Non-Linear Plastic Shrink

I have seen people actually model this in. Come to think of it, it was warp instead of non-linear shrink, but same concept. They made a prototype tool, measured the warp and then made the production mold with the part deformed the opposite direction.

SW does have the ability to do different shrink rates in different directions, which in some cases might serve as an approximation, but the problem is that the direction of flow is never aligned to the origin in the real world.

In practice, it's mostly done at the mold builder. Make all your walls steel safe, then whittle away until you get it right. Very time consuming and expensive.

Mold process is also important, pack pressure, gating, venting, cooling, cycle time. As deterministic as we are, injection molding still has a lot of black science or trial and error involved.

Moldflow software has tools that help you predict anisotropic shrink effects as well as warp and sink, but I don't think it's possible to go electronically back to the CAD model to make corrections. You have to take measurements in Moldflow and take the numbers back to your model to tweek the shape.

snipped-for-privacy@tilikum.com (Bo Clawson) wrote in news: snipped-for-privacy@posting.google.com:

Hi Bo,

First of all I guess I want to know what you mean by `non-linear'. If what you are talking about is differnential shrinkage, then SW already has that capability. It just doesn't happen to be in the cavity feature. What I mean is that if you apply the shrinkage to the molded part first, you have the option of defining different factors in x,y and z. (scale function) Then you make your cavity and you are where you want to be. (at least if I understand you correctly.)

The problem with this is that all of your rounds become splines. This can be an issue for machinablilty. Naturally, for cored holes, you can still put in a round pin, but defining the location will be a bit more work. Also, what were fillets on the original part will now be a more complex shape.

One thing I have done with fairly simple parts is to re-model the part to allow for differential shrinkage. I actually went in and moved features so that I could still use a uniform shrinkage, but they would be `as if' I had applied differential shrinkage. For comples parts, this would be a chore, though.

SolidDesigner from CoCreate had a pretty good solution for this over 5 years ago. When you did your differential scaling you could select features that you wanted to scale isometrically, like round pins. The location would be scaled differentially, but the pin stayed round. It still didn't get around the fact that you really need to be able to define the shrinkage directions locally.

Jerry Steiger Tripod Data Systems

Bo,

In the few instances I've had to deal with non linear shrinkage, I was forced to model the molded part in SWX and manually shrink the sketched dimensions.

Malcontent

Bo:

In 2004 the "Cavity Feature" DOES have "Non Uniform Scaling".

Jacob, thanks for the note from SWks 2004, which I am not using yet.

What I see in my medical cylindrical tapered fittings work (like the ANSI 15mm & 22mm anesthesia/respiratory fittings) where functional diameters need to be met, is that I can't keep a constant wall and the thicker the wall, the more shrink in the Polyethylene & Polypropylene materials. That is different than X,Y, & Z shrink rates being set differently.

The 3D CAD software may never realistically be able to deal with it. Obviously there is a disconnect in being able to have the part drawing directly relate to the tool when the shrink is non-linear. One of the solids is going to be "wrong". Since the part drawing needs to be "right" for inspection of finished parts, the tool solids will remain "wrong" and need the manually noted 'adjustments'.

I suppose a real elaborate FEA program could step in between the part drawing and the cavity shape, but that would introduce lots of decisions that would get real sticky in a hurry and probably require the user to make a lot of judgement choices.

It looks like the solution is manual corrections.

Bo snipped-for-privacy@tilikum.com

If you get numerical data out and have a well-meshed surface model (lots of control points in the right places) an API program should be able to get you pretty close. This is sometimes done with CMM data for stamping & forming die design ... after a bit a predictive history might be built up ... Design iterations can be reduced in any case I think.

John,

I do much the same process when needed. What I do though is create a new configuration in the plastic part model for shrinkage. Then adjust all of the part dimensions for the non-linear shrinkage manually for that configuration only. I will have 1 configuration showing the original part, and 1 configuration showing the part with shrinkage allowance added. I just have to make sure that I have the part set to the correct configuration in my assembly when doing the cavity feature in you steel, and use 0% shrinkage in that feature.

The problem like you said is on complex parts as well as imported parts. Makes it kind of impossible...