ABSTRACT
In line with current seismic design practice, steel structures may be designed to EC8 according to either non-dissipative or dissipative behaviour. The former, through which the structure is dimensioned to respond largely in the elastic range, is normally limited to areas of low seismicity or to structures of special use and importance; it may also be feasible if vibration reduction devices are incorporated. Otherwise, codes aim to achieve economical design by employing dissipative behaviour in which considerable inelastic deformations can be accommodated under significant seismic events. In the case of irregular or complex structures, detailed non-linear dynamic analysis may be necessary. However, dissipative design of regular structures is usually performed by assigning a structural behaviour factor (i.e. force reduction or modification factor) that is used to reduce the code-specified forces resulting from idealised elastic response spectra. This is carried out in conjunction with the capacity design concept, which requires an appropriate determination of the capacity of the structure based on a predefined plastic mechanism, often referred to as failure mode, coupled with the provision of sufficient ductility in plastic zones and adequate overstrength factors for other regions.This chapter focuses on the dissipative seismic design of steel frame structures according to the provisions of EN 1998-1 (2004), particularly Section 6 (Specific Rules for Steel Buildings). After giving an outline of common configurations and the associated behaviour factors, the seismic performance of the three main types of steel frame is discussed. Brief notes on material requirements and control of design and construction are also included. The chapter concludes with illustrative examples for the use of EC8 in the preliminary design of lateral resisting frames for the eight-storey building dealt with in previous chapters of this book. 6.2 Structural types and behaviour factors
There are essentially three main structural steel frame systems used to resist horizontal seismic actions, namely moment resisting, concentrically braced
and eccentrically braced frames. Other systems such as hybrid and dual configurations can be used and are referred to in EC8, but are not dealt with in detail herein. It should also be noted that other configurations such as those incorporating buckling restrained braces or special plate shear walls, which are covered in the most recent North American Provisions (AISC, 2005), are not directly addressed in the current version of EC8.As noted before, unless the complexity or importance of a structure dictates the use of non-linear dynamic analysis, regular structures are designed using the procedures of capacity design and specified behaviour factors. These factors (also referred to as force reduction factors) are recommended by codes of practice based on background research involving extensive analytical and experimental investigations. Before discussing the behaviour of each type of frame, it is useful to start by indicating the structural classification and reference behaviour factors (q) stipulated in EC8 as this provides a general idea about the ductility and energy dissipation capability of various configurations. Table 6.1 shows the main structural types together with the associated dissipative zones according to the provisions and classification of EC8 (described in Section 6.3 of EN 1998-1). The upper values of q allowed for each system, provided that regularity criteria are met, are also shown in Table 6.1. The ability of the structure to dissipate energy is quantified by the behaviour factor; the higher the behaviour factor, the higher is the expected energy dissipation as well as the ductility demand on critical zones.The multiplier a
u /a1 depends on the failure/first plasticity resistance ratio of the structure. A reasonable estimate of this value may be determined from conventional non-linear ‘pushover’ analysis, but should not exceed 1.6. In the absence of detailed calculations, the approximate values of this multiplier given in Table 6.1 may be used. If the building is irregular in elevation, the listed values should be reduced by 20 per cent.The values of the structural behaviour factor given in the code should be considered as an upper bound even if in some cases non-linear dynamic analysis indicates higher q factors. For regular structures in areas of low seismicity having standard structural systems with sections of standard sizes, a behaviour factor of 1.5-2.0 may be adopted (except for K-bracing) by satisfying only the resistance requirements of EN 1993-1 (2005, EC3).Although a direct code comparison between codes can only be reliable if it involves the full design procedure, the reference q factors in EC8 appear to be generally lower than R values in US provisions (ASCE/SEI, 2005) for similar frame configurations. It is also important to note that the same force-based behaviour factors (q) are proposed as displacement amplification factors (qd). This is not the case in US provisions where specific seismic drift amplification factors (Cd) are suggested; these values are generally lower than the corresponding R factors for all frame types.
