Jun Kanda

Engineering role in failure cost evaluation for buildings

Cost benefit analysis is a common tool for decision making on safety. The appropriate degree of safety in structural design can be discussed based on a simple formula for optimum reliability including parameters such as coefficient of variation of lifetime maximum load, cost-up constant and normalized failure cost. In many cases, the degree of safety is determined as specified in codes or regulations. However, when additional information is available, such as owners demands, engineers are responsible for providing quantitative values for various parameters, in particular for failure cost. This paper examines sources of failure costs and relations between the engineers role in structural safety and discusses failure costs.

Proposed formulae for the power spectral densities of fluctuating lift and torque on rectangular 3-D cylinders

New empirical formulae are presented for the spectra of lift and torque for square and rectangular section buildings with different side ratios and their validity and applicability are discussed comparing with experimental data in three different boundary layers. As a results, it is shown that the new empirical formulae may provide more accurate input forcing functions for preliminary design of isolated tall buildings with various geometrical conditions than existing formulae.

Correlating dynamic wind force components on 3-D cylinders

Abstract Dynamic wind force components on 3-D cylinders with typical sectional shapes, i.e. square, rectangle, triangle and diamond, were measured in three simulated boundary layers in a wind tunnel. Force coefficients, power spectral densities and root-coherence results are presented and their representative characteristics are summarized with respect to the sectional shape. Computed responses are also presented to examine the significance of considering the torsional component and the coherence between components.

Reliability of gust response prediction considering height dependent turbulence parameters

Abstract Parametric studies on the alongwind response of tall buildings in typical terrains have been performed, taking the height dependence of the turbulence characteristics into account. The reliability of the gust response prediction is examined in terms of the coefficient of variation of the gust response factors allowing reasonable variations for parameters. The relative contribution of these parameters especially height dependence parameters to the response is investigated and the required accuracy for turbulence parameters is discussed. Detailed assessment of gust response prediction may not be necessary for buildings with the height less than 100 m in Urban terrain from the reliability point of view. On the other hand the height dependence parameters could be significant contributors for the gust response prediction of super-highrise buildings.

Probability-based seismic safety evaluation of existing buildings

A probabilistic second-moment seismic safety measure is proposed and the seismic safety of eleven existing buildings is evaluated. These buildings are analyzed using lumped-mass models and the seismic hazard at each building site is estimated by utilizing an available probability-based seismic hazard map. Results for the distribution of the safety measure with story level suggest that significant damage may be concentrated only in a few stories in some buildings. Second-moment seismic margin indices for some buildings are relatively high, while lower seismic margins result for others, perhaps due to underestimation of the ultimate limit state resistance. The feasibility of the procedure for evaluating seismic safety of existing buildings is demonstrated.

Full-scale measurements of wind actions on Chiba Port Tower

Chiba Port Tower is 125 m high with a rhombus-shaped plan. On the top of the tower, a ‘Tuned Mass Damper’ (TMD, hereafter) was installed in order to reduce the vibration caused by strong winds and earthquakes. A lot of data during strong winds has been recorded since August 1987. From this data, it was found that the TMD works effectively when the response of the tower exceeds 0.5 cm/sec2 and that the TMD can reduce the wind-induced response of the tower by nearly 50 percent for all wind directions.

Distribution of seismic hazard in Japan based on an empirical extreme value distribution

Abstract A probabilistic model of seismic hazard intensity in Japan is proposed which is based on an empirical extreme value distribution with lower and upper bound limits. There have been quite a few seismic hazard studies in recent years, but most of them discuss only the expectancy in certain recurrence intervals but not the coefficient of variation nor the probability distribution. This paper utilizes a fairly long-term Japanese earthquake catalogue to determine realistic upper bound limits for the extreme value distribution and uses a recent 300-year catalogue to establish the seismic hazard probability distribution which is presented in terms of contour maps for the mean and c.o.v. of 50-year maximum peak velocity at bedrock.

Recent developments in full-scale wind pressure measurements in Japan

Abstract Since 1968 quite a few full-scale wind pressure measurements have been conducted in Japan. Sixteen wind pressure measurement projects are examined and their scope and published results are summarised. The number of projects increased in the early 1970s, then suddenly dropped to zero for some years. The variation in the number of presentations at A.I.J. (Architectural Institute of Japan) annual meetings has also followed a similar trend, but with some time lag. In the early stage of projects wind pressure coefficients and the qualitative tendency of the power spectral density distribution where discussed, then the probability density distribution, r.m.s. wind pressure coefficients and coherence characteristics were reported. The most recent two projects, which have started after a six years pause, focused on the quantitative findings on both power spectral density distribution and coherence characteristics, with some explanation of the possible behavior of turbulent flow around tall buildings. In spite of the amount of observed data, it is still difficult to discuss the accuracy of the wind pressure coefficient. As a result most published papers concluded the importance and the need for further full-scale measurements with a focus on the internal pressure estimation and for comprehensive wind speed profile data on individual site conditions.

Flow simulation by wind tunnel with computer-controlled multiple fans

Abstract This paper presents the simulation of turbulent flow by controlling multiple fans in a two-dimensional wind tunnel. The main purposes of this study are to generate a large turbulence scale and simulate a turbulent flow fitted to specified power spectrum and co-coherence in a low frequency range, both of which are not easily accomplished only by the means of roughness blocks, spires and grids. It was relatively easy to obtain a large turbulence scale by controlling multiple fans numerically. According to the proposed procedure, turbulent characteristics of wind flow were successfully simulated to satisfy a specified condition for the target power spectrum and co-coherence.

Comparison of correlated Gumbel probability models for directional maximum wind speeds

Abstract In this paper we present a comparison of two probabilistic models for directional annual maximum wind speeds to clarify the characteristics of directional maximum wind speeds. First, two models for directional annual maximum wind speeds are summarized. These models are applied to recorded 10-min average wind speeds at several meteorological stations, and the consistency of models and effects of correlation coefficients between directional annual maximum wind speeds on the relation between the CDF for direction-free annual maximum wind speed and those for individual directional wind speeds are presented. The effects of correlation coefficients are clearly seen at some stations, and the usefulness of the proposed models is demonstrated for these sites.