Feature

A different approach to hotfill PET bottle design

A smaller diameter finish will permit more design flexibility and cut costs


Print this page

February 1, 2001 by Lindsay Mulholland



Amcor PET Packaging, Mississauga, Ont.

Driven by safety and lightweighting issues, PET continues to replace glass in hotfill juices, isotonics and sauces. Our industry standard hotfill finish (the threaded part of the bottle) is currently 43 mm — this is the overall outside diameter as seen on 64, 32 and 20 oz. hotfill bottles. Unfortunately, this size imposes severe challenges on bottle design. To understand why, a review of the hotfill process is necessary.

Hotfilling is not employed to reduce microbial counts in the product, but to sterilise the internal surface of the container and closure. Bottles from any reputable manufacturer will have very low microbial loads since the melt and reheat-blow processing sterilizes them. Subsequent handling should maintain cleanliness. The product has already been pasteurized by a thermal treatment before filling in most modern processes. Hot liquid, typically at 185F (85C), is introduced to the bottle by a fill tube which is withdrawn leaving a headspace, and the bottles are conveyed to a capper where more headspace is created by sloshing. Our industry specification is that bottles must be able to withstand a headspace of 15 mm depth.

Bottles must be designed to withstand the hotfilling and the subsequent vacuum when the liquid cools. This is why hotfill PET bottles have panels and domes. The panels are designed to flex on cooling and take up vacuum, while the rest of the bottle must remain rigid and meet topload, shrink and ovalization specifications. On cooling after capping, the water vapor in the headspace condenses contributing to the vacuum. The liquid also thermally contracts. We have always believed that the major factor in creating vacuum pressure was the condensation in the headspace but closer examination reveals some surprises.

As it turns out, headspace volume imposes severe challenges on bottle design and the larger the headspace, the larger the problem. Headspaces are defined by the internal diameter of the finish and the volume is calculated by, V = * r2h, where V is the headspace volume, r is the internal radius and h is the height of the headspace. Obviously, the internal radius has a large effect on the volume.

The composition of the headspace must now be considered. Estimates of the water vapor pressure were obtained from tables and the balance, to one atmosphere, is air. In practice, the headspace cools during the time that elapses between filler and capper and the air content is higher than at 85C. We have measured this cooling in the laboratory and find that sitting still for 7 seconds causes the headspace to cool to 75C. This is good news as it reduces vacuuming, but bad news for oxygen sensitive products since more oxygen is trapped in the headspace.

Correcting for temperature and compositional changes leads to a decrease in volume due to condensation of water vapor. Surprisingly, however, this is not the major cause of vacuuming. Simply cooling the liquid from 85 to 23C causes a much more significant decrease in volume. Expressing this as a fraction of the total volume of the package shows us why it is more difficult to design a smaller container for hotfill than a larger one. (See Figure 2.)

In conclusion, smaller finish sizes are necessary if we are to design successful smaller-volume containers. This will lead to lighter-weight (lower cost) containers and lower cost closures. Although this study led to the conclusion that most vacuuming is due to liquid contraction and not vapor condensation, the implications are very serious for oxygen- sensitive liquids. There is significantly more oxygen in the headspace than we had previously thought and this may be the subject of a forthcoming article. We at Amcor have developed technology to address this issue and we look forward to sharing this with Canadian Plastics readers.