2.5. Determination of Residence Time Distribution

RTD of the continuous powder blending was determined by impulse disturbances. MCC was fed to the twin-screw blender with the Brabender twin-screw feeder. The feeder was used in gravimetric mode, set to the examined total mass flow. The powder transport was fast enough in the powder blender with the chosen rotational speeds to avoid any accumulation at the hopper as that would have affected the residence time distribution.

ASA was introduced to the system by manually adding pure ASA directly into the hopper of the blender. Different impulse sizes were tested in the range of 0.5–2.5 g ASA (1.5 g). To determine the size of the disturbance on the system, the mass hold-up was calculated from the MRT of the system and the mass flow with Equation (1) and the spike was compared to the hold-up with Equation (2).

where m_hold-up is the hold-up mass in the blender, MRT is the mean residence time, m˙ is the mass flow, d is the disturbance rate, and m_impulse is the mass size of the impulse disturbance.

The impulse response was evaluated with NIR spectrometry after the powder blender. The time period between the spectral acquisitions was intended to be reduced in order to obtain the most amount of information about the RTD of the system. For this purpose, we used only four scans, after that the acquisition time was determined mainly by the waiting time between two measurements, so further reducing the number of scans does not result in significantly higher sampling frequency. NIR spectra were evaluated in real-time and repeated impulse disturbances were made when the ASA concentration settled at zero.

The effect of the CPPs on the blending (mass flow and screw rotation speed) was studied using DoE studies. The mass flow in general is determined by the equipment’s capacity and the capacity need. The minimum screw rotational speed is determined by the mass flow, as the screw rotation has to be capable of transporting the material without accumulation. These two factors were changed at two levels according to a two-factor 22 experimental design. In order to study the scale-up of the system, two different mass flow ranges were examined with two sets of DoEs. Three repetitions were made at each experimental setting. The main parameters of the distribution were evaluated by numerical integration of the responses according to Equations (3)–(5), and the effects of the factors were determined with ANOVA tests.

where E(t) is the probability distribution function (PDF) of the process in t as time, Var is the variance of the PDF, and σ is the standard deviation of the PDF.