Thin-Wall Warpage

Share This Page • • Free Access to 80,000 Datasheets

There’s much to consider in runner, gate, and feedpoint sizing when working with skinny parts.

The runners and parts I picked for this month’s article demonstrate two important points to be aware of when trying to fill and pack
thin-wall parts.

The molder was having problems with warpage in this flea and tick medication container for dogs. The part is made of an easy-flow grade of low-density polyethylene (LDPE). It could be colored by any one of several methods, but it looks like it’s being colored at the press with a color concentrate loader (hopefully gravimetric).

The part wall thickness varies a bit. The wall being gated into at the neck of the part is .025-.030 inch, depending on where I measure it. The part’s shoulders are .020 inch, close to the neck section. The design then goes into a 45° angle, which measures .025 inch. The straight sidewalls also vary in dimension somewhat, between .025 and .015 inch.

These dimensions lead me to the first point to consider when running thin-wall parts: In most cases, a part should be designed so material flows from thick to thin. In this case, it flows from thick to thin to thick again.

I’m guessing that this variation in wall thicknesses can be attributed to one or more of the cores shifting during injection, depending on injection pressure and speed. Thin core shift problems like this are why a part should be gated at the base of the cores instead of at the top. Although the top of the core is generally considered to be the best gate location for uniform walls, the cores often do not have a way of directing the material evenly around them. The result is a core pushed to one side or the other during injection, which leads to part wall thickness differences.

Putting a flat or a dimple on the end of the core pin is commonly used to avoid deflection during injection. It’s also common to slow down the injection speed to keep the core from wobbling. Most molding technicians I know reduce the injection speed to lock the core pin in place slowly, and then increase the injection speed to fill out the rest of the part.

Runners and gates
The runner dimensions are a little bit off also. The main runner and subrunners are a half-round shape, with a runner depth of .125 inch. To maximize the runner system’s performance, the subrunner depth should be .025 inch less than the main runner depth.

With this project, we can fix the runner sizing in two ways: either increase the main runnerdepth by .025 inch or decrease the subrunner by .025 inch—it’s the toolmaker’s choice. If we don’t want extra regrind, or if the amount of material available in the injection unit is limited, it might make sense to recut the runner plate so the subrunners are smaller than the main runner. If you don’t mind the regrind and have enough injection unit capacity, then you can increase the main runner depth. The best news here is that the half-round runners are perfect for feeding material through a three-plate runner system, but it would not be good for an edge gate. Whoever worked on this mold must be pretty good at his job because I see he offset the sucker pin locations in the runner by about .100 inch so they aren’t restrictions to flow during injection.

The drop dimensions are .155 inch where they attach to the runners and taper down to .060 inch just before the .040-inch gate diameter. Yes, I know, the gate diameter is usually only 50% of the wall thickness being gated into (.030 inch in this case), but we can actually size a subgate or a three-plate gate like this one anywhere from 50-150% of the wall thickness of the part, depending on the volume requirements needed to fill and pack.

What looks like a gate that is too big for the parts being molded is actually the size needed. The side issue here is how to keep the gates from stringing when the parts are ejected. This is where the mold really has to be optimized. If the barrel heats are too high because of nozzle orifice, sprue, runner, or gate sizes being out of balance, we will see gate stringing almost every time.

You might notice I didn’t mention packing out these parts. With walls this thin, there isn’t much packing involved—just filling.

Feedpoints, sprues, and waterlines
The second important point I want to highlight has to do with the feed point to the runner, which is .075 inch in diameter. As with this part, the feedpoint to a runner in a three-plate mold is often undersized. I think this is a major contributor to the problems we have been handed this month, since the feedpoint to the runner should be larger in diameter than the depth of the half-round main runner (.125 inch). In this case I would increase the feedpoint diameter from .075 inch to .150 inch, and also increase the general-purpose nozzle orifice to the same size as the flow tube diameter in this heated sprue bushing.

But wait a minute: Is the sprue bushing hot or cold? The cone shape of the feedpoint to the runner looks like what I usually see in the recessed sprue bushing being sold by National Tool & Mfg., in which the nozzle fits clear down inside the sprue bushing, where it seats up on the cone. I would guess that this sprue bushing design is being sold as an economical substitute for a heated sprue bushing. It’s kind of a poor man’s approach to reducing the regrind or slow cycle times of a cold sprue without spending the money on a heated sprue bushing. There are two ways to address the improper sizing of the recessed sprue bushing: increase the overall cone diameter but maintain the ratio of the taper, or buy a new one.

One last thing to check is the waterline details. I would ask the molder to get rid of the quick disconnects and eliminate any hose jumpers, mostly to see if better water circulation helps speed up the cycle. I recommend moving to hose barbs with hose clamps on each water circuit and using straight in-and-out waterline connections.

Optimum cycle time and material choices
When runners are improperly sized and flow is restricted by an undersized runner feedpoint and nozzle, often a molder has to use higher barrel melt temperatures to overcome the resulting pressure losses. Once the restrictions to flow and the pressure losses are reduced or eliminated, we should be able to bring down the barrel and nozzle heats to help stop any stringing or drooling issues.

Then, if we get rid of the waterline quick disconnects and jumper hoses, we should see better heat extraction numbers from the mold steel and stop any tendencies of the parts to warp, as well as speed up the cycle. If he can speed up the cycle by 25%, which is not that hard to do, he will be able to earn an extra $78,000/year on a machine he bills out at $50/hr.

I did not get a question from the molder as to what the cycle time should be on these parts, but in case he wants to know, I will include it in my report. I just multiply 225 (for this type of sprue bushing; use 250 for a cold sprue and 200 for hot runners) by the nominal wall and add 1 second for each 100 tons of molding machine clamp capacity to get the target cycle time. For this part, my estimate is 7.75 seconds (225 x .030 + 1 [100-ton press]). I bet his cycle time is 10 seconds or more.

What’s left to do? The LDPE this molder has been running seems a little stiff when I flex the part’s sidewall. I would typically use a higher-melt-flow material when running parts like these. So I went to www.ides.com and used the free portion of Prospector database to see what LDPEs are available for this molder to try in the future. At least he will have an option for a higher-melt-flow material if and when he has time for sampling.

The Prospector website is so easy to use and so quick to get an answer from that in just a couple of minutes I had my answer. I will recommend that this molder switch from his present 8-melt material to a 24- or 30-melt grade from his current supplier. Aren’t computers great?