The Troubleshooter, Part 56: Even fill with flow restrictors

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.

I received a package this week from Alsons Corp., a Masco Co., in Michigan containing six sets of runners and parts. Alsons is a company that makes shower heads and related items for the plumbing industry. The engineer who sent the parts, Joe Phillips, was concerned because the handles on these shower heads were not filling out evenly—they violated the thick-to-thin rule. The material was gated into a .140-inch-thick wall, entered a .550-inch-thick section, and then narrowed back to .140 inch.

He wanted the part to fill out evenly so that when the ABS ended up at the end of the handle, he would be able to push all of the air trapped in the mold out the parting line vents and not have any knitline-causing air trapped on the side of the handle. To compound the issue, Alsons planned to chrome-plate these parts, so any surface defects would be considered a reject.

I looked the runner over and could see that the sprue O-diameter (.250 inch) was a little bit small for the runner system (.265-inch main runner). A sprue O-diameter of .312 inch should feed the .250-inch-deep trapezoidal main runner with a nozzle orifice diameter of .290 inch feeding the cold sprue.

I couldn't tell if the runner was vented, but that would be on my list for Joe to look into since venting is crucial to parts that are going to be plated. Sometimes we even have to use vacuum venting on a mold that produces parts that will be plated, such as compact mirrors. Anything that plates out on the cavity or core gives chrome-plated parts a grainy look in the area where the plateout occurs.

The sprue and nozzle orifice size are also critical for plating grades of ABS since we don't want any shear to occur in the flow path of the material. Shear can cause the additive package to break out of the material as gas or volatiles and plate out on the polished surfaces of the cavity or core.

Joe was using a trapezoidal runner to feed each of the two subgates, which is OK. Edge gates need to be fed from the center of a full-round runner, but subgates can be fed by either a full-round, deep-half-round, or trapezoidal runner. The round subgate diameter was OK for ABS at .110 inch. Joe first needed to check on the runner venting and enlarge the sprue and nozzle orifice, but none of this would help him fill out the handles more evenly. These changes were purely housekeeping items and needed to be done to eliminate any shear points in the flow path. They would also allow the barrel melt temperatures to be set at 450 to 500F and the mold at 120 to 150F, in keeping with the requirements of plating-grade ABS. It also helps with cycle time, but Joe hadn't mentioned needing help with that.

Flow Restrictors
Joe called back a few days later and told me that after he made the changes I recommended, he was running lower injection pressures. He set his barrel melt temperatures at 500F by using setpoints of 465 to 480F on the barrel heaters. He wasn't having any trouble with streaks or surface defects now, but he still wasn't filling the handle out any more evenly than before. He was still trapping air and causing a knitline on the side of each handle with every shot.

He said he tried gluing some pieces of sectioned parts onto the core side of one of the handles in an effort to slow down the travel of the ABS on one side of the handle so it would fill out more evenly. When I asked him how well this was working he hesitated and said it helped a little bit but not enough to correct the problem. I have seen this done before but without much success.

However, I hate to give up on a new idea, so I told Joe I haven't been in favor of flow restrictors in the past, but if he wanted to try to make them work, I would do what I could to help him. As we talked, I started to see how we might be able to make these flow restrictors work a little better. (I think my biggest problem with flow restrictors is their use in runner systems where, in theory, they act to reheat the material as it flows through the restricted area, or to choke down the flow of material through a particular runner so smaller parts in a family mold can't be overpacked.)

I could see what caused Joe's early attempts to fail. He was trying to block material flow—like putting up a dam—but the material flowed over the top of the block. These dams were too thin and too short to do any good. What Joe should have done was make the restrictor longer and wider so that the wall thickness in the restricted area was less than the rest of the part wall.

On this Alsons part, this thinning of the wall would cause the material flow to move more slowly on one side of the core than on the unrestricted side. It wasn't being used as a dam, but as a slowdown area.

Joe seemed to buy in and said he couldn't wait to give it a try. He glued some longer and wider pieces of parts to the side of the core where the material was flowing better than on the other side. He then made the ABS pieces about half as thick as the wall on which he was trying to slow down the flow and wide enough to do some good.

It took five or six tries, but he finally got what he was looking for. He had the toolroom permanently attach a piece of tool steel to the cavity in the same shape as the ABS piece. With the permanent restrictor in place, he was getting even flow down all sides of the core and moving the trapped air in the mold to the parting line vents. The secret was in thinning the wall to slow the material flow instead of trying to build a dam to hold the flow of material back.

Plating Problems
Joe called again a few days later and said he was molding evenly filled parts now, but he could see defects in the plating that were unacceptable. He said he was still getting some sink that he couldn't pack out, and these sinks really showed up after the part was plated. I told him that when I teach seminars on this subject I explain that sinks and voids are directly related. You either get a void in the thick section of molded parts or you have sink on the surface of the part, just above the void. A sink is actually a void that has collapsed and allowed the surface of the part to cave in.

This is why a lot of molders like to drop parts right out of the mold into a water bath, which is not a good idea if you are going to plate the parts later. They want to get the surface of the part to set up quickly so the voids will form in the thick sections and not let the surface cave in to form a sink. The problem with this method is that the plating process will relieve the stress caused from the water bath. My approach has always been to fill and pack the part first, and then use hold pressure at a setting equal to the injection pressure, or sometimes as much as 20 percent higher, to fill out any voids as they form in thick sections.

Of course, you have to keep the flow path open so the material continues to move when you start pushing with the hold pressure. An undersized gate depth or overly long gate land causes these thin sections (such as a small-diameter sprue or an incorrectly shaped and sized runner) to set up before you can get the voids filled. These thin sections do not allow the material to flow through the part walls and fill out the voids. The result is voids in the thick sections of your molded part.

Joe asked me to review one more time the best heats, speeds, pressures, and mold temperatures for molding plating-grade ABS parts. I first told him that I suspected his plating rejects were being caused by molded-in stress. It goes without saying that he needed to properly dry the ABS in a well-maintained desiccant dryer. He should run his melt temperature at 450 to 500F in the barrel and use a warm (120 to 150F) mold to keep from molding in stresses that would show up after the parts are plated.

I told him to use a medium to medium-fast injection speed and try to keep the injection pressure in the 800-psi region, with the hold pressure pretty close to the same if he was having sink problems. The screw rpm had to be run at a medium speed and the backpressure minimal—probably 50 to 100 psi for most ABS applications, and a little more if he encountered dispersion problems.

I also suggested to Joe that he look up one of my earlier articles, which discusses testing for molded-in stress of plating-grade ABS parts by using glacial acetic acid as a stress promoter ("The Troubleshooter, Part 24: Plating Problems with Platable ABS," June 1998 IMM, pp. 102-108).

This was enough for Joe to stop talking and start molding parts. I heard from him again in about a week and he told me that the parts were looking great, but he was still getting a little bit of sink that he couldn't eliminate. He decided to pursue blowing agents and is now exploring that approach to remove the last vestige of sink.