The Troubleshooter, Part 23: Dimensional problems

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.

The other day, I visited a good-size molding shop with molding machines that ranged from 500 to 1500 tons, so I knew I was going to see some big parts. The molder, Dave, has been kicking around in the injection molding business for some 25 years. Together, we went out to the shop floor where a 700-ton press was running a big hopper-shaped part with all kinds of problems.

Where Troubleshooting Starts
According to Dave, the mold had been running just great in this same machine, until recently. Now it wouldn’t run anywhere close to the required dimensions. Now the first question anyone doing troubleshooting on something that used to work should ask is, “What has changed?”

Has there been barrel and screw wear? Maybe the barrel heat controllers haven’t been calibrated lately. Maybe the computer-based process controller was fried by lightning. You never know. I asked Dave all the pertinent questions, but everything we looked at seemed to be working OK.

I then asked if the material had been changed lately or if the regrind content was different than it had been. Again, no help. Next was the mold question. Had the mold been worked on recently? Yes, Dave answered, as a matter of fact, it had just come back from having some changes made. Now, I thought, we might be onto something.

What were the changes? The mold had been sent out to have some boss sizes changed, to radius a few sharp corners, and to have a couple of other things done, Dave explained.

Before we pulled the work order to find out what exactly had been done, I thought we should investigate what part of the mold had been worked on. We shut the press down and shut off the pumps (I hate to work in a mold with the pumps running), and stuck our heads up between the mold halves to see what was going on.

With my vantage point and bifocals, I couldn’t see much. I asked Dave if he could get a toolmaker over to the press to see if they could spot any changes that might be causing the dimension problems. Mike, the toolmaker, arrived ready to help. He has been working in one toolroom or another most of his adult life and his favorite sayings are “Keep it steel safe” and “It’s only steel, we can weld it or machine it; one way or another we will fix whatever is wrong.”

I told Mike that I liked his attitude and willingness to help out. Together, the three of us pondered the dilemma for about half an hour. We reviewed the machine and material questions with Mike and then eased into the discussion about what changes had been made to the mold in question.

Mike knew the tool shop that had done the rework on the mold, so he gave them a call and spent the next 45 minutes on the phone. Meanwhile, Dave and I went to the lunch room to get a cup of coffee. We were a couple of donuts and a cup of coffee into talking about whatever molder types talk about when Mike rejoined us. The mold repair people had told him what rework had been done to the mold, so we all went out to the molding machine to see if we could retrace their steps. I was tempted to stay in the air conditioned break room eating donuts and drinking coffee instead of going out on the molding floor where the temperature was about 125F, but we needed to resolve the problem quickly.

Mike got positioned up inside the mold with his handy high intensity light and started reviewing the changes one by one. However, none of the changed areas showed any sign of what could be causing the dimensional change problems. Again Mike called the tool shop.

Detective Work
This time his questions were directed more at how the work had been done instead of what was done. He quickly discovered that work had begun with the plugging of each of the water lines to keep metal chips from getting in and plugging them up. Interestingly, when Mike asked if they had taken the plugs out when they were done making chips they replied that they were pretty sure they had.

Well, pretty sure wasn’t quite good enough for Mike. He went directly to the molding machine and checked each of the water lines, and found one with a pipe plug still in it. That, of course, meant that at least one circuit was not getting any water circulation.

Mike went back to the toolroom, got a wrench, and removed the plug. He put a quick-disconnect fitting into the hole (the quick disconnects with the 3/8-inch ID, not the ones with the 1/8-inch ID), and hooked the appropriate water lines back up. Dave started the machine, got it back into cycle, and waited for the mold heats to stabilize so he could check the dimensions on the part.

It took about 30 minutes for the mold heats to stabilize; by then, the parts were back in dimension. At this point we thanked Mike for his “above and beyond the call of duty” efforts, and he headed back to the toolroom. Dave and I went back to the air conditioning and discussed ways to keep this from being a recurring problem in every mold that was sent out for rework. From our discussions, we came up with a couple of things.

Preventive Steps for Water Problems
First, we could use one of the EMI manifolds with the little triangles in the tubes to monitor water flow. Or, one of the Cito digital-readout water line monitors that some molders are using could work. In either case, the lack of water flow in the water circuit would have been spotted and the trouble avoided.

Water problems aren’t something I run into very often, mostly due to the optimizing that I recommend. My approach to correcting dimension problems is most often to open the flow path to get the barrel melt temperatures down. This allows the water lines to remove the heat from the steel more quickly. The more heat you put into the mold, the longer it takes to get it out. Still, all the tool optimizing in the world won’t correct a plugged water line.

My techniques also won’t work as well if the water lines are hooked up using jumper hoses, usually on the back side of the mold. These cause the water in any given circuit to heat up too much. Jumpers get you into trouble. Just use delivery lines on one side of the mold and return lines on the other side.

My rules of thumb for water lines are pretty simple. No water line should exceed 41/2 ft in length, and the water flow should be turbulent, not laminar. I don’t recommend unreinforced water lines if hot water is running through the system because the rubber might plug up the bubblers or the normal cooling channels. Little bits of rubber that break out of the hoses can do that.

Also, I don’t recommend cascade-type bubblers either. I prefer the baffle design. Even better are spiral flow channels on the cores of parts. I also don’t like to see more than a 5 deg F rise in water temperature between the mold inlets and outlets.

Between monitoring the water temperatures, pressure drops, and the actual mold temperature on the mold faces, I have found that parts can be molded that are pretty much free of stress and defects. Also, a fairly quick cycle time can usually be run for a given wall thickness. It doesn’t really matter what material is being used because they all respond well to this approach.

Dave called a week later to tell me that they were now regularly getting good parts. They also implemented some other suggestions. Now the mold produces a part not only with the correct dimensions, but also with an improved cosmetic appearance, eliminating gate blush, glossy spots, and a bit of warpage that their customer had been complaining about. What suggestions did I make? Just the usual ones: open up the sprue, runners, gates, and nozzle orifice; and vent the mold, especially the runner.