Application of polyvinyl chloride for nanocomposites (analysis and optimization of quality indicators)

Introduction. The application of nanocomposites in various industries has increased in recent years due to their unique properties and performance characteristics. Enhancing the quality performance of nanocomposites presents a formidable challenge, primarily attributable to the intricate interactions among their constituent elements. This paper presents an overview of the methods used to analyze and calculate the quality performance of nanocomposites using PVC as the base material. Methods and materials. To calculate technological parameters that are not directly measurable, we conducted a thorough analysis of literature sources related to the specific process under consideration. Through this analysis, we established appropriate relationships between empirical data in accordance with the fundamental principles of thermodynamics and mass transfer processes. In the course of the research, the method of neural networks was applied in order to describe the process of vinyl chloride polymerization carried out by suspension method. To solve this problem, a cascade network with forward propagation of the signal and backward propagation of the error was applied. Network composition: in the hidden layer - ten sigmoid neurons, in the output layer - two linear neurons. Results and Discussion. Throughout this research, it was observed that the heat flux during polymerization exhibits temporal variation, contingent upon the concentration level of the initiator. Further the dependencies obtained can be used in controlling the flow rate of refrigerant into the reactor cooling jacket, to ensure that the entire process is isothermal. It was found that by varying the stirrer speed, it is possible to change the particle size and hence the molecular weight distribution of polyvinyl chloride. The developed neural network was tested. The obtained results have minimal error and are close to the real values, from which we can conclude that the network is trained correctly and the dependence between the data is found.