実験値に基く条件付き最小値の推定法に関する基礎的研究
スポンサーリンク
概要
- 論文の詳細を見る
The estimating method of the possible lower limit, that is, the irreducible minimum of the strength of material is proposed herein on the basis of scores of data given by the strength tests. In general, for these data, a specified distribution is assumed based on the test of hypothesis of the distribution. The normal distribution is mostly choosen for the specified distribution, and in this case, thrice of the standard deviations becomes the lower limit of the strength. However, the assumption of the distribution may not be desirable, since the test of hypothesis of the distribution does secure nothing. Therefore, the distribution for the population of the strength is determined based on the problem of variation so as to minimize the lower limit and so as to satisfy several conditions extracted adequately from the test data. These conditions are mean, variance, minimum, maximum, mean of the lower S data, mean of the upper T data, etc.. Calculations of the distribution are made for various combinations of these condition, and numbers of conditions increase from only two to eight. Between two and eight conditions, seven conditions are selected for practical use. This technique is applied to the data given by the test of the steel buckling strength, steel tensile strength and cement-mortar bending strength. And, the propriety of this technique is interpreted in detail. Further, a simple estimating method which needs no calcluation is proposed in order to increase its applicability.
- 社団法人日本建築学会の論文
- 1975-07-30
著者
関連論文
- 鋼製煙突の地震強度に対する一検討(続報)
- 鋼製煙突の地震強度に対する一検討
- 極東アジアの木造屋根構造における「トラス度」の数値的推定
- 極東アジアの木造屋根構造における「トラス度」の数値的推定
- 鋼橋の人物史談(3)
- 鋼橋の人物史談(2)
- 鋼橋の人物史談(1)
- わが国の戦後の鋼橋の歴史
- 構造安全性解析への極値的なアプロ-チ
- 測定値の誤差を考慮した信頼性解析