Manufacturing · November 2025
Silicone Impeller Process Redesign
Rebuilt the molding and cure process for an HCR silicone impeller, taking scrap from roughly 10% to near zero.
The problem
A high-consistency-rubber (HCR) silicone impeller was scrapping at roughly one part in ten. The failures were not one mode. Some parts came out short, some tore at the hub on demold, and some passed inspection but failed downstream life testing.
What I did
- Characterized the incoming material and the cure window with tensile testing to ASTM D412 Die C geometry, across multiple lots.
- Ran a catalyst concentration investigation to separate cure-rate effects from fill and demold effects.
- Redesigned the cure schedule and handling steps around what the data actually showed, rather than around the historical recipe.
- Rewrote the work instruction so the process was reproducible by an operator who had not been trained by the person who invented it.
Result
Scrap on the redesigned process runs near zero. Test time on the validation loop dropped as a side effect of not having to re-run failed lots.
What I would do differently
I chased the tear mode for longer than I should have before instrumenting the cure. The lesson that stuck: measure the process variable before you argue about the failure mode.