Automated image analysis to characterize the melt densification stage of polymer sintering processes

Knowledge of melt densification during sintering is generally acquired by visualization methods, recording bubble formation and dissolution with time. Only manual visualization methods have been reported for polymer studies, which restrict the availability of data to a few discrete moments in time for the overall process. An automated vision system is presented in this paper to provide an improved level of analysis on densification, validated against a manual method. The machine vision technique was applied in the analysis of sintering for various polyethylenes of differing melt flow index and particle size. The automated technique was found to be very accurate and capable of collecting bubble size distributions on a timescale of seconds, which is an improvement in data collection. The method was prone to underestimate bubble numbers (∼20% error), especially those less than 100 μm in size, till the bed temperature rose significantly above the melting temperature of the polymer.