A Novel Soft Sensor for Real-Time Monitoring of the Die Melt Temperature Profile in Polymer Extrusion

Polymer extrusion is the most fundamental technique for processing polymeric materials, and its importance is increasing due to the rapid growth of worldwide demand for polymeric materials. However, the process thermal monitoring is experiencing several problems resulting in poor process diagnostics and control. Most of the existing process thermal monitoring methods in industry only provide point/bulk measurements, which are less detailed and low in accuracy. Physical thermal profile measurements across the melt flow may not be industrially compatible due to their complexity, access requirements, invasiveness, etc. Therefore, inferential thermal profile monitoring techniques are invaluable for obtaining detailed, accurate, and industrially compatible measurements and, hence, to achieve improved process control. In this paper, a novel soft sensor strategy is proposed to predict the real-time temperature profile across the die melt flow in polymer extrusion for the first time in industry or research. It is capable of determining the melt temperature at a number of die radial positions only based on six readily measurable process parameters. A comparison between the simulation results of the novel melt temperature profile prediction soft sensor and the experimental measurements showed that the soft sensor can predict the real-time melt temperature profile of the die melt flow with good accuracy. Therefore, this will offer a promising solution for making real-time melt temperature profile measurements noninvasively in polymer extrusion, and also, it should be applicable to other polymer processes only with a few modifications. Moreover, this technique should facilitate in developing an advanced process thermal control strategy