During my days as a structural package designer, when designing a new bottle, the last step for our department was to produce the CAD drawing of a new design. Then it went over to Engineering, who would tweak our CAD to produce the tooling drawing. Then we were ready to "cut steel," i.e. have the molds made for the blow-molding machine. That last step took months, and cost five figures.
That was in the '90s, and today you can have steel cut for a new bottle mold in about 4 weeks. However, Pepsi has figured out how to get that time down to just 48 hours, and with a cost savings of up to $10,000. It's staggering.
The way they do it is by not cutting steel at all. Instead they use a universal mold shell, then they 3D-print an insert that goes inside the mold.
People have messed around with this before, but the problem has been durability and cost. A 3D-printed ABS insert is nowhere near as strong as steel, and you can blow-mold maybe 100 bottles before the insert fails. Hardly worth your time and money, particularly since the PolyJet 3D printers required cost $250,000, the Digital ABS material you print with runs $422 per kilogram, and the prints can take three days.
To solve this, a research team led by Max Rodriguez, PepsiCo Senior Manager of Global Packaging R&D, turned to 3D printer manufacturer Nexa3D. The company used their NXE 400 Printer, an ultrafast machine "capable of printing at injection molding repeatability and tolerances, making it the fastest and most accurate production 3D printer in its category," the company says. It also costs just $60,000.
For material, the team used a polymer called xPEEK147, made by Henkel Loctite. xPEEK147 is tough, thermally stable and has high dimensional stability, and rings in at just $308 per kilogram.
The new inserts were able to print more than 10,000 bottles before failing.
And the cost savings are significant:
- Compress prototype tooling development time from 4 weeks to 48 hours
- Slash prototype tooling costs from $10,000 to $350 per mold set
- Create durable tooling that can produce more than 10,000 bottles per mold
- Enable multiple design iterations to allow for timely verification of downstream activities
As a former package designer, I'm torn on this. I think it would've been so cool to be able to produce new bottle runs and design changes for small batches—new bottles for the World Cup and other major events. And I'm super-impressed with the technologies harnessed here. But then I start thinking wait…don't we have a plastic bottle problem?
In any case, you can read the case study for the above here.