This series of articles is designed to help molders understand how a few analytical tools can help diagnose a part failure. Michael Sepe is our analyst and author. He is the technical director at Dickten & Masch Mfg., a molder of thermoset and thermoplastic materials in Nashotah, WI. Mike has provided analytical services to material suppliers, molders, and end users for 15-plus years.

Better flow can mean trade-offs in long-term material performance.

Molders prefer higher-flow materials. They enable processors to fill the same part with less pressure or at a lower melt temperature, which can translate to shorter cycle times and lower clamp tonnage. They can also extend the complexity of the geometries that can be produced. In addition, processors receive plenty of encouragement from material suppliers to use lower-viscosity resins. Special suffixes at the end of the grade nomenclature promise fast cycles, easier molding, and higher productivity. What they often fail to address is the sacrifice in properties that comes with such a selection.

First, it is important to understand the two mechanisms by which flow is increased in a polymer: additive technology and modification of the polymer itself. Both may be used, and it is not always simple to determine the approach a material supplier has taken. However, in most cases the differences in flow from grade to grade within a product line are achieved by reducing the average molecular weight of the material.

While there are many techniques for measuring this molecular weight, the most commonly used method is the melt-flow-rate test. This is being replaced in some circles with melt-volume rate, but the test does not change, only the calculation. When a processor purchases a material, the melt-flow rate or some alternate measure of viscosity is usually an expressed or an implied part of the specification. Those familiar with commodity materials such as PE and PP are accustomed to using the melt-flow rate or melt index (again the same thing) as part of the designation for the material.

In the case of polyethylene, this property is coupled with density to explain the range of properties that can be expected from the material. With polypropylene, the melt-flow rate is usually combined with some reference to polymer structure (homopolymer, random copolymer, and so on) or some key property that is related to structural factors such as an Izod impact number. But there are still many molders who do not comprehend the relationship between melt-flow rate, average molecular weight, and properties.

No Skimping on Property Tests

Part of the reason for the lack of attention to performance is that property changes are not always obvious. Short-term tests that are typically used to characterize a material often miss the alterations that occur when the flow of a material is modified. Table 1 shows the comparison of select properties for three acetal copolymers that are distinguished by their average molecular weight. The first four properties are almost always provided on a data sheet. The first three of these are unaffected by the changes in molecular weight and anyone who has performed a notched Izod test knows that the differences cited above would hardly be considered statistically significant.

As we move down the list, the differences become more evident. Elongation at break can be a good relative indicator of toughness that is not influenced by the artificial stress riser of a notch. Tensile impact provides a higher strain rate assessment of toughness in samples that are also free of notches. Unfortunately, the tensile impact test is one of those superior tests that somehow never caught on to the same degree as the notched Izod test. Fatigue data is very difficult to find.

The table illustrates an important point: If property testing is limited, it is possible to conclude that a higher-flow material has the same properties as its lower-melt-flow counterpart. New advancements in polymerization technology have made it possible to produce high-flow materials that approach the performance of higher-viscosity analogs, but the marketing hype tends to exceed reality. Often, the more demanding the application environment becomes, the more disparate the performance of the low-flow and high-flow choices will be. Here is a case study that illustrates the point in a polypropylene application.

A Falling Dart Hint The part is a large chute deflector. Parts of this type are traditionally made in some variation of a high-molecular-weight impact copolymer, often referred to in the vernacular as a 2-melt (2 g/10 min) 12-Izod copolymer. This part was launched in just such a compound and was successful.

At some point, the molder was persuaded on the strength of some notched Izod results and an incomplete data sheet to change to a material with a nominal melt-flow rate of 15 g/10 min. The decision was not a casual one; the two data sheets showed virtually no difference in strength or modulus and a decline in Izod from only 12 to 11 ft-lb/in. As expected, productivity increased. However, within four months chutes began to come back from the field with cracks.

At first, improper processing was suspected. However, notched Izod and instrumented falling dart impact tests both showed that the 15-melt material was quite ductile. While it did not put up the same numbers as the original material, they were not different enough to cause any concern. In addition, melt-flow-rate tests showed a mere 7% increase from pellets to parts. Table 2 shows the average values for both impact tests. The Izod impact results for the 15-melt material were only 12% lower than the values for the 2-melt resin, and the total energy of the falling dart test was reduced by 14%. In addition, all breaks were ductile.

The first indication of a subtle difference between the two materials can be seen in the detail from the falling dart test. The instrumented striking device used in this test monitors the energy collected during the impact test and plots it so that it is possible to separate the energy generated before and after achievement of maximum load. In most materials, it is instructive to compare the percentage of total energy expended before and after maximum load. The greater percentage used after maximum load, the better the energy management provided by the material.

As the Chains Unravel

A more significant problem became apparent when tensile tests were performed. While tensile tests are run at a significantly slower strain rate than impact tests, the tensile stress-strain curve provides a wealth of information about the full spectrum response of a material from the elastic region to ultimate failure. Table 3 shows the average results of these tests. They confirm the equivalence of strength and stiffness that the data sheets advertised. But they also provide numbers missing from the data sheets—the strain values associated with yield and break. The elongation at break is a relative measure of ductility, and in this test, the 2-melt material exhibits almost three times the elongation. But perhaps even more important is the difference in elongation at yield. One of the shortcomings of traditional impact tests is that they measure the response from a single catastrophic event. But impact failures seldom occur in this manner in the real world.

Instead, impact occurs repeatedly and the first 10 or 20 events may not cause complete failure. However, they do raise flaws in the product that act as notches. These flaws dictate the effects of subsequent impact events until finally a complete overload occurs. A material with a yield point of less than 7% experiences the initial stages of damage sooner and more frequently than a material with a yield strain of more than 11%.

Fundamentally, thermoplastics consist of an entangled network of long chains. Failure occurs when sufficient energy is applied to the system to disrupt this entanglement. Longer individual chains found in higher-molecular-weight materials are more thoroughly entangled, so more energy is required to create this disruption.

Visible failure occurs when the number of events is sufficient to pull apart the matrix. All things being equal, a lower-molecular-weight material with shorter strands will experience more progressive damage than a high-molecular-weight material. It is for this reason that so-called equivalent performance based on short-term properties cannot be relied upon to predict long-term behavior.

Contact information Dickten & Masch Mfg. Co. Nashotah, WI; Mike Sepe (262) 369-5555, ext. 572 dmlab@execpc.com www.dicktenplastics.com