0% found this document useful (0 votes)
27 views21 pages

Tensegrity Truss Mechanical Properties Analysis

Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
27 views21 pages

Tensegrity Truss Mechanical Properties Analysis

Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

European Journal of Mechanics / A Solids 83 (2020) 103998

Contents lists available at ScienceDirect

European Journal of Mechanics / A Solids


journal homepage: http://www.elsevier.com/locate/ejmsol

Equivalent mechanical properties of tensegrity truss structures with


self-stress included
Wojciech Gilewski *, Anna Al Sabouni-Zawadzka
Faculty of Civil Engineering, Warsaw University of Technology, Al. Armii Ludowej 16, 00-637 Warsaw, Poland

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

Keywords: Due to the complex structure of tensegrity systems it is often hard to understand their behaviour and estimate
Tensegrity their mechanical properties. As a result, they are rarely used in real engineering applications. The present paper
Mechanical properties focuses on the evaluation of equivalent mechanical properties of various tensegrity modules based on a con­
Continuum model
tinuum model. The aim of the applied model is to facilitate the identification and understanding of mechanical
characteristics of tensegrities through their comparison with a continuum body with equivalent features. The
model is built by assuming that the strain energy of an unsupported tensegrity module or structure is equivalent
to the strain energy of a solid. The approach enables to estimate the influence of self-stress on deformation of the
structure, identify the influence of cables and struts on the properties of the whole system, determine equivalent
mechanical characteristics, such as Young’s moduli, shear moduli, Poisson’s ratios etc., and conditions that limit
their values. What is more, a qualitative as well as quantitative evaluation and comparison of mechanical
properties for various tensegrity modules and systems is possible with the use of the applied technique. A
comprehensive analysis of typical tensegrity modules with various elastic symmetries is presented. Using the
adopted model, mechanical characteristics are determined and graphs of identified mechanical coefficients for
five typical tensegrity modules are presented in relation to the self-stress multiplier and cable to strut properties
ratio. The analysis can be useful for the development of extreme mechanical properties of smart tensegrity-
inspired 2D or 3D lattices or metamaterials.

1. Introduction (Ptochos and Labeas, 2012) and nanostructures (Odegard et al., 2002).
Few works on continuum models of tensegrity structures can be found.
Numerous works on the development of continuum models of bar In the paper (Luo et al., 2008) an attempt of modelling a continuum
systems as well as bar models of continuum structures can be found in biological structure using a discrete tensegrity grid was made. Kebiche
the literature on mechanics of structures and materials. The aim of the et al. (2008) developed a bar model of a tensegrity structure with
development of such models is usually connected with various self-stress included. Gilewski and Kasprzak (2014) presented the idea of
computational problems and complexity of the analysis of structures a continuum model of tensegrities, which is based on the comparison
and materials. Typically, different homogenization techniques (Ara­ between the strain energy of a tensegrity module and a continuum
bnejad and Pasini, 2013) together with an experimental verification body of a certain volume. They developed a method for obtaining an
(Salehian and Inman, 2008) are used to build such models. One of the equivalent elasticity matrix. However, no practical conclusions can be
common problems is the development of continuum models of bar, drawn from their work, as they did not determine mechanical prop­
plate or shell grids (Noor and Anderson, 1979; Noor and Anderson, erties of the analysed structures. Continuum models of hexagonal lat­
1978; Teughels and De Roeck, 2000) and sandwich plates/shells (Wang tices within geometrically non-linear models are proposed in (Pal et al.,
and Chung, 2011), where equivalent stiffness moduli are determined. 2016a, 2016b).
Continuum models are also used to analyse structures with repeating For the purpose of this paper tensegrities are defined as pin-jointed
geometry (Dow et al., 1985; Ben Kahla, 1995), including micro- systems consisting of isolated compressed elements (struts) inside a

* Corresponding author.
E-mail addresses: [email protected] (W. Gilewski), [email protected] (A. Al Sabouni-Zawadzka).

https://doi.org/10.1016/j.euromechsol.2020.103998
Received 28 July 2019; Received in revised form 9 March 2020; Accepted 23 March 2020
Available online 7 April 2020
0997-7538/© 2020 Elsevier Masson SAS. All rights reserved.
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

continuous net of tensioned members (cables) (Motro, 2003; Skelton 3D elasticity theory (LTE) (Green and Zerna, 1970). The strain energy of
and de Oliveira, 2009; Tran and Lee, 2010; Wroldsen, 2007). Node a tensegrity truss according to the FEM is dependent on the global linear
configuration of these structures ensures occurrence of infinitesimal stiffness matrix KL and the global geometric stiffness matrix KG . The
mechanisms that are balanced with self-stress states (Koohestani and self-stress of the structure is represented by the second of the above
Guest, 2013; Calladine and Pellegrino, 1991). In this paper the authors matrices. The matrices can be also obtained directly (without the finite
focus on a particular type of tensegrities, which are tensegrity trusses. element approximation) (Lewin � ski, 2001; Pełczyn � ski and Gilewski,
Tensegrity systems are complicated regarding both their geometry 2018; Lewin � ski et al., 2019). However the FEM formalism is recom­
and mechanics. Therefore, it is often difficult to determine and under­ mended for the 3D models (Pełczy� nski and Gilewski, 2018) as it is
stand their properties, which makes it hard to use them in engineering simpler to algorithmize and programming. The technique leads to the
applications. In order to understand these untypical, unique features of equivalent symmetric elasticity matrix E ¼ ½Eij �; i; j ¼ 1; :::; 6; according
tensegrity structures and identify their properties, a continuum model to the LTE theory. Altogether, there are 36 coefficients, including 21
(Gilewski and Kasprzak, 2014) is used. The proposed model can be used independent ones. The elasticity matrix can take different, particular
for description of single tensegrity modules as well as more complicated forms, depending on the type of symmetry. It can be proved that there
multi-module structures. The continuum approach for determination of are exactly eight types of symmetry in the linear theory of elasticity
equivalent mechanical properties of tensegrity structures should enable (Chadwick et al., 2001). According to this theory each material is either
to identify their structural properties and understand precisely how they isotropic or anisotropic, and among the anisotropic ones the following
work. This should allow the architects and engineers to apply them in are distinguished: triclinic (fully anisotropic), monoclinic, trigonal,
the design of both single structural elements, such as beams, columns, orthotropic, tetragonal, cubic and transversely isotropic. All these
plates or shells, and more complicated structural systems. Another symmetry types have different number of independent coefficients of the
advantage arising from the continuum model is the possibility of matrix E and different number of symmetry planes (Chadwick et al.,
building active systems that control structural properties of tensegrities, 2001; Green and Zerna, 1970). The applied continuum model is
what can be used in the design of smart structures with the effects of non-linear in the sense of equilibrium equations considered in actual
self-adjustment, self-diagnosis, self-repair etc. (Culshaw, 1999; Fest configuration. Validation of the model for structures with self-stress
et al., 2003; Gilewski and Al Sabouni Zawadzka, 2015; Motro, 2003; included was presented in the annex to (Al Sabouni-Zawadzka and
Sultan, 2014). The continuum model can be successfully used in the Gilewski, 2019).
recently developed area of tensegrity inspired metamaterials and lattices The approach reminded above can be applied to different types of
(De Tommasi et al., 2017; Ma et al., 2018; Rimoli and Pal, 2017; Sal­ structures (one-, two- and three-dimensional tasks) with any degree of
ahsoor et al., 2018; Al Sabouni-Zawadzka and Gilewski, 2018) as well as freedom. It can be used to determine properties of simple tensegrity
for the evaluation of extreme properties of metamaterials (Milton and modules as well as more complicated, multi-module structures. A direct
Cherkaev, 1995; Kadic et al., 2012; Al Sabouni-Zawadzka and Gilewski, area of possible applications is the evaluation of extreme properties
2019). It should be highlighted that the continuum model is not meant (Milton and Cherkaev, 1995) of tensegrity-inspired metamaterials with
to replace the standard approach to the analysis of tensegrity systems – it unimode or multimode characteristics (Kadic et al., 2012; Al
is just a supplement, which can help in understanding and promoting Sabouni-Zawadzka and Gilewski, 2019) as well as a continuum model of
this type of structures in engineering applications. plate-like tensegrity structures with orthotropic properties (Al
The main objective of this paper is a qualitative as well as quanti­ Sabouni-Zawadzka and Gilewski, 2016; Al Sabouni-Zawadzka et al.,
tative estimation of equivalent mechanical properties for a selection of 2016).
typical tensegrity modules analysed with the use of the continuum
model. 3. Application of the continuum model for determination of
The aims of the paper are: mechanical characteristics

� identification of the structural properties corresponding to the Application of the continuum model in the analysis of tensegrity
typical deformation modes (tension, shear) of typical tensegrity structures enables a physical interpretation of elasticity coefficients.
modules, Through determination of mechanical characteristics for a given module
� physical interpretation of the determined mechanical characteristics, or structure and the conditions that limit their values, a proper under­
� estimation of the influence of self-stress and cable/strut character­ standing of structural properties and the behaviour of tensegrity systems
istics on deformation of the structures, is possible.
� comparison of the elastic properties of typical tensegrity modules, In a general case of a fully anisotropic structure the following tech­
� determination of limiting conditions for the values of mechanical nical coefficients can be defined:
characteristics.
� Young’s moduli (E): proportion between normal stresses and strains,
2. Basis of the continuum model � shear moduli (G): proportion between shear stresses and strains,
� Poisson’s ratios (ν): relation between normal strains in perpendicular
The continuum model proposed in (Gilewski and Kasprzak, 2014) is directions,
based on the comparison of the strain energy of an unsupported ten­ � coefficients (μ): relation between shear strains in perpendicular
segrity structure defined with the use of the finite element method (FEM) directions,
(Bathe, 1996; Crisfield, 2003; Hughes, 2000; Zienkiewicz et al., 2005) � coefficients (λ): relation between normal strains in three directions
and the strain energy of a solid determined using the symmetric linear and shear strains in one direction,

2
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

� coefficients (κ): relation between shear strains in three directions and


normal strains in one direction.

Mechanical characteristics can be determined from the inverse


elasticity matrix H ¼ E 1. The analysis of particular stress states leads to
the determination of relations between the values of mechanical char­
acteristics and the coefficients of the matrix H (1):

2 3
1 ν21 ν31 λ11 λ21 λ31
6 E1 E2 E3 G1 G2 G3 7
6 7
6 7
6 ν12 1 ν32 λ12 λ22 λ32 7
6 7
6 E1 E2 E3 G1 G2 G3 7
6 7
6 7
6 ν13 ν23 1 λ13 λ23 λ33 7
6 7
6 E1 E2 E3 G1 G2 G3 7
6 7
H¼6 7: (1)
6 κ11 κ21 κ31 1 μ21 μ31 7
6 7
6 E1 E2 E3 G1 G2 G3 7
6 7
6 7
6
6
κ12 κ22 κ32 μ12 1 μ32 7
7
6 G3 7
6 E1 E2 E3 G1 G2 7 Fig. 1. Blocks 3 � 3 in the matrix H.
6 7
4 κ13 κ23 κ33 μ13 μ23 1 5
E1 E2 E3 G1 G2 G3

Application of the continuum model in structural analysis allows not


only to determine mechanical characteristics of the structure, but also to
v12 v21 <1;v13 v31 <1; v23 v32 <1; v12 v21 þv13 v31 þv23 v32 þv12 v31 v23 þv21 v13 v32 <1;
formulate conditions that limit their values.
The determination of such limiting values is based on the condition (4)
that the strain energy is non-negative definite. The strain energy equals limiting conditions for coefficients μ (principal minors 2 � 2 and 3 �
zero only when no deformations occur in the structure. If the de­ 3 for the block B4):
formations appear, the strain energy has to be positive. Therefore, both
the elasticity matrix E and the inverse matrix H have to be positive μ12 μ21 <1; μ13 μ31 < 1; μ23 μ32 < 1; μ12 μ21 þ μ13 μ31 þ μ23 μ32
definite.
In accordance with the Sylvester’s criterion, if a matrix is positive μ12 μ31 μ23 μ21 μ13 μ32 < 1 (5)
definite, its principal minors are positive. Formally, six conditions In a general case global conditions can be formulated, considering
limiting the values of mechanical characteristics can be formulated. the principal minors 4 � 4, 5 � 5 and 6 � 6 of the matrix H. The ex­
However, it should be noticed that the simultaneous swap of columns pressions are relatively large and will not be presented in this paper.
and rows with the same numbers does not change the positive definition However, all the above conditions are considered in the analysis of the
of the matrix. It also applies to the whole blocks 3 � 3 (Fig. 1). modules in the next section.
Therefore, more than six limiting conditions for the values of me­
chanical characteristics can be formulated: 4. Analysis – typical tensegrity modules
limiting conditions for Young’s moduli E (diagonal of the block B1):
Mechanical properties of five typical tensegrity modules inscribed in
E1 > 0; E2 > 0; E3 > 0; (2) the cube of edge length a are determined and evaluated using the pro­
posed continuum model: a 4-strut simplex, a 3-strut simplex, an octa­
limiting conditions for shear moduli G (diagonal of the block B4): hedron, a tetrahedron and an X-module. The purpose of these examples
is to present a full analysis of equivalent mechanical properties of typical
G1 > 0; G2 > 0; G3 > 0; (3) tensegrity modules.
The analysed structures consist of cables and struts (Koohestani and
Guest, 2013; Calladine and Pellegrino, 1991) with the parameter k ¼
limiting conditions for Poisson’s ratios ν (principal minors 2 � 2 and ðEAÞcable =ðEAÞstrut , ðEAÞstrut ¼ EA where E is the Young’s module of the
3 � 3 for the block B1): material and A is the cross-section. The self-equilibrated system of
forces is represented by the parameter σ ¼ S=EA, where S is a free
multiplier (Koohestani and Guest, 2013; Calladine and Pellegrino,

3
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

1991).
k > 0:184806 σ : (6)
The first module is a 4-strut simplex with geometry presented in
Fig. 2. Graphs of the mechanical coefficients in the range k; σ 2 ð0; 1Þ are
The following matrices are obtained for the module: presented in Figs. 3–7. The areas of positive definiteness of the matrices
2 3 are to the right of the dotted lines.
e11 e12 e13 e14 0 0
6 7 The Young’s moduli E1 and E3 have similar characters and vary
6
6 e11 e13 e14 0 0 7
7 slightly with the maximum values reached for k→1; σ →0.
6 7
6 e33 0 0 0 7 The module G1 reaches maximum value for k→1; σ →0, G2 varies
6 7
E¼6
6
7
7 between 0.8 and 1.2EA=a2 , maximum values are observed for k→1; σ →1.
e 0 0
6 7 The Poisson’s ratios ν12 ; ν13 are nearly constant and ν31 is slightly
12
6 7
6
4 e13 0 7
5 variable.
sym: e13 The distributions of coefficients in Fig. 6 are nearly constant.
2 3 The second module in the analysis is a 3-strut simplex with geometry
1 ν12 ν31 λ11 presented in Fig. 7.
6 0 0 7
6
6
E1 E1 E3 G1 7
7 The following matrices are obtained for the module:
6 7 2 3
6 ν12 1 ν31 λ11 7 3e12 e12 e13 0 e15 e16
6 0 0 7
6 E1 E1 E3 G1 7 6 7
6 7 6
6 7 6 3e12 e13 0 e15 e16 7
7
6 7 6 7
6 ν13 ν13 1 7 6 7
6 0 0 0 7 6 e 33 0 0 0 7
6 E1 E1 E3 7 E¼6 7
6 7 6
E 1
¼6 7 6 e12 e16 e15 7
7
6 7 6 7
6 κ11 κ11 1 7 6 7
6 0 0 0 7 4 e 13 0 5
6 E1 E1 G1 7
6 7
6 7 sym: e13
6 7 2 3
6 1 7
6 0 0 0 0 0 7 1 ν12 ν31 λ21 λ31
6 G2 7 6 E 0
6
6
7
7 6 1 E1 E3 G2 G2 7 7
6 7 6 7
4 1 5 6 7
0 0 0 0 0 6 ν12 1 ν31 λ21 λ31 7
G2 6 0 7
6 E E1 E3 G2 G2 7
6 1 7
2EA 6 7
e11 ¼ ð0:314815 þ 1:39827⋅k 0:0794978⋅σÞ; 6 7
6 ν13 ν13 1 7
a2 6 0 0 0 7
6 E E E 7
6 1 1 3 7
EA E 1¼6 6
7
7
e12 ¼ ð0:296296 þ 0:707107⋅k 0:0134742⋅σÞ; 6 1 μ μ 7
a2 6 0 0 0 21 31 7
6 G G G 7
EA 6 1 2 2 7
e13 ¼ 2 ð0:740741 þ 0:357771⋅k þ 0:17247⋅σ Þ; 6 7
6 7
a 6 κ12 κ12 μ 12 1 7
6 0 0 7
EA 6 E1 E G G 7
e14 ¼ 2 ð 0:2222222 0:0808452⋅σÞ; 6 1 1 2 7
a 6 7
6 7
2EA 4 κ13 κ13 μ13 1 5
e33 ¼ 2 ð0:592593 þ 1:43108⋅k 0:17247⋅σ Þ: 0 0
a E1 E1 G1 G2

Slight anisotropy is observed with coupling in the xy plane. It is


EA
possible to express the technical elastic coefficients of the module using e12 ¼ ð0:0957031 þ 0:595459⋅k 0:0400226⋅σÞ;
a2
the elements of the matrix E, however in the present paper these co­ EA
efficients are presented graphically. e13 ¼ 2 ð0:492188 þ 0:142302⋅k þ 0:16009⋅σ Þ;
a
Taking into account the fact that the matrices E and H are positive EA
definite, the limiting conditions for mechanical characteristics must be e15 ¼ 2 ð0:182677 þ 0:0770235⋅σ Þ;
a
fulfilled. In general the conditions are nonlinear (polynomial) functions EA
of the parameters k and σ. However the curvatures of the lines are slight e16 ¼ 2 ð 0:117187 0:0237171⋅k 0:0415049⋅σÞ;
a
and all the conditions can be sufficiently estimated by the linear 2EA
equation e33 ¼ 2 ð0:632813 þ 1:28072⋅k 0:16009⋅σ Þ
a

Fig. 2. Geometry of the 4-strut simplex module.

4
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

Fig. 3. 4-strut simplex – Young’s moduli E1 ¼ E2 ; E3 (to be multiplied by EA=a2 ).

Fig. 4. 4-strut simplex – shear moduli G1 ; G2 ¼ G3 (to be multiplied by EA=a2 ).

5
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

Fig. 5. 4-strut simplex – Poisson’s ratios ν12 ¼ ν21 ; ν13 ¼ ν23 ; ν31 ¼ ν32 .

Fig. 6. 4-strut simplex – coefficients λ11 ¼ λ12 and κ11 ¼ κ21 .

6
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

Fig. 7. Geometry of the 3-strut simplex module.

The condition for positive definiteness of the matrices can be esti­ 2 3


e11 e12 e13 0 0 0
mated as 6 7
6 e22 e23 0 0 0 7
k > 0:180605 σ: (7) 6
6
7
7
6 e33 0 0 0 7
6 7
The module is strongly anisotropic with coupling in the xz and yz E¼6 7
6 0 7
planes. 6 e12 0 7
6 7
Graphs of the mechanical coefficients in the range k; σ 2 ð0; 1Þ are 6
4 e13 0 75
presented in Figs. 8–13.
sym: e23
The Young’s moduli E1 and E3 have similar characters and vary 2 3
slightly with the maximum values of E1 for k→1; σ→0 and E3 for k→ 1; 1 ν21 ν31
0 0 0 7
6 E E2 E3
σ →0. 6 1 7
6 7
The shear module G1 varies slightly with the maximum value 6
6 ν12
7
7
1 ν32
reached for k→1; σ→0, the module G2 is highly variable close to the 6
6 E 0 0 0 77
E2 E3
domain line with the maximum value for k→1; σ →1. 6
6
1 7
7
6 7
The Poisson’s ratios are relatively highly variable close to the domain 6 ν13 ν23 1 7
6 0 7
line. 6 E E2 E3
0 0 7
6 1 7
The coefficients μ12 ; μ13 vary quite regularly, the distributions of μ21 ; E 1
¼6
6
7
7
μ31 responsible for shear deformations in the xy plane are less regular 6
6 0 0 0
1
0 0 7
7
close to the domain line. 6
6 G1 7
7
6 7
The coefficient λ21 is dominated in the domain, so the structure is 6 7
6 1 7
flexible with regard to the extensional strain in the direction x against 6 0
6 0 0 0 0 77
G2
shear stress in the xz plane. 6
6
7
7
The levels of both coefficients κ12 ; κ13 are comparable, so the struc­ 6
4 1 5
7
ture is flexible with regard to the shear deformation caused by tension/ 0 0 0 0 0
G3
extension in the directions y and z.
The third module is an octahedron with geometry presented in e11 ¼
2EA
ð1 þ 1:52325⋅k þ 0:129225⋅σÞ;
Fig. 14. a2
EA
The module is orthotropic with the following matrices. e12 ¼ 2 ð0:845615⋅k 0:105243⋅σÞ;
a
EA
e13 ¼ 2 ð1:26604⋅k 0:153207⋅σÞ;
a
2EA
e22 ¼ 2 ð1 þ 1:35912⋅k þ 0:137028⋅σÞ;
a
EA
e23 ¼ 2 ð1:51283⋅k 0:168813⋅σÞ;
a
2EA
e33 ¼ 2 ð1 þ 0:921194⋅k þ 0:16101⋅σÞ
a

7
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

Fig. 8. 3-strut simplex – Young’s moduli E1 ¼ E2 ; E3 (to be multiplied by EA=a2 ).

8
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

Fig. 9. 3-strut simplex – shear moduli G1 ; G2 ¼ G3 (to be multiplied by EA=a2 ).

Fig. 10. 3-strut simplex – Poisson’s ratios ν12 ¼ ν21 ; ν13 ¼ ν23 ; ν31 ¼ ν32 .

9
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

Fig. 11. 3-strut simplex – coefficients μ12 ; μ21 ; μ13 ; μ31 .

Fig. 12. 3-strut simplex – coefficients λ21 ¼ λ22 ; λ31 ¼ λ32 .

10
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

Fig. 13. 3-strut simplex – coefficients κ12 ¼ κ22 ; κ13 ¼ κ23 .

Fig. 14. Geometry of the octahedron module, with the adopted parameters:k=K ¼ 0:65; l=L ¼ 0:30; m=M ¼ 0:56:

11
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

Fig. 15. Octahedron – Young’s moduli E1 ; E2 ; E3 (to be multiplied by EA=a2 ).

Fig. 16. Octahedron – shear moduli G1 ; G2 ; G3 (to be multiplied by EA=a2 ).

12
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

The matrices are positive definite if 2 3


e11 e12 e13 0 e15 0
k > 0:124457σ : (8) 6
6 e22 e23 0 e23 0 7 7
6 e33 0 e15 0 7
Graphs of the mechanical coefficients in the range k; σ 2 ð0; 1Þ are E¼6
6 e12 0
7
e23 7
6 7
presented in Figs. 15–17. 4 e13 0 5
The values of Young’s moduli E1 ; E2 ; E3 slightly differ from each sym: e23
other because of the differences in geometry in each direction. The
changes are regular with the maximum values reached for k→ 1; σ→ 1. 2
1 ν21 ν31 λ21
3
The shear moduli G1 ; G3 are 50% higher than the module G2 . The 6 E 0 0 7
6 1 E2 E3 G2 7
changes are regular with the maximum values reached for k→ 1; σ→ 0. 6 7
6 ν12 1 ν32 λ22 7
The Poisson’s ratios are relatively highly variable close to the domain 6
6 E 0 0 77
line with the maximum values for k→1; σ→1. 6 1 E2 E3 G2 7
6 7
The fourth module analysed in this paper is a tetrahedron with ge­ 6 ν13
6 ν23 1
0
λ23
0 7
7
6 E 7
ometry presented in Fig. 18. 1 6 1 E2 E3 G2 7
E ¼6 7
The tetrahedron module is orthotropic with regular layout of the 6
6 0 1 μ31 7
7
0 0 0
coefficients and the following matrices: 6
6 G1 G3 77
2 3 6 7
e11 e12 e12 0 0 0 6 κ12 κ22 κ32 1 7
6 0 0 7
6 7 6 E1 E2 E3 G2 7
6 e 11 e 12 0 0 0 7 6 7
6 7 6 7
e 0 0 0 4 1 1 5
E¼6 6
11 7
7 0 0 0 0
6 e12 0 0 7 G2 G3
4 e12 0 5
sym: e12 2EA
2 3 e11 ¼ ð0:0441942 þ 0:110947⋅k 0:00988212⋅σÞ;
1 ν12 ν12 a2
6 0 0 0 7
6 E1 E1 E1 7 EA
6 7 e12 ¼ 2 ð0:0883883 þ 0:0340207⋅k þ 0:0395285⋅σÞ;
6 ν12 1 ν 7 a
6 12
0 0 0 7
6 E E1 E1 7 EA
6 1 7
6 7 e13 ¼ 2 ð0:0968935⋅k 0:0197642⋅σÞ;
6 ν12 ν 12 1 7 a
6 0 0 0 7
6 E E1 E1 7 EA
1 6 1 7 e15 ¼ 2 ð 0:00850517⋅kÞ;
E ¼6 7
6 1 7 a
6 0 0 0 0 0 7
6 G 7 2EA
6 1 7 e22 ¼ 2 ð0:0441942 þ 0:193041⋅k 0:0197642⋅σÞ;
6 7 a
6 1 7
6 0 0 0 0 0 7
6 G 7 EA
6 1 7 e23 ¼ 2 ð0:0340207⋅kÞ;
6 7 a
4 1 5
0 0 0 0 0 2EA
G1 e33 ¼ 2 ð0:0625 þ 0:0484468⋅k þ 0:00988212⋅σ Þ:
a

e11 ¼
2EA
ð0:836131 þ 0:728014⋅k þ 0:0709611⋅σÞ; The conditions for positive definiteness of the matrices can be esti­
a2 mated as
EA
e12 ¼ 2 ð0:232405 þ 0:696312⋅k 0:0709611⋅σ Þ: k ¼ 0:258058 σ : (9)
a
The matrices are positive definite for all the considered values of the Graphs of the mechanical coefficients in the range k; σ 2 ð0; 1Þ are
parameters. presented in Figs. 21–26.
Graphs of the mechanical coefficients in the range k; σ 2 ð0; 1Þ are The Young’s moduli E1 ; E2 vary regularly with the maximum values
presented in Fig. 19. at k→1; σ →1. The module E3 is nearly constant.
The Young’s moduli and shear moduli vary regularly with the The shear moduli G2 ; G3 vary regularly in a high range with the
maximum values at k→1; σ →1 or k→1; σ →0 respectively. The Poisson’s maximum values at k→1; σ→0. The module G1 is slightly variable in
ratios are relatively highly variable and reach maximum values close to respect of σ and seems to be constant against k.
the domain line. The Poisson’s ratios increase or decrease slightly. An irregular dis­
The last structure of the selection is an X-module with geometry tribution can be observed for the ratio ν32 .
presented in Fig. 20. The coefficients μ13 is constant and μ31 varies regularly with the
The module is relatively highly anisotropic with the following maximum value reached for k→1; σ →1.
matrices: The coefficients λ21 ; λ22 ; λ23 vary highly close to the domain line and
are nearly constant for a higher value of the parameter k.
An interesting distribution is observed for the coefficients κ12 ; κ22 ; κ32

13
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

Fig. 17. Octahedron – Poisson’s ratios ν12 ; ν13 ; ν23 ; ν21 ; ν31 ; ν32 .

14
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

Fig. 18. Geometry of the tetrahedron module.

Fig. 19. Tetrahedron – Young’s moduli E1 ¼ E2 ¼ E3 (to be multiplied by’EA=a2 ), shear moduli G1 ¼ G2 ¼ G3 (to be multiplied by EA=a2 ) and Poisson’s ratios ν12 ¼
ν21 ¼ ν31 ¼ ν31 ¼ ν23 ¼ ν32 .

Fig. 20. Geometry of the X-module.

15
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

Fig. 21. X-module – Young’s moduli E1 ; E2 ; E3 (to be multiplied by EA=a2 ).

Fig. 22. X-module – shear moduli G1 ; G2 ; G3 (to be multiplied by EA=a2 ).

16
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

Fig. 23. X-module – Poisson’s ratios ν12 ; ν13 ; ν23 ; ν21 ; ν31 ; ν32 .

17
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

Fig. 24. X-module – coefficients μ13 ¼ 1; μ31 .

Fig. 25. X-module – coefficients λ21 ; λ22 ; λ23 .

18
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

Fig. 26. X-module – coefficients κ12 ; κ22 ; κ32 .

Table 1
Numerical values of mechanical properties of 4-strut Simplex tensegrity module. Table 3
4-strut Simplex Numerical values of mechanical properties of Octahedron tensegrity module.

Version 1 Version 2 Version 3 Version 4 Octahedron

k ¼ 0.29 k ¼ 0.98 k ¼ 0.29 k ¼ 0.98 Version 1 Version 2 Version 3 Version 4


σ ¼ 0.001 σ ¼ 0.005 σ ¼ 0.5 σ ¼ 0.5 k ¼ 0.29 k ¼ 0.98 k ¼ 0.29 k ¼ 0.98
E1 ¼ E2 [MPa] 52.15 159.77 40.28 151.44 σ ¼ 0.002 σ ¼ 0.005 σ ¼ 0.5 σ ¼ 0.5
E3 [MPa] 72.04 193.66 51.17 178.33 E1 [MPa] 158.64 256.29 167.33 266.89
G1 [MPa] 22.29 53.33 18.88 51.93 E2 [MPa] 152.03 228.64 161.57 240.88
G2 ¼ G3 [MPa] 47.54 61.47 52.38 66.27 E3 [MPa] 136.67 176.68 147.94 190.39
ν21 0.25 0.25 0.22 0.24 G1 [MPa] 13.79 46.61 10.84 43.68
ν31 ¼ ν32 0.43 0.25 0.50 0.28 G2 [MPa] 20.65 69.78 16.35 65.52
λ11 ¼ -λ12 0.24 0.09 0.30 0.11 G3 [MPa] 24.67 83.39 19.94 78.69
ν21 0.06 0.08 0.05 0.08
ν31 0.11 0.20 0.09 0.19
ν32 0.15 0.28 0.12 0.26
Table 2
Numerical values of mechanical properties of 3-strut Simplex tensegrity module.
3-strut Simplex

Version 1 Version 2 Version 3 Version 4


Table 4
k ¼ 0.29 k ¼ 0.98 k ¼ 0.29 k ¼ 0.98 Numerical values of mechanical properties of Tetrahedron tensegrity module.
σ ¼ 0.001 σ ¼ 0.004 σ ¼ 0.5 σ ¼ 0.5
Tetrahedron
E1 ¼ E2 [MPa] 27.50 90.79 20.75 84.83
Version 1 Version 2 Version 3 Version 4
E3 [MPa] 83.16 195.87 61.38 181.87
G1 [MPa] 9.96 33.48 7.57 31.18 k ¼ 0.29 k ¼ 0.98 k ¼ 0.29 k ¼ 0.98
G2 ¼ G3 [MPa] 19.80 31.17 18.69 33.66 σ ¼ 0.002 σ ¼ 0.007 σ ¼ 0.5 σ ¼ 0.5
ν21 0.38 0.36 0.37 0.36
ν31 ¼ ν32 0.50 0.23 0.62 0.27 E1 ¼ E2 ¼ E3 [MPa] 109.50 151.04 114.89 156.92
μ21 0.46 0.21 0.58 0.24 G1 ¼ G2 ¼ G3 [MPa] 24.44 51.46 22.45 49.49
μ31 0.68 0.27 0.89 0.34 ν21 ¼ ν31 ¼ ν32 0.17 0.23 0.16 0.22
λ22 ¼ -λ21 0.34 0.13 0.45 0.17
λ31 ¼ -λ32 0.23 0.10 0.29 0.12

19
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

which are very highly variable and extreme close to the domain line. 6. Conclusions
Also the coefficient κ22 is much higher than κ12 ;κ32 , what means that the
module is highly flexible with regard to the transverse deformation in Tensegrity structures are complicated regarding both their geometry
the xz plane against normal stress in the direction y. and mechanics. In order to understand their properties and identify
mechanical characteristics, a continuum model is applied for evaluation
5. Numerical verification of mechanical properties of selected typical tensegrity modules. The
model was built by assuming that the strain energy of an unsupported
Numerical verification of the presented models is provided for fixed tensegrity structure is equivalent to the strain energy of a solid. In a
material properties and possible self-stress values of tensegrity modules general case the solid can be arbitrary – it can be a cube, a sphere, a
inscribed into a cube of side dimension a ¼ 1. Four options were rectangular prism etc. However, taking into account the geometry of
considered. In the first variant, the modules are made of Halfen steel typical tensegrity modules, the authors decided to use a cube. As far as
profiles (www.halfen.com/en/). The struts are pipes with 210 GPA the qualitative analysis is concerned, the characteristics of the structure
Young’s modulus, external diameter 60.3 mm and internal diameter do not depend on the type of a solid into which the tensegrity is
57.4 mm 10 mm diameter cables were adopted. The k factor in this case inscribed.
is 0.29, while the σ coefficients range from 0.001 to 0.002 depending on A continuum analysis of typical tensegrity modules was presented in
the module. In the second variant, the cables were made of carbon fiber this paper. Five modules were analysed: 4-strut simplex, 3-strut simplex,
with a diameter of 10 mm and a Young’s 700 GPa module (Montouri and octahedron, truncated tetrahedron and X-module. Using the applied
Skelton, 2017). The k factor is 0.98 and the σ coefficients range from model, mechanical characteristics in function of two parameters
0.004 to 0.007. A significant increase in σ coefficients is possible only k ¼ ðEAÞcable =ðEAÞstrut ;ðEAÞstrut ¼ EA and σ ¼ S=EA were determined and
with the use of high performance materials, or a specific strut con­ graphs of all coefficients were plotted and described. The presented
struction (cf. e.g. (Zhang et al., 2018)). The value 0.5 is considered for examples show that the continuum analysis of tensegrity systems en­
comparison in the third and fourth variant. Numerical values of all ables to estimate the influence of self-stress on the deformation of the
mechanical properties for five tensegrity modules are presented in structure, identify the influence of cables and struts on the properties of
Tables 1–5. the whole system, determine mechanical characteristics and conditions
Tables 1–5 show that the equivalent mechanical properties of ten­ that limit their values. What is more, a qualitative as well as quantitative
segrity modules largely depend on the parameter k. The influence of the evaluation and comparison of various tensegrity modules and systems is
self-stress parameter is much smaller. All parameters obtained meet the possible with the use of the proposed technique.
restrictive conditions. Interesting properties have the module-X, in The following equivalent material symmetries are identified for the
which there are some negative Poisson’s ratios together with relatively selected modules:
small Young’s and Kirchhoff’s modulus.
� 4-strut simplex – slight anisotropy with coupling in the xy plane and
5 independent mechanical parameters,
� 3-strut simplex – strong anisotropy with coupling in the xz as well as
yz plane and 5 independent mechanical parameters,
� Octahedron – orthotropy with 6 independent mechanical
Table 5 parameters,
Numerical values of mechanical properties of tensegrity X-module. � Tetraheron – orthotropy with 2 independent mechanical parameters,
X-module � X-module – strong anisotropy with coupling in the xz as well as yz
Version 1 Version 2 Version 3 Version 4 plane and 7 independent mechanical parameters.
k ¼ 0.29 k ¼ 0.98 k ¼ 0.29 k ¼ 0.98
σ ¼ 0.001 σ ¼ 0.004 σ ¼ 0.5 σ ¼ 0.5 Physical interpretation of mechanical properties of the tensegrity
modules is given through identification of equivalent Young’s moduli,
E1 [MPa] 5.67 13.49 3.60 12.75
E2 [MPa] 7.55 22.98 4.47 20.75
shear moduli, Poisson’s ratios and other parameters determined using
E3 [MPa] 8.29 10.70 9.00 11.52 the inverse matrix H ¼ E 1. Analysis of the graphs presented in this
G1 [MPa] 4.98 4.98 6.09 6.09 paper makes it possible to estimate equivalent elastic properties of the
G2 [MPa] 1.55 5.19 0.98 4.64 tensegrity modules for fixed parameters k and σ .
G3 [MPa] 0.50 1.36 0.51 1.43
Limiting conditions for the values are identified and presented in the
ν21 0.65 0.40 0.82 0.48
ν31 0.22 0.31 0.19 0.30 graphs through the estimated dashed line.
ν32 0.07 0.02 0.08 0.03 A comparison of the modules inscribed into a cube (with fixed
μ31 0.10 0.27 0.08 0.24 geometrical and physical parameters of cables and struts) leads to the
λ22 0.06 0.07 0.10 0.08 notice that the most flexible and anisotropic is the X-module. The ranges
0.03 0.01 0.06 0.02
of Young’s and shear moduli for other structures are similar. Diversity of
λ21
λ23 0.02 0.03 0.02 0.03
Poisson’s ratios is observed depending on the directions of deformations
and stresses. The most predictable properties are estimated for ortho­
tropic modules: octahedron and tetrahedron.
Similar analyses can be performed for other types of modules as well

20
W. Gilewski and A. Al Sabouni-Zawadzka European Journal of Mechanics / A Solids 83 (2020) 103998

as more complex, 2D plate-like and 3D metamaterial-like multi-module Kadic, M., Buckmann, T., Stenger, N., Thiel, M., 2012. On the practicability of
pentamode mechanical metamaterials. Appl. Phys. Lett. 100, 191901.
structures if the modules are placed in a consistent way.
Kebiche, K., Kazi Aoual, M.N., Motro, R., 2008. Continuum models for systems in a
The proposed analysis should stimulate the imagination of engineers selfstress state. Int. J. Space Struct. 23, 103–115.
and architects to understand how the tensegrity structures work and Koohestani, K., Guest, S.D., 2013. A new approach to the analytical and numerical form-
therefore, make it possible to apply them in the future as both single finding of tensegrity structures. Int. J. Solids Struct. 50, 2995–3007.
Lewi� nski, T., 2001. On algebraic equations of elastic trusses, frames and grillages.
structural elements, such as beams, columns, plates or shells, and more J. Theor. Appl. Mech. 39, 307–322.
complicated structural systems. Moreover, the information obtained Lewi� nski, T., Sok� oł, T., Graczykowski, C., 2019. Michell Structures, Chapter 1. Selected
from the analysis could be used in development of smart systems that Problems of Statics. Springer International Publishing AG.
Luo, Y., Xu, X., Kumar, S., Ingber, D.E., 2008. A multi-modular tensegrity model of an
actively control the structures, tensegrity-inspired metamaterials and actin stress fiber. J. Biomech. 41, 2379–2387.
extreme metamaterials. Ma, Y., Zhang, Q., Dobah, J., Scarpa, F., Fraternali, F., Skelton, R.E., Zhang, D., Hong, J.,
2018. Meta-tensegrity: design of a tensegrity prism with metal rubber. Compos.
Struct. 206, 644–657.
Declaration of competing interest Milton, G., Cherkaev, A.V., 1995. Which elasticity tensors are realizable ? J. Eng. Mater.
Technol. 117, 483–493.
The authors declare that they have no known competing financial Montouri, R., Skelton, R.L., 2017. Globally stable tensegrity compressive structures for
arbitrary complexity. Compos. Struct. 179, 682–694.
interests or personal relationships that could have appeared to influence Motro, R., 2003. Tensegrity: Structural Systems for the Future. Kogan Page Science,
the work reported in this paper. London.
Noor, A.K., Anderson, M.C., 1978. Continuum models of beam- and plate-like lattice
structures. AIAA J. 16, 1219–1228.
References
Noor, A.K., Anderson, M.C., 1979. Analysis of beam-like lattice trusses. Comput. Methods
Appl. Mech. Eng. 20, 53–70.
Al Sabouni-Zawadzka, A., Gilewski, W., 2016. On orthotropic properties of tensegrity Odegard, G.M., Gates, T.S., Nicholson, L.M., Wise, K.E., 2002. Equivalent-continuum
structures. Proc. Eng. 153, 887–894. modelling of nano-structured materials. Compos. Sci. Technol. 62, 1869–1880.
Al Sabouni-Zawadzka, A., Gilewski, W., 2018. Smart metamaterial based on the simplex Pal, R.K., Ruzzene, M., Rimoli, J., 2016a. A continuum model for nonlinear lattices under
tensegrity pattern. Materials 11, 673. large deformations. Int. J. Solids Struct. 96, 300–319.
Al Sabouni-Zawadzka, A., Gilewski, W., 2019. Soft and stiff simplex tensegrity lattices as Pal, R.K., Rimoli, J., Ruzzene, M., 2016b. Effect of large deformation pre-loads on the
extreme metamaterials. Materials 12, 187. wave properties of hexagonal lattice. Smart Mater. Struct. 25, 054010.
Al Sabouni-Zawadzka, A., Gilewski, W., Obara, P., 2016. Continuum model of Pełczy� nski, J., Gilewski, W., 2018. An Extension of Algebraic Equations of Elastic Trusses
orthotropic tensegrity plate-like structures with self-stress included. Eng. Trans. 64 with Self-Equilibrated System of Forces. ECCM VI, Glasgow, UK.
(4), 501–508. Ptochos, E., Labeas, G., 2012. Elastic modulus and Poisson’s ratio determination of
Arabnejad, S., Pasini, D., 2013. Mechanical properties of lattice materials via asymptotic micro-lattice cellular structures by analytical, numerical and homogenization
homogenization and comparison with alternative homogenization methods. Int. J. method. J. Sandw. Struct. Mater. 28, 597–626.
Mech. Sci. 77, 249–262. Rimoli, J.J., Pal, R.K., 2017. Mechanical response of 3-dimensional tensegrity lattices.
Bathe, K.J., 1996. Finite Element Procedures in Engineering Analysis. Prentice-Hall, New Compos. Part B 115, 30–42.
York. Salahsoor, H., Pal, R.K., Rimoli, J.J., 2018. Material symmetry phase transitions in the
Ben Kahla, N., 1995. Equivalent beam-column analysis of guyed towers. Comput. Struct. three-dimensional tensegrity metamaterial. J. Mech. Phys. Solids 119, 382–399.
55, 631–645. Salehian, A., Inman, D.J., 2008. Dynamic analysis of a lattice structure by
Calladine, C.R., Pellegrino, S., 1991. First-order infinitesimal mechanisms. Int. J. Solids homogenization: experimental validation. J. Sound Vib. 316 (1–5), 180–197.
Struct. 27, 505–515. Skelton, R.E., de Oliveira, M.C., 2009. Tensegrity Systems. Springer, London.
Chadwick, P., Vianello, M., Stephen, C.C., 2001. A new proof that the number of linear Sultan, C., 2014. Tensegrity deployment using infinitesimal mechanisms. Int. J. Solids
elastic symmetries is eight. J. Mech. Phys. Solids 49, 2471–2492. Struct. 51, 3653–3668.
Crisfield, M.A., 2003. Non-linear finite element analysis of solids and structures. In: Teughels, A., De Roeck, G., 2000. Continuum models for beam- and plate-like lattice
Essentials, vol. 1. Wiley, Chichester. structures. In: IASS-IACM 2000. Fourth International Colloquium on Computation of
Culshaw, B., 1999. Smart Structures and Materials. Artech House, Boston. Shells and Spatial Structures. Chania-Crete, Greece.
De Tommasi, D., Marano, G.C., Puglisi, G., Trentadue, F., 2017. Morphological Tran, H.C., Lee, J., 2010. Initial self-stress design of tensegrity grid structures. Comput.
optimization of tensegrity-type metamaterials. Composites Part B 115, 182–187. Struct. 88, 558–566.
Dow, J.O., Su, Z., Feng, C.C., 1985. Equivalent continuum representation of structures Wang, H.-X., Chung, S.W., 2011. Equivalent elastic constants of truss core sandwich
composed of repeated elements. AIAA J. 23, 1564–1569. plates. J. Pressure Vessel Technol. 133, 041203-1-6.
Fest, E., Shea, K., Domer, B., Smith, I.F.C., 2003. Adjustable tensegrity structures. Wroldsen, A.S., 2007. Modelling and Control of Tensegrity Structures. Ph.D. dissertation.
J. Struct. Eng. April, 515–526. Department of Marine Technology, Norwegian University of Science and
Gilewski, W., Al Sabouni Zawadzka, A., 2015. On possible applications of smart Technology.
structures controlled by self-stress. Arch. Civ. Mech. Eng. 15, 469–478. www.halfen www.halfen.com/en/.
Gilewski, W., Kasprzak, A., 2014. 3D Continuum Model of Tensegrity Modules with the Zhang, Q., Zhang, D., Dobah, Y., Scarpa, F., Fraternali, F., Skelton, R.E., 2018. Tensegrity
Effect of Self-Stress. WCCM XI. ECCM V, Barcelona, Spain. cell mechanical metamaterial with metal rubber. Appl. Phys. Lett. 113, 031906.
Green, A.E., Zerna, W., 1970. Theoretical Elasticity. Courier Dover Publications, Zienkiewicz, O.C., Taylor, R.L., Zhu, J.Z., 2005. The Finite Element Method: its Basis and
Mineola, New York. Fundamentals. Butterworth and Heinemann, Oxford.
Hughes, T.J.R., 2000. The Finite Element Method: Linear Static and Dynamic Finite
Element Analysis. Dover Publications, Mineola, New York.

21

You might also like