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. Mike has provided analytical services to material suppliers, molders, and end users for the last 12 years. He can be reached at (414) 369-5555, ext. 572.
One of the hottest markets in injection molding today is the overmolding of a flexible elastomer onto a rigid substrate to create the soft-touch feel in a finished product. The final component can be created using a traditional insert molding approach, where the rigid material is placed in the mold and the elastomer is shot over it, or it can be done as a two-shot process. Regardless of which process is used, the success of the final product depends upon the bond created between the rigid substrate and the elastomeric surface. This bond can be achieved by mechanical means, but the most successful products rely on a true adhesion of the two materials.
While the heat and pressure of the molding process are obviously a great help in promoting this adhesion, a good deal of the needed resistance to peeling and separation is a matter of compatible chemistries. Elastomers based on olefins like polyethylene and polypropylene, the so-called TPEs, will naturally bond well to substrates made of polyethylene and polypropylene. Styrenic-based elastomers will bond well with styrenic rigid materials like general purpose and high-impact polystyrenes and styrene-butadiene copolymers like K-Resin. Flexible nylons will hang on for dear life to any rigid nylon, and PBT polyester-based elastomers like Hytrel and Riteflex have an affinity for, you guessed it, rigid PBT polyesters.
There are, of course, some less obvious unions that result in products with an excellent bond. Polyurethane elastomers, for example, bond very well to polycarbonate. And polymer chemists have managed to adapt the chemistry of some more popular, cost-effective elastomers so that they will adhere to materials like nylon.
In general, however, the range of products that can be mixed and matched with ideal results is limited. Consequently, when contamination occurs in one of the two materials in the assembly, a very sensitive balance can take a turn for the worse. This was the case with a product that was sent to us for examination after a successful soft-touch product suddenly developed problems.
The part was a knob with a rigid substrate made from a 30 percent glass-fiber-reinforced polypropylene. The elastomer, which was applied by an insert molding operation, was actually based on polystyrene. However, a substantial part of the chemistry in this particular grade relied on an ethylene-butylene component that had good affinity with the polypropylene. The part had been working without problems when quite abruptly a particular lot of parts failed to produce any bonding at all. In the part that was sent to us it was very easy to rotate the elastomeric covering and slide it off of the polypropylene base. We were also provided with two samples of virgin PP from two different lots. One retain came from a good lot that had been manufactured before the problems arose. The other sample came from the lot used to make the parts that were failing.
DSC Finds the Problem
We began by running a DSC test on a piece of the rigid molded part and quickly came upon the problem. As the material was heated in the DSC we detected the typical melting point for the polypropylene near 164C (327.2F). But as we continued the heating process, we found another ingredient that did not belong in the molded part. The second melting point at 221C (429.8F) belonged to a nylon 6. Judging from the intensity of the curve related to the melting of the nylon 6 and the dilution of the polypropylene melting event, it appeared that the nylon 6 constituted 40 to 60 percent of the molded part.
Figure 1 shows the result of the DSC. The elastomer would have no affinity for the nylon, explaining the problem with the final product. We also ran a TGA to check for filler content, assuming that the nylon that had found its way into the mix probably did not have the same filler content as the polypropylene.
Ideally, we hoped that the nylon contaminant would be unfilled, and the reduction in glass fiber content would allow us to determine precisely the extent of the contamination that could only be approximated by the DSC results. Instead, we found that the filler content was actually within specification at 32.5 percent as can be seen in Figure 2. The polymer decomposition part of the TGA does not show the presence of the nylon because both materials decompose at similar temperatures.
This led us to one of two possibilities. First, the grade of nylon contaminating the polypropylene contained the same amount of glass fiber as the polypropylene. This is rare, but it does happen. If you go all the way back to the second article of this series we talked about a case where a 30 percent glass-reinforced PPO was contaminated by a 30 percent glass-reinforced PET polyester. In that case, our client had performed its own glass content determination as a first step in investigating the reasons behind a brittle product and had come up with no deviation from the specification.
The other possible scenario was that the polypropylene and the nylon had been mixed together prior to compounding and that the correct amount of glass had been added to an already contaminated polymer. This would have shifted the attention from the molder to the compounder, a company that worked extensively with both nylon and polypropylene.
To determine the source of the contamination, it was necessary to turn to the two lots of raw material that had been provided. We first examined the pellets themselves. It is rare that two materials have exactly the same pellet appearance. Something always varies such as pellet shape or gloss or the way the ends are cut. If all of the pellets look the same, then this all but eliminates a physical mixing of two materials.
Both samples were of uniform appearance. Even with this assurance, we did not want to perform our analysis on the as-received pellets because we were still concerned about a nonuniform mixture of the polypropylene and the nylon. DSC samples are very small, usually less than the size of one pellet. If the contamination is not homogeneous, a sample can look good simply because the wrong fragment was selected for testing. For this reason, we always prefer a molded part to a sample of raw material when looking for contamination because the part is made up of many pellets that have been blended into a more or less homogeneous mixture.
Improvising to Pinpoint Cause
The sample we had been given did not give us enough resin to purge a molding machine properly and then mold test samples. So we did the next best thing. We used the melt-flow tester as our mixer. We were not looking for a quantitative viscosity value in this case; we simply wanted to put 15 to 20g of material through the heated barrel and produce an extrudate that would represent a larger sampling. We also used the low shear and precise temperature control of the instrument to our advantage. Instead of running the material through the instrument at the conventional temperature for polypropylene of 230C (446F), we reduced the temperature to 190C (374F).
A look at Figure 1 will tell you why. At 190C a clean polypropylene will still flow freely through the instrument because it will be completely melted. In fact, at least one European polypropylene supplier performs its melt-flow tests at 190C. However, nylon 6 is still solid at this temperature and should not flow. If we had run the test at 230C, both polymers would be molten. So as we produced our mixture, we were also looking for evidence of solid material coming through our instrument or even stopping the flow altogether.
We saw neither. Figure 3 shows the DSC result for the extrudate produced from the suspect lot of material. It was a clean polypropylene—no signs of even a trace of nylon were found. In this case, the contamination had occurred in the processor’s plant and the presence of the nylon interfered with normal adhesion. It was a coincidence that the filler content and type in the nylon happened to be the same as that of the polypropylene.
As is often the case, the last step in a process like this is the isolation of the bad product. Obviously, it is not practical to run DSC tests on thousands of parts, nor is it necessary. As in that case of PET mixed with PPO, the unwanted material is almost always of a different density. In this case, a 30 percent glass-reinforced nylon 6 has a density of 1.35g/cu cm while a polypropylene with the same amount of glass has a density of 1.13. This will cause the part weight to vary well outside of the normal limits and provides a fast and nondestructive way of separating the good parts from the bad parts.
The emphasis on multistep processes such as overmolding, two-shot molding, and inmold decorating is driven by the need to reduce cost and provide a higher-quality product or impart a distinctive look to a product. But as the molding process becomes more complex, the need to eliminate the simple mistakes such as contamination becomes that much more important. The costs associated with these traditional mistakes are greatly amplified when they are made while running these advanced processes.