多元线性回归优化聚偏氟乙烯/乙酸纤维素共混微滤膜成膜因素

Optimization of preparing PVDF/CA blend microfiltration membrane by linear multi-regression

首先分别以纯水通量、平均泡点压力减最大泡点压力、纯水通量加最大泡点压力减平均泡点压力作为考察PVDF/CA共混微滤膜的性能指标,设计了9因素4水平的正交实验表L32(49)并进行实验,其次采用标准化系数矩阵的多元线性回归模型对实验数据进行处理。结果表明:当以纯水通量为目标函数时,共混比、固含量、添加剂含量、凝胶浴温度是影响因子;当以平均泡点压力减最大泡点压力为目标函数时,固含量是影响因子;当以纯水通量加最大泡点压力减平均泡点压力为目标函数时,固含量是影响因子。结合环境扫描电镜照片对实验结果的综合分析后确定,当共混比、固含量、添加剂含量和凝胶浴温度分别为4∶1、12%、2%～3%和30～35℃时,即可制得性能较好的PVDF/CA微滤膜。

Firstly, the pure water flux, average pressure of bubble point minus maximum pressure of bubble point, the pure water flux plus average pressure of bubble point minus maximal pressure of bubble point were selected as the performance indices of poly（vinylidene fluoride） （PVDF） /cellulose acetate （CA） blend microfiltration membrane. An orthogonal table with nine factors and four levels L32 （4^9） for the PVDF/CA system was designed to study the performance of PVDF/CA blend microfiltration membrane. Secondly, the linear multi-regression model with standardized coefficients matrix was used to study the effects of nine factors. The results showed that： when pure water flux was taken as the objective function, blend ratio of PVDF/CA, solid content, additive content and gelation temperature were the major influence factors; when average pressure of bubble point minus maximum pressure of bubble point was taken as the objective function, solid content was the major influence factor; when the pure water flux plus average pressure of bubble point minus maximal pressure of bubble point was taken as the objective function, solid content was the major influence factor. After analyzing the SEM photographs and experimental results, the optimum condition for preparing the PVDF/CA blend microfiltration membrane was as follows： blend ratio of PVDF/CA 4 ： 1, solid content 12%, additive content 20%-30%, gelation temperature 30-35℃.