Part 1 of this two-part series introduced a fishbone diagram showing many of the parameters requiring control to successfully make injection molded parts. Part 2 offers more details on some of these, broken down again by the four main categories: machine, method, material, and man.
Platens. The condition of the platens and how they come together affects not only mold longevity, but also mold performance. Flat platens properly support the mold under clamp load and injection pressure. These can be checked with a large straight edge. Platens sag when worn and can move during clamp lockup, typically upwards. Check this with indicators on the mold and the indicator base on the frame or some other place that does not move or flex under tonnage. Using full clamp tonnage on a mold base that is too small can not only crush a parting line but also can allow the platens to wrap around the mold a little. When paying attention to minimum mold sizes, a rough rule of thumb is not to use full clamp pressure on mold bases covering less than half the distance between the tiebars. Clamping on molds that are too small can give you flash in a mold that might check out OK on the bench.
Reciprocating screws. Understanding how the screw design affects the resin being processed can help dramatically. As a rough rule of thumb, assuming a proper shot size, about 50% of the heat required to melt the resin comes from screw-generated shear. Nylons might get 80% of the heat from the screw. Semicrystalline resins require more energy to break down the crystal structures. Amorphous resins such as polycarbonates like long transition zones to avoid degradation. Barrier screws help separate the melted plastic from the unmelted plastic to prevent degrading and to concentrate the screw-generated shear energy on the unmelted pellets. You can have screws optimized within families of resins—for instance, different screws for HDPEs of different average molecular weights.
Custom molders have it hard in this regard because they need to process many different resins with typically a limited set of equipment. At a minimum, it would help to get a handle on the screw designs in your machines—usually the compression ratio, the percentage of the total length used by the feed, transition and metering zones, the metering zone depth, and the length-to-diameter (L/D) ratio. Many resin manufacturers will publish the best screw conditions for their resin. These guidelines are also available from some screw manufacturers.
As far as screw wear goes, a rough rule of thumb for new screws and barrels is that there should be 0.001 inch of clearance between the screw flights and barrel ID for every inch of screw diameter. Problems with worn screws or barrels include the melted resin slipping back over the screw flights for extra shearing and added recovery time to bring the screw back. Other things roughly being equal, slowly increasing screw recovery times on a job can indicate screw or barrel wear.
Barrels. If a barrel is having an ill effect on first-time-correct parts, it’s probably because it’s worn. Barrels don’t wear evenly, so don’t think that your barrel is OK just because you don’t have the equipment to measure along its entire length. Barrels wear for many reasons, and probably more often than we think due to temperature settings that are too low in one or more zones.
Nozzles. Nozzles are sized 1/32 inch under the sprue O-dimension to prevent slight mismatch from keeping the cold sprue from coming out. Remember the nozzle size when reviewing shear rates in your melt delivery system and that the highest-shear point for nylon nozzle tips is inside. A 1/8-inch nylon nozzle that matches a 5/32-inch (0.156 inch) sprue O will actually have a 0.090-inch orifice inside. Take advantage of the different nozzle types and specify the nozzle used for each mold so this does not vary from run to run.
Tiebars (where applicable). The amount all the tiebars stretch during clamp lockup should be within about 10% of each other. Above that, you might run the risk of putting excess strain on the molding machine, damaging a mold, or just flashing in one area of the part.
Oil. Just like in your car, the condition of the oil can prevent your machine from working correctly. Unlike your car that costs $29.95 to completely start over with fresh oil and filter, some machines hold a lot of oil, and it is expensive and time consuming to change. Make sure you have a good filter and, if you don’t already, consider getting the oil tested [http://www.plasticstoday.com/imm/articles/how-stop-varnish-it-costs-you] to know when it’s time to change.
Valves (where applicable). Valves are less of a concern if the oil has been properly maintained and the correct oil used. But over time these can varnish up and the increased or variable response time can affect the processing.
Mold/water temperature controls (thermolators)
Simply put, they need to have enough power to supply enough water at the proper temperature. For most types of molding, each waterline should only pick up 3-5 deg F between the “in” and the “out.” If you have a number of mold water circuits hooked up to a common source, like a manifold, bear in mind that the water will follow the path of least resistance. If all your waterlines are pressure-balanced so that no lines are starving due to pressure loss, and you are still picking up too much heat, you might need more psi at the control to push more gpms of water.
Water circuit hookups. With all the issues noted under mold/water temperature controls, don’t leave the task of plumbing the mold to be reinvented every time. Pick a way to document how the water gets hooked up and make sure it is done the same way every time.
There are many variables here, but a list of highlights for things that can vary shot-to-shot or prevent you from molding good parts immediately includes:
Pressure-balanced water circuits. This was covered a little under “Mold temperature controls.” Most shops run a different mix of molds every day, which means the draw on the water supply varies every day. Some days the water pressure available at an individual mold might not be enough to feed all circuits properly, but you need to monitor this to find out. This is certainly one of many gremlins running around when the mold techs say, “I set it up exactly the same way as last time but the results are different.”
Enough steel (or aluminum, etc.) wrapped around the cavities and cores to prevent excessive bowing, particularly on deep-draw parts. Among other things, molds are pressure vessels that withstand tens of thousands of psi of plastic pressure. Put an indicator on the mold side and watch it during injection and packing to find out if your mold is flexing too much. Do this in different areas to map the flexing. More than 0.003-0.005 inch could be a problem.
Side-to-side shifting of molds during molding. The press platens hold the mold shut in the line of draw. What if the part has an imbalance of projected area side to side? There are at least two things to watch for here and it is easy to do with indicators measuring from one half to the other. One is mold shift on clamp lockup, which tells you how well your parting line is shutting off. The other is a shift during injection and packing, which tell you how well your mold is resisting molding pressure. Your mold might have one or both.
Variations in resin dryness can cause a lot of headaches. A lot of newer dryers have features that help you dry resin properly, particularly when it comes to monitoring the supply and return air and reacting accordingly. One item to be wary of is the possibility that resin is not flowing through the hopper LIFO, or is “rat-holing,” where resin slips down a channel from the top of the hopper to the bottom. There are experiments you can run to check for this. Start with a layer of colored pellets placed on top of a full hopper, and then unload it at a rate intended to empty the hopper in 4 hours. This should get checked at least two ways—with the hopper being continuously replenished and running the hopper all the way out.
The airflow through the hopper should be checked for variations with the hopper empty and full. Another key area is matching dryer throughput with mold throughput. Don’t put a dryer on a mold that consumes resin faster than the dryer can dry it. Classify your dryers and put them on the BOM for each part/mold.
These need to be calibrated, properly programmed, and used in an environment conducive to performing what is essentially a laboratory test.
Reduce process variation through consistent particle size. One primary goal in any molding process is to get a consistent, homogenous melt to the mold. Big variation in particle size makes this much harder because they don’t melt the same. Set the proper knife gap. Use different knife angles for different plastics. Roughly speaking, use sharper, more acute angle grinds for more flexible resins to slice them and more wedge-like angles for shattering higher-impact plastics.
Hot runner manifolds
Pressure loss in any melt delivery system prevents you from tightly controlling your part. There are a couple ways to look at it. In general, no melt delivery system should have more than a 4000-psi pressure drop. Even on a press with a maximum of 17,000 psi available at the nozzle, that still leaves a lot for part control.
Another general guideline is to limit the pressure loss of your runner as a percentage of the total pressure loss. If it takes 4000 psi to get resin through your runner and you have a small part requiring only 1500 psi to fill, remember that the machine will have a harder time “seeing” the part as it pushes on the squishy rope of plastic in your melt delivery system. Cavity pressure sensors may help overcome this by telling the machine what to do based on what is happening in the mold.
Another common mistake is with wiring thermocouples, particularly when repairing them. A thermocouple works by having two dissimilar metals touching each other at the point of temperature sensing. If you splice in some other metal, you might get erroneous measurements.
Methods for analyzing moisture should include handling the resin in sealed jars and making sure the program runs a proper time/temperature profile that drives off moisture and not other volatiles affecting the moisture calculation. Also, make sure your sample is cool enough prior to testing. Running the tests too warm, particularly an issue if the drying temperature is high, can skew the results. Not-to-exceed moisture limits are the primary checks. Running some resins with moisture too low can cause other issues. Nylon’s viscosity can increase significantly. PPA becomes more brittle. If not properly diagnosed, the “cures” to these issues can make matters worse.
Quality, by one definition, is “conformance to requirements.” Are the customer’s requirements clear and commonly understood? Are physical limit samples available that the customer approved? If you run cosmetic parts, you might decrease your scrap rate simply by working with your customer to get a common understanding of what is acceptable.
Gauge repeatability and reproducibility. GR&R is a measurement systems analysis technique that tells you if the people and the gauges measuring your parts can do it reliably. Some parts have features that are difficult to measure. Some measuring tools are harder to use than others. And some less-experienced people don’t draw clear distinctions between a c-clamp and a micrometer. If you have high scrap rates due to dimensional issues, this is one area to check. Reduce the measuring variation.
Control plans. These are basically the guidelines for what to check on each part and how to check it. They might include verification of the resin, moisture content, what critical features to measure and how, cosmetics, and so on. Most molders use these. They can affect your ability to produce parts correctly the first time by serving as a checklist and a reminder for what goes into making any part correctly.
Material cleanliness comes from a number of areas. Good habits include covering gaylords and bags. Using air blasts to clean hoppers, filters, feedthroats, machines, and so forth is a terrible practice. Remember, the purpose of the plant’s HVAC system is to circulate plant air. The dust kicked up using an air blast can carry farther than most people think. For resins that require drying, keeping the resin sealed from moisture not only saves energy required to redry, but also reduces the variation inherent in dealing with wet resin.
Residence time. Many resin manufacturers publish maximum residence times for their resins sensitive to heat history. Remember that this is very much a function of melt temperatures. Most resins will degrade much faster at the high end of the melt temperature range than at the low end. Residence time can be shortened by delaying the extruder/start of screw-back until necessary. If you need to process on the high end of the melt temperature and your residence time is long, it is worth having clear procedures for every startup. These might need to go beyond how many parts you throw away before boxing parts again by including the size of the purge puddle.
Heater band profiles. For better melt consistency, these can be adjusted as a function of shot size and resin type. Keep the front zone at the target melt temperature and move the middle and rear zones up to improve the melt consistency and minimize screw and barrel wear. Also, make sure the heater bands are kicking in and not overriding. If they are overriding, that could mean that all the heat is coming from shearing, and that might not be good for your resin.
Robust process. If your mold can’t run parts in the manufacturer’s suggested processing window, take the time to figure out why not. Run a design of experiments (DOE) to figure out the critical parameters on your part and put control limits on those. Use statistics to figure out if gate seal is required and remember it is a function of time as well as pressure.
Regrind control has a lot of opportunities to reduce scrap. It is difficult to control what generation of regrind ends up in any part, which is how the commonly used 25% maximum figure came to be. Statistically, first- and second-generation regrind is the highest content percentage and should not have seen enough history to significantly compromise the properties. When you calculate out the content percentage of the nth generation regrind, it should be insignificant enough that even if it is degraded, it will not cause ill effects. Typically, 50% and 100% regrind are reserved for parts and/or resins where the product performance is well within the nominal resin’s capabilities. Adding regrind in a homogeneous mixture is required and is another good opportunity for gravimetric blending.
Cavity-pressure sensing can overcome a number of these issues. If you use part weight to determine quality, be sure to understand the variations in specific gravity, particularly for filled resins. Remember, there are tolerances on filler content and if the specific gravity of the filler is significantly different from the base resin, your part weight will naturally vary accordingly.
The effects people have on the process are many and varied and this area has been the subject of any number of books both in our industry and certainly across any industry, particularly when you move from “what” to “why” things get done, or don’t. For the most part, if proper procedures are being established and followed, things will go better. The Plan-Do-Study-Act (PDSA) cycle, when used, will move you away from variation that leads to scrap.
One interesting area where perceptions are unique to individuals is color acuity—i.e., the individual’s ability to tell if a molded part is the same color as a master sample. People perceive color differently and that ability changes over time.
While this is by no means an exhaustive list, hopefully we have provided some ideas on how to improve operations and better serve customers.
Author Mike Miller is director of engineering at contract molder New Berlin Plastics (New Berlin, WI).