Braced frame structures are used for low-rise steel buildings with large space such as school gymnasium, and angle braces are often adopted as the main seismic component. These buildings are used as emergency public shelters in event of disaster. Therefore, it is very important to secure the sufficient seismic performance. However, a reconnaissance of the recent earthquake reports that the brittle fracture at the effective cross-section of an angle brace. Single-angle brace has the inevitable eccentricity between the brace and its gusset plate, and this detail is a disadvantage in terms of connection strength. Therefore, in the case of double-angle brace, it is known that the removal of the out-of-plane eccentricity leads to the improvement of the ultimate strength at the effective cross-section. In addition, the previous paper reports that Z-shaped double-angle brace removing both out-of-plane and in-plane eccentricities brings further increase in the ultimate strength at the effective cross-section. On the other hand, the recent studies on double-angle brace point out the decrease in cyclic deformation capacity. A purpose of the present study is to clarify the effect of connection detail on not only strength at the connection but also cyclic deformation capacity of angle brace.
In the present paper, both connection tests and component tests were carried out. Specimens of the tests are angle braces connected to gusset plates using high strength bolts. In the connection test, the monotonic tensile loading was employed in order to investigate the effect of connection detail on the ultimate strength. The main parameters were angle cross-section (L65x6, L75x6 or L90x7), thickness of gusset plate (PL9, PL16 or PL19), connection detail (single, double-angle or Z-shaped) and the number of bolts (2 to 5). On the other hand, in the component test, the cyclic loading was employed to verify the influence of connection detail on cyclic deformation capacity of the angle braces. The test parameters were angle cross-section, thickness of gusset plate (PL9, PL12 or PL16), connection detail and loading protocol (incremental or constant amplitude).
The result of the connection tests indicates that the effective yielding leg ratio of single-angle brace is independent on the number of bolts, which are approximately constant at about 0.5. However, those of both double-angle and Z-shaped double-angle brace increase in proportion to the number of bolts, whose values are roughly from 0.2 to 1.1. Therefore, it is found that the out-of-plane eccentricity has a great influence on the yield strength at the connection of angle brace. Meanwhile, the ultimate strength at the effective cross-section tends to increase in proportion to the number of bolts. However, the effective leg ratio of the cross-section of both double-angle and Z-shaped double-angle brace is 1.2 times to that of single-angle brace. It means that the removal of out-of-plane and/or in-plane eccentricity has a relatively small effect on the ultimate strength at the effective cross-section. On the other hand, the results of component tests indicate that the cyclic deformation capacity of both double-angle and Z-shaped double-angle brace is much lower than that of the single-angle brace. In the case of L65x6 cross-section, a cumulative plastic ductility ratio of both double-angle and Z-shaped double-angle brace is decrease to roughly 5% to 20% of the single-angle brace. Conclusively, although the removal of the out-of-plane eccentricity brings the increase in the connection strength of angle brace, it causes the decrease of the cyclic deformation capacity as the whole member.
