Application of shrinkage feature to a part design: reference CSYS

[tbraxton]

Creo Parametric 4-12. Mold design/Casting application of shrink by scale for anisotropic shrink.

When applying a shrink by scale feature a reference csys is used to apply the shrink. Can anyone offer some insight into how/why the selection of this csys is important? What are the criteria that should be used to determine the "best" csys for use with anisotropic shrink applied to a design part (reference model).

Can anyone offer best practices based on theory or practical experience? I am focused on injection molded polymers in this context if that matters. I would default to the use of a csys located by determining the centroid of the reference part being shrunk, but this is an educated guess.

[KenFarley]

I've not had to do this kind of thing, but I would think you'd HAVE to have a CSYS to specify what the shrink factors are for the different directions. For example, if we are making a part by casting aluminum impregnated with ceramic fibers, and the fibers are aligned in a specific direction, the shrink along the fiber direction would be very different from the non-fiber direction.

I wonder how good the calculations are in Creo with respect to predicting actual final size. Sounds like you're doing some fun stuff.

[tbraxton]

Fundamentally, I suppose I am actually asking for details of the algorithm used to apply the shrink to the geometry. I have inquired with PTC employees but so far have not gotten details that allow me to understand how it works.

The scenario you reference would be dealt with by using Moldflow simulation to see the fiber orientation as well as the shrink/warp resulting from a given set of processing conditions within a mold. I should have been clearer in my original post that I am asking this in the context of the resulting geometry created in Creo as a result of using this feature and not as a mold design question. I am looking at using the shrink feature to deal with a design problem of geometry and constraints to dependent models. I am not able elaborate on the real-world problem in this forum, hence my use of generalities. The geometries are irregular and complex forms.

Another way to consider this is to ask how the geometry of a part is affected as the result of varying the reference csys when using shrink by scale functionality. Is the resulting geometry expected to be different if I use the centroid csys vs a csys located in space away from the geometry of the part? I may end up having to test this out to learn the results by creating some test cases.