Optimization Of PET Bottleblow Molding Processand Prediction Of Molding Time

With the change of people's living habits,the application range and demand of PET material in the field of plastic bottle packaging are growing day by day.Various PET bottles have emerged with the change of consumers' preferences."Injection,stretch and blow molding" is one of the plastic bottle processing methods that has developed rapidly so far.Due to the complex shape changes of the material during the stretching process,there are many rheological parameters that are diffficult to control,and likely to result in a series of problems such as uneven wall thickness distribution and long molding time in the processing process.Therefore,this dissertation focused on the blowing molding process of PET bottles to carry out relevant research.The main research contents are as follows:Firstly,the research results of molding equipment,bottle blowing process and numerical simulation made by domestic and foreign scholars in recent years are summarized and reviewed and some problems are pointed out through the specific analysis.Based on the method of numerical simulation,the process parameters are adjusted to complete the design of blow molding scheme,so as to achieve the research objectives of improving the performance of molding bottles,shortening the processing cycle and reducing the cost of enterprises.Secondly,taking POLYFLOW as the research platform,the actual process parameters are recorded and sorted out and the influence of the moving speed,delay time and blow molding pressure on the distribution of PET bottle materials are mainly focused.The deformation characteristics in the process of drawing and blowing molding are observed and the trend of thickness distribution of molding bottle is naturally explained by numerical simulation.More importantly,the feasibility of the numerical simulation method is proved by the comparative experiment of PET bottle wall thickness measurement.Thirdly,based on the Taguchi experimental design idea,several technological conditions related to the molding quality in the blowing molding process of PET bottles are comprehensively considered and the influence of meshing on wall thickness uniformity and molding time is analyzed.By means of SNR mean analysis and variance analysis,the influence weights of different factors on the test target are obtained.Through the adjustment of the existing scheme design,the final goal of optimization process is realized.By comparing the geometric and mechanical properties of PET bottles before and after the optimization treatment,it is proved that the numerical simulation method can effectively improve the thickness distribution and vertical load capacity of PET bottles.Finally,the accurate grasp of the molding time is conducive to better control the operation rate of the bottle blowing machine and realize the high-speed operation of the bottle blowing machine.The multiple regression prediction model is obtained through data fitting in MATLAB,and the regression coefficient is adjusted by eliminating abnormal points to obtain the prediction model with relatively small errors.Using Python language to complete the design and construction of the BP neural network prediction model,constantly modify the number of iterations to improve the neural network,to obtain a prediction model with strong learning ability.Also,the classical support vector regression machine is used to predict the molding time of PET bottles.By comparing the predicted results of aforementioned three methods and the mean square error between the predicted value and the real value,the quality of the prediction model is evaluated and complete the prediction exploration of the molding time in the end.