提高混炼胶分散性及预测精度的研究

使用自制的ＭＧＫＪ型密炼机混炼微机监控系统，采用日常生产混炼工艺过程的有关参数，剔除数据中的奇异点，根据泰勒开原理进行变量变换后，数学模型仍欠理想，只有进一步根据橡胶混炼机理，减去压砣提起时的能量和时间后，数学模型的复相关因数Ｒ才等于０．９５。有关检验均已通过，预测相对标准误差达到１０％以内，这一结果同时也证明，密炼机混炼的高功率条件下对炭黑附聚体的分散才最有效。

Study on Improving Mix Dispersion and Predictive Accuracy

A mathematic model for improving the mix dispersion and predictive accuracy was established by using the relative parameters in routine production mixing recorded with a self made computer monitoring system for MGKJ internal mixer,rejecting the abnormal points from the data and varying the variables according to the Tailor developing principle,but the model was not ideal enough and needed to be improved.The complex correlation coefficient R of the model was equal to 0 95 only when the energy and time for lifting the ram were reduced.The new model passed various examinations,and the predictive standard relative error was less than 10%.The results showed that the internal mixing was most effective for the dispersion of carbon black agglomerates only running with high power.