Influence of sample shape and size on self-ignition of a fat-filled milk powder

Abstract

A lecithinated, 30% tallow fat-filled, milk-product powder, commercially-manufactured in Ireland, was used as the test material to assess the effect of sample shape and size on self-ignition in a thermostatically-controlled oven. Cubes of edge 25, 38 and 51 mm, spheres of diameter 26, 50, 77 and 102 mm, cylinders with diameter equal to length of 51, 76 and 100 mm and layers of 100 mm diameter and thickness of 10, 15 and 20 mm were tested. The minimum ignition temperature (MIT) varied from 206°C for the thinnest layer to 131°C for the largest sphere. For each shape, the results fitted the linear relationship ln( δ c T 2 r 2 ) = M− (N/T) , with close correlation. However, contrary to theory, the four straight lines did not coincide. Those for the spheres and layers were very close while those for cubes and cylinders had approximately 25% greater values for their slope (N). These results are shown to introduce significant differences when the values of M and N for the two most divergent lines within the test range are used to extrapolate estimates of critical layer thickness for the lower operating temperatures in parts of a milk spray dryer. M and N obtained from spheres gave the lowest estimates while those for cubes gave values up to 2·32 times, and those for cylinders, 2·85 times greater at 50°C. These results could also affect the safe design of silos for the storage of powder in the 40–50°C range to avoid self-ignition and allow the contents to cool to ambient temperature.