Our Lab schedule is usually fully packed with customer trials, but with travel restrictions in place we have now one complete week to perform some internal test to support R&D department. So, I have decided to organize a race on the new Prexima 300T for soup cube manufacturing.
The machine has been fully tested and debugged, demonstrating that it can reliably match the 1,000 tablets/minute output of Corazza I10 wrapper. But how fast can we go? The idea is to challenge both the machine compression structure and its feeding system, to understand if there is some process aspect that could be worth a further innovation effort. We’ve asked a couple of our best customers to send us a few hundred kilos of some different recipes among the most challenging they have, and now here they are. As I look at the different colours and smell the different aromas of the various powders I feel very curious: who will be the winner?
I have fitted the machine with the IMA 32 tooling, 23×31 mm, and now I take some time to get to know the contenders: to avoid waste of product during machine setup, I’ll try to foresee which feeding parts and process parameters will suite at most their characteristics, Prexima has so many options to try!
We have four products: two vegetable based blends, one chicken based and one beef based.
Flowability (IC %)
Water content (LOD)
Flowability for all blends turns out to be from fair to good, with one on the two vegetable blends to be excellent. Water content is acceptable for this kind of products, where also fat plays its role in keeping the tablet together.
Particle size distribution is very similar for all blends. For all products, I notice the presence of lumps bigger than 1,180 µm that might cause some instability in the process: this is typical for product with fat content, during shipment and storage some lumps creation is quite common. My Prexima is strong enough to compress them too, but in case they would be too big it would be advisable to sieve them out. In this specific case all in all it’s not so bad, and I expect stable compression processes and high output.
The challenge looks interesting, and we’re ready to start: ready, steady, go!
I decide to start with the Veg1 blend: flowability is excellent and I am optimistic. With these product characteristics the high die feeder is the best solution, so the powder will gently flow according to its weight. Round standard profile for the paddles should be ok. For a soup cube of 10 mm height, considering powder density, I could use the 20 mm loading cam, but let me keep the 25 mm, so I will have more setting possibility when I will increase the speed. Because I will!
The machine has already been extensively tested at 1,000 tablets/minute, and in fact production runs smooth, with good quality tablets and low standard deviation. I can increase to 1,500 tablets/minute. As expected the tablet weight decreases slightly, as the dies have less time to be filled under the feeder. Increasing a little bit the dosing chamber and adjusting paddle speed will bring back the tablet to the target… done!
After a quick visual inspection, tablets do not present any capping or lamination, but at this speed product loss is increasing a lot: product residues are building up in process area, in the long term they could affect process reliability, not to consider product waste. Let’s take punch penetration from the initial 1.5 mm gradually to 5 mm, and see what happens. I keep increasing the speed with small adjustments on the dosing cam up to 1,800 tablets/minute. Great! With deeper penetration product losses are back to usual acceptable value, and no lamination is occurring. I feel with this powder there might be a potential to go even faster 1,800 tpm, but… oh no, product is finished! Well, no surprise. At this speed my Prexima is consuming almost 20 kg/min!
It’s time to try the second contender. Maybe is better to continue with Veg2, the second vegetable based blend. I assume it could have similar characteristic of the Veg1, even with a worse flowability as Carr Index clearly states. Supposing I can use most of the process optimization done with Veg1 powder, I try to start directly with a speed of 1,600 tablets/minute. Perfect, the process is very similar to the previous product. However, due to the lower flowability of the product I am using 23 mm of dosing with a 25 mm filling cam. It’s better to change the filling cam with a 30 mm one: to increase the speed I will need to fill more.
Now I can increase the speed: I reach 1,700 and then 1,800 tablets/minute without any major issues, but the screw feeder can’t keep pace in loading powder to the machine. I must accept 1800 tpm as the maximum value for this blend as well. From a first visual inspection, the tablets seem perfect. No capping or lamination. From a functional point of view is amazing to see how the weight trends at different speeds are overlapping each other. The machine autoregulation system follows the production changes maintaining the process stable.
In addition, the hardness is constant during the production at the same speed, meaning that the robust structure makes the machine reliable during the time without any fluctuation. The change of the filling cam has improved the hardness as I was assuming. Great!
Compression Technologist at IMA Active Process Development R&D Laboratory