ON REDUCTION OF COMPUTATIONAL COST IN TRANSIENT ANALYSIS OF STRUCTURES AGAINST NEAR FIELD EARTHQUAKES

Abstract eng:
Time integration is a powerful tool in time history analysis against seismic excitations. Nevertheless, the computational costs are considerable and inaccuracies exist for the computed responses. In 2008, a technique is introduced for reducing the analysis computational cost, with negligible additional inaccuracy. The key idea is a convergence-based replacement of the digitized seismic excitation with excitations recorded at larger steps such that the new step size is larger than the original step size by a positive integer and the scale of the enlargement satisfies an accuracy-based inequality. Accordingly, the technique is applicable merely when a positive integer satisfies the inequality, and then the accuracy of the responses after implementation of the technique should be acceptable. Since these applicability and performance issues are not theoretically established in a rigorous way, they are studied numerically; the results of implementation of the technique in time integration analysis of many structural systems, e.g. 2D and 3D frames of regular and irregular buildings, bridges, silos, space structures, a power plant, a telecommunication tower, etc. against different far field earthquakes, and by different time integration methods, are successful (very recently, it is also displayed that the technique can be effectual in reducing the computational cost of static analyses). To extend this versatility and in view of the special features of near field earthquakes, in this paper, the applicability and performance are tested regarding transient analysis of a telecommunication tower against a near field record by three time integration methods, while once disregarding the P-Delta effect and once taking it into account. The observations were all satisfactory evidencing that the computational cost reduction technique proposed in 2008 might be successfully applicable to transient analysis against near field earthquake records.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket