Size (Shrinkage and Warp)

When molten plastic is injected into a mold and begins to cool, it shrinks. Therefore, the size of a plastic part is determined by the dimensions of the mold cavity and the shrink characteristics of the plastic. Indeed, the first things a mold maker wants to know is the part size and material shrink rate.

Unfortunately, plastics shrink at different rates depending on plastic type, part thickness, part geometry, and production machinery settings. To make things even more difficult, the plastic may shrink at different rates in different areas of the same part causing the part to warp.

These issues, in addition to tolerances of mold fabrication, produce an uncertainty in the final dimensions and shape of a molded part that should be planned for in the design of the part and assembly. Indeed, this uncertainty is the cause for the majority of costly changes made to new tools.

Knowledge of the specific plastic behavior and how part geometry will affect shrinkage helps the designer anticipate problem areas. Thus the designer can insure that the part will function within an appropriate range of uncertainty or that the mold design will allow adjustment of the part dimensions in a cost effective manner.

If the dimensional uncertainty derived from the shrinkage uncertainty plus the fabrication tolerance is less than the part tolerance for the given dimension, the mold may be produced with reasonable confidence that the part will conform to specification. However, if this is not the case, there is a risk that the molded parts will not conform to the design specification. In this case, the mold is typically constructed in a “steel safe” manner such that if the part does not meet tolerance, the mold can be modified by removing material, thus bringing the dimension into specification. For this reason, there is the potential for higher tooling costs when tighter dimensions are required.

It should be understood that while the uncertainty of a dimension may be high, the repeatability from part to part may be very good. This is because the shrinkage observed for a given process and part geometry is, in general, very consistent. The difficulty lies in the prediction of the shrink behavior before the first part is made.