The Materials Analyst, Part 39: The hidden effects of color (Part 1)

One of the major benefits molded plastic has over most competitive materials is the ability to mold in the desired color of the product. Even in markets such as automotive that have traditionally painted large plastic parts, molded-in color is an attractive goal that can reduce cost and eliminate regulatory issues. However, the degree of additional difficulty that adding color introduces to the plastic molding process is often not understood. This month we will highlight one unanticipated effect of adding color to a product. Next month we will discuss a second common problem.

   This month' s case involves a thin-walled product for the handheld electronics market. This project involved a set of parts designed in polycarbonate. In addition to a clear and a black product, the OEM wanted to produce the parts in transparent blue, red, and green. The products were specified in a medium-viscosity polycarbonate with a melt flow rate of 9 to 12 g/10 min. The OEM was familiar with the negative effects that abusive processing can have on the properties of polycarbonate, and was also aware that some of the design features in the parts left something to be desired. Consequently, the specified increase in melt flow rate from pellets to molded parts was not to exceed 20 percent. 

In past articles we have talked about the rules regarding good processing and have quoted an increase of 30 percent as a good boundary to follow. However, it is a fact that as the nominal wall of a part becomes thinner, design properties such as impact are reduced. The polymer must have greater integrity to perform the same function and molecular weight retention becomes that much more important. In this case, impact tests suggested that the traditional marker of 30 percent was not going to be good enough.

   This places the processor in a difficult spot. In order for parts to function at the desired level, the molecular weight of the material must be adequate. But the demands of filling a thin-walled part make the use of a high-molecular-weight material challenging. Processors get around this in one of two ways. First, they gravitate towards a material with a higher melt flow rate. Unfortunately, this can cause a significant decline in properties, particularly impact strength. The other route is to use the higher-molecular-weight material but process it aggressively. Most often this entails using higher melt temperatures to reduce viscosity, a technique that is effective with polycarbonate because it has good thermal stability.

Problems With Colors
As this particular product line launched, different parts presented challenges as a result of their configuration, but a pattern of problems began to emerge in meeting the desired melt flow rate specifications. When molding the unpigmented clear material, the molder had no trouble maintaining a melt flow rate increase well below the target of 20 percent. But to varying degrees all of the colors, particularly the tinted transparent ones, were hard to manage. Moisture checks on the raw material confirmed that it was being dried to appropriate levels. Repeated measurements with a Karl-Fischer-based moisture analyzer showed that the material was consistently below .02 percent (200 ppm) at the time of processing.

   Using one of the more difficult parts in the assembly, technicians examined the melt temperature as the next possible cause of an excessive reduction in the molecular weight. Parts were molded in four colors using the standard melt temperature. The melt temperature was then reduced as far as possible without producing short shots and a new set of samples was produced. Parts from both processes were examined for increases in the melt flow rate, with pellets from the specific lot being sampled serving as the basis for the calculations. Table 1 shows the results in terms of the percent increase from pellets to parts.

   The reduction in melt temperature clearly benefited all of the colors, with some improving more than others. At the lower temperature two of the four colors had been brought into the desired range, but the other two were clearly still in trouble. In addition, the pressure required to move the material at the lower melt temperatures was becoming a problem.

Residence Time
The problem in this case was one that negatively affects a lot of processes. A relatively large barrel was being used to process the material. Custom molders tend to purchase machines with larger barrels just in case a job comes in for quote that requires a larger shot size for a given clamp tonnage. This does two things that negatively affect the processing of smaller shot sizes and thin-walled parts. First, the larger barrel increases the residence time of the raw material in the barrel. This acts to increase the effect of the melt temperature on the molecular weight of the raw material.

   Second, shot-size capacity is achieved for a given size of injection molding machine by increasing the diameter of the screw and barrel. The diameter of the ram that generates the hydraulic pressure required to move the screw does not change with the barrel diameter on a majority of the machines built today. Therefore, as the barrel becomes larger the mechanical advantage (the intensification ratio) that translates hydraulic pressure into plastic pressure decreases.

   It is possible to have an injection molding machine of a given clamp tonnage that will generate 30,000 psi of plastic pressure and have another machine with the same specifications but a larger barrel that can only generate 20,000 psi. Many molders do not even notice the difference, but if you are involved in thin-wall molding it is impossible to compete without being painfully aware of this problem.

   There is an important side note here regarding the topic of residence time. If you ask most molders to calculate the residence time for material in the barrel, they will take the shot capacity of the injection unit, divide it by the shot size for the job, and multiply that quotient by the cycle time. For example, if we have a 24-oz machine running a 6-oz shot on a 30-second cycle the calculation will be that there are four shots in the barrel and the material is in the heating cylinder for 2 minutes. This gives processors a false sense of security and leaves them scratching their heads when their materials degrade.

   There are two problems with this calculation. The first is a failure to account for the differences in specific gravity between polystyrene and the resin that is being processed. Shot capacities are rated for polystyrene, with a specific gravity of 1.04 g/cu cm. If the molder is running a 30 percent glass-filled PET with a specific gravity of 1.57 g/cu cm, the barrel holds 50 percent more material. This, in turn, adds 50 percent to any calculation of residence time.

   But the biggest error comes from the assumption that the barrel can only hold the amount of material listed in the shot capacity specifications. In fact there is a substantial amount of additional material in the screw flights and much of it is molten. In order to obtain a realistic calculation of residence time, introduce a material of substantially different appearance to a feed throat where the screw flights are just becoming visible while the machine is on cycle. Then stand back and count the number of shots before that material appears in the molded part. You may be surprised to find that it takes two and a half to three times longer than you expect. This is the true residence time. If you add the two sources of error, the supposed 2-minute residence time can be as high as 7 minutes, and for many materials this is the difference between life and death.

Material Stability
Back to our problem. Traditionally, polycarbonate is a material with excellent thermal stability in the melt. While the improvements the processor made in retained melt flow rate were obvious when the melt temperature was reduced, neither of the two temperatures appeared to be so aggressive that they would result in the types of shifts that we were seeing. In particular, the same process was leading to a very different result depending upon the color.

   Here is where the analytical part comes in (in case you were wondering). We have talked extensively about the use of the melt flow rate test to assess degradation in molded parts. But the same device can also be used to measure the relative stability of the resin itself. This is done simply by running the standard melt flow rate test on the pellets using the standard preheat time of 5 minutes and then repeating the test using a longer preheat time. In this case we used a 20-minute preheat. The extended preheat time simulates an increased barrel residence time. It cannot simulate the effects of shear, and the specified temperature for the melt flow test is only 300C (572F) so it is hardly an aggressive treatment of the material. If problems appear during this type of evaluation, it is virtually certain that they will only get worse in the real world of processing.

  This melt stability test was run on all four colors and the unpigmented clear material, and the results are a little startling (see Table 2). The clear material lives up to the reputation that polycarbonate is a material with excellent thermal stability. But the addition of color has a significant negative effect on this thermal stability. In the case of green and blue, just exposing the material for an additional 15 minutes to a temperature of 300C was enough to produce a change in the melt flow rate that was greater than the end user' s specified allowance. No wonder the increases in melt flow rate associated with the process were so severe. The problem was not with the process; the thermal stability of the polycarbonate was compromised significantly by the introduction of colors.

   This is not the type of revelation that is welcome in the middle of a product launch, especially in the fast-paced world of personal electronics. However, once the molder understood the nature of the problem, it recovered admirably. Moving the molds to machines with smaller barrels reduced residence time and provided the additional injection pressure needed to fill the parts at lower melt temperatures. Some work was also required on runners and gates to modify the shear rates being generated in the molds. Tests comparing the melt flow rate of runners and parts showed that some of the degradation was shear induced.

   Once these changes were in place, the molder was able to turn out parts that consistently exhibited shifts in melt flow rate that were well below the target value of 20 percent and in many cases were less than 10 percent. He continues to do so months after initial product introduction.

   This study illustrates how the introduction of color can alter the traditional properties of a raw material. The tests to uncover this problem are quite simple, but an understanding of the problem and its relationship to the real world of processing, mold design, and part design are critical to a solution. In this case the molder overcame a material compromised by pigments and a melt flow rate specification that many consider too tight. In the process, it learned important lessons about material behavior and process control. Next month we will look at a different type of pigment effect on a semicrystalline material—polypropylene.