Cable-stayed bridges - Computational Analysis and Design of Bridge Structures - page 360

pk

e4

gk

pj gj

gi

e2

e1

e3 e5

pi

Figure 11.28 Vertical supports in 2D model.

reaction between these two nodes cannot be obtained directly. If the bearing
reactions are of interest, a similar connection truss element has to be built
in a 2D model. As shown in Figure 11.28, a separated node gj is added.
Girder elements e1 and e2 connect nodes gi and gj and gj and gk, respec-
tively. Pylon elements e3 and e4 connect nodes pi and pj and pj and pk,
respectively. Truss rigid element e5, which simulates the bearings, con-
nects nodes pi and gj. The bearing reactions can be obtained from the
internal force of element e5.
For cable-stayed bridges in which longitudinal semifloating systems
or damping systems are used as shown in Figure 11.21, connections
between girder and pylons should be simulated carefully when the dis-
placement is beyond its allowed movements. In general, a horizontal
truss element can be added between the girder and pylon. Its stiffness
has to be modified to a very small value when the displacement is within
its limit and to the correct stiffness of the links when the displacement
is beyond its limit.

11.4.4 Cables
In most analysis scenarios, each cable can be simply modeled as one truss
element by its two anchor points. When sag effects have to be considered,
the equivalent Young's modulus has to be calculated according to Equation
11.10 either manually or automatically by the analysis package. As men-
tioned in Section 11.2.7, the equivalent Young's modulus changes when
the erection phase changes; a special FEA package with the capability to
handle this issue is always preferable.