The Troubleshooter, Part 28: Polycarbonate molding problems

I received some parts in the mail from a molder back east. The part was a cover for something, probably a thermostat, and it had a rectangular window in the top surface. Not a terribly intricate part, but it was causing problems for this molder. His biggest issue was with cosmetics just inside each gate, down the side of the part, and at each end of the part where the end-of-fill weld lines were forming.

I called him and discussed the part problems. I told him all the cosmetic problems I could see were probably caused by the small diameter sprue and the trapezoidal runner feeding an edge gate.

Debunking the Myth
It still amazes me that toolmakers and molders don't get together to discuss some of these molding myths. Myth number one is that a big sprue causes the cycle to slow down so that the sprue doesn't break away from the runner when the mold opens up. (If it did, this obviously would mean the molding tech would have to back off the injection unit and pound the sprue out with a brass rod.)

But it's just not true! I suggest opening the sprue bushing up to the correct size for the material being used and for the volume of material needed to fill and pack the parts. Then we can bring the melt temp of the material down to the lower end of the manufacturer's recommendations, at which point the sprue won't break off the runner and stick in the sprue bushing.

The diameter of the sprue needs to be bigger so the orifice in the nozzle can be bigger. That will eliminate a shear point in the flow path, which can be so detrimental to polycarbonate and many other engineering resins.

There's another myth that should be dispelled. There are those who think the larger sprue and runner will be a significant consumer of material. Yet in a two-plate mold, the difference is only a few percentage points, maybe the difference between 15 percent and 17 percent regrind. But when the scrap is reprocessed, that regrind is certainly better quality because the material runs at lower heats.

The runner should always be full round when feeding an edge gate so the edge gate can come off the center of the runner and fill each cavity through an abrupt transition instead of through a ramp design-a design that usually causes a blush or flow mark just inside the gate on the surface of the part.

The last point I made was to remind him that with polycarbonate he should make sure the edge gate depth is 90 percent of the wall thickness he is feeding into. The width is based on the volume of material being injected into each part, which means for a small part the gate should be about as wide as it is deep. For a medium sized part, the width of the gate would be twice as wide as it is deep, and for a bigger part, the width of the gate would be about 3 times wider than it is deep. I ended by reminding him the land length of the gate should be half the depth but should not exceed .030 inch, no matter what material he is using.

I took out my verniers and measured his part. I found the wall thickness to be .070 inch, and the part itself is what I consider a medium-sized part. Therefore, his gate dimensions for this part should be .063 inch deep, .126 inch wide with a .030-inch land length-pretty easy when you think about it. I suggested he go .065 inch deep and .125 inch wide with the .030-inch land because toolmakers get crabby when you use dimensions like .063 inch and .126 inch.

We finished our conversation by discussing the venting requirements for the runner and the parting line of each part. Polycarbonate runner vents should be .003 inch deep, as wide as the runner itself, then out .060 inch from the parting line into a .040-inch-deep channel to atmosphere. The vent lips should be polished to an A1 or mirror finish to make them self-cleaning.

Showing Improvement
This conversation took place on Friday. By the following Tuesday, I received more parts. He had indeed opened up the sprue bushing, made the runner full round, and changed the gate to the dimensions we discussed. I picked up the phone and called him to see what he wanted to do next. I could still see a flow pattern originating at the gate area, and I could see a dullness at the parting line that is typical of polycarbonate parts when the parting line vents aren't deep enough. It is just the trapped air at the parting line being heated up by the air compressed in the air trap areas. The hot air is slightly burning the polycarbonate, just enough to get that dullness.

When the molder got on the phone, he said he was tickled to death with the parts just like they were and, better yet, his customer was completely satisfied. He thanked me for the help and was getting ready to hang up when I mentioned that I wasn't as satisfied with the improved parts as he was.

Well, you could have heard a pin drop for a few seconds. Then he asked me what I meant, and I told him I could still see a pattern radiating outwards from the gate and could still see some of the dullness running along the edge of the part at the end of fill areas. He could see what I was talking about, but his customer was happy. I told him we were so close to making it right we should finish the job and not give up until we were successful.

Well, I must have hit his hot button because he agreed and wanted to know what else we could do to get the job finished. I looked at the part again and checked off each of the trouble points that we changed to get the parts to look better in the first place. I was talking kind of under my breath, saying things like, "Well, we opened up the sprue, changed the shape of the runner, shortened the land, vented the runner down the ejector pins, added vents to the parting line, what else could it be?"

Then I saw it: The end of sprue, where the nozzle seats up, had an interesting look to it. What I could see were little tiny bubbles on the end of the sprue, no doubt caused by the nozzle being too hot. The end of the sprue looked like what you get when you burn a piece of polycarbonate trying to identify the material. You put the hot end of a lighter to the plastic, and it burns with a black, sooty flame. The material kind of bubbles up when you burn it, kind of like little moisture bubbles, and you know it's polycarbonate because that is the signature of polycarbonate. That's exactly what this looked like.

So why was the nozzle being run so hot? It had to be due to the nozzle having a small orifice. To keep it from freezing off, the molding technicians were running the nozzle heats 50 deg F or so higher than the front zone. I got my verniers out again and measured the sprue O diameter and the diameter of the vestige in that area left by the nozzle orifice.

Our newly sized sprue had a .300-inch diameter, and the nozzle orifice was only .125 inch from what I could see. That was it. When the molder heard me talking out loud about all these conclusions, he yelled out to someone to go out on the floor and look at the nozzle on the machine to see if I was right or not.

I had to be right because my verniers don't lie, but he came back on the phone 25 or 30 seconds later with the confirmation about the nozzle still having the 1/8 inch orifice in it that they used when the sprue O diameter was only .150 inch.

I told him he needed to change the nozzle tip to a bigger size or find another nozzle that better matches up to the .300-inch O diameter in the sprue bushing. Then he should run another sample and let me know how everything turns out. He said he would and ended our conversation.

He called back in a few minutes and said he didn't have any nozzles with orifices bigger than 1/8 inch and, if he bought a new one, what size orifice should he get? I told him to get a .275-inch orifice to match up to the .300 inch sprue O diameter. He said he would do so and get back to me with the results.

A Learning Experience
It took a week or better for him to get back to me, but when he did, he sounded pretty excited. He couldn't believe the difference the new nozzle made, and this time, he said the parts looked absolutely gorgeous. I asked him if he thought it was the nozzle that made the difference or all the work he did leading up to the new nozzle being used. He guessed it might have been all the changes coming together to make the parts look as good as they do now.

I had to agree, especially because he could not have used the bigger orifice nozzle until he opened up the sprue diameter in the first place. I felt good about solving his problem and letting him learn from the experience.

What had he learned? He found out you can't restrict the flow path without running into cosmetic problems on the part, not to mention the longer cycle times. Typically, most molders like to raise the barrel heats to get the material to flow more easily instead of making the changes to the mold, which means they have to lower the mold temperatures to keep the parts from warping. This, in turn, makes them add time to the cooling portion of the cycle, and that means they will end up with parts that have high levels of molded-in stress as well as cycle times that are longer than quoted. In this case, the cycle time was reduced from 40 seconds to 27 seconds after all the optimization had been done.

I see it all the time. Restricted flow paths and molds that are way under vented, which cause higher melt temperatures for the material, all of which may or may not cause a residence time problem in the barrel. I see molding techs that are slowing the injection speed down to give air time to get out and cold mold temperatures being run to keep parts from warping.

I've said it before, and I'll say it again; this is an upside-down situation. You will not make any money on jobs being run this way. Don't give away your profits! Open up the flow paths to where they should be for the material being run and vent the mold, especially the runner. Warm the mold up, and you will be molding parts with a lot less molded-in stress. And the cycle times will be a lot closer to the numbers you quoted to get the job in the first place. As for the "excess" scrap on this part: The new part with the larger sprue and runner weigh only 1.9 oz. compared to 1.8 for the original, smaller-sized flow path. Not much for a significant reduction in scrap parts!