Molded vs Machined Gears

The link in your story isn't about "Molded vs Machined" gears, it's about backlash -- the title of the article is "Sorting out gearhead semantics."

Additionally, I take issue with the comment (regardless of its source), "With virtually no way to accurately predict the amount of shrinkage of plastics [sic] gears for a given application, the gear mold manufacturer must fabricate one virtually perfect cavity and process each gear from that cavity within tolerance for every product made."

Not only is shrinkage predictable in plastic (or metal, for that matter), it is well known for long-used materials. And because common gear geometries are scalable up or down within limits, percent shrinkage or growth are scalable. It is only when extremes of sizes (far from sizes from which data has been acquired empirically) are approached that shrinkage or growth of materials may become non-linear.

Bill's remarks on the comment are entirely valid. Moreover, any competent moldmaker will know to target the small side of the tolerance band for the finished part; after measuring early samples, fine adjustments can be made in specific areas by polishing away metal (thus enlarging the gear in that area). The common term is "steel safe". Finally, if you have the ability to measure that "virtually perfect" mold cavity, why would you not replicate it? I began working with molded plastic gears in large production quantities in 1977, and we were building timing devices quite successfully, including backlash control, in reversing geartains, back then. There have been improvements in several of the technologies involved, in the interim.

Although the thread Chris refers is only about backlash, I must agree with him about molded gear shrinkage. I wrote an article in Gear Technology about molded gears. It can be viewed at http://www.geartechnology.com/pa/members/julaug06/section7.pdf. One cannot predict the exact shrinkage of a newly designed gear. There are too many unkowns. First of all, wall thickness changes shrinkage considerably. A filled gear will generally shrink much less but knit lines and fiber orientation will become dominant variables. We always assume that two iterations of shrinkage will be required to reach the best molded conditions. I have been doing this for quite a few years, and I'm pretty sure I haven't missed any remarkable revelations in that regard.

Concerning Ron's remarks about steel-safe cavity construction, I would disagree with that as well. The problem is that tooth width is small. If the normal tooth thickness is .030 inches and the tool maker goes steel safe by .005 inches, he has altered tooth thickness by 1/3 for his sampling. This tremendous change will affect the shrinkage dimension and the sampling shrink will not equal the final shrink. If he polishes the cavity, he is modifying a precision shape by guess. This is even a bigger problem.

Of course we can measure a mold cavity precisely and we can measure the molded gear precisely. The trick is to size the mold cavity to one exact dimension so that the molded gear is fabricated to another exact molded dimension. N'est pas?