Acta Astronautica 185 (2021) 132–139

Contents lists available at ScienceDirect

Acta Astronautica
journal homepage: www.elsevier.com/locate/actaastro

Three-dimensional analysis of a counterweight type space elevator

Taiki Okino a, Yoshiki Yamagiwa a, *, Shoko Arita a, Yoji Ishikawa b, Kiyotoshi Otsuka b
a
Shizuoka University, 3-5-1 Johoku, Hamamatsu City Naka-ku, Shizuoka, Japan
b
Obayashi Corporation, Tokyo, Japan

A R T I C L E I N F O A B S T R A C T

Keywords: Space elevators studied thus far have been mainly the climber type, in which a cable connects the ground to a
Space science space station and the payload is transported by climbers ascending and descending along the cable. However,
Space technology this type of system has problems, such as the difficulty of supplying energy to climbers and a short lifespan owing
Space elevator
to the abrasion of the cable and climber wheels during operation. To avoid these problems, in this study, we
investigate a novel counterweight-type space elevator. This system consists of two cables: a guide cable that
withstands the tension applied to the structure and a moving cable that connects two gondolas, one at either end
of the cable, attaches to a driving wheel in the space station, and transports the payload in the gondolas by
driving the wheel. In this study, we analyzed the cable dynamics by using the point mass cable model developed
by our group when the counterweight type is applied between the station and the ground, and we calculated the
energy necessary for actual operation. As a result, when the counterweight type is applied between the Mars
gravity center (altitude 3900 km) and the ground, while the climber type is applied at altitudes above that, the
system uses less energy than conventional climber-type space elevators.

supply energy to climbers, however it is difficult to maintain high effi­

1. Introduction
A space elevator [1] is an unprecedented large-scale space trans­
portation system in which a cable extends from the ground to a space
station and a payload carrier (climber, gondola, etc.) moves along the
cable. A material that can withstand more than 50 GPa of stress is
required for realization, and until several decades ago the system was
not expected to be realized soon. However, in recent years the devel­
opment of space elevators has become a reality with the invention of a
lightweight and very strong material called carbon nanotubes (CNTs) [2,
3].
The type of space elevator system that has been examined since
Edwards’ 2003 study [1] is majorly the climber type [4–7], in which a
cable (or tether) connects the ground to space stations and the payload is
transported by climbers ascending and descending along the cable by
using the friction between the rollers of the climber and the cable
(Fig. 1). Such a system is simple and, because there is only a single cable,
it avoids the risk of cables colliding with each other. However, this type
of system has serious problems to overcome. A major technological
problem is the difficulty of supplying energy to the climbers. Wireless
energy transmission using microwaves or lasers is potential method to
ciency over long distance [8]. Edwards and Westling [9] describe the
diameter of the power receiving antenna as 3 m, but Takano [10]
showed that the diameter of the receiving antenna would be 780 m for
microwaves and 360 m for lasers by considering the beam divergence
with the diffraction limit angle if 770 kW is transmitted to GEO. In
addition, the research and development of wireless power transmission
is active at present, but the transmission distance is still limited on the
order of 10 km [11], and the technology involved in the extremely
long-distance transmission required for space elevators is still immature.
Another large technological problem is the short lifetime of the
climber owing to the abrasion of the cable and climber wheels during
operation. Wheel wear has been recognized as a large problem through
Space Elevator Challenges [12].
Transportation methods for space elevators other than the common
climber type have been proposed in the past, for example, the electro­
static climber, which uses the principle of a linear motor rather than the
common roller climber [13], or the loop space tether transportation
system, which uses two tethers rather than a single tether [14], but these
proposed systems are essentially variations of the climber type, in which
the tether does not move and the payload is transported by
self-propelled climbers, and the problems associated with the climber

* Corresponding author.
E-mail addresses: okino.taiki.15@shizuoka.ac.jp, okino.taiki@gmail.com (T. Okino), yamagiwa.yoshiki@shizuoka.ac.jp (Y. Yamagiwa), arita.shoko@shizuoka.ac.
jp (S. Arita), ishikawa.yoji@obayashi.co.jp (Y. Ishikawa), otsuka.kiyotoshi@obayashi.co.jp (K. Otsuka).

https://doi.org/10.1016/j.actaastro.2021.04.040
Received 22 December 2020; Received in revised form 26 March 2021; Accepted 29 April 2021
Available online 5 May 2021
0094-5765/© 2021 IAA. Published by Elsevier Ltd. All rights reserved.
T. Okino et al. Acta Astronautica 185 (2021) 132–139

Nomenclature

m mass
M mass
t time
ω angular velocity
r position
G gravitational constant
l0 equilibrium length
k elastic modulus
a acceleration
ac acceleration coefficient
vdep winding velocity
T tension
F force
ξ damping ratio
c critical damping coefficient

Subscripts
i i-th mass
g gondola
U ascent
D descent

Fig. 2. Diagram of a normal elevator used on the ground.

Fig. 1. Conceptual figure of climber-type space elevator.

type remains.
Given this background, we investigated the possibility of a coun­
terweight type mechanics to a space elevator. This is the commonly-used
method for ground elevators, and it transports the payload by moving
the cable itself. Although it has not yet been studied academically as a
space elevator, it may make it possible to avoid some of the problems Fig. 3. Conceptual figure of a counterweight-type space elevator.
with the climber type.
In the present study, we investigated the counterweight-type space employed between the GEO station and the ground. Based on the results
elevator with particular attention to its dynamics and the energy of the dynamics in this first-step analysis, the power required for the
required for its operation. Specifically, as a first step, we analyzed the operation was calculated by performing simulations for cases where the
dynamics of the cable, using the 2-D point mass cable model that our counterweight type is applied to different altitude ranges of the space
group developed previously [4], when a counterweight type system is elevator in two types of sections in order to propose the optimal design

133
T. Okino et al. Acta Astronautica 185 (2021) 132–139

2. Constitution of a counterweight-type space elevator

The counterweight type is used as a normal elevator on the ground.

The normal elevator consists of a gondola, counterweight, cable, and
driving wheel (Fig. 2). By balancing the counterweight against the
weight of the gondola, it is possible to reduce the energy required to
drive the gondola.
To apply this type mechanics to space elevators, we considered a
system consisting of two cables, that is, a guide cable that withstands the
tension applied to the structure and a moving cable that connects two
gondolas, one at each end of the cable, instead of a counterweight at one
end in a normal elevator, via a driving wheel located in the space station.
The payloads in the gondolas are transported upward and downward
simultaneously by operating the driving wheel (Fig. 3). The energy
required to operate such a system is supplied only to the driving wheel,
and can be supplied easily from the energy source on the station. Thus,
the technological problem of energy supply that exists in the case of the
climber type elevator can be avoided in this system. In addition, main­
tenance will be easier with this system than with the climber type,
because the part most likely to show wear in this system is the driving
wheel on the station, not the parts in the moving climber, and it will thus
be easier to detect wear.
Furthermore, the operating energy itself can be reduced by using the
potential energy of the upper gondola. However, interference between
Fig. 4. Constitution of a space elevator.
cables may be a problem in this system since it requires two cables for its
operation, and the dynamics of the cables in this system must be clarified
in order to determine the likelihood of realizing this type of system.

3. Analytical methods

3.1. Constitution of space elevator in this study

Fig. 4 shows the constitution of the space elevator considered in this

study. This constitution is based on a plan proposed by Obayashi Cor­
poration [15].

3.2. Analytical model

As shown in Fig. 5, a rotational coordinate system is applied to the

analysis where the system rotates around the z-axis at an angular ve­
locity ω with the origin at the center of the earth. The guide cable has a
tapered structure so that the internal tensile stress is constant along the
cable. On the other hand, we assumed that the cross section of the
moving cable is constant. The analytical model is a lumped mass model,
Fig. 5. Analytical model. and each mass is connected by a spring and a damper. Basically, there
acts tension inside the cable during operation, because gravity is always

( ) ( )
d 2 ri dri 2 GMearth mi |Δri− 1 | − l0 |Δri | − l0 dri dri− 1 dri+1 dri
mi = − 2m i ω × + m i r i ω − r i − ki− 1 Δri− 1 + k i Δr i 　 − ξci − + ξci −
dt2 dt ri 3 Δri− 1 Δri dt dt dt dt
( )
GMsun mi GMsun mi
+ risun − rsun (1)
risun 3 rsun 3
( )
GMmoon mi GMmoon mi
+ 3
rimoon − 3
rmoon
rimoon rmoon

conditions for the counterweight type, and compared them with those of
the climber type. Then, after determining the optimum operating sec­
tions, we analyzed the cable dynamics and confirmed the occurrence of
the interference between cables in that section by using a 3-D upgraded
dynamics model.
acting on the gondola and the cable is expected always to be in a
stretched state. However, the tension of the cable is set to zero when the
cable slackens due to abrupt deceleration. It is assumed that the gon­
dolas touch the guide cable in moving along the guide cable (we ignore
the contact resistance), and the tension in the moving cable also acts on
the guide cable, in the same way as for the gondola and the station.
Therefore, the dynamics of the moving cable influence the dynamics of

134
T. Okino et al. Acta Astronautica 185 (2021) 132–139

Table 1
Analytical conditions.
Distance between mass points [km] (Guide cable) 100
Distance between mass points [km] (Moving cable) (Ground ~ GEO 100
station))
Distance between mass points [km] (Moving cable) (Range 1 & 25
Range 2)
Timestep [s] 0.05
Cable density [kg/m3] 1300
Young’s modulus [GPa] 1000
Taper ratio (Ground: GEO: CW) 1 : 2.4 : 1.7
Guide cable cross section [mm2 ] (Ground) 30
Gondola mass [t] 20
Anchor mass [t] 6564
GEO station mass [t] 4000
Other station mass [t] 100
Distance between moving cables [km] 0.1
Gravitational constant [m3 kg− 1 s− 2 ] 6.674 ×
10− 11
Earth mass [kg] 5.973 × 1024
Solar mass [kg] 1.989 × 1030
Lunar mass [kg] 7.35 × 1022

Fig. 6. Winding velocity of moving cable.

Fig. 7. Winding acceleration of the moving cable in the case in which the
counterweight type is operated between the ground and the Mars grav­
ity center.
Fig. 8. Time variation of cable motion in the case where the counterweight
the guide cable. type is applied to the range from the ground to GEO station.

Coriolis force, elastic force, solar tidal force, and lunar tidal force. In Eq.
3.3. Equation of motion
(1), the first term is Coriolis force, the second is centrifugal force, the
third is gravity, the fourth and fifth are elastic forces, the sixth and
The equation of motion used in this study is shown in Eq. (1). In this
seventh are damping terms, and the eighth and ninth are the tidal forces
study, the motion of the cable is analyzed by solving the equation of
of the sun and moon, respectively.
motion for each mass point considering gravity, centrifugal force,

135
T. Okino et al. Acta Astronautica 185 (2021) 132–139

Fig. 9. Time variation of tension of descending moving cable.

Fig. 10. Power required for driving wheel.

Fig. 12. Relationship between required power and onboard power.

Table 3
Gondola total mass and climber total mass.
Range 1 Range 2

Number of motors 28 8
Climber total mass [t] 31 23
Gondola mass [t] 20 20

Table 4
Ratio of required energy for gondola
ascent to effective available potential
Fig. 11. Power required for driving wheel. energy.
Range 1 2.3
Range 2 10.0
Table 2
Specifications of gondola and climber.
Gondola and climber basic mass [t] including payload 20 Table 5
Motor output (1 unit) [kW] 302 Power consumption. [MWh].
Motor mass (1 unit) [t] 0.393
Range 1 Range 2

Counterweight type 232 271

Climber type 246 102

136
T. Okino et al.
Fig. 13. Time variation of tension from the moving cable.
Fig. 14. Tidal force.
3.4. Analytical conditions
The equation of motion shown in Eq. (1) is solved by the Run­
ge− Kutta method. The guide cable is arranged on the equatorial plane
and fixed to the ground. As an initial condition, the guide cable and
moving cable are straight in the vertical direction. Gondolas are ar­
ranged east− west at the first step analysis, then arranged north− south
at the following analysis with respect to the equatorial plane. The
physical properties of the cable material were based on carbon nano­
tubes [3]. Table 1 shows the main analytical conditions.
In this study, analysis was conducted when the counterweight type
was applied to altitude ranges between the ground and the GEO station
at the first step, then between the ground and Mars gravity center
(altitude 3900 km) (hereinafter, this range is called “range 1”) and be­
tween the Mars gravity center and the Lunar gravity center (altitude
8900 km) (hereinafter, this range is called “range 2”). The analysis was
performed only in the two-dimensional east− west direction at the first
step, then performed in 3-D for the following sections 1 and 2 cases.
3.5. Winding velocity of moving cable
Fig. 6 shows the assumed time variation of the winding velocity of
the moving cable, and Fig. 7 shows the time variation of acceleration/
deceleration of the winding velocity. Since the sine curve is often used
for the acceleration and deceleration of a normal elevator on the ground,
we also applied this variation in this study. If the velocity variation from
the ground to the GEO station is that shown in the figure, the range of
constant velocity time is too long for the acceleration range, and it is
difficult to grasp the variation characteristics of acceleration. Therefore,
Fig. 7 shows the acceleration variation from the ground to the Mars
137
Acta Astronautica 185 (2021) 132–139
gravity center as an example.
3.6. Gondola velocity
The gondola velocity is regulated so as not to exceed the assumed
limit, where the acceleration is determined by the difference between
the tension of the moving cable and the gravitational force of the
gondola as shown in Equation (2). In this analysis, the acceleration co­
efficient was set to 5.0 × 10− 5 and the limit of Gondola velocity was set
to ±110 km/h. These values were determined by doing several pattern
analyses so that the speed change of the gondola was small and abrupt
acceleration and deceleration did not occur.
( )/
a = T + Fg Mg × ac (2)
3.7. Power of the driving wheel
The power of the driving wheel changes depending on the shape of
the cable, but in the present study, it was simply calculated by Eq. (3).
Note that slippage of the cable is ignored.
P = (TU − TD ) × vdep (3)
4. Analytical results
4.1. Analysis in a case where the counterweight type is applied to the
range from the ground to the GEO station
4.1.1. Cable dynamics
Fig. 8 shows the cable dynamics in the east− west direction during
the operation of transportation from the ground to the GEO station.
Since the Coriolis force acts in the right direction to the moving direction
in the rotating coordinates, the moving cable moves in a direction away
from the guide cable. Therefore, interference between cables during
operation does not occur. However, after the operation is stopped, the
moving cable librates greatly in the east− west direction. This is due to
the restoring force by the elastic force of the cable. Fig. 9 shows the time
variation of tension of the descending moving cable. Tension continues
to change for several hours after the moving cable has stopped. This
libration of the moving cable is due to the restoring force. This causes
interference between cables as shown in Fig. 8 (c). The possibility of
interference between cables increases with the length of the moving
cable, because the amplitude of cable libration due to the Coriolis force
increases with the increase of the cable length. So, it is considered
preferable to shorten the moving cable to avoid cable interference. In
addition, such libration occurs in the east− west direction and the Co­
riolis force does not act in the north− south direction, so it is considered
that if the gondolas are arranged north− south with respect to the
equatorial plane, cable interference may be avoided.
4.1.2. Power required for operation
Fig. 10 shows the time variation of the power required for the driving
wheel. The positive region is the power required to wind the moving
cable. The negative region is the power regenerated by braking the
driving wheel. Considering the operation in the section from the ground
to the GEO station, the maximum power of the driving wheel is about
100 MW. The power required to wind up the moving cable depends on
the weight of the counterweight type operates in the whole altitude
range from the ground to the GEO station. Therefore, it is necessary to
shorten the section where the counterweight type is applied, or slowing
the winding velocity of the moving cable for the operation with realistic
power.
T. Okino et al.
Fig. 15. Time variation of cable dynamics in the case where the counterweight type is operated between the ground and Mars gravity center.
4.2. Analysis in a case where the counterweight type is applied to the short
altitude range
From the simulation in a case where the counterweight type is
applied from the ground to the GEO station, it was found that it is better
to arrange the transportation gondolas north− south with respect to the
equatorial plane and to make the moving cable as short as possible.
Therefore, we considered operating an elevator system by dividing the
section from the ground to the GEO station into several ranges.
In the following of this section, we analyzed the performance of a
space elevator system when the counterweight type is applied to short
altitude ranges between the ground and Mars gravity center (altitude
3900 km) (range 1) and between the Mars gravity center and the Lunar
gravity center (altitude 8900 km) (range 2) with the gondolas arranged
in the north− south direction as the initial analysis of a segmented
counterweight-type space elevator, then compared the results with those
of the climber-type case.
4.2.1. Power required for operation
Fig. 11 shows the power required for the driving wheel during
operation of the counterweight type in ranges 1 and 2. The positive
region is the power required to wind the moving cable, whereas the
negative region is the power regenerated by braking the driving wheel.
138
Acta Astronautica 185 (2021) 132–139
The maximum required power was larger in range 1 than in range 2.
The regenerative power was also larger in range 1. These are explained
by the fact that gravity is stronger in range 1 than that in range 2, which
is closer to the earth.
4.2.2. Comparison of power consumption
Table 2 shows the parameters used for power calculation based on
current technology [16]. We assume that the basic masses of a gondola
and climber including payload are the same, but the climber is heavier
than the gondola because it must be equipped with a motor in order to
run on its own. Therefore, the total climber mass is the sum of the
gondola mass and the motor mass. The power required for climbers
increases as gravity increases. The onboard power source of a climber is
required to supply its maximum power at its lowest altitude, and the
mass of a climber must correspond to this maximum power. Fig. 12
shows the relationship between the required power for climbing and the
onboard power of a climber. Table 3 shows the gondola mass and the
climber total mass for the ranges 1 and 2 cases.
In the case of the counterweight type, the potential energy of the
descending gondola is used efficiently, and it is not necessary to supply
much energy to the driving wheels to lift up the upward moving
gondola. The efficiency is estimated by the ratio of consumed energy to
regenerative energy. It can be said that the more closely this ratio
T. Okino et al.
approaches 1, the more efficiently the potential energy of the descend­
ing gondola is used. Table 4 shows the ratio of required energy for
gondola ascent to the effective available potential energy in range 1 and
range 2.
Table 5 shows power consumption during operation when the
counterweight type is applied to ranges 1 and 2 as well as the power
consumption by a climber when the climber type is used in the same
ranges. For comparison under the same conditions, the time variation of
the climber velocity was matched with the time variation of the winding
velocity of the moving cable shown in 2.5. The power consumption was
calculated for the assumed masses of climber and gondola shown in
Table 3.
For range 1, the counterweight type consumes less power for trans­
portation than the climber type. From Table 4, it is considered that this is
caused by the efficient use of the potential energy of the descending
gondola to an ascending gondola. On the other hand, in range 2, the
climber type consumes less power than the counterweight type, because
the potential energy of the descending gondola cannot be used so much
in the higher altitude range where gravity is small.
From this result, it is concluded that the total power consumption for
transportation using a space elevator can be reduced by applying the
counterweight type in the lower altitude range and the climber type in
ranges above that.
4.2.3. Cable dynamics
Fig. 15 shows the time variation of the cable dynamics while the
counterweight type is operated from the ground to the Mars gravity
center. Three-dimensional results are shown separately as the results of
the x− y plane (east− west direction) and the z− y plane (north− south
direction).
The guide cable librated in the east− west direction and the north­
− south direction by the tidal forces of the moon and sun, but the
libration due to the change in the tension from the moving cable was
more dominant in the east− west direction when the gondola was lifted.
Based on Figs. 13 and 14, we see that the force that acts on the guide
cable from the moving cable is larger than the tidal force. This force is
the major cause of cable libration in the east-west direction because it
pulls down the entire system and produces a Coriolis force that causes
libration in the east-west direction. The moving cable also librated due
to tidal forces in both the east− west and north− south directions. During
operation, the Coriolis force was also generated in the east− west di­
rection when the cable and gondolas ascended and descended, and these
forces complicated the libration of cables.
In this analytical condition, there was no interference between cables
during operation, but interference occurred due to the tidal force and the
restoring force after the operation was stopped.
5. Summary
In this study, we analyzed the performance of a counterweight-type
space elevator. We found that strong interference between cables occurs
and that the power required for the driving wheel becomes unrealisti­
cally large when the counterweight type is applied to the whole altitude
139
Acta Astronautica 185 (2021) 132–139
from the ground to the GEO station.
However, we also found that the total power consumption for
transportation can be reduced by using the counterweight type in the
low altitude range from the ground to the Mars gravity center (altitude
3900 km) where the potential energy of gondola can be effectively used
for operation, compared with the system where the climber type is
applied to the whole altitude.
The cable dynamics of the counterweight type in the low altitude
range were also analyzed by using a 3-D dynamics simulation model.
The results show that, as results of motion of the cable and the gondola,
cable libration becomes complicated by libration in the north− south
direction caused by tidal forces in addition to libration in the east− west
direction caused by the Coriolis force. Thus, there is a risk of cable
interface.
In the future, it will be necessary to develop a method to prevent
interference between the cables, such as by improving the gondola
operation strategy, or by increasing the distance between the guide
cable and the moving cable.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
References
[1] B.C. Edwards, The Space Elevator, NIAC Phase II Final Report, 2003. http://www.
niac.usra.edu/files/studies/final_report/521Edwards.pdf#search=’The+Space+
Elevator%2C+NIAC+Phase+II+Final+Report.
[2] S. Iijima, Helical microtubules of graphitic carbon, Nature 354 (1991) pp56–58.
[3] S. Yahachi, Introductory Materials Science of Carbon Nanotubes, 2005, pp. 37–41.
[4] Y. Ishikawa, K. Otsuka, Y. Yamagiwa, H. Doi, Effects of ascending and descending
climbers on space elevator cable dynamics, Acta Astronaut. 145 (2018) 165–173.
[5] D.D. Lang, Space Elevator Dynamic Response to In-Transit Climbers, 1st
International Conference on Science, Engineering, and Habitation in Space, Space
Engineering and Science Inst., Albuquerque, NM, 2006. Paper 10152148.
[6] S.S. Cohen, A.K. Misra, The effect of climber transit on the space elevator, Acta
Astronaut. 64 (2009) 538–553.
[7] P. Williams, W. Ockels, Climber Motion Optimization for the Tethered Space
Elevator, AIAA/AAS Astrodynamic Specialist Conference and Exhibit, 2008,
pp. 1–17. AIAA paper 2008-7383.
[8] E. Fujiwara, Transmission technology by microwave, JIEEF 129 (2009) 418F.
[9] B.C. Edwards, E.A. Westling, The Space Elevator: A Revolutionary Earth-To-Space
Transportation System, 2003, 0-9746517-1-0.
[10] T. Takano, Study about energy supply method to space elevator, in: Report of
Review Committee for Orbital Elevator, Japan Society for Aeronautical and Space
Sciences, 2015 (in Japanese).
[11] B. Hu, H. Li, T. Li, H. Wang, Y. Zhou, X. Zhau, X. Hu, X. Du, Y. Zhao, X. Li, T. Qi,
M. Helaoui, W. Chen, F. Ghannouchi, A long-distance high-power microwave
wireless power transmission System based on asymmetrical resonant magnetron
and cyclotron-wave rectifier, Energy Rep. 7 (2021) 1154–1161.
[12] Y. Aoki, Basic research on space elevator climbers and prototype design of self-
supporting probe climbers, J. Jpn. Soc. Aeronaut. Space Sci. 62 (2014) 325–330.
[13] A. Bolonkin, Electrostatic Climber for Space Elevator and Launcher, 2007,
pp. 1–12. AIAA-2007-5838.
[14] G. Li, Z.H. Zhu, G. Shi, A novel looped space tether transportation system with
multiple climbers for high efficiency, Acta Astronaut. 179 (2021) 253–265.
[15] Obayashi Corporation, Quarterly Ohbayashi 53 (2012).
[16] TOYO DENKI. https://www.toyodenki.co.jp/products/transport/train/motor.php.