F. Danesh
1 Introduction
The results obtained from previous investigations clearly indicated that the semi-rigid-connection is feasible and indeed more economical than the rigidly-connected frame, and also show that the optimum seismic design solution is not necessarily a rigidly connected frame.
Experimental work done by Elnashai and Elghazouli (1992) indicated that semi-rigid frames do exhibit a ductile and stable hysteretic behaviour. Although the stiffness and capacity of semi-rigid frames are both lower than similar rigid frames under monotonic and cyclic loading, the response under earthquake loading largely depends on the dynamic characteristics of both the frames and the input motion. In many cases, the response of semi-rigidly connected frames may be superior to rigid frames, provided that stable hysteretic behaviour is ensured.
Kuwamura and Suzuki (1992) indicated that any welded joints, whatever excellent workmanship is employed, eventually suffer fatigue fracture. Many modern steel buildings that suffered from local failures during the 1994 Northridge earthquake, illustrated that the dominant local failure occurs in welded beam-to-column connections (Bertero et al., 1994).
Today, the seismic behaviour of semi-rigid steel frames is a new field in research and represent that this type of connection has some benefits against rigid connection. But there is not enough information to predict the real behaviour of semi-rigid connections and few tests have been done in this area.
The main objective of this experimental and analytical work is to study the performance of semi-rigid steel frames under cyclic and dynamic loads. For this purpose, five two-storey steel frames, one rigid and four semi-rigid steel frames were tested. To compare experimental results with analysis and to investigate the dynamic characteristics of the frames and ground motions, many frames with different configuration are subjected to different records. The results of experimental and analytical investigation are presented and discussed in the following sections.
2 EXPERIMENTAL INVESTIGATIONS
2.1 Testing procedure
Five two-storey steel frames have been tested under cyclic and dynamic loads. These comprise one rigid and four semi-rigid structures, the later with different connection strengths. The pseudo-dynamic testing method was used for the dynamic test.
To facilitate testing, the frame was placed horizontally. The different parts of the assembly were adequately prestressed to form a self-reacting frame. An actuator which imposes displacements was connected to the end of the beam at the second floor level whilst another actuator operated in force control was fixed to the end of the beam at the first floor level. A third actuator with two hinges to allow for in-plane movement was installed at the top of the column to apply an axial load. In the rigid steel frame, the flanges and the web of the beam were fully welded to the flange of the column. Top and seat angles with double web angles were used for the connections in the semi-rigid frames.
To reduce the risk of shear failure of the panel zone, two stiffeners were welded to the flange and the web of the column. In all the models a constant axial load was applied at the top of the column by means of hydraulic jack
The details of the connections for each semi-rigid model are given in Table 1. The connections in these models were designed as partial strength semi-rigid connection with a moment capacity which varied between 30% and 70% of the plastic moment of the connected beams.