• Back
• Next

日本語

A “Shell and Spatial Structure” has a curved surface structure like a shell (dome, cylinder, etc.), and can cover a huge column-free space with thin and narrow parts (Fig. 1). Therefore, shell and spatial structures are widely used in industrial structures such as tanks, school gymnasiums, sports arenas, and others. This type of structure has a relatively high public interest and requires a high earthquake resistance. In particular, it is expected that school gymnasiums are to be used as evacuation facilities after earthquakes, and earthquake resistance performance evaluation for nonstructural materials, including ceiling materials, is gaining attention.

Our laboratory is conducting research on buckling design and earthquake-resistant designs for shell and spatial structures. Shell and spatial structures often use thin and narrow parts. Also, if compression is applied to the entire structure like a dome, a buckling phenomenon will occur, which means that an elongated part rapidly bends out in a lateral direction when compression is applied from both ends. Furthermore, when designing shell and spatial structures, it is necessary to take account of both the total buckling (shell buckling), in which the entire structure buckles, and individual buckling (Euler buckling of the columns), in which the parts buckle. We propose new analytical and design methods that take full account of these buckling phenomena.

For earthquake-resistant designs of shell and spatial structures, we are developing analysis software and conducting research on analysis methods to consider the questions, “How do buildings shake when an earthquake is occurring?”, and “At what intensity of shaking will buildings collapse?” We also are proposing a shell and spatial structure with base isolation devices and vibration dampers, in order to realize spatial structures that are safe, even when larger earthquakes occur.

In addition, at our laboratory, we are developing a seismic performance evaluation method based on earthquake risk analyses. In conventional earthquake-resistant designs, a design method that takes into consideration the yield strength and the deformability of the main structural members (columns, beams and earthquake-resistant walls) are adopted. Nowadays, the need for an evaluation of earthquake resistance performance that combines the damage of non-structural materials, including ceiling materials, and the function maintenance performance of buildings is increasing. The Earthquake Risk Analysis is a method of stochastically obtaining earthquake hazard (Seismic hazard curve) according to the construction position of the buildings and the amount of damage inflicted by the earthquake (Seismic loss function) from building strength and deformability. This method allows us to evaluate in consideration of economic losses due to the damage of nonstructural materials, in addition to damages to major constructional materials and losses due to the cessation of the building functionality.

By adopting this method: 1) it is possible to design buildings in such a way as to take account of the expenditure that the building will require during its service period (LCC, Life Cycle Cost. 2) It is possible to quantitatively measure the reduction in costs which results from installing base isolation devices and vibration dampers and to judge the effectiveness of seismic strengthening. 3) It can be used for analyzing earthquake function maintenance performances in school gymnasiums, factories, medical facilities, etc.

Furthermore, a faster computer is required to perform high-precision numerical analyses and earthquake risk analyses of large-scale structures such as shell and spatial structures. For example, in the seismic risk analysis, a large number of calculations (so-called Monte Carlo simulation) have to be performed in order to consider earthquake inputs and variations in the yield strength in structures. Therefore, we position parallel grid computing (Fig. 3), an HPC (High-Performance Computing) method, as the basic technology for our numerical analysis method and are also conducting research on simple HPC construction methods and how to utilize them in the field of construction civil engineering. Specifically, we are proposing optimization methods using a grid system and a genetic algorithm, and large-scale structural analysis methods using a grid parallel computing system.