The Troubleshooter, Part 27: Mold fill programs do a pretty good jobThe Troubleshooter, Part 27: Mold fill programs do a pretty good job

Parts arrived this week from an old acquaintance in Minneapolis. Tony Boire runs a company in the twin cities called Emplast. The parts he sent in are clear, fairly large, and except for a little swirl on the end of each part, they are in pretty good shape.

The runner is one of the old style "H" pattern balanced runners with the cold slug well at the end of each runner. I've told Tony many times the cold slug well doesn't function as a cold slug catcher; it's just a place to vent the runner.

I figured he was just playing around to see what I would say, but I wasn't going to bite. This mold was probably built in a shop where they cut the runner plate on an older mill with only X-Y axis capability, and the cold slug wells are just where the cutter runs out after cutting each runner channel. Otherwise, they would have used newer equipment to cut curved runners to feed material from the sprue to each of the subgates.

A trapezoidal runner design feeds into the subgates, which is OK. The use of full round runners for an edge gate and either full round, deep half round, or trapezoidal runners for subgates is correct. Anyway, the runners and gates are OK for a clear material. Clear materials are amorphous, so big runners and big gates are the way to go.

The depth of the trapezoidal runners looked a little undersize, but after checking with the digital measuring stick, I'm pretty sure they will work. I can see bubbles in the runner, some of which are round and some are elongated. The round ones are caused by low levels of moisture, and the elongated ones represent pressure problems. Doing a little dryer maintenance will correct the round bubbles, and increasing the size of the sprue bushing will eliminate the elongated airless voids.

The sprue O diameter is currently .280 inch feeding a .250-inch main runner, which feeds .220-inch subrunners into a .210-inch subrunner. That feeds a .180-inch stubrunner that feeds into the subgate. The subgate diameters look to be about .060 inch.

Optimum Sizing
If I were sizing the runner, I would start with a .225-inch subrunner feeding the gate. I would keep the dimension at .225 inch for the subrunner back to the first level subrunner. Then I would go to .250 inch for the first level subrunner, .275 inch for the main runner, and .312-inch sprue O diameter with a .290-inch orifice in a full taper nozzle for this clear material.

Not that these recommendations represent a big difference, but the runner system is just a conveyor for getting the material from the barrel to the gate. For some stiffer flowing materials, it could make the difference between filling and packing the parts or just molding a nice looking runner and ending up with bad looking parts.

The gate size is based on the part wall thickness and flow requirements of the material. The wall thickness of the parts is a nominal .060 inch, and if I look at the wall thickness divided into flow length (L/T), I can determine if this wall thickness and flow length is OK for the material being used.

I had to call Tony to find out what the material is, and he told me it was an easy-flow grade of polycarbonate. Now that I have been told the material is polycarbonate, I know I will need a flow number of 200 or less when I divide the wall thickness into the flow length to be able to fill and pack the parts.

A .060-inch wall thickness divided into a 12-inch flow gives me the 200 I'm looking for. This part is only 6 inches from each gate to the end of fill, so we should easily fill these parts without having to use higher heats or pressures which will cause molded-in stress and slower cycles.The .060-inch gate diameters are also about right to use on a part with a .060-inch wall. The nice thing about subgates is that you can go larger in diameter with the gate than the wall is thick. Not that I always do, but you can.

If you keep the barrel melt temperatures in the 525F range for this easy-flow grade of polycarbonate and if you eliminate the shear condition of injecting this material through a small diameter gate, you will not have a serious blush condition on the part at the gate area.

The part itself looked pretty good. The sharp corners were all radiused, and the corners appeared to be vented well enough. At least, I couldn't see any dullness in the corners that indicated a heat buildup from compressing air into the corners due to poor venting. Even the runners appeared to be well vented at each of the cold slug wells.

I still couldn't explain the swirls at the end of fill areas on each part. The swirl wasn't at the parting line, and it didn't look like air was being trapped in the knit lines. It had to be that material was being injected into the mold too fast, and it was being torn apart by sharp edges on the core geometry in that end of fill area.

So I called Tony again and gave him my review of his mold and parts. He was somewhat amused and told me he was happy with the parts just the way they were. He said he should have told me during our previous phone call, but he wanted to see if I could find anything he had missed.

Proving a Point
He said all they had to do to get away from the swirls was to slow the injection speed down near the end of the fill time. He slowed it down at about 90 percent of fill instead of the normal 98 percent and switched over to pack pressures.

He said he was really trying to prove to me that you can design parts and runners with a good mold filling program and get results good enough to take to the tool room and then to the molding floor and mold acceptable parts the first time you put the mold in the press.

Tony is right about one thing. If you use a decent design program like Pro/E and run the STL file through a mold flow program like MoldFlow, you will get pretty good results.

I even see a lot of benefit from using the simpler and quicker programs from C-Mold and the MoldFlow people. There are many good mold flow programs available today. At least, they can predict where the knit lines and flow patterns will be as well as predict temperature changes during the filling process. This is much better than in past years.

So Tony had his fun, and I was able to come back with slight improvements. My suggestions would allow him to bring his barrel melt temperatures down to the 525F level, and he should be able to cycle at about 18 to 20 seconds, which he isn't able to do now. He's currently running his melt temperature at 550F, so his cycle is around 30 seconds.

The more I travel the more confidence I have in runner and gate sizing. Not everyone has the same urgent need to run optimum cycles, but the custom molders do. All they have to sell is machine time and expertise.

For Tony to speed up his cycle from 30 seconds to 20 seconds would add at least 15 percent profit to his bottom line. Sometimes, we just let these kinds of profit opportunities slip through our fingers. Personally, I've always been too cheap to give machine time away.