Development of Group Contribution Method for Predicting Liquid Viscosities Applicable to High Pressure from Low Pressure

作为预测常压下运动粘度的基本方程，用绝对反应速度理论ln(νM)=Σx_iln(ν_Im_i)-ΔG^E/RT。已经确定了与CH2、Arch、CyCH、OH、H2O、CO、CoO、CCl3和CCl4相关的基对参数。预测了含有烷烃、芳香烃、醇、水、酮、酯和氯化物的体系的运动粘度。对于二元体系和三元体系，实验值与预测值的绝对算术平均偏差分别为4.15%和5.03%。为了提出利用基团贡献预测纯组分粘度的方法，使用以下Andrade方程：lnη=B(1/T-1/T_0)，用神经网络方法预测参数B和T_0。对烷烃、烯烃、烷烃、芳香烃、氯化物、溴化醇、酮、酯、醚、醛和有机酸的粘度进行了预测，提出了用低压粘度预测高压粘度的方法。基本方程是Peng-Robinson状态方程和超额自由能混合规则。对醇+水体系的高压粘度进行了预测，平均偏差为2%。

As the fundamental equation for predicting kinematic viscosities at normal pressure, the following absolute reaction velocity theory has used.ln(νM)=Σx_iln(ν_iM_i)-ΔG^E/RTThe ASOG-VSICO was proposed in order to obtain the excess Gibbs free energy using ASOG group contribution method. The group pair parameters have been determined for the group pairs relating to CH2, ArCH, CyCH, OH, H2O, CO, COO, CCl3, and CCl4. The kinematic viscosities were predicted for the systems containing paraffines, aromatic hydrocarbons, alcohols, water, ketones, esters, and chlorides. The absolute arithmetic average deviations between experimental and predicted values are 4.15 and 5.03% for binary and ternary systems, respectively.To propose the predictive method of pure component viscosities using group contribution, the following Andrade uquation was used.lnη=B(1/T-1/T_0)The parameters B and T_0 were predicted using neural network method. The viscosities can be predicted for the paraffines, olefins, alkyns, aromatic hydrocarbons, chlorides, bromides alcohols, ketones, esters, ethers, aldehydes, and organic acids.The predictive method to predict high- pressure viscosities using low-pressure viscosities was proposed. Fundamental equations are Peng-Robinson equation of state and excess free energy mixing rule. The high-pressure viscosities for alcohols + water systems were predicted with the average deviations of 2%.