Civil Engineering Reference
In-Depth Information
0.004
0.004
0.003
0.003
0.002
0.002
0.001
0.001
0.000
0.000
0
5
10
15
0
5
10
15
-0.001
-0.001
-0.002
-0.002
-0.003
-0.003
-0.004
-0.004
Time (seconds)
Time (seconds)
Figure 4.21
Angle of torsion time history at all columns in the fi rst fl oor of the full-scale frame in Figure 4.2
(Montenegro 1979 - Herceg Novi, 0.2 g, bidirectional loading
members shows the same angle of torsion at every corner, thus satisfying the assumption of an in- plane
rigid fl oor.
Computer programs do not generally accommodate the modelling of rigid inclined decks, which are
typically used as roofs. Inclined diaphragms can be discretized in a number of different ways. The most
common models include equivalent one-dimensional elements, e.g. trusses or beams, and two-
dimensional elements, e.g. shells. The choice of model for inclined rigid diaphragms depends on the
degree of accuracy sought. One-dimensional elements are a compromise between accuracy and
economy, especially for inelastic dynamic analyses as those illustrated in Sections 4.6.1.2 and 4.6.1.3 .
Evaluation of local stresses requires the use of shell elements. The mesh should be refi ned gradually
towards openings and other regions where high stress gradients are expected.
4.5.3.4 Infi lls
Masonry and concrete block infi lls are frequently used for interior partitions and exterior walls in steel,
RC and composite frames. Infi lls can either be isolated or connected to the bounding frame. In the latter
case, the interaction between walls and frame should be taken into account. Infi lls can be discretized
using one of two approaches:
•
Micro -
or
refi ned models
;
•
Macro -
or
simplifi ed models
.
Micro-models generally utilize three types of FEs to discretize the frame, infi lls and the interaction
between the two. One-dimensional elements are used to model the frame while walls are idealized using
triangular and rectangular plate or shell elements. Interface elements are essential for the simulation of
cracking, which may take place between the frame and infi lls. They also account for the likelihood of
separation and describe friction conditions where contact remains. It is diffi cult to defi ne accurately the
boundary conditions at the interface between infi ll and frame, a region affected by materials, details
and construction method. For instance, in reinforced masonry walls, if the beams of the bounding frame
are supported by the masonry when cast
in situ
, the interaction is full. Conversely, when infi lls are built
after the frame and a separation gap is used between them, the interaction is negligible. Micro- models
can be employed to assess global and local response of frames with infi lls provided that the properties
of the interface are clearly defi ned. Macro-models are an attractive alternative to detailed models
because of computational simplicity and effi ciency. Diagonal struts with appropriate mechanical char-
acteristics may be used to simulate the presence of infi lls (e.g. Stafford-Smith, 1968; Mainstone, 1974 ;
