Digital image processing for measurement of residence time distribution in a laboratory extruder

A non-labor intensive, non-destructive method based on image processing was developed to measure residence time distribution in a laboratory extrusion process. The a∗ value (of L∗a∗b∗ color space) was measured using image processing instead of a colorimeter to determine concentration of tracer in the product. The method was faster and less subject to human error. The a∗ values obtained by the two methods correlated well with r2 > 0.98 and MSE < 4.70. The parameters of the residence time distribution (RTD) were found to be similar. Effects of screw speed and barrel temperature on RTD as determined by the image processing method correlated well with the previously reported results. Increases in screw speed and temperature resulted in decreases in the mean residence time and increases in degree of mixing. A 3 × 2 factorial design with screw speeds of 80, 120 and 160 rpm and barrel temperatures of 120 and 160 °C was used for the experiment.