Engineering Structures 229 (2021) 111623

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Modeling verification of an advanced torsional spring for tracked vehicle

suspension in 2S1 vehicle model
Tomasz Nabagło a, *, Andrzej Jurkiewicz b, Janusz Kowal b
a
Faculty of Electrical Engineering, Automatics, Computer Science and Biomedical Engineering, AGH University of Science and Technology, Al. A. Mickiewicza 30, 30-
059 Krakow, Poland
b
Faculty of Mechanical Engineering and Robotics, AGH University of Science and Technology, Al. A. Mickiewicza 30, 30-059 Krakow, Poland

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

Keywords: In the article, the authors present an advanced model of hyperbolic torsion spring based on the Finite Element
Torsion spring Method (FEM). This model takes into account the contact forces between the spring arms, which allows you to
Tracked vehicle suspension simulate its operation even in the case of a large twisting angle. On the basis of this FEM model, the spring
Finite element method
stiffness characteristics were constructed. Based on this characteristics, the existing 2S1 track platform model has
ADAMS
Flexible body
been rebuilt. In this model, a much higher mass of the body is allowed than in the original model. On the basis of
the elastic spring model, parameters such as the static deflection of the suspension have also been determined, as
well as its range of mobility compared to the original suspension based on torsion bars. In addition, spring
deflection phases have been presented in terms of internal stresses, which are important for the strength analysis
of the spring. The originality of the described solution lies in the unique design of the so-called a hyperbolic
spring. This design has not been previously described in the literature.

1. Introduction
The solution of the original passive suspension system of the 2S1
tracked platform is based on torsion bars. These bars are problematic
elements. Problems with these elements are associated with their length
and location of them in the vehicle. Because of their reactive character,
which during explosion of an anti-tank mine may destroy the suspension
system, they should be replaced with another solution. This destruction
could be also harmful for a crew of the vehicle. An alternative solution
may allow to replace reactive elements from under floor space to the
vehicle sides. One of possible solutions of the problem is application of
the helical springs. They were analyzed by Taktak et al. [17], with usage
of the Finite Element Method (FEM). FEM computer applications are a
good solution also for the analysis of other kinds of springs. This is very
popular method of flexible bodies analysis, but there are also others, like
method applied by Mellouli et al. [10] for free vibration analysis, the
mesh-free Radial Point Interpolation Method. The helical springs could
be problematic in application to the tracked vehicle, which was previ­
ously based on torsion bars. The suspension of the vehicle would require
more interference in its structure than in the case of a torsional system
with other parameters. The FEM model application for various kinds of
springs was developed in collective work edited by Shimoseki et al. [16].
A solution worth considering is the use of torsion springs. At this point, a
question appears. This question is associated with the kind and con­
struction of above mentioned torsion springs. The answer will be put
forward by us in the first section of our paper. But before this, we have
analyzed constructions of full tracked vehicles and also modeling solu­
tions, which may be applied in the tracked vehicle model. A complex
vehicle model was described by Campanelli and Shabana [3]. For
analysis of the track link natural frequencies mode they use FEM. Less
complex models of tracked vehicles were described by Nabagło et al.
[11,12]. The model developed in [12] was based on proving ground
measurements described by Jurkiewicz et al. [4]. It included many
important elements of tracked vehicle model. One of the most important
problem in this area is associated with joints between track segments.
Ryu et al. [15] take under consideration single pin track link. He de­
scribes the so-called rubber bushing model, which connects two track
links. Lee [6] in his article describes contact forces between wheels,
track links and ground surface. There is also explained a numerical
method of force contacts calculation. Details of this connection forces,
contact forces between wheels, track links and road surface and also
FEM model construction of the flexible body are described in ADAMS
Documentation [2]. Selected parts, with high fidelity of real parts
shapes, may be designed in other CAD programs. Łukaszewicz [7] in his

* Corresponding author.
E-mail address: nabaglo@agh.edu.pl (T. Nabagło).

https://doi.org/10.1016/j.engstruct.2020.111623
Received 11 April 2020; Received in revised form 27 September 2020; Accepted 17 November 2020
Available online 13 December 2020
0141-0296/© 2020 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
T. Nabagło et al.
article, explains multi-body CAD designing problem for simulation ap­
plications. For springs, the failures and fatigue damage are very serious
problem. The expression of this problem was described by Prawoto et al.
[13]. They analyze not only construction but also materials from which
the springs are built. As it was said, the basic approach to the modeling
solution is also important. The authors of the presented article used tools
provided by MSC.ADAMS and MATALB programs, but they were also
familiar with other solutions. Because the contact between spring arms
is a serious problem in the torsional springs models, very valuable in­
formation was put forward by Lankarani et al. [5]. They described a
detailed parameterized model of the contact force. This model is usually
applied in the ADAMS program. Above mentioned models of subsystems
may be connected as parts of tracked vehicle model applied in the
dynamical simulation, similar to the simulation described by Wong [18].
In our paper we try to show that the use of hyperbolic springs in­
creases the safety of the crew, allows to significantly increase the weight
of the vehicle, and at the same time to maintain the stability of the
vehicle at the same, or even slightly higher than in the case of the sus­
pension based on torsion bars. We would like to emphasize that the very
idea of a hyperbolic spring is so new and unique that at the present stage
of the research presentation we do not discuss the topic of improving the
vehicle dynamics, but only show that it could replace the torsion bars,
used so far. The optimization of the spring structure in terms of
improving the dynamics of the vehicle is planned in further research.
This is the first work on full view of construction of the hyperbolic spring
which we have published. For comparison, other solutions of spiral
torsion springs are presented [3] (see Figs. 1 and 2). The mechanical
structure of the hyperbolic spring discussed earlier has not been found in
the literature.
2. Analysis and development of existing suspension systems
During research many constructions of the torsion spring were
analyzed, such as the Archimedes spiral shape spring (see Fig. 1a),
logarithmic spiral shape spring (see Fig. 1b), and here presented hy­
perbolic shape spring. The torsion springs package assembly system
compared to the torsion bar assembly system is shown in the Fig. 2. The
main aim of research was to construct a torsion spring with stiffness
characteristics as similar to initially used torsion bar characteristics as
possible. Because of failures and fatigue damage of previously tested
springs, it was not easy. Problems were associated with construction but
also with mounting of the springs external ends. Finally, we have used
an advanced mounting system, which solved these problems. This
mounting system was associated with the sliding contact between spring
arms and pins placed in triangular grooves on their ends. This solution
caused the stiffness characteristics to be close to the torsion bar char­
acteristics but in a very narrow range of suspension displacement. Above
an angle of approximately 20 degrees, in the spring, the contact forces
between spring arms appeared. This phenomena caused change in the
spring characteristics, turning it into a nonlinear stiffness characteris­
tics. In the case of logarithmic shape springs, effects of the fatigue pro­
cesses appeared, which finally caused damage of the spring arms in close
Fig. 1. Archimedes spiral shape (a) and logarithmic spiral shape (b)
springs [4].
2
Engineering Structures 229 (2021) 111623
Fig. 2. Mounting system of the torsion bar (a) and the torsion springs package
(b) [4].
neighborhood of the suspension arm axle.
Problems of too early appearing contacts between spring arms were
overcome by the application of hyperbolic shape springs, which will be
described in detail in the next section.
3. Modeling of the suspension system
Finally, the designed hyperbolic shape spring should be applied and
verified in the tracked vehicle suspension system. This system was
already built and described in articles [4,11,12]. In the articles the full
tracked vehicle models are described. There are models with torsion
bars, and also with torsion spring packages. In the article a concept of
the springs package is applied because of final construction solution of
the spring. It is assumed that the final spring is built from a series of
thinner springs put in the parallel way. This construction was many
times tested in the previous springs applications described before. If the
packages of the springs are in an oil environment, the oil conducts heat
out from the springs as a kind of spring cooling system. As a supplement
to the cooling system, usage of the piezoelectric surface, proposed by
Maleek et al. [8], might be interesting.
To get a complete picture of the modeling problem, we used various
and complementary simulation environments in the modeling process.
During the simulation tests, ADAMS/View software as also ADAMS/Flex
module were used. The ADAMS/View and Marc Mentat software were
used to develop the stress analysis results. In the process of engineering
calculations, the Matlab program was used for the initial calculation of
the model parameters.
3.1. Modeling assumptions
To construct the stiffness characteristics of the torsion bar, the
characteristics reduced to the axle of the road wheel was measured. This
was measured on real suspension unit of the 2S1 vehicle. This charac­
teristics initially has so called tolerant area which was associated with
hysteresis loop. This loop was the result of dry friction, which caused an
additional damping effect. Therefore the mean values from this tolerant
area are equal to the values of stiffness characteristics of the suspension
system. The result is shown in Fig. 3a, as the compression spring stiffness
characteristics. Because this compression spring is theoretically placed
between wheel axle point and the vehicle body, the real torsional stiff­
ness had to be calculated (see Fig. 4 a and b). As a result of this calcu­
lation the torsional spring stiffness characteristics was presented in
Fig. 3b. This characteristics has become a reference for the later ob­
tained simulation results.
The torsional stiffness characteristics in the Fig. 3b was calculated
according Eqs. (1) with data presented in the Fig. 3a. Shown here lover
limit tolerance area and upper limit tolerance area are part of a whole
hysteresis loop measured during experiment on a test stand including a
real 2S1 vehicle. Stiffness characteristics in the Fig. 3a is arithmetic
mean from this hysteresis values.
T(α(z) + α0 ) = d⋅cos(α(z))⋅F(z) (1)
( )
z− z0
where: z0 = d⋅sin(α0 ) and α(z) = arcsin d
T. Nabagło et al. Engineering Structures 229 (2021) 111623

Mu sprung mass of the single suspension unit.

Mu is part of the sprung mass associated with a single suspension

unit. Because of 14 road wheels, the vehicle has 14 units like this. The
unit sprung mass is described by Eq. (3)
ku
Mu = (3)
4π2 f 2n

When linear stiffness, based on the characteristics (see Fig. 3a), is

equal to 50 kN/m, then single unit sprung mass is equal to Mu = 810 kg.
This mass must be multiplied by 14 and the whole sprung mass of the
vehicle should be equal to 11.35⋅103 kg. Finally this mass was assumed
for the vehicle model with the torsion bars.
Based on simulation results of hyperbolic spring (see Fig. 5b), the
stiffness characteristics of this spring was calculated. As it was said
before the problems with too early appearing contacts between spring
arms were overcome by the application of so called hyperbolic shape
spring with advanced system of leading links (see Fig. 5). These links
lead external ends of the spring arms in order to achieve an appropriate
stiffness characteristics. The shape of this characteristics is identical to
characteristics of the torsion bar multiplied by constant equal to 2.6. In
other words, the stiffness of the torsion spring is 2.6 times bigger than
the stiffness of the torsion bar, but with the same nature of changes,
when the angle of deflection increases. Torsion bar characteristics (blue)
in the Fig. 5a was calculated based on the real torsion bar suspension
deflection and Eq. (1), but torsion spring characteristics (red) was taken
from the hyperbolic torsion spring model constructed as MSC.ADAMS

Fig. 3. Suspension stiffness reduced on the wheel axle (a) stiffness character­
istics of the torsionbar (b).

Fig. 4. Representation of the suspension stiffness (a) geometric representa­

tion (b).

Based on the original 2S1 parameters, the natural frequency of the

sprung mass of the vehicle should be approximately fn = 1,25 Hz. This
frequency is in range of optimal frequency of automotive suspension
construction for the human body. Reimpell et al. [14] state that for the
sake of comfort, f ≈ 1 Hz is recommended for the natural frequency of a
vehicle and they give an acceptable range of 1 to 1.5 Hz. On the base of
Eq. (2), the whole sprung mass of the vehicle was calculated.
√̅̅̅̅̅̅
1 ku
fn = ⋅ (2)
2π Mu

where:

fn natural frequency of the system,

ku suspension stiffness coefficient, Fig. 5. Comparison of the stiffness characteristics (a) and hyperbolic shape
spring (b).

3
T. Nabagło et al. Engineering Structures 229 (2021) 111623

flexible model. Contact forces appear above an angle of 40.42 degrees equal to 10 kNs/m. This damping is reached after exceeding a pene­
(see Fig. 8), but without noticeable influence. Radical increase of stiff­ tration, which is equal to 0.1 mm. Details of this contact model was
ness, because of contact influence, should appear after exceeding the described by Lankarani et al. [5]. These contact parameters are
value of 60 degrees. approximately equal to parameters of the contact between two surfaces
The simulated model is described in the Section 3.2. The spring of steel elements. This contact model neglects the friction phenomena.
system shown in Fig. 5b is built with the hyperbolic spring and sets of This situation is close to real conditions when steel torsional springs are
guide links connected between spring and its housing through revolute immersed in oil and friction may be neglected. The FEM algorithm in its
joints. The housing in the form of a rigid drum is rigidly connected to the generalized form is described below. It realizes subsequent steps:
vehicle hull. The torsion axle of the hyperbolic springs is also the axle of
suspension deflection (see Fig. 6). As is shown in Fig. 5a the hyperbolic 1. Discretization and meshing,
shape spring is 2.6 times stiffer than the torsion bar. Then the linear 2. Formulation of stiffness matrices for structural elements,
spring on the wheel axle, for the hyperbolic spring system is reduced to 3. Assembly of the global stiffness matrix,
130 kN/m. In continuation of this way of thinking and based on Eq. (3), 4. Assembly of the global load vector,
the unit sprung mass is equal to 2, 107 kg and the full sprung mass 5. Definition of boundary conditions.
should be equal to 29.5⋅103 kg. These parameters may be applied in the 6. Solving the system of equations,
full vehicle model. Static displacement with such mass and stiffness is 7. Determining internal forces and reactions.
described by Eq. (4) and it is about 0.16 m.
Given the stiffness matrix K, the static equation has the form (6).
Mu ⋅g
zst = (4)
ku K⋅x = F (6)
(z − z )
αst = α0 − arcsin 0 st
(5) where:
d
K – stiffness matrix,
where:
x – displacements vector,
F – force vector.
g gravitational acceleration, 9.81 m/s2 .
Such kind of model may be simulated in the Marc program, but as
This translational displacement is adequate to the angular displace­
was said before, based on elements of this model also ADAMS flexible
ment, which is about 30 degrees (see Eq. 5). These parameters are
body model may be performed. Chosen steps of the simulation of the
important for estimation of preload position in the full vehicle model.
above described ADAMS flexible model are shown in Fig. 7. It begins
Contrary to the characteristics of the torsion bar, which were
with the initial position of the torsional spring for angle 0 degrees (see
calculated from the real suspension system, only a computer model of
the hyperbolic spring was performed.

3.2. Simulation of the FEM spring model

In the presented paper, dynamic analysis of flexible body in the

ADAMS program was performed, but in order to realize the FEM anal­
ysis, the authors also performed a co-simulation in the MSC.MARC and
ADAMS/View programs. To estimate technical parameters of the hy­
perbolic shape spring, the FEM model of the spring was built and
simulated. Based on this model also ADAMS flexible model was per­
formed. For the physical parameters of the spring, there was assumed
that it was made of 50HS steel. The following parameters were assumed
for this steel: Mass density is 7850 kg/m3 , Young’s modulus is 200 GPa,
Poisson’s ratio is 0.3. Traditional FEM model as also the ADAMS flexible
model of a spring is problematic, because during simulation when the
spring is twisted, the subsequent arms of the spring interpenetrate. The
model presented in our paper is enriched with contact elements between
subsequent arms. The contact stiffness is equal to 100, 000 kN/m with
force exponent coefficient is 2.2, which is described in ADAMS/Solver
Documentation [2]. This contact has also damping coefficient, which is

Fig. 6. Torsion spring in suspension system unit (a), torsion spring construction
details (b). Fig. 7. Subsequent steps of simulation of the ADAMS flexible model.

4
T. Nabagło et al. Engineering Structures 229 (2021) 111623

Fig. 7a). In the figures from 7b to d, the results of the classic flexible body
simulation are presented. In Fig. 7d, the material deformation may be
noticed, especially on the external ends of the spring arms.
After Fig. 7e, where angle of 40 degrees was gained, the influence of
the contact forces appears. Vectors of these forces are described with red
arrows and the arrow length is proportional to the values of the forces. It
may be noticed in comparison with Fig. 8, where the stiffness charac­
teristics of the torsional spring is presented, the contacts have negligible
influence on the stiffness characteristics shape. This is only true to the
moment when external the ends touch the closed spring. This condition
is fulfilled when the angle exceeds 60 degrees. After exceeding this
angle, stiffness should radically increase, which may be adequate to the
situation when the system is blocked. In Fig. 7f, high deformation are
visible. They are visible on the external ends of the spring arms. It may
suggest the necessity of making some changes in the spring construction.
From the simulation effect of the above model we obtained the stiffness
characteristics of the hyperbolic spring.
The FEM model implemented in the MARC program allowed for the
stresses analysis in the hyperbolic spring [9]. In this model, like in the
previous flexible model, the contact forces between the spring arms are
implemented. Fig. 9 shows the visualization of stresses for a suspension
deflection angle of 50 degrees. We can observe here the most stressed
parts of the spring, which can fail due to material fatigue during long-
term operation.
The co-simulation model with the ADAMS/View software allowed
for the presentation of the forces affecting the suspension mechanism
from the hyperbolic spring [1] (see Fig. 10). The both figures show the
result for the angle of 50 degrees because it was the maximum angle
obtained during the simulation, which allowed to observe both the Fig. 9. Equivalent of stress presented on the FEM model.
greatest forces and stresses in the mechanical system of the spring.

3.3. Full vehicle model

In the full vehicle model many factors have to be taken into

consideration. There are model parameters but also environment fac­
tors, like for example ground stiffness. This element has huge influence
on the ground-track contact parameters, which is tightly associated with
the vehicle dynamics [18]. In the model, it is assumed that the contact
forces have linear characteristics. The contact forces appear between
wheels and track links and also between the track links and the ground.
The contact is characterized by two main parameters, namely stiffness
and damping. The stiffness parameter of contacts between wheels and
tracks is equal to 200 kN/m, the damping is equal to 2 kNs/m. The
stiffness parameter of contacts between tracks and ground is equal to
1000 kN/m, the damping parameter is also equal to 2 kNs/m.
In the final torsional spring element, because of the heat exchanging

Fig. 10. Forces acting from the spring FEM model.

problem, the above described hyperbolic shape spring will be divided

into several parallel connected torsional springs. They will be connected
on the so called spring packages. In the full vehicle model, the packages
are mounted in place of the removed torsion bars. The spring packages
are mounted outside the hull, close to cooperating suspension arms.
Since the number of torsion bars present in the base model is 14 (see
Fig. 11a), they have been replaced in the modernized vehicle model by
hyperbolic spring packages, the number of which is also 14 (see
Fig. 11b).
Because in the full vehicle model springs are installed at preload
position which is known from the static equilibrium position, the au­
thors decided not to test the single spring at preload. From static
Fig. 8. Moment of first contact between spring arms.

5
T. Nabagło et al.
Fig. 11. Model of the full tracked vehicle.
equilibrium position we know, that an angle at preload is about 30 de­
grees. In relation to this, as is apparent from the suspension deflection in
the static equilibrium state, this deflection is reached for an angle of 30
degrees. From Sections 3.1 and 3.2 we know, that a dramatic increase of
the stiffness appears after exceeding an angle of 60 degrees. Between
these two angle positions, the range of 30 degrees exists. Therefore, the
deflection of 30 degrees is an angle from the static equilibrium position
to the position of the bump stops. Thus, the mentioned increase of
stiffness, perhaps may even support the effects of bumpstops activity.
In the full vehicle model a stiffness characteristics taken from the
spring flexible model was used. Exact characteristics with its extrapo­
lation used in simulation is presented in the Fig. 12 (red/upper curve).
The same was done in the case of the basic vehicle model, where the
stiffness of the torsion bars are based the blue/lower characteristics,
from the same figure. This figure expressed also, how stiffness is
extrapolated in the full vehicle model. Red/upper characteristics de­
scribes stiffness of the hyperbolic torsion spring, but blue/lower one
describes characteristics of the real torsion bar. Extrapolation in the
Fig. 12. Torsion stiffness characteristics.
6
Engineering Structures 229 (2021) 111623
spring model, contrary to appearances, is very important. The lack of it
during the simulation of the full vehicle model may cause problems with
implementation of the simulation.
Thus additional free space is provided. This space could be used for
installing better floor armor or other equipment. This solution improves
also safety, because all reactive elements of the suspension system are
removed from the floor plate and they are placed at both sides of the
hull, where they may be better covered. As was said before, in the model
with torsion bars, the mass of the hull was assumed 11.35⋅103 kg. Ac­
cording to calculations from Section 3.1, the hull mass increased to 29.5⋅
103 kg and was applied in the modernized model (see Fig. 11b). In the
previously described models, other forces were applied. Idler wheel
mechanism acts on the track with a constant force, which is equal to
39 kN. This force acts in longitudinal axis of the vehicle, as is shown in
Fig. 11 a and b. This force is used for maintaining a constant tension of
the tracks. This is a simplified version of real idler wheel mechanism
which is based on an eccentric mechanism. Forces, acting on the idler
wheel from the track side during acceleration or deceleration, may be
higher than the idler force value. In this situation these forces cannot be
balanced with 39 kN. As a result, the idler wheel may be displace to­
wards the front of the vehicle. For this reason, a bumpstop is used to stop
the movement of the idler wheel. This stops the idler after a displace­
ment of 100 mm from the construction position. These basic and
modernized suspension system models will be applied for future simu­
lations of the behavior of the full vehicle models with torsion bars and
hyperbolic shape springs.
3.4. Simulation results of the full vehicle model
As it was said previously, the suspension stiffness increased after
hyperbolic spring application. Torsion bars are replaced with hyperbolic
springs which give additional space inside the 2S1 vehicle, and lowers
overall mass. According to the models calculation, mass of single torsion
bar made from steel is about 35 kg, mass of single Archimedes spiral
spring system is about 21 kg but mass of single hyperbolic spring system
is only about 13.5 kg. Aggregate mass of the hyperbolic are less than
40% of the aggregate mass of torsion bars. In comparison, 14 torsion
bars weigh approximately 490 kg, when hyperbolic springs are used, the
weight of springs system decreased to 189 kg. The 14 torsion bars are
elements of the suspension system of the full vehicle model presented in
Fig. 11a. It should be remembered that the entire suspension stiffness
increases about 2.6 times when torsion bars are replaced with the hy­
perbolic springs, which is observable in Fig. 12, where torsion stiffness
characteristics are compared.
Due to increased stiffness, stability of the vehicle should be
increased. This may be demonstrated when simulating each vehicle
traversing a 71 mm triangle obstacle under the right track with a vehicle
velocity of 20 km/h (see Fig. 13).
Result of this simulation is shown in Figs. 14 and 15. In the Fig. 14,
roll rate decreased and pitch and yaw rate are comparably the same.
This demonstrates another advantage of hyperbolic spring usage when
stiffness and stability of the vehicle may increase when the entire mass
of the vehicle suspension system decreases.
Fig. 15 demonstrates the vertical forces on road wheel axles for the
vehicle model with the torsion bars and the model with the hyperbolic
springs. It should be remembered that because the stiffness of suspension
with the hyperbolic springs increased 2.6 times, the hull body mass was
also increased about 2.6 times. Therefore, the forces on the wheels of a
vehicle with hyperbolic springs will have much higher values. There­
fore, we only analyze the nature and shape of these forces. But even on
the basis of such a simple observation, the stabilizing effect of the sus­
pension system based on the hyperbolic springs can be noticed.
We would like to emphasize that our main intention is to focus on the
hyperbolic torsion spring in the context of its application to a tracked
vehicle. If there is stabilization effect to the suspension with the
T. Nabagło et al. Engineering Structures 229 (2021) 111623

hyperbolic springs, this is not intended. Our intention is to show that the
new solution of hyperbolic springs can successfully replace torsion bars,
bringing with it additional benefits.

4. Conclusion

Summarizing the results of the simulation, the pros and cons of the
presented solution of the hyperbolic torsion spring should be put
forward.
Undoubtedly, the advantages of the solution are the dimensions and
the stiffness characteristic, which reflects the shape of stiffness charac­
teristics of the torsion bar. These torsion bars were originally mounted in
the vehicle. Because of the higher stiffness of the torsion spring packages
over the stiffness of the traditional torsion bars, the constructors can
increase the vehicle hull mass. This is another advantage of the hyper­
bolic springs. Because of their higher stiffness, which allows to signifi­
cantly increase the weight of the vehicle body, the vehicle will be able to
be used as a platform for heavy guns or as a transport vehicle for
transportation of heavy equipment. As is known in the case of military
vehicles, this is a huge advantage. The hyperbolic springs packages al­
lows also to resign from axial offsets of the road wheels, so the right and
left wheels may be mounted on the same axles. Usage of the spring
packages gives more space between wheels. This space may be used for
additional equipment or even additional armor. Usage of spiral springs
Fig. 13. Single obstacle being crossed.
also increases the safety level, because the torsion bars solutions are
associated with reactive elements of suspension under the floor of crew
cabin. During explosion of a charge placed under the vehicle, these
reactive elements could be harmful for the crew. The spiral spring
packages are placed on both side of the vehicle and can be covered. As
there was shown in the case of the full vehicle model simulations, the use
of hyperbolic springs can even slightly improve vehicle stability.
A disadvantage of the proposed torsion springs system is the limited
range of spring torsion, which can reach a maximum angle of 60 degrees.
Nevertheless, this range is obtained before obtaining a bumpstops po­
sition. For this reason, the aforementioned disadvantage may be negli­
gible under conditions assumed in the full vehicle models. A
phenomenon, which should be noted, is the material tension, which
appears on the ends of the spring arms. But it can be significantly
reduced by using the appropriate design solutions.

CRediT authorship contribution statement

Tomasz Nabagło: Methodology, Software, Writing - original draft.

Andrzej Jurkiewicz: Conceptualization. Janusz Kowal: Supervision.
Fig. 14. Angular velocities comparison.

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] ACSI Adams ver. 2016, Help Documentation assoc. with Adams Co-Simulation
Interface. MSC.Software, 2016.
[2] ADAMS ver. 2016, Help Documentation assoc. with ADAMS/View and Solver.
MSC.Software, 2016.
[3] Campanelli M, Shabana A, Choi J. Chain vibration and dynamic stress in three-
dimensional multibody tracked vehicles. Multibody Syst Dyn 1998;2:277–316.
https://doi.org/10.1023/A:1009758701296.
[4] Jurkiewicz A, Nabagło T, Kowal J, Apostoł M. A new suspension system of an
autonomous caterpillar platform. J Theoret Appl Mech 2014;52:857–67.
[5] Lankarani HM, Nikravesh PE. Continuous contact force models for impact analysis
in multibody systems. Nonlinear Dyn 1994;5:193–207.
[6] Lee K. A numerical method for dynamic analysis of tracked vehicles of high
mobility. KSME Int J 2000;14:1028–40.
[7] Łukaszewicz A. Modelling of solid part using multibody techniques in parametric
Fig. 15. Comparison of vertical forces on wheel axles. cad systems. In: Solid State Phenomena. Trans Tech Publ.; 2009. p. 924–9.

7
T. Nabagło et al.
[8] Mallek H, Jrad H, Wali M, Kessentini A, Gamaoun F, Dammak F. Dynamic analysis
of functionally graded carbon nanotube–reinforced shell structures with
piezoelectric layers under dynamic loads. J Vib Control 2020;26:1157–72.
[9] Marc Mentat ver. 2016, Help Documentation assoc. with Marc and Mentat. MSC.
Software, 2016.
[10] Mellouli H, Jrad H, Wali M, Dammak F. Free vibration analysis of fg-cntrc shell
structures using the meshfree radial point interpolation method. Comput Mathe
Appl. 2020;79:3160–78. http://www.sciencedirect.com/science/article/pii/
S0898122120300316, doi: 10.1016/j.camwa.2020.01.015.
[11] Nabaglo T, Kowal J, Jurkiewicz A. Construction of a parametrized tracked vehicle
model and its simulation in msc. adams program. J Low Freq Noise Vib Active
Control 2013;32:167–73.
Engineering Structures 229 (2021) 111623
[12] Nabagło T, Kowal J, Jurkiewicz A. High fidelity model construction and its
verification based on 2s1 tracked vehicle. In: Innovative control systems for
tracked vehicle platforms. Springer; 2014. p. 205–15.
[13] Prawoto Y, Ikeda M, Manville S, Nishikawa A. Design and failure modes of
automotive suspension springs. Eng Failure Anal 2008;15:1155–74.
[14] Reimpell J, Stoll H, Betzler JW. The automotive chassis: Engineering principles,
1995.
[15] Ryu H, Bae D, Choi J, Shabana A. A compliant track link model for high-speed,
high-mobility tracked vehicles. Int J Num Methods Eng 2000;48:1481–502.
[16] Shimoseki M, Kuwabara T, Hamano T, Imaizumi T. FEM for Springs. Springer
Science & Business Media; 2003.
[17] Taktak M, Dammak F, Abid S, Haddar M. A finite element for dynamic analysis of a
cylindrical isotropic helical spring. J Mech Mater Struct 2008;3:641–58.
[18] Wong J. Dynamics of tracked vehicles. Vehicle Syst Dyn 1997;28:197–219.