Nastaj Andrzej, Wilczyński Krzysztof
Polymer Processing Department, Faculty of Production Engineering, Warsaw University of Technology, Narbutta 85, 02-524 Warsaw, Poland.
Polymers (Basel). 2021 May 12;13(10):1547. doi: 10.3390/polym13101547.
A review paper is presented on optimization and scale-up for polymer extrusion, both single screw and twin screw extrusion. Optimization consists in obtaining a multidimensional space of process output variables (response surface) on the basis of an appropriate set of input data and searching for extreme values in this space. Scaling consists in changing the scale of the process according to specific criteria, that is, changing the process while maintaining the scaling parameters at a level that is as close to the reference process parameters as possible. It consists in minimizing the differences between the parameters characterizing the reference process and the resulting process. This may be obtained by using optimization techniques leading to the minimization of discrepancies between the parameters of scaled processes. In the paper, it was stated that optimization and scale-up based on process simulation are more effective than those based on experimentation which is time consuming and expensive. The state-of-the-art on extrusion process modeling which is the basis of optimization and scale-up has been presented. Various optimization techniques have been discussed, and the Genetic Algorithms have been identified as powerful and very efficient. Optimization and scale-up based on the process simulation using Genetic Algorithms have been broadly reviewed and discussed. It was concluded that, up to date, there is a lack of optimization studies on the counter-rotating twin screw extrusion, although the global models of this process are known. There is also a lack of process simulation-based scaling-up studies, both on the counter-rotating twin screw extrusion and on the starve fed single screw extrusion. Finally, development perspectives in this field have been discussed.
本文发表了一篇关于聚合物挤出（包括单螺杆挤出和双螺杆挤出）的优化与放大的综述论文。优化在于基于一组适当的输入数据获得过程输出变量的多维空间（响应面），并在该空间中搜索极值。放大则是根据特定标准改变过程的规模，即在保持缩放参数尽可能接近参考过程参数的水平上改变过程。这包括使表征参考过程和最终过程的参数之间的差异最小化。这可以通过使用优化技术来实现，从而使缩放过程的参数之间的差异最小化。论文指出，基于过程模拟的优化和放大比基于实验的优化和放大更有效，因为实验既耗时又昂贵。本文介绍了作为优化和放大基础的挤出过程建模的最新进展。讨论了各种优化技术，并确定遗传算法功能强大且非常高效。对基于遗传算法的过程模拟进行优化和放大进行了广泛的综述和讨论。得出的结论是，尽管已知反向旋转双螺杆挤出过程的全局模型，但迄今为止，缺乏对反向旋转双螺杆挤出的优化研究。在反向旋转双螺杆挤出和饥饿进料单螺杆挤出方面，也缺乏基于过程模拟的放大研究。最后，讨论了该领域的发展前景。
