Ever wonder how many shots are made each day via injection molding? Anybody have any idea? A huge number no doubt, yet not all shots produce a part you can actually sell. Many are short, flashed, warped, splayed, etc. These common part problems cost companies lots of time and money. Causes are debatable and often vary with individual conditions. The situation is further complicated during troubleshooting when egos and politics supplant objective analysis. Any analysis must investigate the five key components of making a successful plastic application, as defined by my colleagues Glenn Beall and Mike Sepe:
Each of these comprise hundreds of variables, and they all interact, so analysis is complex. Processing is usually the first to be investigated, as it is the easiest to target and/or change. But shouldn’t we first make sure the machine is working properly before we play with the process? The processor can do several machine checks relatively quickly to provide data on whether the machine is process-capable. We will discuss the first six of the following nine machine checks, which should take no more than 2 hours altogether and should be done at least once per year as part of an annual machine check-up. (The last three tests will be covered next month.)
These first three tests are done as the machine is running on cycle and take about 10 minutes. Is the screw moving as the mold is opening, ejecting, and then closing? If so, the shot size will be inconsistent. A machine we were working on recently drifted forward 3 to 5 mm as the mold opened, ejected parts, and closed. With that machine’s 55-mm barrel, that amounted to more than 6 cc (0.9 cu in.) variation in resin volume. It is also possible for the screw to move backward during mold open/close.
As the machine is cycling, note the position of the screw at the end of screw recovery. Is it identical on every shot? Also check that the machine is switching from first to second stage as the screw hits the set cutoff position—assuming you are cutting off via stroke position. At high injection rates, the screw may miss the cutoff position a bit, but it should be no more than 0.05 mm (0.002 in). If the machine fails any of these tests, the shot size is fluctuating beyond acceptable levels. Part dimensions, warpage, shorts, flash, gloss variations, or plugged gates may be result. (Note: These problems could still exist even if the machine passed these first tests, due to a faulty check valve or worn barrel.)
For this test, make first-stage or fill-only shots. Take hold pressure off, or set it as low as possible. Leave time on the hold timer and run on automatic, making 90-99% full parts that can be ejected. Once the machine is at steady state, mold 10 to 20 shots and compare them for size and/or weight. If you have a worn check valve or barrel, these short shots will not be uniform.
Static testing for a worn check valve or barrel is a bit more difficult and has potential to stick parts. Remember safety first, so make sure you know what you are doing and work with low pressures.
First mold a regular full and packed part, and leave it in the mold by going to manual just before hold time ends. Do not open the mold or take out the runner or sprue; leave them in to block the flow path. Now, set the second-stage pressure equal to your normal second-stage pack/hold pressure (or a little lower to be on the safe side). Make sure the second-stage pressure does not overshoot the pressure setpoint when you manually inject by watching the pressure as you push inject with the screw beyond the cutoff position, but not bottomed.
Manually rotate the screw until you get it back to about 75% of the barrel capacity. After verifying that second-stage pressure will not exceed your normal second-stage pack and hold pressure, manually inject with the cutoff position set at above 75% of barrel capacity for about 10 to 15 sec and watch to see if the screw moves forward. Since the system is filled with plastic, the screw should compress the shot, then stop and not drift forward. Next, do the same procedure for 50%, 25%, and 10% of barrel capacity. Be careful that the part in the mold and the sprue or runner do not shrink so much that plastic can find a flow path to overpack the mold, sprue, or runner. This can be a problem with certain resins, like soft-touch TPEs.
If the shot at 75% of barrel position holds during manual injection at low pressure but does not hold at any of the other barrel capacity tests, you most likely have a worn barrel and the check valve is okay. The barrel must be repaired or replaced. If the screw drifts forward during manual injection at all of the barrel capacities, you cannot tell whether the problem is the barrel, check valve, or both, but you know you have a problem that must be fixed. Remove the screw and measure all components to see if they meet specifications. Replace what is necessary.
Yes, all this means downtime, but you’ll lose more money by not fixing a root cause. It’s not easy to accurately measure barrels, screws, or check valves, so have your favorite screw/barrel supplier do this to get good data. Even more important, if your measurements are used for replacements and there is a mistake, you are responsible. If your supplier made the measurements, they’ll have to make good on any new component that’s not to specification. How many of you have seen a new screw, barrel, or other component that was too long, short, or out of round? It happens more often than anyone likes to admit.
As safety note, whenever you reassemble the barrel, make sure there is a clearance between the end cap and screw tip. The acceptable range is around 0.060 in. (1.50 mm). Do this by bringing the injection screw to its most forward position. Put Silly Putty, Plastigage, or soft solder on the tip of the screw and bolt on the end cap. Now remove the end cap and check the thickness of the putty, solder, or Plastigage. Make sure the screw tip does not touch the end cap. If for any reason, the total assembly is too long, you could blow off the end-cap and do serious damage to the machine and personnel.
To check the screw bottom or zero point, switch the cycle to manual with low pressure and bring the screw full forward. Check the screw position reading: It should be zero. If not, recalibrate the position sensor. Would anybody like to admit how many times they thought they had a cushion for pack and hold when in actuality the screw was bottomed at 5 mm (0.200 in.)? Part problems arising from this would be short shots, dimensional variations, warp, gloss variations, and sink.
The load sensitivity test checks a machine’s ability to keep fill time constant by adapting to viscosity changes, such as temperature cycles, resin lot changes, color changes, and other normal process variations. This is a difficult but critical test for a machine to ensure quality production. The idea is to vary the load requirement on the injection unit and note the change in fill time.
Start up the machine following your normal safety procedures, making sure you are operating with an appropriate Delta P, (see Plastics Technology, January 2010, p. 13 or at ptonline.com) and switching from first stage to second stage via stroke position. You will be making short shots by reducing second-stage pressure as low as possible while leaving some time on the hold timer.
Observe the fill time and pressure at transfer to make this 90-99% full shot. The fill time is FT1 and the pressure at transfer is PK1. Now, either manually or in automatic cycle using screw decompression before screw rotation, pull the screw back (without screw rotation) to the exact same position used to make your 90-99% full part. That is, you have a shot of air, no plastic in front of the screw. Before you make this air shot, make sure you can easily dig out a very short shot. Now, on semi- or fully automatic cycle, shoot air into the mold and note the fill time and pressure at transfer. This is FT2 and PK2. The fill time should be the same, with a significantly reduced pressure at transfer. The following equation calculates the machine’s load sensitivity:
(FT1 – FT2) ÷ FT1) ÷ (PK1 – PK2) ÷ 1000) × 100 = % error/1000 psi.
If you are testing an electric machine, divide the pressures by 10 before entering them into the equation.
The percent error should be near zero; slightly negative is ideal. Machines having greater than +5% error are not properly load compensated. Test at three different injection velocities—slow, medium, and fast. If your machine passes this severe load-variation test, it will keep fill time constant during production.
These tests should not take an excessive amount of time and may save you hours of troubleshooting. Plus, you’ll have some data rather than just opinions on how well your machine functions. We’ll continue with more testing procedures next month
John Bozzelli is the founder of Injection Molding Solutions (Scientific Molding) in Midland, Mich., a provider of training and consulting services to injection molders, including LIMS, and other specialties. E-mail firstname.lastname@example.org or visit scientificmolding.com.
Last month we started defining some basic tests to ensure your injection molding machine is working ...