Sponsored By

June 20, 2002

13 Min Read
The Troubleshooter, Part 55: Thick-wall parts

This article continues our series of troubleshooting reports from one of the leading on-the-spot problem solvers in the molding industry. Bob Hatch is manager of technical service and customer support for Prime Alliance, the Des Moines-based resin distributor. Before his present assignment, Bob managed a molding operation for 25 years.

pulley.jpg

For such a thick part, this pulley had surprisingly few problems for the Troubleshooter to correct. Made of glass-filled nylon 6, it molded without voids but did require some modifications to the runner system to increase flow. (see photo below)

Things had been slow for a couple of days when, out of the blue, I got a phone call from Mike Gebel, who runs the toolroom for Midwest Industries in Ida Grove, IA. You might know the company best for the ShoreLand'r boat trailers and ShoreStation boat hoists it produces. Mike has been a moldmaker for quite a few years, and has come up with some interesting innovations for the ShoreLand'r and ShoreStation product lines.

He said he had a new idea for gating a very thick pulley they were getting ready to use in conjunction with their ShoreStation boat hoist product line. He wanted me to take a look at his concept and see if I could find any problems with the design and production development for this new pulley.

To say the least, I was intrigued by the opportunity to look at a new design for a thick-wall pulley, so I told him I had a couple of days free the following week and would stop in to see him. Being able to mold thick-wall parts without any voids is something all molders wish they could do.

When I got there, it didn't take Mike long to pull up the designs on the Pro/E screen that he uses for almost everything he does these days. Then I started asking questions about whether he could explode some of the drawings we had just reviewed so I could get a better look at how the different components go together to make the molded pulley.

After a few minutes, I could see what Mike was up to and had to admit I had not seen anything like it in my many years on the job. Mike said he had not seen anything like this either and that was why he wanted me to check out this new design.

Mike has built pulley molds before, but this was uncharted territory for him because of the thicker walls, the molded-in insert, and his desire to use a disk gate. He had used a similar design previously, but not one in which the disk gate fed in material around a brass insert. Mike's number one issue was that he could not have any voids under the grooved section of the pulley. A high-tensile aircraft cable goes around these pulleys to lift big, heavy boats out of the water and then support that weight for days, weeks, or even months at a time.

One Tough Part
The material used in this application was a glass-filled, impact-modified nylon 6 that required a good surface finish to maintain the quality standards of the product line and still be strong enough to perform the function needed by the part. The advantage of that material in this application was the good cold weather impact strength and the better surface gloss compared with most of the glass-filled 6/6 nylons.

Mike and I reviewed the samples they had molded a few days before, and I could see why he was concerned. Some of the sections were 1.450 inches thick, and even though he had cored out some sections between the spokes, it still left a lot of room for problems with voids. To have voids in the 1.450-inch-thick sections directly under the .650-inch groove that the cable rides in during the boat hoist's lifting and holding function would be totally unacceptable. The boat's weight would force the groove to cave in on these voids, destroying the pulley.

runner.jpg

Section

Old

New

Sprue O-diameter

.275 inch

.343 inch

Sprue at disk gate

.335 inch

.403 inch

Disk gate wall at sprue

.220 inch

.300 inch

Feed point width

.125 inch

.175 inch

Middle of disk

.180 inch

.240 inch

Top of skirt

.100 inch

.140 inch

Bottom of skirt

.060 inch

.090 inch

Mike had already sectioned several of the molded parts with his trusty band saw and neither of us could find any voids or other forms of porosity in the part's thick sections. The key to molding thick parts void-free is to use a pack pressure at least as high as the injection pressure and most often just a little higher. All this sounds well and good, but it means the flow path has to stay open for material to flow for as long as it takes to get the part filled and packed; then pack pressure has to be used to fill any voids that might form in the thick sections.

I told Mike that, in my opinion, he could not have done a better job molding these parts. He said that a sister company of Midwest Industries was doing the molding. The molding manager, Ron Schirrmacher, is one of the best molders he has ever worked with, so Mike didn't want me to give him all the credit when Ron did the work. I know Ron well so I have to agree with Mike. It takes a good relationship between a molding person and an equally good toolmaker to make a molding shop hum.

Eliminating Warpage
I went on to ask Mike about any post-mold warpage that he might be getting due to the thick sections of the part. Mike said it was funny I should ask because he was getting some warpage that wasn't acceptable for the function of the part. It was the type of warpage that causes the center hub of the pulley to cool in a position that makes it wobble when it turns.

My guess was that he had a couple of choices here. He could drop the parts in water to set up the surface and just let the material in the thick areas take as long to cool as they needed, but that often causes shrinkage voids. I'm not talking about packing voids in this case—just thermal shrinkage voids—but all voids are unacceptable for this application.

p51_blurb.gifAnother approach was to cool the parts in stages. The 2-minute cycle time was long enough to take the parts right out of the mold and put them into a water tank heated to 150F. Then, after a few minutes, the parts could be moved into 100F water, and then into room temperature or slightly cooler water for final cooling. This is a good way to keep the surface of the part from cooling quickly and trapping the hot material in the center of the walls. It basically lets the part cool more evenly from the center of the thick walls to the outer surface of the pulley. However, this method requires a lot of extra handling, so most molders don't like to use it.

Another approach, and perhaps my favorite, is to open the flow path in the mold to allow the glass-filled nylon to flow at a lower temperature than the typical 540F or higher temperatures that many molders use for glass-filled nylons. The temperature shouldn't be so low that the nylon sets up and twists the nose cone off the end of the screw, but low enough so we don't have to remove as much heat from the molded part after it is ejected. Using stainless or D-2 steel in the gate area to eliminate abrasive wear from the glass filler in the material was also a good idea.

We would have to use a pyrometer to verify the actual melt temperature, but setpoints of 460F or so would do a pretty good job for this material. Factor in the shear heat generated by the screw as it recovers and the 460F setpoint typically reads an actual melt temperature of 490F. Since nylon 6 sets up at 475F it's important to have a little difference in melt temperature as a cushion.

This is a good time to explain the difference between nylon 6 and 6/6 when it comes to cycle time. Nylon 6 sets up at 475F and 6/6 at 500F. If you are trying to take heat out of the barrel, it is easier to do with nylon 6 than with 6/6. The less heat we put into the part, the less time it takes to get to a decent demolding temperature.

mold.jpg

In this drawing of the pulley mold, the molder's spreader insert (gray) is used to mold around the brass insert.

A Whole Lot of Flow
The next question was how Mike was going to increase the flow path of the material through the disk gate. First, he needed to increase the diameter of the sprue from .275 inch at the sprue O-diameter and .335 inch where it attached to the disk portion of the gate to something bigger. I estimated that .343 inch at the sprue and .403 inch where it attached to the disk gate would be about right. We needed to increase this diameter so the sprue could handle the added volume requirements we would need when we thickened the walls of the disk gate. Needless to say, the general purpose nozzle orifice would need to be increased to an orifice diameter of .325 inch after we opened up the sprue.

In addition to making the sprue bigger, we needed to increase the wall thickness of the disk gate to handle the increased volume of material that would be running through the sprue. The disk gate wall was .220 inch thick where the sprue attached to the disk; then the wall thickness gradually tapered down to the skirt portion of the disk gate where the wall thickness was .100 inch. The wall thickness at the top of the skirt was .100 inch, and the skirt wall gradually tapered down to .060 inch where the skirt attached to the part.

This follows the rule of flowing material from thick to thin, but this entire design was on the thin side for a part with 1.450-inch-thick walls. Think of this as a material starvation problem for the cavity due to this flow restriction.

The top portion of the disk gate, where the sprue attached to the disk, needed to be increased to .300 inch from .220 inch. The middle portion of the disk had to be bumped up to .240 inch from .180 inch, and the top of the skirt needed to be changed from .100 inch to .140 inch. The .060-inch wall thickness of the skirt, where it attached to the part, could be increased to .090 inch to finish the changes needed to open up the restricted flow path.

I explained to Mike that I thought these changes would allow them to run the material a little cooler and thus avoid the warpage issues. I also told him he might want to use staged cooling or put the parts on stainless steel fixtures and cool them under water to get rid of every last bit of warpage.

Mike then asked about the memory problem if he used fixtures to hold a part flat until the thick sections set up. He was mostly concerned about molded-in stress. I reminded him that the interior of thick walls is still molten for some length of time after the outer walls set up.

trouble_shooter.gif

The memory he was talking about begins when the molten material finally sets up. If the part walls set up while in the fixture, then it is that dimension that will become its memory position. It is the little points, like this one, that make thick-wall part molding a whole different ball game, at least when compared to more normal injection molding projects.

I suggested to Mike that he not only thicken the disk gate walls, but also increase the cross section of the points in the disk gates that transition the material from the sprue to the disk gate walls. These points were only .125 inch or so in width and would need to be increased to at least .175 inch at each of the four points to handle the extra volume of material when the sprue diameter was increased.

p53_blurb.gif

Spreader Inserts
Now, we got to the meat of the visit. How was Mike spreading out the material in the disk gate and still molding around the brass insert that served as the pulley's hub? The real question was how to do it without filling the seal groove in the brass insert with nylon. Mike set up the mold so the press operator would place the brass insert over a core in the center of the mold (see drawing, above). This holds the insert in place. Then a spreader insert would be placed on the end of the brass insert so that when the mold closes, the modified sprue bushing just touches the spreader insert to hold the spreader in place. Then the material would be injected through the sprue bushing, through the four feed points, and onto the spreader insert, which in turn feeds the material through the disk gate into the pulley's cavity.

After filling and packing out the pulley, the hold pressure setting would be responsible for packing out any voids that formed in the part's thick sections. This was the unique part of this design because the spreader insert would now be trapped between the molded-in brass insert and the disk gate. This posed no problem, though, because as soon as the operator cut the gate away from the part, the spreader insert would be set free, to be returned to the operator's table and used again.

Mike has a half dozen or so of these spreader inserts made up to allow for the 2-minute cycle, gate cutting time, and a little extra time for the spreader insert to cool before being used again. When he opens the flow path and brings the barrel heats down, he'll have to make more spreader inserts just to keep up with the new, faster cycle times.

Molding thick-wall parts is not extremely difficult. It's just that sometimes you have to use your imagination and, of course, you want to draw on past experience when you can. Mike does this all the time. I have to admit, I feel fortunate to have been a part of this new and interesting technique for molding void-free, thick-wall parts.

Sign up for the PlasticsToday NewsFeed newsletter.

You May Also Like