
whole span of the cable are calculated. However, in order to better represent its behavior,
masses corresponding to the vehicles are added for the computation of eigenmodes. These
masses are removed before the reduction and replaced by pendulums for the assembly step.
The modes are computed with a finite difference algorithm. Figure 6 shows the variations
of the frequencies of the first eigenmodes with respect to the positions of the vehicles. The
frequency of the first mode of the cable without any vehicle is f0 .

15

10
f/f 0

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
x/L c

Figure 6: Frequencies of the first ten modes of the track rope with the vehicles as punctual
masses with respect to the progression of the track rope

4.2 Hauling rope

The hauling rope is moving with the vehicle with a velocity C. Convection terms appear in
its dynamic equations [2, 3]:
 2
∂2u 2
   
∂ u 2∂ u dC ∂u ∂ ∂u dU
µ + 2C +C + = µg sin α + T + (21)
∂t2 ∂t∂s ∂s2 dt ∂s ∂s ∂s ds
 2
∂2v 2
   
∂ v 2∂ v dC ∂v ∂ ∂v dV
µ + 2C + C + = −µg cos α + T + (22)
∂t2 ∂t∂s ∂s2 dt ∂s ∂s ∂s ds

8
By considering a constant speed of the cable, subtracting the static terms and replacing the
dynamic tension with the Green-Lagrange strain, we obtain:
2 !
∂2u ∂2u dV d2 V ∂2u
  
dTS dC ∂u dV 2
µ 2 + 2µC = + 2ES 2
−µ + TS + ES − µC
∂t ∂t∂s ds ds ds dt ∂s ds ∂s2
(23)
dU d2 V dV d2 U ∂v dU dV ∂ 2 v
 
+ ES + + ES
ds ds2 ds ds2 ∂s ds ds ∂s2
2 !
∂2v ∂2v dV d2 V ∂2v
  
dTS dC ∂v dV 2
µ 2 + 2µC = + 2ES 2
−µ + TS + ES − µC
∂t ∂t∂s ds ds ds dt ∂s ds ∂s2
(24)
dU d2 V dV d2 U ∂u dU dV ∂ 2 u
 
+ ES + + ES
ds ds2 ds ds2 ∂s ds ds ∂s2

The mixed partial term creates an antisymmetric damping, resulting in complex modes.
The shape of the first mode of a portion of a hauling rope of 100 m is represented on figure 7
for two values of C.

× 10-3
5

0 Static equilibrium
C = 0 m/s
C = 50 m/s
t = 0.1 T
y/L

-5
t = 0.2 T
t = 0.3 T
t = 0.4 T
-10 t = 0.5 T

-15
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
x/L

Figure 7: Shapes of the first mode of a portion of a hauling rope for a speed of 0 and
50 m s−1 at different instants corresponding to factions of the period T .

4.3 Vehicle

Figure 8: Model of a vehicle

Each vehicle is modelled by a pendulum consisting of a carriage a hanger and a point

9
mass. The kinetic and potential energies K and U are expressed by:
1 1 1
K= mc (u̇2c + v̇c2 ) + Jc ϕ̇2c + Jp β̇ 2 (25)
2 2 2
1
U = kp δ 2 + g [mc (uc sin α + vc cos α) + mp (uc sin α + vc cos α − (lp + δ cos(β + α))] (26)
2
Where mc , uc , vc , Jc and ϕc are the masses, the displacement in both directions, the moment
of inertia and the rotation angle of the carriage respectively. The variables δ, lp , Jp , βp are
the elongation, the length, the moment of inertia and the angle of the hanger respectively
and mp is the mass of the cabin. By applying the Lagrange equations, the dynamic matrix
equation is obtained:

Mv Ẍ + Cv Ẋ + Kv V = Fg + FNL (β, β̇, δ̇) (27)

The vector X is the vector of the five degrees of freedom of the pendulum:
 T
X = uc vc ϕc β δ (28)

The matrix of mass Mv and the vector of gravitation force Fg are not constant but depend
on the β angle. The non-linear terms are gathered in the nonlinear force vector:

−mp (lp β̇ 2 sin β − 2β̇ δ̇ cos β)

 
−mp (lp β̇ 2 cos β + 2β̇ δ̇ sin β)
 
FNL =
 0 
 (29)
 −2mp lp β̇ δ̇ 
mp lp β̇ 2

4.4 Time integration

At each step p, the track rope, the portions of the hauling rope and the pendulums are
assembled to make the matrices of mass Mp , damping Cp and stiffness Kp . The generalized
degrees of freedom Qp associated are a combination of the modes of the carrying and hauling
ropes and the physical degrees of freedom of each pendulum X. Then, a transient calculation
is performed on a duration of the step ∆t:

Mp Q̈p + Cp Q̇ + Kp Qp = FQg + FQNL (Q) + FI (30)

The external forces FQg , FQNL (Q) and FI are respectively the gravitational forces, the non-
linear terms of the pendulums and the inertial forces due to the quasi-static assumption
calculated with the succession of static equilibriums. The initial conditions of the transient
calculation at the step p are given by the previous step projected from the previous modal
basis Φp−1 to the current basis Φp :

Qp (0) = Φ−1
p Φp−1 Qp−1 (∆t) (31)

5 Results
5.1 Start with an angle
In this case, a train of three vehicles cross the span from left to right. Each vehicle arrives
on the track rope with different angles: 10°, 20° and 30°. Figure 9.a shows the variations of
the angles of the cabins. The acceleration felt by the passengers increases with the growth
of the start angle.

10
 (a) Hanger angle
10
0

β (°)
-10
-20
-30
Cabin 1
Cabin 2
(c) Horizontal acceleretion Cabin 3
d 2 u cab /dt 2

(d) Vertical acceleration

d 2 v cab /dt 2

0 5 10 15 20 25 30 35 40 45 50

Figure 9: (a) Angles of the hangers of the vehicles (b) Longitudinal acceleration of the
cabins (c) Vertical acceleration of the cabins

5.2 Emergency braking

The objective now is to simulate an emergency braking of the ropeway. With the quasi-static
assumption, it is not possible to make a deceleration on the hauling rope because at each
step, the vehicles are not moving in the model. To represent the deceleration, a braking
force is applied on both extremity of the hauling rope. Figure 10 shows the accelerations
felt by the passengers after the emergency braking.

(a) Braking force

FB

(b) Hanger angle

β

(c) Horizontal acceleretion

d 2 u cab /dt 2

(d) Vertical acceleration

d 2 v cab /dt 2

0 5 10 15 20 25

Figure 10: (a) Braking force (b) Angles of the hangers of the vehicle (c) Longitudinal
acceleration of the cabin (d) Vertical acceleration of the cabin

6 Conclusion and perspectives

In this article, a dynamic model of a ropeway is made. The assumptions of quasi-static
advance of the cabins and small displacement allow to use a strategy of step by step compu-
tation. At each step, the quasi-static equilibrium is determined with a non-linear calculation.
Then a modal basis of the track rope with masses is computed. The eigenmodes of each part
of the hauling rope are calculated in considering its velocity. The vehicles are represented by

11
non-linear pendulums. After assembling all the parts, a transient calculation is performed
to know the dynamic behavior of the ropeway according to the load case. Two examples
of load cases are given: a vehicle with a start angle and an emergency braking. The model
could be improved by considering several spans, pylons as flexible structures and introducing
the third dimension.
The I2TC project has been funded by the FUI I2TC, the Rhone-Alpes Region, the Ile-
de-France Region and the Paris Department in collaboration with the Ecole Centrale de
Lyon, EIFFAGE, POMA, RATP, CD VIA, The Vibrant Project, the University of Paris 1
Pantheon-Sorbonne and the ENSTA.

Figure 11: Contributors and funders

References
[1] H. M. Irvine, T. K. Caughey, and Gerald Beresford Whitham. “The linear theory of
free vibrations of a suspended cable”. Proceedings of the Royal Society of London. A.
Mathematical and Physical Sciences 341.1626 (1974), pp. 299–315. doi: 10.1098/rspa.
1974.0189 (cit. on pp. 2, 7).
[2] R A Sack. “Transverse oscillations in travelling strings”. British Journal of Applied
Physics 5.6 (1954), pp. 224–226. doi: 10.1088/0508-3443/5/6/307 (cit. on pp. 2, 8).
[3] A. Simpson. “On the oscillatory motions of translating elastic cables”. Journal of
Sound and Vibration 20.2 (1972), pp. 177–189. doi: 10.1016/0022-460X(72)90420-8
(cit. on pp. 2, 8).
[4] M.S. Triantafyllou. “The dynamics of translating cables”. Journal of Sound and Vi-
bration 103.2 (1985), pp. 171–182. doi: 10.1016/0022-460X(85)90231-7 (cit. on p. 2).
[5] S. S. Sergev and W. D. Iwan. “The Natural Frequencies and Mode Shapes of Ca-
bles With Attached Masses”. Journal of Energy Resources Technology 103.3 (1981),
pp. 237–242. doi: 10.1115/1.3230844 (cit. on p. 2).
[6] F. Rosenthal. “Vibrations of slack cables with discrete masses”. Journal of Sound and
Vibration 78.4 (1981), pp. 573–583. doi: 10.1016/S0022- 460X(81)80126- 5 (cit. on
p. 2).
[7] S.-P. Cheng and N. C. Perkins. “Free vibration of a sagged cable supporting a discrete
mass”. The Journal of the Acoustical Society of America 91.5 (1992), pp. 2654–2662.
doi: 10.1121/1.402973 (cit. on p. 2).
[8] M. Al-Qassab and S. Nair. “Wavelet-Galerkin method for the free vibrations of an
elastic cable carrying an attached mass”. Journal of Sound and Vibration 270.1-2
(2004), pp. 191–206. doi: 10.1016/S0022-460X(03)00490-5 (cit. on p. 2).
[9] B. Portier. “Dynamic phenomena in ropeways after a haul rope rupture”. Earthquake
Engineering & Structural Dynamics 12.4 (1984), pp. 433–449. doi: 10 . 1002 / eqe .
4290120402 (cit. on p. 2).
[10] James M.W. Brownjohn. “Dynamics of an aerial cableway system”. Engineering Struc-
tures 20.9 (1998), pp. 826–836. doi: 10.1016/S0141-0296(97)00113-2 (cit. on p. 2).
[11] Alba Sofi and Giuseppe Muscolino. “Dynamic analysis of suspended cables carry-
ing moving oscillators”. International Journal of Solids and Structures 44.21 (2007),
pp. 6725–6743. doi: 10.1016/j.ijsolstr.2007.03.004 (cit. on p. 2).

12