Plastic types
Much of what mikes919 and JR said is very similar to what I've heard from a variety of other sources. Unless anyone has additional information, the corresponding plastics (not including additives) appear to be:
DX/D-Plastic: Polyethylene
Pro/Elite X: Hytrel*
Champion/Z/E/Opto: Polyurethane
Star/ESP: Hytrel*-Polyurethane blend
*Hytrel is a registered trademark of DuPont. There are many different kinds of Hytrel (see http://www2.dupont.com/Plastics/en_US/Products/Hytrel/hytrel_amer_data_sheets.html for a list) Does anyone know which particular Hytrel is used by disc makers?
Note: mikes919 suggested that Nylon might also be used in some of these plastic blends.
Another question: When the disc makers use a blend, do they blend it themselves? If so, how well homogenized is the mix before it goes into the hopper?
______________________________________________________________________________
Shrinkage
mikes919 said:
...Generally for a specific plastic and temperature you will assume a constant shinkage rate around the whole mold. For example, I used to design parts in 3D and then blow them up by a set percentage, then use the blown up model to create a mold. There are exceptions, where you learn from experience that certain shapes might shrink more or less than the general rule.
Discs appear to be one of these exceptions, owing to their asymmetry, causing the wing to droop upon cooling, thus lowering the PLH (Parting Line Height). A couple of guys I talked to who used to mold discs for Innova showed me a Firebird that they let cool in the mold for ~30 minutes, so that its shape stayed very similar to that of the mold. The top was very flat, the PLH was very high, and the disc is so over-stable and hooks so strongly that even the biggest of arms can only throw it 250'. They also said that the dome of a disc is produced entirely during shrinkage.
What this information implies is that the mold itself must have a more over-stable shape than the desired shape of the disc it produces, since shrinkage lowers the PLH. Mechanically, this makes sense since the flight plate partly braces the upper rim, while the underside of the rim is free to shrink inward. The differential shrinkage at the top vs the bottom of the rim then determines the final PLH.
The degree of upper rim shrinkage should change depending on whether the flight plate gives way easily (e.g., if it domes up easily) or if it is more rigid. In other words, whether the flight plate is a strong or weak brace against radial contraction. I think this explains why the early Blizzards (with the bubbles in the flight plate) turned out to be so over-stable: the bubbles made the flight plate so flexible that it could no longer brace the upper part of the rim against shrinkage, causing the flight plate to dome up instead of resisting radial contraction, so that the upper vs lower rim differential shrinkage was more uniform, thus producing a higher PLH than would a stronger flight plate.
Also, different plastics will shrink to different degrees, because they have different thermal expansivity and elasticity. When I looked at thermal expansion coefficients, the typical value for polyethylene was significantly greater than that of polyurethane. Therefore shrinkage of DX-type discs will be greater than for Champion-type discs. Thus the under-stabilizing effects of shrinkage will be greater for DX than for Champion, which would explain why Champion plastic discs tend to be more over-stable than DX plastic versions from the same mold. Of course, thermal expansivity isn't the only property that matters, it is actually the temperature change upon cooling times the thermal expansivity, and the operating temperature of different plastics is also probably different. Also, the elasticity of the plastic also matters, as well as the glass transition temperature. The only way to sort out all these competing effects is to do some models of the shrinkage process.
mikes919 said:
...I imagine they learned a lot in the early runs at the disc companies...
I recall throwing some of the first discs produced by small start ups (e.g., the Medusa from Skyquest), and they were often very under-stable (more than the designer probably intended). Perhaps this is because they forgot to consider the effect that uneven shrinkage has on the stability characteristics of the disc, particularly the lowering of PLH.
mikes919 said:
Discs are so large that they have quite a bit of residual heat when they come out of the mold, and the final shape can be strongly influenced by the conditions under which they are cooled. I know for a while at gateway they were cooling discs in special racks to make sure they cooled to an extra-overstable shape. That's not generally practical, so that's one reason why you see so much variation from run to run of the same disc.
No doubt, while a faster rate of production is more cost-effective, it also gives rise to greater shrinkage artifacts and opens the door to stronger variations from run-to-run. The only sure-fire way to make a consistent disc is to allow it to anneal completely in the mold, but that costs a lot of money since the production rate drops significantly. These kinds of trade-offs could also be modeled and made quantitative, I wonder if any of the disc makers have done anything this sophisticated (such a thing is straightforward for a scientist or engineer).