The Many Causes of Sink in Injection Molding

By Bob Hatch

Runners, valves, waterlines—oh, my! Which was the primary culprit?

The package that caught my attention this month contained a runner system with ABS parts attached. The first thing I noticed was how spread-out the runner system was for the four small parts being molded in this multicavity tool.

A note in the box from the molder mentioned he was having problems with sink about 5% of the time—most of it in the part’s thick section, which is pretty much straight across from the gate.

I measured all the dimensions that are important to me:

• Sprue bushing: O-diameter, .210 inch; nozzle orifice that feeds the sprue bushing, .200 inch; cold sprue length, 3.365 inches.
  
• Runners: full-round main runner, .255-inch diameter; full-round subrunners, .260 inch.
  
• Gates on each of the four parts: .160 inch deep, land of .180 inch tapered from the runner to the edge of the part, width of .130 inch.
  
• Part wall thickness: .200-.335 inch.

Possible problems

	The sprue, runner, and gates were within normal tolerances from normal dimensions, although the sprue diameter was rather small for the dimensions and length of the runner system. So, what else could be involved? It could be a check valve leaking material back over the screw flights, or it could be a ball check nonreturn valve instead of a ring and seat design. The ball check valve is used for polyethylene and polypropylene, not amorphous materials like ABS. A waterline could be plugged up or the mold waterline 1?2-inch hoses could be hooked up incorrectly—either all the time or just once in a while—which could result in hotter water flowing through the water circuits on one or more sections of the mold; this could easily cause sink on the parts.
Parts off of this long, meandering runner system exhibited an intermittent sink problem. An examination of the entire process revealed the solution.	If we go with the waterline theory, it could be that more than just the hoses are hooked up incorrectly. The quick disconnects on each of the hoses could be undersized due to the shutoff feature; larger quick disconnects might be just the ticket to more consistent cooling for the mold water circuits.
	Or, maybe the mold installers used jumpers on the waterlines to connect the outs to the ins on the back side of the mold. This simple mistake could cause the looped waterlines to pull too much heat out of the mold steel and raise the temperature of the circulating water by more than 5 deg F between the inlet and outlet points of a particular water circuit.

Remember that the total water circuit length, which is the length of any circuit from the water entry point to the exit point where it goes back to the heating or cooling system, should not exceed 4-41?2 ft (54 inches). I reduce the 54 inches depending on the pumping pressure and the mold water temperature.

If the temperature of the water circuit outlet is higher by more than 5 deg F over the inlet temperature, I start getting rid of loops and quick disconnects and shorten the overall length of each waterline circuit until I get the amount of heat extracted from each circuit down to workable limits.

I also monitor the pressure drop on each circuit and try to keep it to within 5-15 psi. Every molder should install gauges in the waterlines to monitor temperature, pressure, and gallons per minute of flow. It’s also possible that this molder didn’t have pack and hold pressures set high enough. An important point to remember here is that the injection pressure should be set to fill out the parts and control parting line flash, while the pack and hold settings should pack out any sinks and voids that show up in the thick sections.

Another option is inconsistency in the material’s melt flow that varies from one shipment to another. The material could be a wide-spec or what I hear called a “pencil prime.” (Not all wide-spec materials are bad in the melt flow area, but quite a few of them are.)

Bye-bye, sink
Now that you know the potential issues, here’s what we did to fix it. First, we replaced the check valve on the end of the screw. Next, we resized the sprue bushing, runners, and gates. The sprue O-diameter increased from .210 to .312 inch. The nozzle orifice that feeds the sprue bushing went from .200 to .290 inch. The main runner diameter increased from .255 to .275 inch, but the subrunner stayed at .260 inch. We left the gate depth at .160 inch deep, changed the gate width from .130 to .200 inch, and shortened the gate land length from .180 to .030 inch.

We got rid of the waterline quick disconnects and eliminated the waterline hose jumpers in favor of hose barb connectors. But the problem was solved by the new check valve. The rest of the changes were just frosting on the cake.

The molder called back and thanked me for all the help, but he said he had just one more question: “Where do I find a good plater that can electroplate plastic parts without warping them?” I first told him he was on his own for that answer. Nobody wants to tell you who their favorite chrome plater is; it’s like giving out the name of your favorite tool shop. If you start sharing your best sources with everyone, then the plater or moldmaker will get too busy to do work for you.

However, I went ahead and gave him the names of a couple plating companies that I used to do business with, but who knows if they are still in business. Hopefully they are.

November, 2006 - Reprinted with permission from Injection Molding Magazine. Copyright © Canon Communications LLC.