Structures 71 (2025) 107891

Contents lists available at ScienceDirect

Structures
journal homepage: www.elsevier.com/locate/structures

Investigation of internal forces of a truss structure

Han-Lin Sun , Zhi-Rui He , Jing-Shan Zhao *
State Key Laboratory of Tribology in Advanced Equipment, Department of Mechanical Engineering, Tsinghua University, Beijing 100084, China

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

Keywords: Unlike the traditional beam-column structures, continuous truss is a good candidate to build a bridge of larger
Internal forces span with lighter self-weight. By taking the advantages of the symmetry and periodicity of the continuous truss
Screw statics structures, a recursive force transfer method is proposed to study the internal force of each rod in a continuous
Truss structure
truss structure in this paper. First, the support force of the pylon and the internal forces of the rods on both sides
Recursive equation, cable-stayed bridge
of the truss structure are derived. Then, a recursive equation for the internal forces between neighboring truss
units is developed. Based on the transfer equation, the internal forces of all rods can be obtained. Finally, an
equivalent external load model was presented to analyze the internal force of the truss structure in a bridge.
Through the recursive algorithm, the optimum rod material might be determined for the selection and fabri-
cation of the truss structure by combining the equivalent external load and the strength theory of material
fracture.

1. Introduction sections. All the equilibrium equations can be reorganized as a system of

linear equations. It requires a lot of manual calculation and checking.
Continuous truss structure mainly consists of straight rods with Since space truss is a vast and elaborate structure. The force system is
triangular or other geometrically stabilized units. This structural form very complicated when the rods are disassembled for internal force
can effectively disperse the load and enhance the bearing capacity of the analysis, many redundant forces need to be considered [10]. Meanwhile,
overall structure. It is widely used in the underside of bridges, building the theoretical mechanics method is available to find the instantaneous
roof structures, cantilevered structures, and spacecraft, which is capable solutions during manual computation, but not fit for real-time compu-
of covering larger spans [1–4]. tation because of its complexity. In addition, some methods reduce the
When continuous truss structures are applied to bridges, the bridge complexity of the model by dividing the entire truss structure into a
deck (stiffening beam) is generally laid on the upper surface of the truss number of finite structure units [11,12]. For example, the equivalent
[5]. When vehicles pass through the cable-stayed bridges, they directly dynamic model of a circular truss structure was studied [13]. However,
act on the truss structure, resulting in high internal stress. As the moving the nonlinearities between trusses may be neglected, which might result
load on the bridge deck changes, the internal force of each rod of the in inaccurate calculations.
continuous truss structure will also change. When the internal stress By modeling and adding constraints in FEM software, the internal
exceeds the allowable one of the materials, damage will occur [6]. forces of trusses seem to be easy to obtain [14–16]. However, the finite
Therefore, it is necessary to develop an algorithm to obtain the internal element method presents certain limitations when applied to calculating
force of the rod in real-time. internal forces in continuous truss structures. First, there are a large
At present, the calculation methods of the internal force of contin- number of rods in the continuous truss structures. This makes it take a
uous truss structure mainly include analytical method (Theoretical lot of time and computational resources to build an accurate finite
Mechanics method) and finite element method (FEM) [7]. The theo- element model. In addition, there are convergence problems for the
retical mechanics method is directly based on Newton’s equations, finite element method when dealing with nonlinear problems. Truss
which is very cumbersome in dealing with the continuous truss with structures may exhibit nonlinear behaviors during the stressing process,
many rods and complicated structures [8,9]. It includes the method of and these nonlinear factors may lead to convergence difficulties in finite
joints and method of sections. It is the main idea to obtain the equilib- element analysis, which ultimately leads to some unavoidable errors. It
rium equations of internal forces on truss rods by selecting some joints or is still a challenging task of the scientists and engineers to find an

* Corresponding author.
E-mail address: [email protected] (J.-S. Zhao).

https://doi.org/10.1016/j.istruc.2024.107891
Received 16 May 2024; Received in revised form 10 November 2024; Accepted 19 November 2024
Available online 2 December 2024
2352-0124/© 2024 Institution of Structural Engineers. Published by Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and
similar technologies.
H.-L. Sun et al. Structures 71 (2025) 107891

Fig. 1. The geometry of continuous truss structure.

Fig. 2. Representation of truss units.

algebraic algorithm for continuous truss structures that is equivalent to

the finite element method but much simpler and more efficient in
calculation. Previously, the calculation of the continuous truss structure
was obtained by obtaining its equilibrium equation for each rod [8–10].
Regrettably, analyzing internal forces in the staggered structure of
spatial truss structures entails extensive computational requirements.
This paper presents a recursive algorithm for determining the in-
ternal forces of continuous truss structures. Compared with theoretical
mechanics methods, it offers lower computational complexity and is
easier to program and solve. In Section 2, the geometry of the continuous
truss structure is introduced, the concept of force screw is reviewed, and
the main framework of the algorithm is presented. In Section 3, the
recursive equations of internal force vectors and internal force matrices
for each rod are rigorously derived based on the continuous truss Fig. 3. Continuous truss structures in bridge design: case study of the Tian-
structure. In Section 4, an equivalent external load model is developed to xingzhou bridge in Wuhan.
validate the performance of the algorithm by using cable-stayed bridge
as an example. In a continuous truss, each face of a cube unit may be a force-bearing
surface, and the diagonal braces are responsible for transmitting forces
2. Statics reduction of a truss structure between these force-bearing surfaces. This design enables the structure
to withstand forces from multiple directions. Meanwhile, due to the
2.1. Truss structure and its geometry introduction of diagonal braces, the structure can better disperse the
loads when subjected to external forces, reducing the concentration of
The continuous truss is a stable construction form, and its sub-units local stresses, and thus improving the overall strength of the structure.
are composed of a cube structure shown in Fig. 1. Each cube structure An important geometric property in the continuous truss structure is
serves as a basic unit, with a diagonal brace on each of its two specific its symmetry. As shown in Fig. 1, the truss is divided into two halves, the
faces. These diagonal braces are connected to the edges of the cube left one and the right one are perfectly symmetric about the green center
structure and together form the framework of the truss. This design al- plane. To take the good advantage of symmetry, the ground coordinate
lows the continuous truss to maintain the stability and rigidity of the system is particularly set on point O1 .
cube while increasing the load-bearing capacity and deformation resis- Another geometric property is its periodicity. Each truss unit is
tance of the structure through the introduction of diagonal braces. The identical. The truss units are connected sequentially to form a contin-
function of diagonal braces is to transfer forces from one cubic to uous structure. With reference to Fig. 2, for the sake of notation, the kth
another, thereby forming a stable force network throughout the entire truss unit on the right half is denoted by letter A(see Fig. 2a). Similarly,
structure. the kth truss unit on the left half is denoted by letter B(see Fig. 2b).

2
H.-L. Sun et al. Structures 71 (2025) 107891

Fig. 4. Parallel theorem of force.

Continuous truss structures have been widely used in the fields of [ ]

bridges and machinery. The unique construction form of the bridge e
$A = F (2)
rA × e
structure shown in Fig. 3 enables it to maintain good stability even for
large span and heavy load [17]. Similarly, the screw of a couple can also be derived. It consists of two
forces of equal magnitude and opposite direction. Suppose these two
2.2. Screw statics forces act on points A and B respectively. According to the additive
property of linear spaces, there should be
The screw originates from line geometry and is a six-dimensional $AB = $A + $B
dual vector in space. It can simultaneously represent the position and [ ] [ ]
e − e
direction of a straight line [18,19]. In line geometry, screw vectors are =F +F
divided into three types, screw vectors, line vectors, and coupling vec- rA × e rB × (− e) (3)
tors. This corresponds to force, couple, and force screw, respectively. As [
0
]
shown in Fig. 4, the pure force FA acting on point A on the rigid body can =
be equivalently represented by a dual vector. This dual vector consists of rBA × FA
two parts, the main part and the auxiliary part. The main part is the force According to [18], a force screw is the combination of a force F and a
Fo acting on the point of the object that coincides with the origin of the couple of moment M = pF parallelling to it in direction. Assuming that
coordinate frame, and the secondary part is the torque Mo about the this force screw acts on point A of the rigid body, its screw expression is
origin of the coordinate frame. This dual vector is called a force screw. [ ]
The screw of a pure force can be therefore expressed by F
$A = (4)
[ ] rA × F + pF
Fo
$A = (1)
rA × F A where p is the pitch of the force screw.
Based on the above analysis, we can find that any external load can
where the pure force acting on a single rigid body can be equivalently be represented by a force screw. Therefore, when a multi-rigid body
replaced by a dual vector acting on the origin of coordinate frame. This system is in equilibrium, its resultant screw must satisfy:
also illustrates the parallel theorem of force. Fo and FA are identical in
n
magnitude and same in direction, but different points of action. ∑
$i = 0 (5)
Assuming the magnitude of the force is F and its unit direction vector is i=1
e, then there is

Fig. 5. The main framework of the algorithm.

3
H.-L. Sun et al. Structures 71 (2025) 107891

Fig. 6. Simply-supported three-span cable-stayed bridge.

where n is the total number of force screws. Eq. (5) is also known as the To simplify the analysis, some assumptions are made.
equilibrium equation of screw statics. By using Eq. (5), we may deduce
the internal forces of large space truss structures. (1) All cables are ideal flexible lines that transmit only axial tensile
loads. The cross-section area of a cable is independent of the
change in tension, and its axial tensile stiffness is always constant.
2.3. Algorithm for a truss structure (2) The truss consists of rods articulated together. All rods are
regarded as rigid two-force bars.
It is our core ultimate target to find the internal force of each truss (3) The forces on the truss include the pressure from the bridge deck,
rod. Each truss unit consists of 10 rods. The internal forces of these 10 the tension of the cables, and the support of the pylons.
rods (see Fig. 2) form an internal force vector FAk or FBk . Due to the (4) To transferring the external loads, the stiffening beam is laid on
continuity of the truss structure, there is a recursive equation for the continuous truss units. All loads from the bridge deck act directly
internal force vectors of two neighboring truss units. The recursive on the hinges of the truss through the stiffening beams.
equation is expressed as
{ ( )
FAk+1 = f FAk , Righttrussunit 3.2. Internal forces of truss structures
( ) (6)
FBk+1 = f FBk , Lefttrussunit
Based on the main framework of the algorithm in Section 2.3, the
When the recursive equation is obtained, the other internal forces are internal forces of the continuous truss structure are solved for all rods in
easy to find. the bridge.
Fig. 5 illustrates the general framework of this algorithm. First, the
support forces of the pylons are determined. By analyzing the forces in 3.2.1. Support force from pylon to truss structure
the entire continuous truss structure, the statics equations are listed, The first step is to solve the support of the tower on the truss struc-
enabling the calculation of the pylon support forces. Second, the internal ture. As shown in Fig. 7, the cable tension acting on the right truss unit is
forces in boundary units are computed, providing boundary condition. TA(i, j) . Due to the load acting on the truss joints is a planar force system,
Next, a recursive equation for the internal forces between each pair of the angles between the tensile cables and the vertical loads are θ0 ,θ1 ,…,
adjacent truss units is established. Using the internal force vectors from θn . The outermost sides of the truss structure act as its supports by the
the boundary truss units, the internal forces across the truss structure pylons.
can be derived sequentially through the recursive equation. Finally, the As shown in Fig. 7, the two support forces acting on the hinge points
internal forces of all rods on the central face are obtained. at the boundary trusses satisfy
{
NA(n, 1) = NA(n, 4)
3. Internal force investigation of a continuous truss structure (7)
NB(n, 1) = NB(n, 4)

3.1. Theoretical model and condition assumptions Since the two supporting forces and their equivalent forces are
identical, there are
Fig. 6 illustrates a three-span continuous cable-stayed bridge with ⎧
1
high loading capacity and considerable stiffness [20–22]. The ⎨ NA(n, 1) = NA(n, 4) = NA
⎪
⎪
2
cable-stayed bridge often uses cables to connect the bridge deck to the (8)
⎪ 1
pylon. The continuous truss is usually the main load-bearing structure of ⎪
⎩ NB = N B = N B
the bridge, effectively transmitting the load from the deck to the pylon
(n, 1) (n, 4)
2
through the cables. This structural form enables a cable-stayed bridge to After knowing the force on the overall truss structure, an equilibrium
span longer distance while maintaining its lightness and aesthetics. equation can be obtained

Fig. 7. Force analysis of the cable-stayed bridge.

4
H.-L. Sun et al. Structures 71 (2025) 107891

Fig. 8. Force analysis of the truss units: (a) the nth unit in the right half, (b) the joint A(n,3) , (c) the joint A(n,4) .

n
∑ n
∑ From Eq. (12), the difference of the support forces of the two pylons
Tk cos θk − Pk + N = 0 (9) is
k=0 k=1

where Tk is the tension of the kth cable, Pk is the kth external force, and
N is the supporting force of pylons on both sides to the truss structure.
The projection of Eq. (9) onto the y-axis is

( )
n (
∑ ) n
∑
TA(k,2) + TA(k,3) + TB(k,2) + TB(k,3) cos θk − PA(k,2) + PA(k,3) + PB(k,2) + PB(k,3)
k=0 k=1
(10)
− PO2 − PO3 + NA + NB = 0

⎛∑ n ( ) ⎞
From Eq. (10), the sum of the support forces from the two pylons is PA(k,2) + PA(k,3) − PB(k,2) − PB(k,3) k
1 ⎜ k=1
⎜ ⎟
⎟
N A − NB = ⎜ ⎟ (13)
⎡ n (
∑ ) ⎤ n⎝ ∑n ( ) ⎠
PO2 + PO3 − TA(k,2) + TA(k,3) + TB(k,2) + TB(k,3) cos θk − TA(k,2) + TA(k,3) − TB(k,2) − TB(k,3) k cos θk
⎢ ⎥ k=1
⎢
NA + N B = ⎢ ( k=0 ) ⎥
⎥
⎣ ∑n ⎦
+ PA(k,2) + PA(k,3) + PB(k,2) + PB(k,3) Eq. (11) and Eq.(13) provides an expression in scalar form. In matrix
k=1 form, we have
(11)
NA + NB = PO2 + PO3 + eTones,n P[ 1 1 1 1 ]T − ΘTcos T[ 1 1 1 1 ]T
The moment equilibrium equation of the truss about the axes O2 O3 (14)
(see Fig. 1) is

n (
∑ ) n (
∑ ) n (
∑ )
− TB(k,2) + TB(k,3) kl cos θk + TA(k,2) + TA(k,3) kl cos θk + PB(k,2) + PB(k,3) kl
k=0 k=0 k=1
n ( ) (12)
∑
− PA(k,2) + PA(k,3) kl − NB nl + NA nl = 0
k=1

5
H.-L. Sun et al. Structures 71 (2025) 107891

⎛ ⎡ ⎤ ⎞
⎡ ⎤ ⎡ ⎤ 0 0 0 ⋯ 0 ⎡ ⎤
1 0 ⋯ 0 1 ⎢0 1
⎜ 1 0 ⋯ 0⎥ ⎟
1⎜ ⎢0 2 ⋯ 0⎥ ⎢ 1 ⎥ ⎢ ⎥ ⎢ 1 ⎥⎟
NA − NB = ⎜eones,n ⎣
⎜ T ⎢ ⎥P⎢ ⎥ − ΘT ⎢ 0 0 2 ⋯ 0⎥⎥T⎣ − 1 ⎦⎟
⎢ ⎥⎟ (15)
n⎝ ⋮ ⋮ ⋱ ⋮⎦ ⎣− 1⎦ cos ⎢
⎣⋮ ⋮ ⋮ ⋱ ⋮⎦ ⎠
0 0 ⋯ n − 1 − 1
0 0 0 ⋯ n

PA(1,2) PA(1,3) PB(1,2) PB(1,3) Finally, all the support forces can be obtained similarly
⎡ PA(2,2) PA(2,3) PB(2,2) PB(2,3)
⋮ ⋮ ⋮ ⋮
⎢P PA(n,3) PB
where P = ⎢ A(n,2)
⎣
⎞ ⎤ , T =
⎟ ⎥ 3.2.2. Internal forces in boundary units
(n,2⎟PB ⎥
⎠ (n,3) ⎦ After obtaining the support forces, we may calculate the internal
n×4
forces of the boundary truss unit (outermost truss unit) with the external
⎡ ⎤
TA(0,2) TA(0,3) TB(0,2) TB(0,3) forces. The force analysis of the nth unit in the right half is shown in
⎢ TA(1,2) TA(1,3) TB(1,2) TB(1,3) ⎥ Fig. 8a. The tension of the cable, the external force of the vehicle, and
⎢ ⎥ , eones,n = [ 1 1 ⋯ 1 ]T n×1 , and
⎣ ⋮ ⋮ ⋮ ⋮ ⎦ the support of the pylon act on the four hinge points on the right-side
TA(n,2) TA(n,3) TB(n,2) TB(n,3) (n+1)×4

⎧ [ ]
⎪ P O2 + P O3 1 T 1 T 1 T 1 T 0 0
⎪ NA = + eones,n Peones,4 − Θcos Teones,4 + eones,n CPe4 − Te4
⎪
⎪ Θ
⎪
⎨ 2 2 2 2n 2n cos 0 C
[ ] (16)
⎪
⎪
⎪ P O2 + P O3 1 T 1 1 T 1 0 0
⎩ NB = + eones,n Peones,4 − ΘTcos Teones,4 − eones,n CPe4 + ΘTcos Te4
⎪
⎪
2 2 2 2n 2n 0 C

Θcos = [ cos θ0 cos θ1 ⋯ cos θn ]T . truss face. The angle between the tension and the external force is θn .
Associating Eqs. (14) and (15), we obtain the support forces directly: Ref [18] provides a method to solve the internal forces of truss
structures by using the screw statics. For solving the internal forces on
⎡ ⎤
1 0 ⋯ 0 the boundary truss structure shown in Fig. 8, it is necessary to write
⎢0 2 ⋯ 0⎥ screw statics equations only for some critical joints. From Eq. (5), the
where C = ⎣
⎢ ⎥ and e4 = [ 1 1 − 1 − 1 ]T .
⋮ ⋮ ⋱ ⋮⎦ statics screw equation for joint A(n,3) in Fig. 8b is
0 0 ⋯ n

⎧ [ ]
⎪ PO2 + PO3 1 T 1 T 1 T 1 T 0 0
⎪
⎪
⎪ NA = NA(n,4) = + eones,n Peones,4 − Θcos Teones,4 + eones,n CPe4 − Θ Te4
⎪
⎨ (n,1) 4 4 4 4n 4n cos 0 C
[ ] (17)
1 T 0 0
⎪
⎪
⎪ PO2 + PO3 1 T 1 T 1 T
⎪
⎪ N B = N B = + e Pe ones,4 − Θ Te ones,4 − e CPe4 + Θ Te4
⎩ (n,1) (n,4)
4 4 ones,n 4 cos 4n ones,n 4n cos 0 C

Fig. 9. Recursive relationships within the truss units: (a) force analysis of a truss unit, (b) force analysis of the joint A(k,3) , (c) force analysis of the joint A(k,4) .

6
H.-L. Sun et al. Structures 71 (2025) 107891

Fig. 10. Force analysis of the center plane: (a) the center plane, (b) the joint O3 , (c) the joint O4 .

$PA(n,3) + $TA(n,3) + $FA + $FA + $FA =0 (18) The solutions of Eq. (20) are
(n,2) (n,3) (n,7)
⎧
The solutions of Eq. (18) are ⎪
⎪
⎪ FA = 0
⎨ (n,4)
⎧ FA(n,8) = FA(n,3) cot α + NA(n,4) cot α (21)
⎪
⎪
⎪ FA = 0 ⎪
⎪
⎨ (n,2) ⎩ FA
⎪ = − FA cscα − N A(n,4)cscα
FA(n,3) = TA(n,3) cos θn − PA(n,3) (19) (n,10) (n,3)
⎪
⎪
⎩ FA = T
⎪
(n,7) sin θ
A(n,3) n The internal forces in the 8th and 10th rods can be found by
substituting Eq. (19) into Eq. (21):
Similarly, for the joint A(n,4) (see Fig. 8c), the statics screw equation is ⎧
⎨ FA = TA cos θn cot α − PA cot α + NA cot α
+ $NA(n,4) = 0 (20) (22)
(n,8) (n,3) (n,3) (n,4)
$F A + $FA + $FA + $F A
(n,3) (n,4) (n,8) (n,10) ⎩ FA(n,10) = − TA(n,3) cos θn cscα + PA(n,3) cscα − NA(n,4) cscα

Fig. 11. Equivalent external load acting on the truss unit. (a) Equivalent external load model of bridge, (b) Equivalent external load model of a single truss unit.

7
H.-L. Sun et al. Structures 71 (2025) 107891

Similarly, the internal forces in the 1st, 5th, 6th, and 9th rods can be
$FA + $FA + $F(k+1,8) + $F(k,8) + $F(k,10) = 0 (28)
solved (k,3) (k,4)

⎧ The above equation can be solved as

⎪
⎪
⎪ FA(n,1) = TA(n,2) cos θn − PA(n,2) ⎧
⎪
⎪
⎪
⎨ FA = TA cos θn cot α − PA cot α + NA cot α ⎪ FA = 0
⎪
⎪
(23) ⎨ (k,4)
(n,5) (n,2) (n,2) (n,1)

⎪
⎪
⎪ FA(n,6) = TA(n,2) sin θn FA(k,8) = FA(k+1,8) + FA(k,3) cot α (29)
⎪
⎪ ⎪
⎪
⎩ FA(n,9) = − TA(n,2) cos θn cscα + PA(n,2) cscα − NA(n,1) cscα
⎪ ⎪ FA
⎩ = − FA cscα
(k,10) (k,3)

Therefore, the internal forces in boundary units can be expressed by A necessary supplementary equation is provided as

FAn = A1 F 1 (24) F(k,3) = T(n,3) cos θk − P(k,3) − F(k+1,10) sin α (30)

Combining the above equations, FA(k,8) and FA(k,10) can be obtained as

where FAn is the internal forces in boundary units, A1 is a parametric
matrix and is related only to the angle of the force, F1 is a force vector
that includes cable forces, external loads and support forces. Expanding
Eq. (24) yields

⎡ ⎤
⎢ FA(n,1) ⎥
⎢ ⎥
⎢ FA(n,2) ⎥ ⎡ ⎤
⎢
⎢
⎥
⎥ − 1 0 cos θn 0 0 0
⎢
⎢ FA(n,3) ⎥ ⎢
⎥ ⎢ 0 0 0 0 0 0 ⎥
⎢ ⎥ ⎥⎡ ⎤
0 − 1 0 cos θn 0 0 ⎥ ⎥ PA(n,2)
⎢ ⎥ ⎢
⎢ FA(n,4) ⎥ ⎢
⎢
⎢
⎥ ⎢
⎥ ⎢ 0 0 0 0 0 0 ⎥ ⎢ PA(n,3) ⎥
⎥⎢ ⎥
FA(n,5)
⎥⎢ TA(n,2) ⎥
⎢ ⎥ ⎢ − cot α 0 cos θn cot α 0 cot α 0 ⎥ ⎢ ⎥
⎢
⎢
⎥ ⎢
⎥=⎢ ⎢ TA(n,3) ⎥ (25)
⎢ FA(n,6) ⎥ ⎢ 0 0 sin θn 0 0 0 ⎥ ⎥⎢ ⎥
⎢ ⎥ ⎢
0 0 0 sin θn 0 0 ⎥ ⎥ NA(n,1)
⎢ ⎥ ⎣ ⎦
⎢
⎢ FA(n,7) ⎥ ⎢
⎥ ⎢ 0 − cot α 0 cos θn cot α 0 cot α ⎥ N
⎢ ⎥ ⎢ ⎥ A(n,4)
⎢ FA(n,8) ⎥ ⎣ cscα 0 − cos θn cscα 0 − cscα 0 ⎦
⎢ ⎥
⎢ ⎥
0 cscα 0 − cos θn cscα 0 − cscα
FA(n,9)
⎢ ⎥
⎢ ⎥
⎢ ⎥
⎣ ⎦
FA(n,10)

⎧
Similarly, the internal force vector of the left side boundary truss ⎨ FA = FA(k+1,8) + TA(k,3) cos θk cot α − PA(k,3) cot α − FA(k+1,10) cos α
structure can be solved by the same method. (k,8)
⎩ FA(k,10) = − TA(k,3) cos θk cscα + PA(k,3) cscα + FA(k+1,10)
3.2.3. Recursive algorithm for internal force vectors (31)
The recursive equation is used to establish the relationship between
Similarly, the internal forces in the 1th, 6th, 5th, and 9th rods can be
FAk and FAk+1 . The internal force vector is a column vector consisting of
obtained as
the internal forces of the 10 rods in the truss unit. Hence both and FAk
⎧
FAk+1 are internal force vectors. ⎪
⎪ FA(k,1) = TA(k,2) cos θk − PA(k,2) − FA(k+1,9) sin α
⎪
Fig. 9a illustrates the force diagram of a truss unit. Let us consider ⎪
⎪
⎪
⎨ FA = F A
(k+1,6) + F (k+1,9) cos α + TA(k,2) sin θk
A
truss units k and k + 1, and assume that FAk+1 is known and solve for FAk . (32)
(k,6)

⎪ F(k,5) = F(k+1,5) + TA(k,2) cos θk cot α − PA(k,2) cot α − F(k+1,9) cos α

⎪ A A A
Then, the static equations of some key joints are established. The key ⎪
⎪
⎪
joints should include as many acting forces as possible, such as cable ⎩ FA(k,9) = − TA(k,2) cos θk cscα + PA(k,2) cscα + FA(k+1,9)
⎪
tension or external load, which will provide more known information for
the next step in the calculation. Based on the above analysis, the recursive equation can be obtained
There is no doubt that joint A(k,3) and A(k,4) are a good object. For the in the adjacent truss structure as
joint A(k,3) as shown in Fig. 9b, the screw statics equation is established ( )
FAk = AFAk+1 + Bk LAk , k = 1, 2, …n − 1 (33)
as
$F A + $FA + $FA + $FA + $FA + $T(n,3) + $F(n,3) = 0 (26) where
(k+1,10) (k+1,7) (k,2) (k,3) (k,7)

Eq. (26) can be solved as

⎧
⎪ F(k,2) = 0
⎪ A
⎪
⎨
FA(k,3) = TA(k,3) cos θk − PA(k,3) − FA(k+1,10) sin α (27)
⎪
⎪ Table 1
⎩ FA(k,7) = FA(k+1,7) + FA(k+1,10) cos α + TA(k,3) sin θk
⎪
Structural parameters of continuous truss structures.

For the joint A(k,4) as shown in Fig. 9c, the screw statics equation is Number of unilateral units Rod length α Material

given as 15 1m π/4 high-strength steel

8
H.-L. Sun et al. Structures 71 (2025) 107891

⎡ ⎤
0 0 0 0 0 0 0 0 − sinα 0
⎢0 0 0 0 0 0 0 0 0 0 ⎥ FBn = A1 F2 (34)
⎢ ⎥
⎢0 0 0 0 0 0 0 0 0 − sinα ⎥
⎢
⎢0 0 0 0 0 0 0 0 0 0 ⎥
⎥ where A1 is just the same matrix in Eq. (24) and F2 is expressed as
⎢ ⎥
⎢0 0 0 0 1 0 0 0 − cosα 0 ⎥ [ ]T
A=⎢
⎢0
⎥, F2 = PB(n,2) PB(n,3) TB(n,2) TB(n,3) NB(n,1) NB(n,4) (35)
⎢ 0 0 0 0 1 0 0 cosα 0 ⎥ ⎥
⎢0 0 0 0 0 0 1 0 0 cosα ⎥
⎢ ⎥ Eq. (33) is the recursive equation for the right-side truss units. By
⎢0 0 0 0 0 0 0 1 0 − cosα ⎥
⎢
⎣0
⎥ using the same method, we can obtain the recursive equation for the
0 0 0 0 0 0 0 1 0 ⎦
left-side truss units as
0 0 0 0 0 0 0 0 0 1
( )
⎡ ⎤ FBk = AFBk+1 + Bk LBk k = 1, 2, …n − 1 (36)
− 1 0 cosθk 0
⎢ 0 0 0 0 ⎥
⎢ ⎥ where A and Bk are just the same matrix in Eq. (33), and LBk is expressed
⎢ 0 − 1 0 cosθk ⎥
⎢
⎢ 0
⎥ as
⎢ 0 0 0 ⎥
⎥
⎢ − cotα 0 cosθk cotα 0 ⎥ [ ]T
Bk = ⎢ ⎥ LBk = PB(k,2) PB(k,3) TB(k,2) TB(k,3) (37)
⎢ 0 0 sinθk 0 ⎥
⎢ ⎥
⎢ 0 0 0 sinθk ⎥ From this, FBn− 1 , FBn− 2 , …, FB1 can be obtained in order.
⎢ ⎥
⎢ 0 − cotα 0 cosθk cotα ⎥
⎢ ⎥
⎣ cscα 0 − cosθk cscα 0 ⎦
3.2.4. Internal forces in the center plane
0 cscα 0 − cosθk cscα
After obtaining the known forces and recursive equations in Section
3.2.1 to Section 3.2.3, the next step is to determine the internal forces in
, and
the center plane of symmetry for the truss rods. The forces on the center
[ ]T
LAk = PA(k,2) PA(k,3) TA(k,2) TA(k,3) . face are shown in Fig. 10a.
Similar to the previous analytical idea, the static equations are ob-
From this, FAn− 1 , FAn− 2 , …, FA1 can be solved in order. tained for joints O3 and O4 , respectively.
Eq. (24) expresses the internal force vector of the rightmost truss As shown in Fig. 10b, the forces acting on joint O3 include the tension
unit, namely FAn . Similarly, the internal force vector of the leftmost truss of the cable, the external forces, and the internal forces of the five rods.
unit can be obtained, namely FBn as For joint O3 , its screw statics equation is given as

Fig. 12. The variation functions of the external loads.

Fig. 13. The support force of pylon.

9
H.-L. Sun et al. Structures 71 (2025) 107891

Fig. 14. The internal forces in boundary units.

Fig. 15. The internal forces of the 10th truss unit.

Fig. 16. The internal force vector of the 11th truss unit.

10
H.-L. Sun et al. Structures 71 (2025) 107891

Fig. 17. Comparison of calculation results by using two methods: (a) Left truss unit, (b) Right truss unit.

$PO3 + $TA(0,3) + $TB(0,3) + $FO2 + $FB + $FA + $FB + $F A + $FO3 = 0 example, for the loads within rectangle A(k− 1,3) A(k− 1,2) A(k,2) A(k,3) (denoted
(1,7) (1,7) (1,10) (1,10)

(38) as SAk ), the forces are

∫
Similarly, the forces acting on joint O4 are only the internal forces of PAk (t) = q(x, y, t)dxdy (42)
the 4 rods shown in Fig. 10c. Its screw statics equation is SA
k

$FO3 + $FO4 + $FA + $FB =0 (39) Then, by the assumption of equal sharing, joint A(k− 1,2) , A(k− 1,3) , A(k,2) ,
(1,8) (1,8)
A(k,3) bears the force of 14PAk separately.
In addition, the static equation for joint O1 can be derived as Therefore, for joint A(k,l) (k = 1, 2, …, n − 1, l = 2, 3), the bridge
$FO1 + $FO4 + $FA + $FB =0 (40) pressure on it is
(1,5) (1,5)

1 1
Based on computer programming to solve Eqs. (38), (39) and (40), PA(k,l) (t) = PAk (t) + PAk+1 (t) (43)
4 4
the forces on the four rods in the center plane can be found. Therefore,
the internal forces of all rods can be obtained. According to Eq. (42), the force acting on each joint is
∫ ∫
1 1 ( )
4. Application on bridge structures q(x, y, t)dxdy = Aq sin wt − bx + q0 dxdy
4 SAk 4 SAk
∫ ∫
4.1. Expression of the external load Aq L (k+1)L q0 L2
= sin(wt − bx)dxdy + (44)
4 0 kL 4
( ( ( ) ) ( ))
The vehicles on the bridge are constantly changing as external loads, Aq L q 0 L2
and it is necessary to find out the real-time changes in the internal forces = cos wt − k + 1 bL − cos wt − kbL +
4b 4
of the trusses for better monitoring. The external loads acting on a
continuous truss structure are uncertain. In order to better analyze the where L is the length of stiffening beam.
internal forces of the truss, an equivalent external load model is Similarly, the force acting on the k + 1th joint is
developed.

∫ ( ( ( ) ) ( ( ) ))
1 Aq L q 0 L2
q(x, y, t)dxdy = cos wt − k + 2 bL − cos wt − k + 1 bL + (45)
4 A
Sk+1 4b 4

With reference to Fig. 11a, an equivalent external load model of the

bridge is provided. As the vehicle acts on the stiffening beams, the
stiffening beams are laid on the trusses. The stiffener beams are con- Based on Eq. (44) and Eq. (45), the force acting on the joint A(k,l) is
nected to the individual truss units by 4 joints. Therefore, it can be
assumed that the external load acts uniformly on the 4 joints of the truss Aq L q0 L2
PA(k,l) (t) = (cos(ωt − (k + 1)bL) − cos(ωt − kbL)) +
unit (Fig. 11b). 4b 4
Let the load set q of the vehicle on the bridge be a ternary function Aq L q 0 L2
+ (cos(ωt − (k + 2)bL) − cos(ωt − (k + 1)bL)) +
with respect to x, y, and time t. For simplicity of calculation, assume that 4b 4
(46)
q is uniform in the y-direction, and let the Aq L q0 L2
= cos((ωt − (k + 2)bL) − cos(ωt − kbL) +
( ) ( ) 4b 2
q = q x, y, t = q0 + Aq sin ωt − bx (41)
Aq L q 0 L2
= sin(ωt − (k + 1)bL)sin(bL) +
It is approximated that the vehicle load on the roadway directly 2b 2
above each unit is shared equally by the four vertices above unit. For
Similarly, the force acting on other joints can be obtained as

11
H.-L. Sun et al. Structures 71 (2025) 107891

⎧
⎪ Aq L q 0 L2
⎪ PB(k,l) (t) = − 2b sin(ωt + (k + 1)bL))sin(bL) + 2 , (k = 1, 2, …, n − 1; l = 2, 3)
⎪
⎪
⎪
⎪
⎪
⎪ (
⎨
Aq L q0 L2
PA(n,l) (t) = cos(ωt − (n + 1)bL) − cos(ωt − nbL) + , (l = 2, 3) (47)
⎪
⎪
⎪ 4b 4
⎪
⎪
⎩ PB (t) = Aq L (cos(ωt + (n + 1)bL) − cos(ωt + nbL)) + q0 L , (l = 2, 3)
⎪
⎪ 2
⎪
(n,l)
4b 4

Fig. 17 shows the comparison of these two methods. Method 1 repre-

The force acting on the joint Ol (l = 2, 3) is sents the recursive force transfer method. Method 2 denotes the theo-
retical mechanics method.
Aq L q0 L2
P Ol = cos ωt(cos(2bL) − cos(bL)) + , (l = 2, 3) (48) Fig. 17 indicates that the results obtained by the two methods are
2b 2
exactly the same, which verifies the correctness of the recursive force
In Table 1, some parameters are given to analyze the internal forces transfer method.
in the rods. The angle between the cable and the vertical direction is Compared with theorical mechanics methods, the advantage of the
arranged in a uniformly decreasing manner. From the middle to both recursive force transfer method is that the computational cost and
sides, it decreases uniformly from 60 to 0 (60 , 56 , 52 , ⋯, 4 , 0 ). The complexity are much less. Comparatively speaking, we get the following
◦ ◦ ◦ ◦ ◦ ◦ ◦

rod is a circular cross-section with diameter of 100 mm. facts:

The sine load concentration of the load is q = q0 + Aq sin(ωt − bx). When using theoretical mechanics methods, the primary computa-
The initial value q0 is 10N/m2 . The amplitude q of the load is 10N/m2 . tional effort involves solving the inverse of a 10n × 10n matrix, where n
The angular frequency is ω = 1rad/s with coefficient of b = 1. Fig. 12 denotes the number of truss units on each side. For recursive force
shows the real-time variation functions of the external loads. In Fig. 12, transfer method, from the force transfer formula (39), it can be seen that
it can be seen that the external load is a sinusoidal function. the maximum dimension of the matrix is 10 × 10, which significantly
reduces the complexity of the calculation. In the finite element method,
computational complexity increases with finer mesh division. This
4.2. Numerical verification method is often used to verify theoretical models, so a detailed com-
parison is not provided here.
After knowing the geometric parameters and external load distri- When solving a system of linear equations, the recursive force
bution of the bridge structure, this section provides the internal force transfer method clearly exhibits lower computational complexity than
calculation results of the rods based on the proposed algorithm. the theoretical mechanics method.
The support force of pylon is determined by using the algorithm
proposed in Section 3.2.1. The calculation results are shown in Fig. 13. 5. Conclusions
The internal force vector of the rods on the boundary truss unit is
calculated by using the method in Section 3.2.2. The calculation results This paper proposes an algorithm for determining the internal forces
are shown in Fig. 14. In Fig. 14, the internal forces change with the in continuous truss structures by an application of a cable-stayed bridge.
external loads. The values of internal forces are not the same for With a large number of truss structure units, both the finite element
different rods, but they all show a sinusoidal variation. In addition, rods method and the equivalent modeling method present certain limitations.
2, 4, 6, 7 are zero-force rods. The recursive force transfer method helps quickly calculate the internal
The internal force vector of the rods on each truss unit in a contin- forces in a continuous truss structure before performing a more accurate
uous truss structure is calculated by using the recursive equation. For the finite element analysis. The required truss material is evaluated based on
sake of representation, the 10th truss unit is selected as the object, and the maximum internal force. From this purpose, a recursive approach is
its calculation result is shown in Fig. 15 which indicates that rods 2, 4 proposed to determine the internal forces in a continuous truss structure.
are zero-force rods. This can also be demonstrated by the force balance This algorithm derives a recursive transfer equation for the internal
equation in the x-axis direction in Fig. 9. force vectors between neighboring truss units, which shows the force
After obtaining the internal force vector on the 10th truss unit, the transfer performance in the truss structure. Based on the recursive
internal force vector on the 11th truss unit can be calculated based on equations, the internal forces of all rods are obtained. Unlike traditional
the recursive equation. The calculation results are shown in Fig. 16. As methods that involve solving numerous linear equations, matrix pro-
can be seen in Fig. 16, rods 2 and 4 are zero-force rods. gramming of the recursive equation can efficiently handle large-scale
From these figures, the zero-force rods are easily found through equations in continuous truss analysis.
recursive algorithms. It can be seen that the calculation process of the
algorithm in solving the internal forces of large continuous truss struc- CRediT authorship contribution statement
tures is simple. In each internal force vector diagram, the maximum
internal force will be used to calculate the maximum stress of the ma- Jing-Shan Zhao: Writing – review & editing, Writing – original
terial. By selecting appropriate materials, the total costs of building the draft, Validation, Supervision, Resources, Project administration,
truss can be greatly saved. Methodology, Investigation, Funding acquisition, Formal analysis,
To verify the reliability of the recursive force transfer method, a Conceptualization. Hanlin Sun: Writing – original draft, Visualization,
comparison is provided between this approach and theoretical me- Validation, Software, Methodology, Investigation, Formal analysis, Data
chanics methods. Unlike finite element method, comparing our curation. Zhi-Rui He: Writing – review & editing, Writing – original
approach with theoretical mechanics not only validates our method, but draft, Visualization, Validation, Software, Investigation, Formal
also provides a detailed explanation of the advantages of our methods. analysis.
Appendix 2 provides the calculation procedure by using theoretical
mechanics method for comparison.
We also used the recursive force transfer method for calculation.

12
H.-L. Sun et al. Structures 71 (2025) 107891

Declaration of Competing Interest Foundation of China under Grant 51575291, in part by the National
Major Science and Technology Project of China under Grant
The authors declare that there is no conflict interest in the paper. 2015ZX04002101, in part by the State Key Laboratory of Tribology,
Tsinghua University, and in part by the 221 Program of Tsinghua
Acknowledgments University.

This work was supported in part by the National Natural Science

Appendix 1

Notation Meaning

PA(k,l) k = 1, 2, …, n;l = 2, 3 Bridge pressure acting on joint A(k,l)

PB(k,l) k = 1, 2, …, n;l = 2, 3 Bridge pressure acting on joint B(k,l)
PO2 Bridge pressure acting on joint O2
PO3 Bridge pressure acting on joint O3
TA(k,l) k = 0, 1, …, n;l = 2, 3 Tension of the cable acting on the jointA(k,l)
TB(k,l) k = 0, 1, …, n;l = 2, 3 Tension of the cable acting on the jointB(k,l)
θk Angle between the cable connected to the joint A(k,2) , A(k,2) , B(k,2) , B(k,3) and the vertical direction
NA(n,l) l = 1, 4 Pylon support force acting on joint A(n,l)
NB(n,l) l = 1, 4 Pylon support force acting on joint B(n,l)
FA(k,1) Internal forces in A(k,1) A(k,2) rod
FA(k,2) Internal forces in A(k,2) A(k,3) rod
FA(k,3) Internal forces in A(k,3) A(k,4) rod
FA(k,4) Internal forces in A(k,4) A(k,1) rod
FA(k,5) Internal forces in A(k,1) A(k− 1,1) rod
FA(k,6) Internal forces in A(k,2) A(k− 1,2) rod
FA(k,7) Internal forces in A(k,3) A(k− 1,3) rod
FA(k,8) Internal forces in A(k,4) A(k− 1,4) rod
FA(k,9) Internal forces in A(k,1) A(k− 1,2) rod
FA(k,10) Internal forces in A(k,4) A(k− 1,3) rod
FB(k,1) Internal forces in B(k,1) B(k,2) rod
FB(k,10) Internal forces in B(k,4) B(k− 1,3) rod
[ ]T
FAk FAk = FA(k,1) , FA(k,2) , …, FA(k,10)
[ ]T
FBk FBk = FB(k,1) , FB(k,2) , …, FB(k,10)

Appendix 2

To verify the reliability of the recursive force transfer method, we used theoretical mechanics method to calculate the internal forces of a
continuous truss structure. The calculation process of theoretical mechanics method is as follows:
Theoretical mechanics method is cumbersome when dealing with continuous trusses with multiple units. Therefore, selecting four truss units is
simpler in calculation. The force analysis of the truss structure composed of four truss units is shown in Fig. A1.

Fig. A1. The force analysis of the truss structure.

.
Firstly, we analyze the forces acting on the two trusses on the right. According to the nodal method, regarding joint A(2,1) , its force balance equation

13
H.-L. Sun et al. Structures 71 (2025) 107891

on the three coordinate axes is

FA(2,4) = 0 (49)

− FA(2,5) − FA(2,9) cos α = 0 (50)

FA(2,9) sin α + FA(2,1) + NA(2,1) = 0 (51)

For joint A(2,2) , it leads to

FA(2,2) = 0 (52)

− FA(2,6) + TA(2,2) sin θ2 = 0 (53)

− PA(2,2) + TA(2,2) cos θ2 − FA(2,1) = 0 (54)

Similarly, for joint A(2,3) , its equilibrium equation in the y and z directions can be obtained as

− FA(2,7) + TA(2,3) sin θ2 = 0 (55)

− PA(2,3) + TA(2,3) cos θ2 − FA(2,3) = 0 (56)

For joint A(2,4) , its results are

− FA(2,8) − FA(2,10) cos α = 0 (57)

FA(2,10) sin α + FA(2,3) + NA(2,4) = 0 (58)

Next, we will formulate the equilibrium equation for the joint A(1,1) on another truss plane. It leads to
FA(1,4) =0 (59)

− FA(1,5) − FA(1,9) cos α + FA(2,5) = 0 (60)

FA(1,9) sin α + FA(1,1) = 0 (61)

For the remaining three joints, the following 7 equations can be obtained. For joint A(1,2) ,
FA(1,2) =0 (62)

− FA(1,6) + TA(1,2) sin θ1 + FA(2,6) + FA(2,9) cos α = 0 (63)

− PA(1,2) + TA(1,2) cos θ1 − FA(1,1) − FA(2,9) sin α = 0 (64)

For joint A(1,3) ,

− FA(1,7) + TA(1,3) sin θ1 + FA(2,7) + FA(2,10) cos α = 0 (65)

− PA(1,3) + TA(1,3) cos θ1 − FA(1,3) − FA(2,10) sin α = 0 (66)

For joint A(1,4) ,

− FA(1,8) − FA(1,10) cos α + FA(2,8) = 0 (67)

FA(1,10) sin α + FA(1,3) = 0 (68)

By organizing Eq. (51) to (60), they can be expressed in matrix form as follows

14
H.-L. Sun et al. Structures 71 (2025) 107891

⎡ ⎤
⎢ A
⎢ F(1,1) ⎥
⎥
⎢ ⎥
⎢ A ⎥
⎢ F(1,2) ⎥
⎢ ⎥
⎢ A ⎥
⎢F ⎥
⎢ (1,3) ⎥ ⎡ ⎤− 1 ⎡ ⎤
0 0 1 00 0
⎢ A ⎥
⎢F ⎥ ⋅ ⋅ ⋅
⎢ (1,4) ⎥
⎢ ⎥ ⎢0 1 0 ⎥
cosα ⎢ 0 ⎥
⎢ FA ⎥
⎢ (1,5) ⎥ ⎢ 1 sinα
⎥ ⎢
− NA(2,1)
⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ A ⎥ ⎢ 1
⎥ ⎢
0
⎥
⎢ F(1,6) ⎥ ⎢ ⎥
⋅ ⎥ ⎢ ⎢ ⎥
⎥ ⎢ ⎥
1 T sinθ
⎢
⎢ A ⎥ ⎢ ⋅ ⎥ ⎢ A 2 ⎥
⎢ F(1,7) ⎥ ⎢ ⎥ ⎢ (2,2) ⎥
⎢
⎢ A
⎥ ⎢⋅
⎥ ⎢ 1 ⋅ ⎥⎥ ⎢
⎢ TA(2,2) cosθ2 − PA(2,2) ⎥
⎥
⎢F ⎥ 1 TA(2,3) sinθ2
⎢ (1,8) ⎥ ⎢ ⋅ ⎥ ⎢ ⎥
⎢ A ⎥ ⎢ 1
⎥ ⎢
⎥ ⎢ TA(2,3) cosθ2 − PA(2,3) ⎥
⎥
⎢F ⎥ ⎢⋅
⎢ (1,9) ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ 1 cosα ⎥ ⎢
⎥ ⎢ 0 ⎥
⎢ FA ⎥ ⎢ ⎥
⎢ (1,10) ⎥ ⎢ 1 sinα ⎥ ⎢
⎥ ⎢ − NA(2,4) ⎥
⎢ ⎥=⎢ ⎥ (69)
⎢ A
⎢ F(2,1) ⎥ ⎢
⎥ ⎢ 1 ⎥ ⎢
⎥ ⎢ 0 ⎥
⎥
⎢
⎢ A
⎥ ⎢
⎥ 1 cosα − 1 ⎥ ⎢ 0 ⎥
⎢ F(2,2) ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ 1 sinα ⎥ ⎢ 0 ⎥
⎢ A ⎥ ⎢ ⎥ ⎢ ⎥
⎢ F(2,3) ⎥ ⎢
⎢ ⎥ ⎢ 1 ⎥ ⎢
⎥ ⎢ 0 ⎥
⎥
⎢ A
⎢F
⎥ ⎢
⎥ ⎢ − 1 1 cosα ⎥
⎥ ⎢
⎢ − TA(1,2) sinθ1 ⎥
⎥
⎢ (2,4) ⎥ ⎢ 1 sinα ⎥ ⎢ TA cosθ1 − PA ⎥
⎢ ⎥ ⎢ ⎥ ⎢ (1,2) (1,2) ⎥
⎢F A ⎥
⎢ (2,5) ⎥ ⎢ − 1 1 cosα ⎥ ⎢ − TA(1,3) sinθ1 ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ A ⎥ ⎢
⎢ F(2,6) ⎥ ⎢ 1 sinα ⎥ ⎢T
A
⎥ ⎢ (1,3) cosθ 1 − P A
⎥
(1,3) ⎥
⎢ ⎥ ⎣0 1 cosα − 1 0 ⎦ ⎣ 0 ⎦
⎢ A ⎥
⎢ F(2,7) ⎥ 0 0 1 ⋅ ⋅ ⋅ sinα ⋅ ⋅ ⋅ 0 0 0
⎢ ⎥
⎢ A ⎥
⎢ F(2,8) ⎥
⎢ ⎥
⎢ A ⎥
⎢F ⎥
⎢ (2,9) ⎥
⎢ ⎥
⎢F A ⎥
⎣ (2,10) ⎦

The truss structure on the left is completely consistent with the analysis approach on the right side. Therefore, the internal force equations of the
left truss can also be obtained as

⎡ ⎤
⎢ B ⎥
⎢ F(1,1) ⎥
⎢ ⎥
⎢ B ⎥
⎢F ⎥
⎢ (1,2) ⎥
⎢ B ⎥
⎢F ⎥
⎢ (1,3) ⎥ ⎡ ⎤− 1 ⎡ ⎤
0 0 1 00 0
⎢ ⎥
⎢ FB ⎥ ⋅ ⋅ ⋅
⎢ (1,4) ⎥
⎢ ⎥ ⎢0 1 0 ⎥
cosα ⎢ 0 ⎥
⎢ B
⎢ F(1,5) ⎥ ⎢
⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ 1 sinα ⎥ ⎢
⎥ ⎢ − N B(2,1) ⎥
⎥
⎢ B
⎢ F(1,6) ⎥ ⎢
⎥ 1 ⎥ ⎢ 0 ⎥
⎥ ⎢ ⎥ ⎢ ⎥
1 T sinθ
⎢
⎢ B ⎥ ⎢⋅ ⎥ ⎢ B(2,2) 2 ⎥
⎢ F(1,7) ⎥ ⎢ ⎥ ⎢ ⎥
⎢
⎢ B
⎥ ⎢⋅
⎥ ⎢ 1 ⎥ ⎢ TB(2,2) cosθ2 − PB(2,2) ⎥
⎥ ⎢ ⎥
⎢F ⎥ 1 TB(2,3) sinθ2
⎢ (1,8) ⎥ ⎢ ⋅ ⎥ ⎢ ⎥
⎢ B ⎥ ⎢ 1
⎥ ⎢
⎥ ⎢ TB(2,3) cosθ2 − PB(2,3) ⎥
⎥
⎢F ⎥ ⎢
⎢ ⎥ ⎢ ⎥
1 cosα ⎥ ⎢ 0
⎢ (1,9) ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ ⎥
⎢ FB ⎥
⎢ (1,10) ⎥ ⎢ 1 sinα ⎥ ⎢ − N B(2,4)
⎥
⎢ ⎥=⎢ ⎥ ⎢ ⎥ (70)
⎢ B
⎢ F(2,1) ⎥ ⎢
⎥ ⎢ 1 ⎥ ⎢
⎥ ⎢ 0 ⎥
⎥
⎢
⎢ B
⎥ ⎢
⎥ 1 cosα − 1 ⎥ ⎢ 0 ⎥
⎢ F(2,2) ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ 1 sinα ⎥ ⎢ 0 ⎥
⎢ B ⎥ ⎢ ⎥ ⎢ ⎥
⎢ F(2,3) ⎥ ⎢
⎢ 1 ⎥ ⎢
⎥ ⎢ 0 ⎥
⎥
⎢ ⎥
⎢ B
⎢F
⎥ ⎢
⎥ ⎢ − 1 1 cosα ⎥ ⎢
⎥ ⎢ − T B(1,2) sinθ 1
⎥
⎥
⎢ (2,4) ⎥ ⎢ 1 sinα ⎥ ⎢ TB cosθ1 − PB ⎥
⎢ B ⎥ ⎢ ⎥ ⎢ (1,2) (1,2) ⎥
⎢F ⎥
⎢ (2,5) ⎥ ⎢ − 1 1 cosα ⎥ ⎢
⎥ ⎢ − TB(1,3) sinθ1 ⎥
⎢ ⎥ ⎢ ⎥
⎢ FB ⎥ ⎢0 1 sinα ⎥ ⎢T
⎥ ⎢ B(1,3) cosθ 1 − P ⎥
B(1,3) ⎥
⎢ (2,6) ⎥ ⎢
⎢ ⎥ ⎣0 1 cosα − 1 0 ⎦ ⎣ 0 ⎦
⎢ B
⎢ F(2,7) ⎥
⎥
⎢ ⎥ 0 0 1 ⋅ ⋅ ⋅ sinα ⋅ ⋅ ⋅ 0 0 0
⎢ B ⎥
⎢ F(2,8) ⎥
⎢ ⎥
⎢ B ⎥
⎢ F(2,9) ⎥
⎢ ⎥
⎢ B ⎥
⎢F ⎥
⎣ (2,10) ⎦

15
H.-L. Sun et al. Structures 71 (2025) 107891

Table A1
Structural parameters and force sizes.
Angle θ0 θ1 θ2

Number of degrees π π 0
3 6
Force PB(2,2) PB(1,2) PO2 PA(1,2) PA(2,2)
Size (N) 5.0 6.0 4.0 3.0 6.0

Force PB(2,3) PB(1,3) PO3 PA(1,3) PA(2,3)

Size 4.0 8.0 3.0 4.0 8.0

Tension TB(2,2) TB(1,2) TB(0,2) TA(0,2) TA(1,2) TA(2,2)
Size 3.6 7.2 9.6 13.2 8.4 4.8

Tension TB(2,3) TB(1,3) TB(0,3) TA(0,3) TA(1,3) TA(2,3)

Size 3.6 6.0 10.8 12.0 7.2 4.8

For ease of comparison, some detail parameters are substituted into the equation above for verification. Some parameters are shown in Table A1.
By substituting concrete parameters into Eqs. (69) and (70), the internal forces of all rods can be obtained as.
⎡ ⎤ ⎡ ⎤
⎢ A
⎢ F(1,1) ⎥
⎥ ⎢ B ⎥
⎢ F(1,1) ⎥
⎢ ⎥ ⎢ ⎥
⎢ A ⎥ ⎢ B ⎥
⎢ F(1,2) ⎥ ⎢F ⎥
⎢ ⎥ ⎢ (1,2) ⎥
⎢ A ⎥ ⎢ B ⎥
⎢F ⎥ ⎢F ⎥
⎢ (1,3) ⎥ ⎡ ⎤ ⎢ (1,3) ⎥ ⎡ ⎤
− 1.07 − 4.16
⎢ A ⎥ ⎢ ⎥
⎢F ⎥ ⎢ FB ⎥
⎢ (1,4) ⎥ ⎢ (1,4) ⎥
⎢ ⎥ ⎢ 0.00 ⎥ ⎢ ⎥ ⎢ 0.00 ⎥
⎢ FA ⎢ B
⎢ F(1,5) ⎥ ⎢
⎥ ⎥
⎢ (1,5) ⎥ ⎢ − 4.74 ⎥
⎥
− 5.83 ⎥
⎥
⎢ ⎥ ⎢ ⎢ ⎥ ⎢
⎥ ⎢ ⎥ ⎥ ⎢ ⎥
⎢ F(1,6) ⎥ ⎢ 0.00 ⎥ ⎢ F(1,6) ⎥ ⎢ 0.00 ⎥
⎢ A ⎢ ⎢ B ⎢
⎥ ⎥
− 3.84 ⎥ − 6.34 ⎥
⎢ ⎥ ⎢ ⎢ ⎥ ⎢
⎢ A ⎥ ⎢ B ⎥
⎢ F(1,7) ⎥ ⎢ ⎥ ⎢ F(1,7) ⎥ ⎢ ⎥
⎢ ⎥ ⎢ 13.52 ⎥ ⎢ ⎥ ⎢ 10.99 ⎥
⎢ A ⎥ ⎢ ⎥ ⎢ B ⎥ ⎢ ⎥
⎢F ⎥ ⎢ 14.52 ⎥ ⎢F ⎥ ⎢ 8.99 ⎥
⎢ (1,8) ⎥ ⎢ ⎢ (1,8) ⎥ ⎢
⎢ A ⎥ ⎢ − 9.51 ⎥ ⎥ ⎢ B ⎥ ⎢ − 7.01 ⎥ ⎥
⎢F ⎥ ⎢F ⎥
⎢ (1,9) ⎥ ⎢ ⎥ ⎢ (1,9) ⎥ ⎢ ⎥
[ ] ⎢ A
⎥ ⎢ ⎢ 1.52 ⎥⎥ [ ] ⎢ B
⎥ ⎢ ⎢ 5.88 ⎥⎥
⎢F ⎥ ⎢ ⎢F ⎥ ⎢
A
F1 ⎥ = ⎢ 6.71 ⎥(N)
⎢ (1,10) ⎥
⎢
⎥ B
F1 ⎥ = ⎢ 8.24 ⎥(N)
⎢ (1,10) ⎥
⎢
⎥
= ⎥ ⎢ − 3.04 ⎥ = ⎥ ⎢ − 2.78 ⎥
A
F2 ⎢ A
⎢ F(2,1) ⎥ ⎢
B
F2 ⎢ B
⎢ F(2,1) ⎥ ⎢
⎢ ⎥ ⎢ 0.00 ⎥ ⎥ ⎢ ⎥ ⎢ 0.00 ⎥ ⎥
⎢ A ⎥ ⎢ B ⎥
⎢ F(2,2) ⎥ ⎢ ⎥ ⎢ F(2,2) ⎥ ⎢ ⎥
⎢ ⎥ ⎢ − 5.04 ⎥ ⎢ ⎥ ⎢ − 1.78 ⎥
⎢ A ⎥ ⎢ ⎥ ⎢ B ⎥ ⎢ ⎥
⎢ F(2,3) ⎥ ⎢ 0.00 ⎥ ⎢ F(2,3) ⎥ ⎢ 0.00 ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ A ⎥ ⎢ − 2.77 ⎥ ⎢ B ⎥ ⎢ − 2.18 ⎥
⎢F ⎥ ⎢ ⎥ ⎢F ⎥ ⎢ ⎥
⎢ (2,4) ⎥ ⎢ 3.78 ⎥ ⎢ (2,4) ⎥ ⎢ 2.84 ⎥
⎢ ⎥ ⎢ ⎥ ⎢ B ⎥ ⎢ ⎥
⎢F A ⎥ ⎢F ⎥
⎢ (2,5) ⎥ ⎢ 3.78 ⎥ ⎢ (2,5) ⎥ ⎢ 2.84 ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎢ A
⎢ F(2,6) ⎥ ⎢
⎥ ⎢ − 4.77 ⎥⎥ ⎢ FB ⎥
⎢ (2,6) ⎥ ⎢
⎢ − 1.18 ⎥⎥
⎢ ⎥ ⎣ 3.92 ⎦ ⎢ ⎥ ⎣ 3.09 ⎦
⎢ B
⎢ F(2,7) ⎥
⎢ A ⎥ ⎥
⎢ F(2,7) ⎥ 6.74 1.67
⎢ ⎥ ⎢ ⎥
⎢ A ⎥ ⎢ B ⎥
⎢ F(2,8) ⎥ ⎢ F(2,8) ⎥
⎢ ⎥ ⎢ ⎥
⎢ A ⎥ ⎢ B ⎥
⎢F ⎥ ⎢ F(2,9) ⎥
⎢ (2,9) ⎥ ⎢ ⎥
⎢ ⎥ ⎢ B ⎥
⎢ FA ⎥ ⎢F ⎥
⎣ (2,10) ⎦ ⎣ (2,10) ⎦

References [4] Wu L, Dai J. A novel tensegrity ortho-triplex derived by the linkage-truss

transformation with prestress-stability analysis using screw theory. J Mech Des -
Trans ASME 2020;27:013302.
[1] Cao S, et al. Extended continuum model for dynamic analysis of beam-like truss
[5] Meriam J., Kraige L. Engineering Mechanics: Statics. 4rd ed. 2016: 174.
structures with geometrical nonlinearity. Aerosp Sci Technol 2020;103:105927.
[6] Liu H. Mechanics of materials I. 6nd. Beijing, 2017.
[2] Liu R, Guo H, Liu R, et al. Structural design and optimization of large cable-rib
[7] Pytel A., Kiusalaas J. Engineering Mechanics: Statics. 3rd ed. 2010.
tension deployable antenna structure with dynamic constraint. Acta Astronaut
[8] Cao S, Huo M, Qi N. Extended continuum model for dynamic analysis of beam-like
2018;151:160–72.
truss structures with geometrical nonlinearity. Aerosp Sci Technol 2020;103:
[3] Chen W, Chu F, Yan S. Integrated optimization design of active vibration control
105927.
system for piezoelectric intelligent truss structures using immune algorithm. Int J
[9] Bedford A. Engineering Mechanics: Statics. 2rd ed. 2008.
Nonlinear Sci Numer Simul 2007;8(4):527–34.
[10] Costanzo F, Plesha ME, Gray GL. Eng Mech: Statics 2010.
[11] Hu Q, Su L, Cao Y, et al. Decentralized simple adaptive control for large space
structures. J Sound Vib 2018;427:95–119.

16
H.-L. Sun et al. Structures 71 (2025) 107891

[12] Voormeeren S, Van Der Valk P, Rixen D. Generalized methodology for assembly [17] Zhang W, Zhang X, Tian G. Determining the hanger force of the suspension bridge
and reduction of component models for dynamic substructuring. AIAA J 2011;49 with short extended spans: an analytical algorithm. Structures 2023;51:1645–54.
(5):1010–20. [18] Zhao J, Sun H, Sun X, et al. Screw statics of spatial structures and mechanisms.
[13] Liu F, Jin D, Wen H. Equivalent dynamic model for hoop truss structure composed Structures 2023;55:411–26.
of planar repeating elements. AIAA J 2017;55(3):1058–63. [19] Zhao J, Sun H. Kinetostatics of a serial robot in screw form. Adv Mech Eng 2023;15
[14] Du X, Du J, Bao H, Sun G. Deployment analysis of deployable antennas considering (6):1–12.
cable net and truss flexibility. Aerosp Sci Technol 2018;82:557–65. [20] Turmo J, Luco J. Effect of hanger flexibility on dynamic response of suspension
[15] Lu G, Zhou J, Cai G, et al. Studies of thermal deformation and shape control of a bridges. J Eng Mech 2010;136(12):1444–59.
space planar phased array antenna. Aerosp Sci Technol 2019;93:105311. [21] Chai S, Wu Q, Wang X. Deformation and force characteristics of double-cable
[16] Nie R, He B, Hodges D, et al. Integrated form finding method for mesh reflector suspension bridges. Structures 2023;54:1705–16.
antennas considering the flexible truss and hinges. Aerosp Sci Technol 2019;84: [22] Park K, Kim D, Hwang E. Investigation of live load deflection limit for steel cable
926–37. stayed and suspension bridges. Int J Steel Struct 2018;18(4):1252–64.

17