Gates & Amorphous Materials

By Bob Hatch

An intriguing mystery part arrives with no note, compelling analysis.

What a morning! I was up to my ears with reports that I had to write for my recent mold trials and other usual activities, and was also getting ready to do an after-dinner talk for the SPE group in Grand Rapids, MI. Then there’s NPE this month. I guess it’s good that I’m busy, because it is this activity that gave me the inspiration for this story.

These parts arrived in an unmarked cardboard box, no note inside, and nothing on the label to tell me who the sender was. I kept the parts and runner on my desk for more than a month and nobody called to ask questions. I figured I had an orphan part that I just needed to put away and forget about. But when I picked up the parts and runner, I noticed something that fascinated me.

	The runner was trapezoidal, which is not correct for an edge gate, and the runner and gate design were unusual for a clear, amorphous material. I guessed it was a hard-surface acrylic, but I didn’t want to burn it to see if the smell gave it away. Being the curious person I am, I reviewed the runner system, gates, and parts, just in case the molder called me up some day wanting to know if I looked at his parts.
With no indication of what the molder wanted improved on these parts, the Troubleshooter found the runner and subrunner sizing for this clear acrylic part to be correct. The sprue O-diameter and nozzle orifice were too small; enlarging them would properly pressurize the runners. Cycle time improvement was possible by changing to an easier-flow acrylic and possibly reducing the mold temperatures.	For starters, the trapezoidal runner was .250 inch deep and .280 inch wide. The little stubby subrunners were .250 inch deep and .300 inch wide. The fact that both the main runner and subrunner depths were .250 inch was fine since the subrunners were so short. That made both of these runner dimensions correct for amorphous materials, such as GP polystyrene, SAN, ABS, ASA, acrylic, and PC. What was wrong here was that the trapezoidal runner system fed into an edge gate that was .110 inch deep and .250 inch wide. An edge gate should always be fed from the center of a full round runner to avoid blush, flow lines, and even jetting from an edge gate in a high-shear condition. This wedge-shaped or tapered gate was designed like a rectangular subgate, but I seldom find this design being used for edge gates—just subgates. So I got to thinking, why are they using this gate design for a part that looks like a coffee coaster or an ashtray?

Ribbon Gate Provides a Clue
After a lot of head scratching, I gave up and went to the lunchroom for some coffee. This distraction was just what I needed. While chatting with someone, my mind wandered to a day many years ago when I saw a Japanese mold running what’s called a ribbon gate using SAN, tinted green. The part was the viewing window for one of the early electronic typewriters, used to look at an LED display of two or three sentences in the typewriter. The lens protected the LED screen.

This ribbon gate was only .020 inch thick, as wide as my thumb, with a 1-inch gate land length (you can see why they call it a ribbon gate). I was fascinated by the design and when we molded parts I was even more blown away by the results. The viewing window was molded very well using this gate design; there was no blush at the gate, no flow lines just inside the gate, and most of all I was really impressed by the lower injection pressures the molder was using to fill the parts.

Without any packing required for the thin typewriter window, I figured it was a unique design used for thin, flat windows that I had not stumbled across previously. I still recommend it once in a while for parts like this but have never used it for thicker parts, such as the thick-wall coaster part.

Why the Trapezoid Works
This brings me to my discovery of the wedge-shaped gate used on our thick part, which had .380-inch wall thicknesses in six areas of the part. It was free of blush and flow marks at or near the gate area.

I am not suggesting my ribbon gate story has anything to do with the parts I received, except that the tapered gate with dimensions of .110 inch deep, .280 inch wide, and a gate land of .250 inch was doing the same thing for these parts as the ribbon gate did for the thin lenses. I considered both gates to be extruding material into the cavities instead of using a regular edge gate of .200 inch deep, .250 inch wide, and a gate land of only .030 inch to flow the material though the gate.

Often with parts this thick, I have used higher pack and hold pressures than the injection pressure used to fill the parts initially. Remember that the injection pressure fills the part and controls parting line flash, while the pack and hold pressures pack out the sinks and voids.

I considered this tapered gate and the ribbon gate to be doing essentially the same thing. They both produced a part free of molding defects without having to use extra heat, speed, or pressure to get the job done.

Checklist
Next I checked the size of the sprue bushing O-diameter to see if it was big enough to pressurize the runners. The sprue O-diameter was only .260 inch with a nozzle orifice of .105 inch—not correct for this runner. With a main runner depth of .250 inch, the sprue O-diameter should have been .312 inch (10-25% larger) and the nozzle orifice should have been 10% less than the sprue O-diameter, or .280 or .290 inch, instead of .105 inch.

Another correction to be made was to change the GP nozzle to a full-taper design—necessary for all amorphous materials, especially acrylic. The full-taper nozzle would eliminate shear points in the nozzle—dead spots just inside the sprue bushing where the nozzle seats itself. If the nozzle orifice was only 10% smaller than the sprue O-diameter, we would have very little pressure loss at that point.

The trapezoidal runner design would probably work with this tapered gate design, but the sprue dimensions needed to be increased so the O-diameter of the sprue would be .312 inch. For the 3-inch-long sprue, the difference between the sprue O-diameter and the diameter of the sprue’s base (where it attaches to the runner) was .075 inch.

That’s a little more than the .017 in/in we use to figure taper dimension differences between the top and bottom of the sprue, but it would work fine. This is important to ensure that the sprue pulls out of the sprue bushing with the runner. The .075-inch taper was unusual for an amorphous or filled material because of their low shrinkage, but is typical of materials such as polyurethane that demand more draft angle differences between the top and bottom of the sprue.

Run faster
This completed my review. If I were to guess, I would say one question the molder might have is what the cycle time should be. My formula for determining cycle time is to multiply the dimension of the part’s thick section by 250. In this case, .380 times 250 yields a cycle time of roughly 95 seconds.

This figure is just the target; the actual cycle could be as low as 60 seconds if barrel heats and mold temperatures stayed on the low side, and if there was adequate mold cooling. This means we don’t want any jumpers attached to the water circuits; straight ins and outs give the best overall cooling.

I’d also recommend getting rid of the quick-disconnect water fittings or at least removing the shutoff mechanism in each of them to improve the flow of water in the mold. Better cooling usually means faster cycle times.

I am sure we could get the cycle down to 60 seconds or less with these changes, and if I could get them to use an easy-flow version of the hard-surface acrylic, we might be able to reduce the barrel heats 20-30 deg F and speed the cycle up to around 50 seconds. I suspect the reason for using this harder-surface acrylic was to resist scratching of the parts. A lower-hardness grade of acrylic would scratch easier, but could reduce barrel heat even more and cycle even faster.

I find it exciting to play these heat, speed, pressure, and mold temperature games to see how fast I can cycle parts of a given wall thickness. In some cases I have cut the target cycle in half by opening the flow path, reducing the barrel heats and mold temperatures, and using all of the optimization techniques available.

I never did get the call from the molder, but perhaps one day I will; and when I do, I am ready to tell my story.

June, 2006 - Reprinted with permission from Injection Molding Magazine. Copyright © Canon Communications LLC.