You have already read this, of course, but repetition is one of the 3 R’s in teaching.

Here is an example of why it is important to set the fill volume. If the eDART does not know where fill is supposed to end then the only guess it can make is to use peak injection pressure. Typically this will indeed be the end of the fast speed portion of the filling of the cavity. As soon as the machine slows down, less injection pressure is needed and the pressure drops off. Thus the peak seems to work much of the time.

Remember that viscosity is the area under the curve during fill. If the eDART has to guess at “fill” then here is what can happen. So try to remember to set fill volume on new processes. Once set the eDART remembers it for that process.

Shot #1: Good Viscosity Number

End of fill is determined by the peak and the peak occurs at the correct time. (this is a short shot while doing a rheology curve)

Shot #2: Next Shot – Truncated Viscosity Number

Here a sticky sprue or cold slug is causing an early peak in the injection pressure. So without the fill volume set the viscosity number is calculated smaller than it should be. If this happens on alternating cycles then the viscosity number on the summary graph goes up and down dramatically and looks like noise.

#3: Peaks Close in Value

Note that the above is a relatively dramatic example. Sometimes the cold slug peak on every shot is almost the same level as the end of fill peak. So the end of fill time can jump back and forth while the curves look almost the same.

---

Note: The “Auto-Set Fill Volume” is standard in Release 8.5 as part of the “Sequence Settings” tool (Machine Fill tab) on the eDART. If the eDART is using peak injection pressure for end of fill this tool sets the fill volume at the first dramatic slow down in speed after 20 stable shots. You can view the document here.

… a “ripe” process being one that you are done fiddling with.

It appears that during process startup some people have been chasing the V1->V2 (in D3) or transfer position (D2) and trying to “Set Fill Volume” every time a change is made. The point here is that the process must be essentially finished or you need to be working on a rheology curve if the Effective Viscosity / Fill value is to have any meaning. Each time you adjust the process you are likely to change the end of filling position (or position of the injection pressure peak). Thus the viscosity number will “bounce” all over (see graphic below).

When you have set up a process:

1. Finish the usual systematic steps: fill only, fill & pack, fill, pack and hold, screw speed etc.
2. Run the process for several shots just as it will continue to run once you leave it – cycle time standard etc.

   Now the process is ripe.

3. Set the fill volume at the slow-down point (V1->V2 on a D3 process or transfer in D2).
4. Monitor Effective Viscosity / Fill on the Summary Graph or on the Light Tree to look for any changes over the long haul.

Further Points:

• To get Effective Viscosity / Fill you must have at least injection pressure (or nozzle pressure). There is no other suitable way to know the work that the machine is putting into filling the mold.

Even though it is possible to get viscosity with just injection pressure…

Try very hard to get the LER-30 stroke sensor attached in order to get a good, stable velocity regardless of how the injection pressure peaks and levels occur.

Try to get at least the Screw Run sequence module input wired up to avoid having to set the zero on the stroke manually.

---

Process Setup: Working on Short Shots

Process Setup: Working on Pack

A “Ripe” Process: Stabilized and Fill Volume Set

Let me repeat: Always Monitor Viscosity

We have covered how to set the eDART’s fill volume eDART to get stable Effective Viscosity / Fill. This is the most fundamental number that represents the characteristics of the incoming material. It can vary with the way the polymer was originally made, dryness, temperatures in mold or barrel, filler content, regrind content etc. As one single indicator of the most un-controlled variable, Effective Viscosity / Fill has no equal.

By monitor I mean:

• If you are saving templates, have Effective Viscosity on the Cycle Values window and look for percent change.
• Add Effective Viscosity / Fill to the Summary Graph.
• If you have a light tree, add Effective Viscosity / Fill to the Alarm Settings tool. Then when the light goes to yellow or red you can preemptively solve a problem before the parts go bad.
• If you have no cavity pressure sensors use Effective Viscosity / Fill to set machine controls to process according to the current material viscosity.

---

Here are a couple of examples

Most of you have seen this one. We have used it for a long time as an example of a few part dropouts in one night. But notice the oscillations in the viscosity all along. Commonly barrel controller PID loops will oscillate at a cycle of a few minutes (1 – 5) if they are poorly tuned. Similarly, mold temperature controllers sometimes do the same. Dryers sometimes show a variation over the drying time (several hours) or when a regeneration bed changes.

This one is 22 minutes. What is causing it?

Addendum from Rod on the 20 minute cycle in viscosity in the first graph: “The 20 minute cycling in this data was found to be the water saver valve on the heat exchanger cycling causing a slight back pressure change over the cycle”

---

The customer was quite surprised to see how much change happened in a night. Why did this happen? Who knows? It is rather slow for a material change. Those usually show up as a step. And it is not cyclical.

The customer was also quite surprised at the process tampering attempted by the night shift. At least, he said, the were tampering in the right direction to get the process back on track (with alarm lines).

---

Here’s what can happen when the mold and machine sit idle for a while. The longer the down time (second stoppage), the worse is the shift in viscosity at startup.