Plastic Material Selection: It's Not About Inventory

In relation to this article, you can apply for membership to the Prospector Plastic Material Selector.

By Michael Sepe, Materials Analyst

It's natural to want to use a tested material that's already being run in your plant, but it has to make a good part.

When a product fails to perform up to expectations, particularly during the new product development cycle, material selection is often the root cause.

Where are several levels to the material selection process. The first is choosing the correct material family: polypropylene, polycarbonate, acetal, etc. Within these general families there are more specific enhancements that can be made. Glass-fiber reinforcement adds strength and stiffness, impact modifiers increase low-temperature toughness, and stabilizers may be added that ensure longer life under specialized circumstances such as prolonged exposure to elevated temperatures or outdoor environments.

Often these characteristics are implied in the specification of a particular grade of material. However, increasingly OEMs are calling out a general description and leaving the selection of the specific material grade in the hands of the molder. This practice arises in part from the assumption that the processor may understand more about the fine points of grade selection and may be able to select from multiple suppliers who provide “equivalent” offerings.

However, this practice is also an attempt to leverage lower material costs. While the contribution of raw material price is very small in most consumer products once they reach the store shelves, it represents approximately half of the selling price of the molded part to the immediate customer and therefore receives a lot of attention at this point in the supply chain.

One of the key components of the material price to the molder has to do with volume. If a processor is using a particular grade of material and a new product comes along that specifies the same general class of material, it is common for the molder to recommend the grade in which it already has a strong position.

While this makes economic sense, it often neglects key aspects of performance. When a product fails and the root cause appears to be related to material selection, one of the first questions that should be asked is, “How did you arrive at this particular grade of material for this application?” If the answer is something like, “We run silos of this stuff for our other customers,” it is time to take a step back and look at the selection on its merits.

Below are three case studies in which the solution required that the molder put another item in inventory for the sake of product performance. In all three cases, the characteristic that was neglected in the original selection was the molecular weight of the polymer.

THINK LONG-TERM
We have covered the importance of polymer molecular weight to part performance a number of times over the last decade in this column, but it continues to be a major factor in product failure. Plastic materials derive their properties from the large size and extended chain shape of the molecules that make up these materials. A certain minimum molecular weight is required in order for the material to be considered a polymer in the first place. Extending the chain length beyond this minimum produces improvements in performance derived from something known as chain entanglement. The longer the individual chains become, the more entanglement results.

These improvements can be difficult to document if one only considers the short-term properties available on most data sheets. Properties such as tensile strength, flexural modulus, and heat deflection temperature may not change as a function of chain length. Even properties that do change, usually those associated with impact, are altered by what seem to be minor percentages. However, studies have shown that long-term properties such as creep resistance, fatigue resistance, and environmental stress crack resistance benefit substantially from longer polymer chains.

This should get our attention as an industry because failure to perform adequately in the arenas of fatigue, creep, and environmental stress cracking accounts for approximately 80% of all field failures of plastic products. Unfortunately, what usually dominates the discussion regarding molecular weight among processors is how the polymer flows. As the chains get longer and more entangled, it takes more energy to move them even when they are in the molten state. This energy requirement is measured as viscosity; higher-molecular-weight materials exhibit higher melt viscosity and can therefore be more challenging to process.

The industry uses a simple test to quantify this difference in viscosity called the melt flow rate (MFR) test, and material suppliers who offer their materials in a wide range of MFRs are essentially providing a choice for the processor and the OEM. This choice represents a trade-off between processibility and product performance.

Polycarbonate is one of those materials that are supplied with a wide range of MFRs, from low single digits to nearly 100 g/10 min. All of these grades have their place, but often they are selected inappropriately because of a lack of awareness of how that MFR number is connected to long-term performance.

Table 1 shows some select properties commonly available from a material supplier for five grades of polycarbonate. A review of these numbers makes it fairly clear why those involved in the business of material selection miss the significance of the differences between these materials. So when a choice comes up between using something already in inventory to the tune of tens of thousands of pounds or ordering a minor quantity of a new and untried grade at a significant increase in cost, the molder can be forgiven for taking the path of least resistance.

CONSIDER THE PRODUCT ENVIRONMENT
Polycarbonate is profiled here because two of the three product failures involve this polymer. The first application involves continuous and rapidly changing rates of cyclic loading in a part first produced in the 25-melt material listed in the table. The part failed in between two and three days of testing, far short of the thousands of hours of service required in the field. The part was then produced in the 10.5-melt flow product and survived beyond the range of test times required for the application, an improvement in fatigue resistance of more than one hundred fold.

The rationale for using the 25-melt resin in the first place was not related to processibility. With minor adjustments the part was moldable in either material. The primary reason for selecting the 25-melt material initially was related to availability at the molder's facility based on usage for other applications.

A second polycarbonate application experienced stress crack failures when produced in a material with a nominal MFR of 20 g/10 min. Alternate grades selected all fell in the MFR range of 19-22 g/10 min, and not surprisingly all of these materials produced parts with similar performance problems.

Here again, the selection process was based on the molder's inventory. The grade or grades being used initially were already mainstream products in the plant where the product was being molded. While these grades of material were suitable for other applications, they were not adequate for the rigors of this particular part. Here again, a change to a material with a nominal MFR near 10 g/10 min provided a part with greater resistance to stress cracking, despite the apparent equivalent performance illustrated by the available data sheets.

PERFORMANCE COMES FIRST
The third application also involved stress crack failures, but the material family used to produce the part was high-density polyethylene (HDPE). This component was a rather large fluid reservoir for an automotive application, and was being produced from a grade of HDPE with a nominal MFR of 4 g/10 min.

Polyethylene materials provide for a greater level of property manipulation than most resins because, along with variations in molecular weight, there is the option of changing the density of the polymer. However, altering the density of the material also changes the balance between load-bearing properties and toughness and requires a more extensive evaluation process. Changing the molecular weight of the material, as we have seen from the table of polycarbonate properties, does little to alter this balance.

In the case of the fluid reservoir, a design review had shown that there were numerous problems with the part design. Stress concentrations were plentiful due to closely spaced ribs and lack of proper radiuses between ribs and nominal walls. The locations of the cracks that developed in the product were clearly related to these deficiencies.

However, this product had been launched and was already eight months past introduction when the problems first appeared in the field. A solution was needed to make the product more robust immediately while design improvements were drawn up. This quick fix was a change to a material with a nominal MFR of 1.2 g/10 min. Initially, the molder predicted that the part would not fill with a material of such a low MFR. However, the pressures required to fill the part did not increase significantly.

The redesign work took nearly six months to complete, and this was followed by the dilemma of whether to shut down production while the tool adjustments were made or to build new molds at a considerable additional cost. However, it was noted that the failure rate from the field had declined rapidly over the ensuing period, and this decline was related directly to the introduction of the higher-molecular-weight polyethylene. While the design problems still existed, the more robust raw material had essentially overcome the weaknesses caused by the poor design, and ultimately no action was taken either to build new tools or to correct the existing molds.

The lesson here is that while logistics and inventory management are certainly important aspects of any business, performance still comes first. This consideration will become increasingly important as manufacturing expands to a truly global scale and the problem of availability of the correct material becomes a question of obtaining the same performance wherever the product is made. Ultimately, the economics of making parts that work is more compelling for both the engineer and the accountant than making parts from materials that happen to be in the warehouse.

August, 2006 - Reprinted with permission from Injection Molding Magazine. Copyright © Canon Communications LLC.