Troubleshooting: The questions to ask

Editor's Note: Consultant Bill Tobin of WJT Assoc. is a regular contributor to IMM and offers these ideas for the right approach to troubleshooting. 

I have always contended that most of us learned troubleshooting by following around the plant the most experienced molder, hoping to absorb via osmosis some of his expertise, knowledge, and wisdom. But troubleshooting shouldn’t be that hard.

   When a mold starts producing bad parts, there are some rudimentary questions that can help you find a solution. First of all, has the mold ever made a good part? If you have had the defect from the initial mold tryout until now, the problem probably exists in the mold or part design. If the problem just came up, and there have been no material, quality requirement, or design changes, the problem probably is in the process.

What Changed?
If the problem is process related, the age-old troubleshooting question is begged: What changed between the time you last made good parts and now? Was it the machine, tool, standards, customer requirements, documentation, resin, or some other variable? Something as subtle as routine mold maintenance or a new quality control inspector at the customer can mean the difference between acceptable and rejected parts, even though there is no discernable difference between a shipment from six months ago and one from last week.

   This leads to a new question and a strategy for more effective troubleshooting in the future. Specifically, what were the original process conditions (from the point of view of the plastic)? A common mistake is to document machine settings and never look at the plastic. Translate the conditions to what the plastic sees. Fill times, pack, and cooling are universal. Time is time.

   You must translate pressures from machine settings to what the plastic sees. To start, determine the diameter of the barrel and the diameter of the cylinder(s) of the injection unit. The difference between the two is called the intensification ratio. If the ratio is 10:1 then 500 psi on the oil generates 5000 psi on the plastic.

   With the use of a simple spreadsheet you can generate a map of each machine and easily translate the pressures. If the original settings produced good-quality parts, then going back to those settings should again produce similarly good parts. But you must record the original conditions the first time so that you can go back and find them.

   Finally, consider the possibility that you are in fact producing good parts. Customers of some molders have been known to use quality standards as a means of inventory control. If your customer has stock that he cannot convert into sales, he may reject your lot and reduce inventory to acceptable levels. To avoid this you must have a mutually agreed upon definition of a good part, even if it contains minor defects. This allows you to reject his reject.

Making of a Defect
Assuming that your customer is not misbehaving and you have identified the molding process as the source of the problem, there is still more work to do. The segment of the process generating the defect must be identified.

   Part defects are made. A short shot is a filling problem. It cannot be solved during the packing, cooling, or ejection phases of the cycle. Looking anywhere other than at the components of the filling portion is a waste of time, money, and energy.

   Where in the molding cycle was the defect made? While this sounds silly, look for the obvious. Wet material is a drying problem. Is the dryer plugged in? Is it turned on? Is the inlet at the bottom of the hopper and the outlet at the top? Has the dryer been off for two weeks? If it has it will take at least 4 hours for it to dry out before it begins to dry resin. While ABS will dry adequately below 200F, polycarbonate will never dry at those temperatures. Find the cause. Eliminate the cause.

   If the check ring system is worn, you usually have difficulty building enough pressure to push the plastic fast enough to fill the mold. This begs the question: Is the equipment malfunctioning? Bad or poorly functioning equipment makes bad parts.

Patience is a Virtue
The most common problem in troubleshooting is patience—or a lack thereof. It’s too easy to make a correction, see the effects show up in five shots, decide you have accomplished your goal, and go back to your office. The process must reach equilibrium before you can be assured that you are making good parts.

   If you determine short shots are caused by material that is too cool and you turn up the heat, it’s easy to watch the defect disappear and walk away. But 20 minutes later you have overheated material, and have traded short shots for flash. Simply put, stay with the machine until it has stabilized before declaring victory.

   The rule in troubleshooting goes two ways: Until you make a consistent flow of good parts the first time, the problem is most commonly in the tool. Once you mold good parts, if you show the plastic the same set of conditions it will almost always give you the same quality parts as you had at approval.