I was recently asked to visit a facility that was a flat sheet producer to determine the cause of a major screw design problem. So, as always, I have to start at the beginning to gather all of the technical information to determine the root cause. This facility had 5 large extrusion sheet lines and they were issues with all 5 extruders.
With the extruder at room temperature, I set up three dial indicators on the discharge flange of the barrel in the X,Y and Z axis. Then turned on the barrel heaters to the standard zone setting to make sure that the barrel thermally expanded in the Z-axis direct as much as it should theoretically and that the X and Y indicators move minimally.
The simple equation to determine the amount of expansion that a barrel should grow is:
ΔL = .00000633 x ΔT x L
- ΔL = The change in Length
- ΔT = The change in Temperature, in this case from room temperature to the barrel zone setting
- L = The heated length of the barrel
Amazingly the barrel grew within about .030” of the theoretical change in length, which in this case was approximately .750”
Then I measured the flight OD on several of the screws for various designs to determine if there was a consistent wear pattern and there was, so that was noted.
Then I gathered all of the process data. This is a very important part of doing CSI on screws. This is where you collect the given throughput rate at a given screw speed against the headpressure during that timed rate check, motor load and melt temperature.
The motor load reading is taken from ammeter on the control panel as is the screw speed taken from the tachometer. If at all possible, it is best to have the customer’s plant manager to check the motor load with a hand held meter to verify that the ammeter on the control panel is reading correctly. As for checking the screw speed, this typically can be done by using a stop watch and counting the rotation of the drive quill at the back of the gearbox.
In this case the control panel ammeter was reading correctly, but the screw speed was not. The customer’s setup sheet showed that their standard setup was to have the extruder operating at 70 rpm, but when I counted the revolutions of the drive quill, I was getting 92 rpm. This is an error of 24%!
I then checked the tachometer on the line next to the one that I was gathering the process data from and the tachometer on it read 86 rpm but when I did the count, it was only rotating at 70rpm. This meter was mis-calibrated by 23%!
So, the moral of the story is, the only thing worse than no data is BAD data. In this case, the customer immediately had their maintenance people re-calibrate all of their control instruments.
NOTE: Sometimes the screw rotation is faster than what a person is able to visually observe. In that case, an old mechanic mentor of mine (who only had a 4th grade education) told me when I was a kid, “Count the clicks”, which I had no idea what he meant until he showed me.
Howard took this machinist scale (a pencil or pen will work) and turned on the chuck of the engine lathe in his shop, then took the scale and let it rub against the chuck, on an extruder it can be a small bolt in the back of the rotating drive quill or the drive key on the shank of the screw. Then with your stop watch in one hand the “clicker” in the other, you can count the number of times that bolt or key, hits the end of the scale, pencil or pen or the number of clicks. “Count the clicks”. Very simple but very effective.
Just make sure that your instruments are calibrated on a regular bases and also do a check and balance when gathering data. Never trust that you think you see the first time.