Nadia GREMER
A contribution to the assessment of the vertical seismic acceleration demand of regular steel frame structures
Despite the fact that in the last decades the knowledge of earthquake induced forces has increased, the assessment of the vertical acceleration demand is still an open issue. Commonly in earthquake engineering it is assumed that buildings under seismic impact behave rigidly in vertical direction. Some studies in the last few years, however, have indicated that the vertical acceleration response increases with the height of a seismically excited regular structure and should not be neglected.
The scope of this thesis is the quantification of vertical peak floor acceleration demands at column lines and along the length of beams of elastic and inelastic moment-resisting regular steel frames subjected to recorded and simulated ground motions. These demands correlate with the maximum strength demands on rigid nonstructural components attached to a frame structure. Nonstructural components have a major impact on the seismic risk. Since it is commonly assumed that buildings behave flexibly in the horizontal direction and rigidly in the vertical direction, the assessment of vertical acceleration demands is typically not considered. Therefore, in this dissertation time history analyses are conducted on regular steel frame structures that are excited with the horizontal and vertical component of ground motions simultaneously to assess the acceleration response in both directions.
The results of this dissertation show that vertical peak floor accelerations of regular steel frames can be up to five times larger than the vertical peak ground acceleration. In contrast, the horizontal peak floor acceleration predictions are only up to three times larger than the horizontal peak ground acceleration for the numerical models used in this study. The most significant amplifications estimated in the vertical direction are found at the center of the girders and the exterior column lines of the considered three bay frame structures. Further investigations on modified steel frames indicate that the story-wise mass distribution has an influence on the vertical acceleration demand. In contrast, the response in the vertical and horizontal direction is only slightly affected by an increase in the flexural stiffness of the beams.
Another scope of this doctoral thesis is the impact of different ground motion sets on the vertical acceleration demand. Thus, the time history analyses were conducted with four different ground motion sets, three recorded ground motion sets and one simulated ground motion set. The results show that all considered ground motion sets yield similar acceleration demands, i.e. also the simulated ground motion set can capture the seismic demands adequately for the investigated steel frame in horizontal as well as in vertical direction.
Finally, this dissertation is concerned with the modeling of different energy dissipation mechanisms for the numerical prediction of the vertical acceleration demand in the considered regular frame structures. One of these issues discussed is the consideration of viscous damping in the structural model. As is shown, well-established Rayleigh-damping may highly overestimate the damping of vertical modes, resulting in much too low vertical acceleration response predictions. The result of a model study provides an appropriate damping modeling strategy that leads to reasonable predictions of both horizontal and vertical frame acceleration demands. Another open question addressed is the effect of inelastic material behavior on the vertical acceleration demand on the considered regular structures. The results of a shell model of a frame structure exposed to high intensity ground motion excitation indicate that inelastic material behavior has virtually no impact on the vertical acceleration demand, while the structural inelasticity leaves the horizontal response significantly smaller compared to the elastic demand. This leads to the conclusion that common frame models, which represent the inelastic horizontal response, but behave purely elastic in vertical direction, are suitable for the computation of the acceleration response.