You are on page 1of 2

e_ect.

On the other hand, it is found that the displacement corresponding to maximum


strength
of each column is not less than 5_y [see Figure 7(a)]. This implies that it is able to design
the
steel box columns with high ductility if we use parameters of Rf =0:35; =_ =3:0; _=0:5,
and
keep __6 0:5.
Copyright ? 2000 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2000; 29:1707{1722
STIFFENED STEEL BOX COLUMNS. PART 2 1715

Figure 10. E_ect of material type of sti_ener.

3.3. E_ect of sti_ener's equivalent slenderness ratio (__


s)
As pointed out in the companion paper, __
s is usually used to measure the ductility of sti_ened
plates. This parameter is a comprehensive index and dependent on the structural
parameters such
as Rf , and _. For a given value of Rf , __
s will be de_nite as long as and _ are determined. In
other words, to meet an expected value of __
s, _ is a key parameter if Rf and are known, or
becomes important if Rf and _ are given. In this section, the e_ect of __
s on the ultimate strength
and ductility of sti_ened box columns is studied.
Figure 8(a) shows a comparison of envelope curves of four columns with di_erent values
of __
s. In these columns, the width (bs) and thickness (ts) are kept constant but the aspect ratio (_) is taken to be 0.5, 0.63, 0.75, and 1.0. As a result, the relative exural rigidity of the
longitudinal sti_ener () is 3.0, 1.88, 1.29, and 0.72 for corresponding columns. __
s obtained is
0.21, 0.26, 0.31, and 0.41. The comparison indicates that the post-buckling behaviour is
improved
as __
s is decreased, although there is no obvious di_erence in the behaviour before the peak.
This
fact implies that for existing columns, adding the diaphragm or transversal sti_eners is
e_ective
for improving the post-buckling behaviour. However, such a retro_tting method is not
attractive if
only the behaviour before the maximum strength is under consideration.
Figures 8(b) and 8(c) show the e_ects of the longitudinal sti_eners on the lateral load and
lateral displacement behaviour. In the case of Rf =0:35; __
=0:35 and _=0:5 [see Figure 8(b)], the
displacement corresponding to the maximum strength is 4.0 in a column with =_ =1:0 (__
s =0:33),
while 5.0 in another column with =_ =3:0 (__
s =0:21). The post-buckling behaviour is greatly
improved when =_ is increased from 1.0 to 3.0. As viewed from __
s, the smaller the value of
__
s, the better the deformation capacity is. As shown in Figure 8(c), this tendency can also
be
observed in the case of Rf =0:35; __=0:35 and _=1:0. Therefore, it can be concluded that
the
seismic performance can be largely improved through reinforcing the longitudinal
sti_eners.
Figures 8(d) and 8(e) illustrate the e_ect of sti_ener's equivalent slenderness ratio
corresponding
to the cases of =_ =1:0 and 3:0, respectively. Aspect ratio of the columns varies from
0.50 to

1.50. It should be noted that the columns are designed by adjusting the value of bs to
meet
=_ =1:0 in Figure 8(d), and =_ =3:0 in Figure 8(e). It can be seen that the maximum
strength
and ductility capacity are greatly improved as __
s increases in the case of =_ =1:0. The same
tendency is also observed in the case of =_ =3:0. This phenomenon does not fully
conform to a
prior _nding, which suggests that the ductility of a compressed plate is inversely
proportional to __
s.
As discussed in a study of compressed sti_ened plates, the de_nition of __
s cannot properly reect
the variation of ductility of sti_ened plates in some cases [9]. Therefore, a modi_ed
sti_ener's

You might also like