Click Image to Enlarge
Last month we started defining some basic tests to ensure your injection molding machine is working properly. We covered the five key components of any plastic application:
- Part design.
- Material selection and handling.
- Tool design and construction.
We then honed in on processing, as it is one of the first areas to be investigated when troubleshooting begins. The idea is to make sure the machine is functioning properly before making processing changes. Part I covered the first six of the nine suggested machine checks, and this column covers just the seventh—velocity linearity.
- Screw does not drift at shot size.
- Shot-size consistency.
- Position cutoff consistency.
- Faulty check valve or worn barrel; dynamic and static testing.
- Screw full-forward position.
- Load compensation.
- Velocity linearity
- Delta P, for velocity control.
- Switchover response.
|Table 1—Data for Velocity Linearity|
|(Shot size: 22.1 mm; Barrel capacity: 40 mm; |
% of Barrel used: 55%; Best if >65%)
Set Velocity, mm/sec
Fill Time, sec
Actual Velocity, mm/sec
% of Set Velocity
The purpose of this test is to compare a machine’s set velocity with what is actually achieved. Before beginning this test, make sure short shots can be ejected easily. If shorts cannot be easily ejected you will have to use second-stage pressure to finish filling the part. And to get good data, the shot should be at least 60% of the barrel capacity. If you find you are using less than 40% of barrel capacity, do not use that mold for this velocity linearity test, as the results may be useless.
The procedure assumes you are using a production press making full parts (it does not matter for this test whether they are acceptable parts or rejects). If it is a new mold or one just installed to begin production, do what you normally do to start up the mold to make full parts. Lower the second-stage, or pack and hold, pressure as far as the machine will allow. At whatever velocity is currently set on the machine’s controller, this should make a short shot. If not, adjust the position transfer to make a part that is short but at least 80% full.
Make sure the press is running under velocity control and you have an appropriate Delta P. If you do not understand velocity control and Delta P, please review my earlier column on the subject (Jan. 2010, p. 13 or at ptonline.com). Now increase the set injection velocity in reasonable (safe) increments until the machine is set to the fastest injection velocity allowable. As you increase the velocity, you may have to adjust the position transfer to ensure you are making short shots, and double check that you are still under velocity control with an appropriate Delta P. Once at the fastest allowable velocity, adjust the position transfer to make a 99% full part.
Once you are making 99% full parts under velocity control, you can start taking data. Note the actual “screw start position”—don’t go by the set shot size and decompression or suckback, if any. Then note the cutoff or stroke transfer position and the fill time (to within 0.01 sec). Without changing anything but injection velocity, reduce the set velocity from as fast as the machine can go to nearly as slow as the machine can go. Make sure you are transferring on position and not time; that is, make sure the overall injection timer has enough time on it to allow you to reach the transfer position. The parts may change in size but do not readjust shot size or position transfer during the test.
Set up the data table to look like Table 1 and if you want, graph the data as shown in Fig. 1. To calculate the actual velocity, subtract the cutoff or transfer position from the actual screw start position and divide by the fill time. Because this method includes the initial inertia to get the screw up to speed, expect the fastest actual velocities to be only 85% to 95% of the set velocities. This inertia can be dealt with if you have good process-monitoring capabilities.
Note that Fig. 1 and Table 1 show at high velocities a significant difference between set vs. actual, actual being a bit slower then set. This is to be expected because of inertia, which we are not accounting for in our calculation. Bottom line: This machine is fine, with no real problems, and if you wanted to get the slow velocities closer to set values, a calibration could be done. But I would not recommend it, as a processor should use only duplicate actual fill times and not depend on set values. Machines wear, amplifier cards burn in, and calibration is not necessary if you duplicate fill times for the fill-only parts. On the other hand, if the graph shows actual velocities hitting a maximum and leveling out, the machine needs repair, possibly of the pumps, controller, or other electrical components.
In our next column we’ll cover the all important switchover response from first to second stage.
About the Author
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.