(T1030-08-47) Characterization and Optimization of the Powder Flow Investigated with Different Feed Frame Setups Used for Continuous Manufacturing

Purpose: Continuous manufacturing (CM) requires a thorough understanding of every production step involved. The feed frame of a rotary tablet press is the last operational unit before tableting and is therefore of high interest. Modification of the feed frame setups may allow an optimization of the tableting process of powder blends. Therefore, the aim of this study was to identify feed frame setups which are best suited for CM by visualizing the powder flow and mixing behavior within the feed frame.

Methods: Experiments were carried out with a Fette FE55 rotary tablet press equipped with 60 pairs of 8 mm faceted flat-faced punches rotated at a turret speed of 9.45 rpm. Five different feed frame setups were examined (Figure 1A). The Fill-O-Matic (FOM) setup “aluminum reversed” is operated with reversed rotational direction of the distribution and filling wheel. The rotational speed of the feed frame was chosen to match the trajectorial velocity of the turret, resulting in a rotational speed of 40 rpm for the FOM setups and 25 rpm for the filling cone setups. Two different placebo powder blends, a white tableting blend and a red tableting blend, were used. First, the hopper was filled with 3 kg of red tableting blend and the tableting process was run up to the point where only the feed frame was still filled with powder. After leveling the powder (Figure 1B), 4 kg of white tableting blend were layered on top. The tableting process was resumed and again stopped as soon as only the feed frame was still filled with tableting blend (Figure 1B). After every five hundred tablets a sample tablet was withdrawn. Starting from the top of the feed frame (Figure 1B), the powder was removed in 0.5 cm thick layers by vacuuming with a stop at the distribution wheel of the FOM. In contrast, the filling cone was vacuumed completely. Pictures were taken of every powder layer. After setting a color threshold, the surface area of the remaining red tableting blend was determined using the image analysis software ImageJ (NIH).

Results: In the filling cone setups no red tableting blend was detected. The results of the FOM image analysis are displayed in Figure 2. The “aluminum galenic” setup resulted in the smallest red surface region in the upper region of the FOM hopper, whereas the “aluminum reversed” setup exhibited the largest red surface region. These results show that the powder flow may be altered by modification of the FOM. It was observed that the red surface region reaches a constant size at a depth of 9.5 cm within the FOM. From a depth of 0 to 9.5 cm, the FOM shape tapers off. Below this depth, the feed frame shows a constant geometry. Three-dimensional sections consisting of red tableting blend found within the FOM indicate transient zones where powder flow is reduced. The size of the transient zones and therefore the powder flow inside the feed frame may be affected by modification of the FOM (Figure 2). The images of the tablet samples taken during the tableting process are displayed in Figure 3. Over the time course of the tableting process it is obvious that the tablet color changes abruptly if using the filling cone setups compared to the fluent color change if using the FOM setups. This observation suggests less intermixing of the powder blends within the filling cone. These findings are consistent with the results from recent residence time distribution studies, where with the filling cone a lower mean residence time as well as variance was found than with the FOM [1].

Conclusion: In CM, a low mean residence time of the feed frame is desired following the first in first out principle. The filling cone shows less powder mixing and no formation of transient zones, suggesting that filling cone setups are more suited for CM. In traditional batch manufacturing, the feed frame is designed to guarantee a constant die filling and to compensate for variations in the composition of the powder blend caused for example by powder segregation in the hopper of the feed frame. This study shows that modification of traditional feed frame setups allows an optimization of the feed frames for CM.

References: [1] M. Dülle et al., Int. J. Pharm., 565, 523-532, 2019.


Figure 1: Schematic representation of the feed frames: (A) Feed frame setups; “Aluminum reversed” setup is operated with reversed rotational direction of the distribution and filling wheel. (B) Start and end of the tableting process.


Figure 2: Red surface regions within the Fill-O-Matic. A depth of 0 cm corresponds to the end of the tableting process as displayed in Figure 1B.


Figure 3: Tablet samples withdrawn after every five hundred tablets: 1) Aluminum FOM standard, 2) Aluminum FOM galenic, 3) Aluminum FOM reversed, 4) Filling cone cuboid, 5) Filling cone cylindrical.