Fig.
2(a) illustrates the reference area corresponding to the test specimens of an I-shaped beam with the grid-purlin in a gymnasium. The distance from the ridge to the eave was assumed to be 13 m. The specimens measured 6.5 m wide and 13 m long [shaded area in Fig.
2(a)]. A built-up I-shaped beam,
, was used for the beam section of each specimen.
and
were used for the purlin sections of Specimens P150 and P75, respectively. The tests were conducted in the Architecture and Building Research Institute (ABRI) in Taipei in 2018. As shown in Figs.
3 and
4, the specimen was anchored to the strong floor and bending moment was applied as illustrated in Fig.
2(b). The actuator at Point C in Fig.
3 applied vertical loads to the end of the specimens. The boundary conditions at the edge of the purlin and the pin-roller support were constructed using the connection details shown in Fig.
5. The grid-purlin was connected to the top flange of the I-shaped beam by welding steel plates, as shown in Fig.
6.
Tables
1–3 presents the dimensions and material properties along with the maximum lateral bracing spacing,
, for moderately and highly ductile beams, where the
is the yield strength,
is the tensile strength, and
is the fully plastic moment. According to ANSI/AISC 341-16 (
AISC 2016a), the bracing of ductile beam members shall have a maximum spacing of
for highly ductile members or
for moderately ductile members, where
is the spacing between points that are either braced against lateral displacement of the compression flange or braced against twist of the cross section,
is the radius of gyration about the minor axis,
is the elastic modulus,
is the ratio of the expected yield stress to the nominal yield stress, and
is the specified yield strength. Calculations of
used the average yield strength between the flange and web instead of
and the average elastic modulus between the flange and web. The fully plastic moment of the I-shaped section,
, was calculated from Eq. (
1)
where
,
,
, and
are the beam width and depth and web and flange thickness, respectively. The
and
are the yield strength of the flanges and the web, respectively (Table
2). Note that the I-shaped beam without grid-purlins did not satisfy the stability bracing requirements for either highly or moderately ductile members. This clearly indicates that the ductility of the I-shaped beam without grid-purlins was not expected. The Japanese Industrial Standard Steel Grade SS400 was used for the I-shaped flange beams, and STKR400 was used for all the purlins. The specimens were subjected to single curvature bending consistent with a loading protocol (Fig.
7). The loading was controlled by the rotations at Point B (Fig.
3). The amplitude was normalized by the fully plastic yield rotation of the I-shaped beam,
, which was computed from the ratio of the fully plastic moment of the I-shaped beam,
, to the flexural stiffness,
. The amplitude was increased from
to
,
,
,
, and
.