非正規確率変数の同時確率密度関数を用いた基準強度の算定

概要

論文の詳細を見る
Joint probability density function (JDF) for non-normal random variables was derived by bivariate normal distribution function. From this JDF, the referenced strength, the 5th percentile of the strength distribution with 75% confidence level, for each grade of stress graded lumbers was calculated, assuming modulus of elasticity (MOE) as Normal distribution and modulus of rapture (MOR) as 2P Weibull distribution. And the referenced strength was also calculated by Monte Carlo simulation using linear regression analysis by weighted residue (LRA). These values were compared. The results are as follows. (1) The difference between these values depends on species. For example, in Douglas-fir and Dahurian larch they are relatively similar because the standard error by LRA of these species is relatively large. But for some species having small standard error, the difference is relatively large. (2) In lower grades than mean MOE, the referenced strength by LRA method is larger than that by JDF method, but in higher grades the situation is reverse. The reason is that the standard error is proportional to √<MOE> and the standard deviation of MOR in lower grades is small and that in higher grades is large. JDF method is really true from the probabilistic and statistical point. So it is hoped to recalculate load and resistance factors for stress graded lumbers in the Recommendation for Limit State Design of Timber Structures(draft) by JDF method.
社団法人日本建築学会の論文
2007-05-30

非正規確率変数の同時確率密度関数を用いた基準強度の算定

スポンサーリンク

概要

著者

関連論文

スポンサーリンク