塑料大气环境老化预测模型研究

目的为塑料老化行为预测提供简易而可靠的途径,充分发挥环境效应数据的内在价值。方法分别运用GM(1,1)模型、幂函数模型、指数函数模型、多项式模型预测9种工程塑料在3种典型大气环境下的拉伸强度、延伸率和弯曲强度等力学性能变化,并与自然环境试验实测结果进行比较。结果拉伸强度、延伸率、弯曲强度等3种力学性能中,弯曲强度预测效果最好,延伸率预测效果最差。4种模型之中,GM(1,1)模型在预测小样本数据方面具有最好的预测效果和最高的正确率,且在数据具有一定波动性的情况下仍能够保持较好的预测效果。结论在挖掘分析塑料大气老化评价指标和试验时间之间的函数关系时,不仅要考虑预测模型的拟合效果好坏、相关系数高低、预测误差和预测正确率高低,还应注意预测模型反映出来的塑料性能变化规律是否符合客观实际,这样得到的预测模型才能尽可能地反映塑料老化程度与试验时间关系的全貌。

Aging Prediction Model of Plastic Exposed in Atmosphere Environments

Objective To provide simple and reliable method for prediction of plastic aging behaviour, and achieve the full value of environmental effect data. Methods The changes of tensile strength, ductility and bending strength for 9 kinds of engi-neering plastics, which were exposed in three types of representative atmospheric environments, were predicted with different models including GM(1,1) model, power function model, exponential model and polynomial model, and the calculated results were compared with actual measurement results of natural environment test. Results Among three mechanical properties, the prediction effect of bending strength was the best; while the prediction effect of ductility was the worst. Of the four prediction models, GM(1,1) model presented the best prediction effect and the highest accuracy. And good prediction effect was achieved when the data possessed the characteristics of fluctuation. Conclusion When the function relationship between atmospheric aging evaluation index and test time is explored and analysed, it is required to consider not only the fitting effect, correlation coefficient, prediction error and prediction accuracy but also the consistency of change law of plastic properties reflected by the prediction model and objective reality. Then the whole relationship between plastic aging degree and test time could be reflected as closely as possible by the prediction model.