Seismic Design of Elevated Silos and Tanks – a Study on Behaviour Factors

Abstract

According to EN 1998‐4:2006 3.4 (4), elevated silos may be designed considering medium or high ductility class for the support structure, abbreviated DCM and DCH respectively, for which an overall behaviour factor q greater than 1,5 may be assumed. Undoubtedly, when a silo or tank supported by a sway frame or a vessel with integrated legs is considered, it is expected that plasticity will be induced due to the overall system's performance and thus, motion amplitudes will be reduced due to energy dissipation. However, the shell upper‐structure is a very thin‐walled class 4 cross section. Hence, it might fail due to buckling at much smaller motion amplitudes (corresponding to acceleration amplitudes) than those, which the support structure can follow without failing. Therefore, from the authors' point of view it is not sufficient to settle with a single behaviour factor for the overall structural system. Alternatively, a two‐step approach is recommended: With this two‐step approach it can be verified, that (sudden) buckling of the shell does not happen prior to the introduction of plasticity at the sub‐structure and in extension to the desired energy dissipation. The proposed two‐step approach is demonstrated by suitable examples with a different range of parameters. Therewith, it will be shown that the assumption of q = 1,5 for elevated silos or tanks can be unsafe. An improvement of the current EC8‐4 rules is proposed in order to enable a practical, safer and more efficient design procedure.

Seismic Design of Non‐Structural Components in Industrial Facilities – a Discussion of the EC8 Approach

“…7.2 is used, having a diameter of 2,52 m, an oscillating mass of 32,9 t with CoG 3,97 m above BE of feet, horizontal drift stiffness 80 kN/cm and a natural frequency of 2,48 Hz. This is close to the eigenfrequencies of the example structure (note that in [12] also an eigenfrequency of 2,83 Hz is given as numerical result, but 2,48 Hz seems to be inconsistent with the given input values).…”

“…7.4 of the present study). Elevated tanks on frame structures or on tube legs depending on the presence of additional cross bracing between the legs they may have a small or large natural frequency (examples are given in [12]; see sect.…”

“…A check with the Rayleigh quotient yields 2,88 Hz, which is considered sufficiently accurate. As an example for a NSC, tank B from [12] sect. 7.2 is used, having a diameter of 2,52 m, an oscillating mass of 32,9 t with CoG 3,97 m above BE of feet, horizontal drift stiffness 80 kN/cm and a natural frequency of 2,48 Hz.…”

“…Dr. Britta Holtschoppen ([1], [2], [11], [12]) was involved as co-author from early phases of this study. Unfortunately, due to time restrictions, she had to withdraw her co-authorship.…”