The Materials Analyst, Part 55: Using the melt flow rate test to optimize the molding process

The qualification process for new injection molded parts has become substantially more formal and complex with the introduction of ISO and QS standards. In spite of all the new detail, one key component of a successful product still falls through the cracks—preservation of the molecular weight of the polymer used to produce the part.

A molding process for a new part is developed by some system of knowledge that combines past experience, material supplier recommendations, and the constraints imposed by the design of the part and the mold. This process has consequences for the changes in molecular weight that occur in the polymer as it travels through the process from pellets to parts. If the process employs aggressive conditions such as an elevated melt temperature, or if the machine selected for the job has an injection unit that holds a large amount of melted material in inventory relative to the shot weight, the resulting quality of parts may be compromised by polymer degradation.

Some materials, like PE and PP, are quite robust and can tolerate a wide range of processing conditions without experiencing excessive reductions in molecular weight. But many high-performance materials such as nylon and polyester are much less thermally stable. Small increases in barrel residence time and melt temperature can produce significant reductions in the molecular weight of the material. In addition, these material families can react with absorbed moisture to produce further reductions in the average length of the polymer chains.

One of the most challenging materials in this group is PET polyester. PET has a thermal sensitivity typical of high-performance semicrystalline materials; the processing window for GP grades in this material family is approximately 80F. For grades that contain additional ingredients like flame retardants or impact modifiers, the range may be even narrower.

In addition, PET undergoes hydrolysis more rapidly than any other commonly used melt-processible polymer. PET suppliers recommend a maximum moisture content of .02 percent at the time of processing. To make matters worse, the traditional cosmetic evidence for the presence of excessive moisture—splay or silver streaking—does not appear in parts molded from wet PET.

There are several methods that can be used to evaluate the molecular weight of a polymer before and after molding. The melt flow rate test is the simplest and most accessible method for processors. This test works extremely well for unfilled materials that undergo straightforward chain scission as they degrade. PE, PP, and PC are just a few of the materials that can be readily evaluated by the technique. Pellets are compared to parts, and the increase in melt flow rate during processing is measured.

Depending on whom you talk to, increases between 30 and 40 percent represent satisfactory preservation of molecular weight. Above this point the process becomes suspect, particularly if parts fail during qualification testing. Below these threshold values part failure can usually be ascribed to other issues such as part design.

Robust designs allow processors to get away with more polymer degradation. We have tested good parts that show a 70 percent increase in melt flow rate from pellets to parts. The parts worked because the part design was excellent relative to the stresses of the application. But with cost reduction the order of the day, thinner nominal wall sections and the use of materials with a reduced and more variable property profile make it more likely that an abused material will cause problems in the field.

Deceptive Fillers
Using the melt flow rate test for PET polyesters can be a problem because almost all commercial grades of PET contain substantial filler. Glass reinforcement can range from 15 to 55 percent, and a series of glass fiber and mineral hybrids complicates the picture even more. The presence of these fillers and reinforcements makes the use of the melt flow rate test more difficult because these inorganic materials suffer a certain amount of mechanical damage during high-shear processing.

As glass fiber lengths become shorter and mineral particles become smaller, they contribute to the viscosity reduction that is normally observed in a molded part. Viscosity reduction is measured as an increase in the melt flow rate. Experience with running the melt flow rate test on highly filled materials has shown that the changes due to filler attrition can dwarf the changes due to polymer degradation. This has led the industry to fall back on other methods that either minimize or eliminate the contribution of the filler in measuring viscosity.

There are two approaches designed to provide a relative measurement of average molecular weight that skirt the problems associated with the melt flow rate test. The first makes use of a capillary rheometer, a device that looks a lot like a melt flow rate tester but has the ability to push the melt at a controlled rate. At high ram speeds, the fillers and fibers tend to orient, reducing their contribution to the viscosity measurement and making it easier to detect changes in the polymer.

An even better method involves dissolving a sample of the material and filtering off the glass and other fillers. The viscosity of the polymer-solvent solution is then compared to the viscosity of the pure solvent to come up with an intrinsic or relative viscosity.

Despite the greater fundamental accuracy of these methods, the melt flow rate test has remained popular because of its ease of use and the low-cost equipment. Capillary rheometers and solution viscometry equipment can cost 10 times as much as a melt flow rate tester and require a much higher level of training to operate effectively. In addition, the solution technique involves a wide array of very expensive and dangerous solvents that discourage most processors from getting involved in the technology. Many large end users also demand melt flow rate information on these materials because they are familiar with the measurement.

Reconciling Methods
Because of these considerations, a lot of work has gone into correlating changes in melt flow rate for highly reinforced materials with changes documented by these more fundamentally sound techniques. The resulting guidelines tend to be specific to a particular grade of material, but a good set of rules has been developed that can correlate melt flow rate increases in highly filled materials with part performance. This article chronicles the use of these rules by one processor to optimize its molding parameters beyond the normal requirements of appearance and dimensional tolerances.

The parts were being molded of a 30 percent glass-reinforced PET polyester. The material supplier had developed a set of guidelines for evaluating the integrity of the molding process using the melt flow rate test. The guidelines stated that parts could be considered good if their melt flow rate was no higher than three times that of the virgin pellets. Parts that increased between three and four times the value of the virgin pellets were considered marginal—usable in some applications but possibly prone to failure in high-demand applications. Above this fourfold increase, parts were considered to be processed poorly.

The processor first supplied us with several lots of raw material. The melt flow rate for these lots fell in a very tight range around an average value of 10 g/10 min. Parts from the initial process had an average melt flow rate of 58 g/10 min, clearly outside the recommended boundaries for good part integrity (see Figure 1, right).

In order to address the possible causes of the large increase in melt flow rate, the processor first submitted two sealed containers of raw material, one from the original drying process used to mold the parts with the high melt flow rate and the second from a new drying process. The first sample, when tested by Karl Fischer titration, had a moisture content of 270 ppm (.027 percent). The second sample had a moisture content of 180 ppm (.018 percent). Parts molded from this second batch of material had a melt flow rate of 44 g/10 min. This was a clear improvement, but still above the marginal threshold of 40 and the ideal goal of 30.

How Dry is Dry?
More attention to the drying unit and the drying process resulted in a new sample of material with a moisture content of 120 ppm. Parts molded from this material on two different days had melt flow rates of 33.5 and 34.4 g/10 min.

This brings up a very important point about the relationship of moisture content and molecular weight retention. Just because the moisture content of a material is less than the maximum recommended limit does not mean that the condition of the material is optimal. Several studies have been performed over the years, particularly on PET polyester, to show that lower moisture contents within the realm of the acceptable range produce parts with better properties.

This was clearly the case here. Unfortunately, many processors are scared away from such thorough drying because moisture removal also increases the melt viscosity of the material. This makes parts a little harder to fill and detracts from the surface finish of parts molded in highly reinforced materials. Many processors invoke the notion of overdrying, a term that is meaningless and has no foundation in scientific study.

The reality is that for materials like PET polyester, which can be broken down by moisture, drier material results in better parts. It also insulates the polymer against the effects of elevated barrel temperatures and extended residence times. In this case the processor gained significantly by reducing the moisture content even though the prescribed moisture content had already been achieved.

With the drying process addressed and the melt flow rate still above 30, this processor now turned his attention to barrel temperatures and residence time. A reduction in the barrel temperature settings and the resulting melt temperature from 565 to 545F produced two separate samples of molded parts with melt flow rates of 25.1 and 24.7 g/10 min.

A month later, having achieved the desired objective, the processor moved the job to a new machine with a smaller injection unit. Triplicate trials over a period of 45 days produced results of 23.2, 20.2, and 19.7 g/10 min. Figure 1 shows the progress of this development. The entire process took approximately five months, although adequate control over the process was achieved in less than three months.

Covering All Bases
This may seem like an inordinate amount of work. But the insurance that this provides the molder should not be underestimated. First, it gives insight into the influences that cause degradation in a polymer such as PET. The interactions between moisture content, melt temperature, and residence time are complex and often poorly understood. If these fundamentals are not addressed, the vintage of the machine and the sophistication of the controller count for very little, and all the ISO procedures in the world will not change the fact that the material is degraded and the parts are brittle.

Beyond this point, it is important to understand that parts in the field are exposed to application environments that can also result in degradation. Once the presence of degradation has been established in a failed part, one of the hardest things to accomplish in material analysis is determining when it occurred. In the absence of evidence to the contrary, the burden of proof and the assumption of responsibility usually falls on the molder. If the molder cannot show some documentation that the process was under control as it relates to molecular weight retention, then it is assumed that it was not. It can take a lot of effort to undo that perception.

So the next time you are filling out that PPAP documentation, consider the importance of documenting the preservation of molecular weight in that process you are certifying. If molecular weight is not preserved, then all the other considerations of product quality are meaningless.