The actual design of seismic resistant steel buildings requires analysis on the most realistic computational models, due to the incident seismic load. Conventional methods of static and dynamic analysis of frame steel structures are based on the assumption of ideal nodal connections, i.e. joints are totally rigid or ideally pinned. However, on the basis of numerous tests, it has been concluded that the connections are semi-rigid and that the moment-rotation relationship at the end of the beam elements is non-linear in almost the entire load range generally for all types of connections. In addition to the established flexibility of the connections, there is also greater or less eccentricity of the connections, which can cause significant influence in the structure.
Within the latest requirements in the seismic calculation as is the control of structural behavior, the importance of designing and installing special structural elements has been determined. They have an important role in reducing or completely eliminating damage to both structural and non-structural elements of the structure. These elements generally play their role by significantly dissipating the energy delivered to the system during seismic loading, thus protecting other structural elements from damage. In order to obtain satisfactory calculation results, it is necessary to apply seismic material-geometric nonlinear analysis of steel frames with semi-rigid, eccentric and viscous beam-column connections.
Particular attention should also be paid to the effects of second-order theory, which can be crucial for flexible structures such as steel frame systems, so that both nonlinearities, geometric nonlinearity of structure and material nonlinearity of connections, which are considered simultaneously, has to be applied. The material nonlinearity of the problem appears only in the connections of the beam elements, while all other parts of the system can be considered in the domain of elastic behavior. This assumption makes the essential rationalization and efficiency of calculation, which is applicable in standard engineering calculations.