The Materials Analyst, Part 91: Crystallinity vs. dimensional stability

Finding a postmold process that helps your part is no replacement for understanding how the polymers are reacting chemically.

In the August issue (immnet.com/articles/2007/August/3310), I discussed the problems with pseudoscience and folklore in the molding industry and the effect they have on achieving a level of competence needed to be globally competitive. Soon after the article came out, I received an e-mail from an OEM who purchases molded parts.

He related in his e-mail that his supplier told him that he achieves a greater level of dimensional stability in the parts he produces by putting them in cold water while they cool. The parts are produced from 30% glass-fiber-reinforced PBT polyester and acetal. The OEM representative wanted to know if there was any downside to this practice.

Without knowing the geometry of the part and the application environment, it is difficult to provide a definitive answer. But we can break down the practice and examine how it achieves the supposed improvement, and then make decisions about how advisable it is.

Strength in crystallinity

We can start with what we know about the base polymers. PBT polyester and acetal are both semicrystalline polymers, acetal more so than PBT. Consequently, when they cool, a portion of their structure naturally organizes into crystals. Not all of the polymer structure will crystallize; however, the two materials will naturally attempt to attain a certain level of crystallinity.

The crystal structure produces some significant benefits. As a family, semicrystalline materials exhibit improvements in a wide range of properties compared to their amorphous counterparts. These include fatigue resistance, stress crack resistance, and chemical resistance.

For example, most polystyrene materials are amorphous. However, if the polymer is synthesized using certain catalysts, this same polymer can organize into a semicrystalline structure. The table above shows examples of the chemical resistance of the two types of polystyrenes. While the semicrystalline variety, known as syndiotactic polystyrene (SPS), is not indestructible, it is clear that there is a substantial improvement in chemical resistance that arises from the presence of a crystal structure in the polymer.

The crystal structure also provides for a retention of useful strength and stiffness above the glass transition temperature (Tg). This is evident in the graph below, which shows the temperature-dependent behavior of polycarbonate, an amorphous material, and nylon 6, a semicrystalline polymer. At room temperature the two materials, in their unfilled form, exhibit comparable stiffness. As the temperature rises, both materials undergo a significant step change in modulus. This event is known as the glass transition, and it arises from the onset of a higher degree of mobility in the amorphous regions of the material.

In a material like polycarbonate, amorphous regions are all we have. So while the polycarbonate exhibits good property retention over a wide temperature range, up to about 135°C, once the transition temperature is achieved, there is nothing left to hold the structure together, and the material softens and effectively loses its load-bearing properties. The transition in the nylon occurs at a substantially lower temperature; however, once it is complete the material retains approximately 20% of its original stiffness and maintains much of that remaining rigidity until it approaches the melting point of the crystal structure at a point some 150 deg C above the Tg.

The amount of crystal formation in the polymer influences the degree of property retention above the Tg. However, in general, the melting point for the crystals in a semicrystalline polymer will be 150-200 deg C above the glass transition temperature for that polymer. If we return for a moment to the polystyrene, we can estimate the effect that the presence of a crystal structure has on the thermal properties of the material.

Amorphous (or atactic) polystyrene has a glass transition temperature of 100°C. At that point the material loses all load-bearing properties and if long-term performance is contemplated, use temperatures may be as low as 75-85°C, depending on the applied loads and expected lifetime of the application. Semicrystalline or syndiotactic polystyrene also has a Tg of 100°C, but because a crystal structure is present, the material does not soften at that point. It will not soften until it reaches the melting point of the crystals, at about 270°C. This significantly extends the range at which the polymer can operate in the real world.

Too much shrinkage

While crystallinity enhances the performance of a material on a number of counts, it also creates some problems for processors. Crystals are relatively well-organized regions and consequently they take up less space than a corresponding amorphous area. Therefore, as the polymer cools from the melt to the solid state, the volumetric change or shrinkage is greater for a semicrystalline material.

We can see this if we refer to the mold shrinkage values for various unfilled materials. Materials like ABS and PC shrink by approximately the same amount, 0.004-0.007 in/in. However, if we look at the same values for semicrystalline polymers like acetal and PBT polyester, we find that they are three to five times greater. In addition, process conditions can create a greater range of final dimensions in a semicrystalline product than in an amorphous compound.

Adding fillers and reinforcements reduces the high shrinkage values in semicrystalline materials, but this can cause additional problems. Glass fibers orient in the direction of polymer flow. Shrinkage is restricted to a much greater degree in the direction of flow than it is in the direction transverse to flow, a condition known as anisotropy. This differential in shrinkage contributes to warpage. This condition can be aggravated by poor part design features such as abrupt changes in the nominal wall thickness.

These tendencies challenge the processor to hold close tolerances and this leads to practices such as fixturing, reducing the temperature of the mold, or trying to cool the parts rapidly by placing them in cold water or running them through some other type of postmold cooling process.

What needs to be understood is that the mechanism for reducing shrinkage in semicrystalline materials involves suppressing the degree of crystallinity that the material achieves. This can result in a reduction in all the properties that we expect to obtain from semicrystalline materials.

In addition, the achievement of the desired part shape can be illusory. If the molded part is exposed to elevated temperatures at some later time, the crystallization process that was supposed to happen during molding will occur anyway, but it will happen once the part is in the field, where structural changes can cause performance problems. In some high-performance polymers like PPS, PEEK, and even some PET polyesters, the difference in crystallinity that occurs when the part is quench cooled can be so significant that it can be the difference between a good and a bad product.

Know what you’re doing

There are several things working in favor of the molder who engages in the practice of quench cooling in materials like PBT and acetal. First, both of these materials crystallize relatively quickly, so the shortfall in crystallinity will not be as severe for these materials as it is for some. Second, the high loading of glass fiber in these materials reduces the overall shrinkage of the materials and consequently reduces the variation.

The last piece of the puzzle is the part geometry. Wall thickness is an important determining factor in the achievement of crystalline structure because it governs the cooling rate. Thick walls take much longer to cool because plastics are relatively poor thermal conductors. Once the outer walls form a solidified skin, the material in the interior cools at a rate that is affected very little by the external conditions to which the part is exposed. Therefore, the final properties of the structure are not substantially affected by the overall cooling rate of the part. However, as the walls become thinner, the cooling rate is more influenced by the mold temperature and by any steps the processor may take to speed up the cooling rate after the part has been ejected.

So the essential question the OEM should ask the processor engaged in quench cooling is this: Do you understand the effects of the practice at a fundamental level or are you just getting away with something? If the molder understands how it arrives at the desired result, it realizes that it is suppressing a natural process and is able to manage the results because of the specifics related to the selected material, the part shape, and the application requirements. This understanding will prevent the molder from adopting the quench cooling process as a general practice and some day apply it to a part where the consequences are more significant.