Logic may not always be the best guide to polymer flow and viscosity. For example, though a mechanical engineer likely assumes that a double bond between molecules equals strength, a chemist knows it’s unstable.
If there has been a recurring theme in these articles over the last nine-plus years, it’s the importance of the polymer’s molecular weight to performance. It is a distinctive characteristic of these materials that we call plastics and without it they do not demonstrate expected properties.
The short-term property that declines first is impact resistance or, in a test involving lower strain rates, properties associated with ductility such as ultimate elongation. The onset of brittle behavior is the proverbial canary in the coal mine, the first sign of trouble. But molecular weight losses have serious long-term consequences as well, including declines in creep resistance, environmental stress crack resistance, fatigue resistance, and barrier properties.
Throughout the discussions on molecular weight, we have emphasized its relationship to melt viscosity. It has long been recognized that polymeric materials made of large molecules exhibit greater resistance to flow than those made of smaller molecules, and this resistance to flow is defined as viscosity. In a tactile, qualitative sense, it is a measure of the material’s stickiness. The large molecules in higher-molecular-weight materials stick to each other more firmly and are harder to separate from one another. As processors, we must use more force to move these materials the required distances needed to make good parts.
This stickiness can be reduced by increasing the velocity with which the material flows. The faster the material flows, the lower its resistance to flow becomes. This is known as non-Newtonian behavior and it fundamentally distinguishes the mechanics of polymer flow from that of classical fluids such as water. This behavior means that the faster a material flows, the harder it is to tell the size of the molecules in that material. So if we are going to make a measurement of viscosity that gives us some insight into the size of the molecules in the system, the measurement is best made at a low flow rate.
They don’t always break down
Enter the melt-flow-rate (MFR) test. This test involves measuring the flow rate of polymer melts at very low velocities. The fact that these velocities are not controlled actually works to our advantage because it exaggerates the effects of changes in molecular weight on the measurements. These measurements are expressed in units of mass melt flow per unit time (g/10 min) and are commonly used in the commercial marketplace to distinguish between grades of material in a product line. They also serve as specifications to which particular grades of material are manufactured. Processors use these numbers, for better or worse, as gauges of how easily a material will process or whether it will process at all in a given part geometry.
In a large number of these articles, we have discussed the MFR test as a means of determining whether or not a material was degraded during the molding process. A large increase in MFR during processing is associated with a significant reduction in the average molecular weight of the material. The likelihood of product performance problems increases in these cases. But the assumption behind this connection between increasing MFR and polymer degradation is that when polymers degrade, the size of the molecule always decreases. In most cases this is a correct assumption. Polymer degradation most often involves a process known as chain scission, where the polymer chain breaks to form two or more smaller polymer molecules. However, there are some cases in which damaging changes can occur in a polymer, and chain scission either is not involved or is not the dominant mechanism. These processes are sufficiently complicated so that they require an extended discussion, so we will treat one type of problem and how to test for it in this session and deal with an interesting twist in the next article.
When two bonds aren’t better than one
For the moment, we will focus on certain types of impact-modified materials. The rubbery materials that are used as impact modifiers in rigid polymers have a variety of compositions, but one of the most frequently used materials is butadiene rubber. This is the modifier in impact grades of polystyrene and it is also the “B” in ABS.
In addition, impact polystyrene is a major component in commercial PPO and PPE compounds. Many impact-modified engineering-grade materials such as nylons and polyesters employ EPDM (ethylene propylene diene monomer) rubber. In organic chemistry, any time you see the root word “diene,” it is a clue that a particular type of structure is present. The crucial component in this structure is a pair of carbon atoms connected by a double bond (C=C). Now, when mechanical engineers see a structure like this, they automatically assume that this is a good thing. After all, two bonds instead of one must indicate twice the strength.
However, chemists know that the second of these two bonds forms because the electron pair responsible for its formation has nowhere else to go. In other words, this is an unstable situation and, given a choice, this electron pair would rather be otherwise occupied.
The orchestrated breaking of this second bond is actually a driving force behind the manufacturing process for many of our most commonly used plastic materials. Ethylene, propylene, styrene, and vinyl chloride are all small molecules that contain such a double bond. Heat, pressure, and catalysts are used to break these double bonds under controlled conditions. This causes the electron pair in each of these molecules to go looking for a better, more stable arrangement. This arrangement involves combining with a neighboring molecule to build up the polymer chains that we then convert into our pipe, tubing, lawn chairs, and so on. Ideally, when we are done building the polymer, we have consumed all of the double bonds that were present in the raw material.
Butadiene is polymerized in much the same way. But, as the name would suggest, butadiene monomer contains two double bonds. So when the polymer chemists work their magic to produce a polymer from the butadiene, one of the double bonds in each molecule is consumed while the other one remains. This remaining double bond represents an Achilles heel in any system into which it becomes incorporated. Ultraviolet radiation, heat, and many chemicals attack this location in the molecule, initiating a range of complex reactions that involve degradation of the polymer.
This is one of the reasons that materials like high-impact polystyrene and ABS do so poorly in outdoor applications. And if you want to get a feel for the effect that the butadiene phase has on chemical resistance, examine a chemical resistance guide for styrene-acrylonitrile (SAN) against one for acrylonitrile-butadiene-styrene (ABS) to see how many chemicals attack ABS but not SAN.
A complex relationship
One of the reactions promoted by the breakdown of butadiene rubber involves crosslinking. Crosslinking is the primary objective in thermoset processing. It is responsible for the distinctive properties of this class of materials, and, as any processor of thermosets can tell you, the crosslinking process involves a significant increase in viscosity.
The reason is quite simple: Crosslinking involves joining together two or more polymer molecules with bonds that are very strong. This results in a rapid increase in molecular weight and a corresponding increase in melt viscosity. If the reaction is carried to an extreme, as it is in thermosets, the viscosity becomes so high that the material can no longer flow and attempts to remelt it are futile. However, in thermoplastic systems like ABS and impact-modified nylon, the crosslinking process represents a minor side reaction rather than an orchestrated objective. So while complete solidification of the material seldom occurs, increases in viscosity are still possible.
This complicates the straightforward reaction between polymer degradation and decreasing viscosity (increasing MFR). In materials like ABS, PPO, and some toughened nylons, it is normal for the MFR of the molded part to be lower than that of the raw material. People familiar only with the principle of MFR increase during processing typically assume that if an increase in MFR is a negative thing, then a decrease in MFR must be beneficial.
If the decrease in MFR can be assigned only to an increase in the size of individual polymer chains, then it is good. In the next article we will show an example of such behavior.
However, as with materials that decrease in molecular weight, the key is to keep the difference between raw material and molded part small. Unfortunately, the phenomenon of MFR decrease has not been studied as thoroughly as that of MFR increase, and well-developed rules that relate these changes to performance do not exist yet.
The answer in notched Izod
But empirical evidence for what is good and what is bad can be collected if we are paying attention. Recently, we received a small sample of molded ABS parts that were cracking in an uncharacteristic manner. A series of tests designed to check for contamination produced no abnormalities, and attention turned to the possibility of degradation. The nominal MFR of the raw material was 6.5 g/10 min, but the material in the cracked molded part produced an MFR result of 1.31 g/10 min.
This either represents a large increase in average molecular weight or it indicates that the wrong grade of ABS was used to make the parts. As usual, it is always best to have a sample of the related raw material for a comparison. When the sample was obtained and tested, the MFR was measured at 5.66 g/10 min. While it was not possible to obtain a specification range from the raw material supplier, a shift from nominal of less than 1 g/10 min is not large. But the failed molded part exhibited a decrease in MFR of almost 80%, or more than a four-fold increase in viscosity. Now the question became one of how to attach a practical significance to what appeared to be an abnormally large change in MFR between pellets and parts. The best way to accomplish this is to mold the material under controlled conditions into a sample geometry that is simple to evaluate and that can be compared to the molded part.
The only other property provided on the material certification was the notched Izod impact. The nominal value for the material is 5.4 ft-lb/in. Notched Izod bars molded in the lab under controlled conditions produced an MFR value of 5.17 g/10 min, a relatively small change compared to the 5.66 obtained for the pellets. This told us that our molding process was quite different from the one used to produce the parts. But how did this difference affect the practical toughness of the material?
It is not always possible to obtain good comparative samples from a molded part that can be used in a standardized test. Often the molded parts are too small or have a geometry too complex to provide the real estate needed to yield a standard test bar. And even if the space is available, factors such as material orientation or nominal wall thickness can influence results. Fortunately, in this case the part was large enough and of the appropriate thickness, and the gate location produced an orientation that allowed for a close comparison between the standard samples and those obtained from the molded part. The table on p. 36 shows the MFR and notched Izod impact results.
A lot of criticism can be leveled at the notched Izod test. However, there is value in repeating a test that is used as a release specification in order to determine how closely the actual product resembles the expectations of the material supplier. In addition, the presence of a notch often faithfully replicates the presence of sharp corners in a part design. An examination of the fractures in the molded product showed that the failures did, in fact, correspond to locations where corner radiuses were smaller than those suggested by good part design rules.
The upside-down relationship between MFR and toughness, as measured by the notched Izod test, is apparent. It is the reverse of what we have come to expect when working with materials that undergo straightforward chain scission when they degrade. The root cause, although difficult to pin down without viewing the actual process, is almost certainly some combination of elevated melt temperature and extended residence time, one of the key mechanisms that can initiate the breakdown of the butadiene rubber phase.
In the next article, we will dig into a more complex version of this unconventional relationship between viscosity and properties as it manifests in materials where no impact modifiers are involved and we will discuss test methods that can unravel the difference between a good increase in melt viscosity and a damaging one.
About
Michael Sepe Michael Sepe has worked in the plastics industry since 1975 in a variety of roles involving both manufacturing and research and development. He is an independent consultant based in Arizona with clients throughout North America. He assists clients with material selection, designing for manufacturability, process optimization, troubleshooting, and failure analysis. Learn
more about Michael Sepe.
