Sponsored By

March 28, 1999

10 Min Read
The Materials Analyst, Part 18: More impact problems—when speed kills


This series of articles is designed to help molders understand how a few analytical tools can help diagnose a part failure problem. 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. He has provided analytical services to material suppliers, molders, and end users for the last 10 years. He can be reached at (414) 369-5555, Ext. 572.

In this day of almost rabid cost cutting, the mantra of faster-better-cheaper that has been popularized by the business gurus of the world has intruded into every aspect of product manufacturing. While this new strategy makes for great copy in the magazines, the story of the cost penalties that arise from an indiscriminate application of it often goes untold. This is due to the fact that the people counting the money hardly ever get down to the production floor where the havoc created by the new strategy plays itself out. The case study for this month is a testimony to what happens when things have finally gone as fast as they can go.

Most of us who have worked with plastic materials for a while have become aware, either intuitively or scientifically, of the fact that these materials behave in unique ways. One of these unique characteristics is an attribute known as rate-dependence. If we apply a load at different rates to a plastic material, the response changes. In the molten state, we take advantage of this property to mold wall sections that might be thought of as impossible to fill. The faster the material flows, the lower its viscosity becomes-an almost magical property we term "shear thinning." In fact, it is impossible to quote a viscosity number for a plastic melt without at the same time mentioning the shear rate at which it was measured. Shear rate is directly tied to flow rate.

In the solid state, the behavior becomes subtler but is no less detectable. Most of us understand that if we come home with a large load of books or groceries in hand and we set them down on a plastic beam such as the top of a television cabinet, the plastic will deflect. If we remove the load in a short period of time, the cabinet top will probably return to its original shape. But if we leave the load in place for a prolonged period of time, we may find that upon removal there is a permanent deformation in the plastic. Those of us who work in the field call this creep. Many of us have also learned that if the plastic is warm, this same deformation will occur in a shorter period of time.

While we may not be surprised to see the permanent bend in the cabinet top, we would probably be a little startled if the plastic cracked under the sustained load. But if we took the same load that produced the deflection, raised it to a sufficient height, and then dropped it, we could certainly imagine a point at which the material would break. What changed? The books or groceries weighed the same in both scenarios, so the load did not change. But the rate at which the load was applied did change. Shear thinning, creep, and impact failure are all manifestations of the rate-dependence of plastic materials. It is a pervasive characteristic of these wonderful materials with which we make our livings. Work with it, and new frontiers are possible. Defy it, and rising costs are guaranteed.

Failure During Assembly
Our case study for this month involves PVC, a material that can be especially rate sensitive at room temperature. The product being made from this material showed no signs of field problems, but the parts were failing during assembly. There were several large parts in the assembly. When the material came in, we performed the usual checks for composition that we have outlined in many past articles. We also ran the melt flow rate test comparing pellets to parts. Everything checked out fine. Yet a product that had been functioning well was suddenly having problems getting through the assembly process.

We decided to perform a direct evaluation of the impact properties of the material. The impact test we use is not the traditional notched Izod test. For reasons too numerous to go into here, the Izod test is not suitable for evaluating the true ability of a material to manage an impact load. Instead, we employ a falling dart configuration with a striking device-called a tup-that is instrumented so it records the impact event. Rather than obtaining just a number at the end of the test, this method gives a picture of how the material fails.


Figure 1. Instrumented impact results on PVC at high speed.

In addition, this method allows the velocity of the impact to be adjusted as part of the test procedure. While this test is gaining in popularity, and some material suppliers are starting to use it in addition to the Izod test, it is still a long way from becoming the industry standard. Most people have never seen the results of such a test or had to wrestle with interpreting the data.

We started with an impact velocity of 11.1 ft/sec, the speed preferred by the ASTM procedure for this test. The parts were large, so it was easy to cut several flat samples from the nominal wall and subject them to the impact test. Figure 1 shows the graphic result for one of the five tests. We are looking at two properties graphed versus time. The red line is the load. This starts at zero, rises to some maximum value as the tup encounters the sample and meets increasing resistance. At some point, the sample can no longer sustain the increasing load, and the sample begins to fail. The load drops off, eventually returning to zero.

The blue line is the energy collected from the sample. It rises from zero as contact is made between the sample and the tup. As the load increases, the rate of energy accumulation increases. Finally, when the sample fails, the energy accumulation stops, and the test is complete. The whole process is over in about three milliseconds, and the energy at the end of the test-in this case just over 15 ft-lb-is reported as the impact strength of the sample.

There is much more to it than that. The way in which it fails is just as important as the number. In any impact test, there are two components. The first is the energy required to initiate a failure. This is represented by the energy needed to get to the load peak, and it has more to do with the rigidity of the material than with true toughness. A very stiff material will resist deformation until a substantial load is applied. The real test of impact resistance, or energy management as some have called it, is what happens after maximum load. As the tup starts to pass through the sample, how is the material responding? Materials with a brittle tendency will fail rapidly past the maximum load point. Sharp pieces will fly out of the sample, cracks will propagate away from the impact point, and energy absorption past the maximum load point will be minimal.

We can see this behavior in our PVC part. While the part absorbs 15 ft-lb of energy, more than 14.5 ft-lb was used to get to maximum load or crack initiation. Beyond that point, failure was catastrophic. Another characteristic of materials with a brittle tendency is the inconsistency of the energy absorption. While the average of five samples was 13.5 ft-lb, the individual samples ranged from 8 to 19 ft-lb. But they all showed the same brittle tendencies.

Temperature and Time
So was this an indictment of the material? Was the material supplier lying when it quoted a very respectable 8 ft-lb/inch on the notched Izod test? Or did the formulation get away from them on this lot in some subtle way that we had not detected? There was a clue in the data sheet.

The Izod test had also been performed at 0F, and the results were not nearly as flattering. The value fell to 1.5 ft-lb/inch. This caused us to think about the rate of the test. One of the not-so-obvious rules arising from this business of rate-dependent behavior is that plastic materials do not distinguish between changes in temperature and time. The same things that occur over a long period of time at low temperature will occur at higher temperatures in a shorter period of time.


Figure 2. Instrumented impact results on PVC at lower speed.

Remember the television cabinet? At room temperature, the cabinet top might take a week before it deforms by a certain amount. But if the TV were on the entire time and the cabinet temperature were 110F, the same deformation would have occurred in less time. In other words, extending the time of loading at a fixed temperature has the same effect as increasing the temperature for a fixed time period. The plastic behaves the same either way.

Because rate is the inverse of time, higher strain rates mimic the effect of reducing temperature. This means the decline in Izod impact at lower temperatures might be duplicated by leaving the temperature constant and increasing the rate of the test. The rate of our test was probably much faster than that of the Izod test. So even though we were testing at room temperature, the high rate of speed was imitating a lower test temperature.

We decided to slow down the rate of our test from 11.1 ft/sec to 5.1 ft/sec. Figure 2 shows the result. Notice that now the load curve is very symmetrical, and the peak is rounded. At this speed, the polymer matrix has time to manage the energy, distributing it throughout the product. The material turns from brittle to tough.

Figure 3 shows a photograph of two specimens tested at the different rates. At the lower speed, the material shows stress whitening, a textbook indicator of ductility. No material has been removed from the sample. At the high rate of speed, a large section of the sample is blown out. A crack runs all the way from the impact zone out to the edge of the test piece. By stretching out the time frame of the test, we had completely changed the response of the material to the event. Not only was the material more ductile, but the total absorbed energy rose from 13.5 ft-lb to 22.2 ft-lb. The variation that had been so troublesome was reduced to a total of less than 1.5 ft-lb from best to worst in a five-specimen group.


Figure 3. Ductile (left) and brittle (right) failure of PVC.

How did this feed back to the customer's problem? The assembly operations involved inserting some metal components that produced a high level of momentary stress. In an effort to increase productivity, the rate of insertion had been increased in order to speed product down the line and reduce cost. In doing so, the rate threshold between ductile and brittle behavior had been crossed, and the material began to crack intermittently. While not all materials are this sensitive to the rate of applied strain, some of the more commonly used materials are notorious for this type of behavior. Among these are polypropylene, polyethylene, and many grades of PVC.

Just as it is impossible to characterize melt viscosity without referring to a shear rate, it is also impossible to characterize toughness without taking into account the impact rate. In some cases, faster is neither better nor cheaper.

Sign up for the PlasticsToday NewsFeed newsletter.

You May Also Like