Analysis of the strength data of wood structures on limit states design III. Reliability index and design point on bending strength of structural lumber

A relationship between reliability indexes (β) and design points was calculated on limit states of structural bending members, second floor beam, third roof rafter and roof purlin. Effect on β of methods, by which parameters of empirical distributions were estimated, was investigated. Estimation methods were method of maximum likelihood, method of moments, least squares by linear model, least squares by non-linear model and Foschi's method. Empirical distributions were Normal, Log-normal, 2P Weibull and 3P Weibull distribution. And also design point of each member was compared to material strength and 5th percentile value with 75% confidence level (R_0.05) determined by a non-parametric method. The results are as follows: (1) The empirical distributions rather than estimation methods affected design points. (2) Values of design points calculated from four empirical distributions were almost the same when β is less than about 1.0 in case of 100% data fitting and when β is less than about 2.5 in case of lower 15% data fitting. (3) Design points of strength on the ultimate limit state were under R_0.05 for species whose strength have a large variation. On the other hand they were about R_0.05 for species whose strength have a small variation. (4) Design points of Young's modulus on serviceability limit state were a little under mean for all species except Sugi and were almost the same as recommended values by Architectural Institute of Japan.