2.6. Sampling and determination of segregation index

After each filling step, the concentration of the filled die is separated into 10 samples. Each sampling unit (Fig. 2), starting from the top, is pulled out of the die and its content is emptied into the sample separator (Fig. 3). The left half of the sample separator contains the powder blend from the far end of the die, and the right half contains the sample from the near end. The step is repeated for each sampling unit, until the bottom is reached, for a total of 5 units. Therefore, a total of 10 samples is collected from each die filling experiment. Each sample is weighed (Entris® Analytical Balance, Sartorius) and transferred into a 100  mL volumetric flask. The determination of ASA concentration is carried out according to the procedure described in §2.4.

The obtained concentrations can therefore be used to examine vertical (top-bottom) and horizontal (near end – far end) segregation at various filling conditions. The die filling results are therefore expressed in terms of average ASA concentration (expressed as mass fraction of ASA to the blend mass), determined for a given sampled volume of the die, as a function of die depth. The die depth is a dimensionless parameter, given by the ratio of the sampling unit distance from the die table (centre of the unit) to the total filling depth (50 mm).

In order to compare the segregation tendency among the three blends, containing different amounts of ASA, it is necessary to develop a concentration-independent indicator. For this purpose, the following segregation index (SI) was introduced:

where c_i is the concentration of the analysed sample, detected by the analytical procedure described in §2.4., c_t is the starting concentration of the blend, determined after blending prior to die filling by the same procedure, and n is the number of sampling units. Therefore, by applying the following equation to each of the 10 samples (5 die levels and 2 halves for each level) produced from each die filling analysis, SI_a for a given suction/filling velocity combination can be estimated. Each die filling analysis is repeated three times (conducted in triplicate), therefore, an overall SI can be obtained as an average of three SI_a that characterises the degree of segregation within the die for each investigated filling condition.