毛竹材强度数据统计与回归分析模型

Data statistic and negressive analysis models of strength relationship of Moso bamboo

为得到毛竹材强度关系和预估强度值,对国内已有竹材强度试验进行了整理,分别采用非参数统计法和参数统计法分析了1173个试验样本数据,回归分析得到了毛竹材顺纹抗压强度与其它强度间关系的函数模型,通过测定顺纹抗压强度预估毛竹材的其他强度。结果表明:经K-S拟合优度检验,毛竹材强度服从偏态分布较好,与对数分布、Weibull分布的拟合优度较高;有竹节顺纹抗拉强度和顺纹抗剪强度在Weibull分布假定下与顺纹抗压强度拟合效果较好;横纹抗压强度、弦向抗弯强度、竹筒顺纹抗压强度、无竹节顺纹抗拉弹性模量以及弦向抗弯弹性模量在对数分布假定下与顺纹抗压强度拟合效果较好;尽管无竹节顺纹抗拉强度在参数统计法下与顺纹抗压强度未获得较好拟合效果,但在非参数统计法下回归模型仍具有较好的精度。

To obtain and predict the strength relationship and estimated strength values of Moso bamboo,the existing bamboo strength tests in China were collated.A total of 1173 test sample data were analyzed by nonparametric statistics and parametric statistics,function models of the relationship between the compressive strength parallel to grain and other strengths were proposed by regression analysis,and the method of predicted strength of Moso bamboo based on compressive strength parallel to grain was proposed.The results show that the strength of Moso bamboo obeys the skewness distribution well by the K-S test,and the goodness-of-fit with the logarithmic and Weibull distribution is good.The tensile strength with nodes and shear strength obey the assumption of Weibull distribution and fit well with the compressive strength of Moso bamboo parallel to grain.The compressive strength perpendicular to grain,tangential bending strength,compressive strength parallel to grain of bamboo tube,tensile modulus of elasticity without nodes,and tangential bending modulus of elasticity fit well with the compressive strength parallel to grain under the assumption of logarithmic distribution.Although the tensile strength without bamboo nodes do not obtain a good fit with the compressive strength parallel to grain under the parametric statistics,the regression model still has good accuracy under the nonparametric statistics.