The Troubleshooter, Part 41: Moldfilling analysis derailedThe Troubleshooter, Part 41: Moldfilling analysis derailed
July 12, 2000
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.
Recently, I worked with a company that had to learn the hard way that a moldfilling analysis is only as good as the moldmaker that builds the mold.
This company came to me with a domed part that was anything but flat, and flat was what it needed to be. The material was a mineral-filled nylon 6/6 and the part was center gated.
Right off I reminded the molder that a center gate keeps a part round and an edge gate keeps a part flat. It’s not easy to keep a center-gated part flat. But the molder insisted it could work, even though he’d been trying for more than six months without success.
The company started out by getting the best moldfilling analysis it could find, and since the cost for the analysis was in the $10,000 range, I would have to say it was also the most expensive. In general, moldfilling analysis costs have come down. Cost for a quick fill analysis ranges from $1000 to $2000, and for a full-blown moldfilling analysis it’s $3000 to $5000. In this case, the analysis printout was impressive, which is to say there were many pages with a lot of color (mostly red).
The analysis determined that the center gate was the perfect place to gate this part. The waterlines were laid out in a circular pattern that would best extract the heat from the mold steel and eliminate any hot spots on the surface of the mold. The mold temperature calculated as most efficient for this part was 180F; the report said this would be the mold temperature that would keep the part flat as it cooled on the table.
Building From Analysis
Once the moldmaker received the moldfilling analysis he began to build the tool, working off and on for more than six weeks. He used H-13, S-7, and A-2 steel in the places he was supposed to, but he had two problems to work out. First, he didn’t receive any information on what size the sprue bushing needed to be, or for that matter, whether it should be a heated or cold sprue bushing. To save money, the moldmaker installed a cold sprue—a mistake he would later have to correct. He also chose to use a sprue bushing with a 3/16-inch ID—yet another mistake that would be realized after the mold was sampled.
The other problem was that the moldfilling analysis did not indicate ejector pin and blade locations. The moldmaker decided to install ejector pins at rib junctions and just inside the edge of the outside ring of the part. Ejector blades were installed on a few of the curved, thinner ribs near the center of the part; a few more ejector pins were put on each side of rib intersections. The moldmaker was determined to ensure even part ejection, and he wasn’t going to let any ribs stick in the mold and break away from the main part during ejection.
With the ejector pins and blades installed he went to work on the waterlines. Several pages of the moldfilling analysis were dedicated to waterline requirements, but, now that the ejector pins were installed, conflicts arose. How to get the proper number of ejector pins in a mold and still put the waterlines where they need to be is a common challenge.
Doing the best he could, the moldmaker in this case put waterlines in and around the ejector pins until he was sure he had enough water to cool the mold.
The effective size of a water circuit is as big as its smallest restriction. |
Sampling Uncovers Problems
When the mold was sampled for the first time, however, everyone discovered just how difficult this problem could be. The 1/4-inch-thick part came out of the mold looking pretty flat, but within 2 to 3 minutes it began to curl up from the center of the part to the outside edges. To counter these effects the molding technicians cooled the mold and lengthened the cycle. Still the problem remained.
It was at about this point that I got called in. I looked at the melt flow, measured the parts to see how serious the warpage was, and found the parts to be about .080 inch away from being flat. After that, the first item of business was to find out how much warpage would be tolerated. The answer was zero.
Since we don’t generally operate at zero tolerances in our business, the problem immediately became more serious. We could, of course, machine the parts to achieve a zero tolerance dimension, but unless the parts are small and the material has a low shrinkage, we normally expect to have a tolerance applied. In this case, we agreed that ±.020 inch would be an acceptable tolerance since this part would be bolted to a piece of steel.
My next step was to review the moldfilling analysis; this is when I discovered the waterlines laid out in a beautiful circular pattern. Unfortunately, the analysis didn’t look anything like the waterlines in the mold itself. When I asked the molder and the customer what happened to the waterlines, no one knew.
Waterline Basics
The next step was to bring in the moldmaker for the next day’s session. The moldmaker arrived the next morning and went straight to the toolroom to look at the mold. He studied the mold intently and soon announced that the mold was exactly as he had built it, and that he had followed the moldfilling analysis recommendations almost to the letter. The only variance he knew of involved the change from a circular waterline pattern to a series of straight-through waterlines.
We next discussed the merits of using circular waterlines to even out heat distribution and heat extraction in the mold. After reviewing the reasons for using a circular waterline pattern, as recommended by the moldfilling analysis, we looked at the mold again. In addition to the waterlines running straight through both halves of the mold, we saw quick-disconnect couplers hooking the outside waterlines to the water channels in the mold. These were not just any quick disconnects, either. They had a small 1/8-inch ID that we would have to overcome. I explained to all assembled that the effective size of a water circuit is only as big as its smallest restriction. In this case, even though we had 1/2-inch hoses and 7/16-inch water channels, we really had only a 1/8-inch-diameter water circuit with which to work.
Not only that, but when the moldsetters installed the waterline hoses, they used jumpers on the back side of the mold. This is not recommended because it causes the water temperature to increase more than 5 deg F between the inlet and the outlet of that particular water circuit. This can create hot sections in the mold, leads to warpage, and generates glossy and dull spots on the surface of the part.
For the moment we were stuck with the waterlines not being circular as suggested, but we could improve the situation by going to 3/8-inch quick disconnects. We could also eliminate the quick disconnects altogether and use hose barb connectors. The jumpers on the back side of the mold could be eliminated with straight-through waterlines; delivery lines would enter on the operator side of the mold, and return lines would exit on the back side of the mold. It was not a perfect solution, but it was worth a try.
Feeding a Big Part
At this point we were closer to understanding why the parts were warping. The part design seemed to be OK, the center gate seemed to be appropriate, so what else was wrong?
I drew everyone’s attention to the fact that there was a small-diameter sprue bushing feeding what amounted to a big part. Surprisingly, not everyone realized this was such a big part. The wall thickness was a nominal .250 inch, and the part was 10 inches in diameter. It also weighed 1.5 lb, making it, in my mind, a fairly big part.
To provide enough material to fill and pack this part with a cold sprue I always recommend that the sprue O diameter be at least equal to the nominal wall thickness. Sometimes, if I am dealing with stiffer flowing materials, I will recommend a sprue O diameter 11/2 times larger than the nominal wall thickness.
Troubleshooter's Notebook |
---|
PART: Domed part with flat base molded of mineral-filled nylon 6/6. TOOL: Single-cavity, all steel cold sprue bushing. SYMPTOMS: Warpage on the flat base part. PROBLEM: Improperly constructed, sized, configured cooling channles; small sprue bushing; no vents. SOLUTION: Increase ID of quick disconnects from 1/8 inch to 3/8 inch; eliminate looping of waterlines; change to heated sprue bushing; drill out nozzle to 1/2 inch; add vents. RESULT: Warp reduced to within customer spec. |
We discussed this and realized we were quite a bit undersized with a 3/16-inch sprue O diameter and decided a sprue with a .312-inch O diameter would be better. But that would mean that the root diameter of the sprue—and this was a long sprue—would be about 1/2 inch, requiring a milling or smoothing operation at the gate once the sprue had been cut away. Everyone agreed that this would be unacceptable. Instead, we agreed that a heated sprue bushing should be used, and the orifice size of the bushing should be the same size required for the depth of an edge gate, or, in this case, .187 inch.
Heating Things Up
Heated sprue bushings effectively open the flow path of the material from what it once was. With this flow path opened, the barrel melt temperatures can be lowered to the material manufacturer’s minimal recommendations, bringing the heats down, in this case, from 590F to around 510F. In my opinion this change has the greatest effect on reducing part warpage.
Of course, there are two important points to cover when changing to a heated sprue bushing. First, the molding machine nozzle has to be drilled out to match the flow tube diameter of the heated sprue bushing. Sized correctly this results in a 1/2-inch flow path straight through the nozzle into the center of the heated sprue bushing. The other area of concern is the orifice of the heated sprue bushing itself. The diameter of the orifice is recommended to be equal to about 75 percent of the nominal wall of the part as a starting point. Though this orifice size can go bigger, it’s best not to when direct gating into a part. It is, however, OK to go larger with this diameter when feeding a runner with a heated sprue bushing.
Since this case involved a direct gate into a part we decided to use the .187-inch-diameter orifice with the heated sprue bushing. We were also required to relieve the land of the heated sprue orifice at a 45° angle, making it a full taper back up from the .187-inch-diameter orifice into the flow path of the material.
Venting Required
Venting was another area of concern. Once I got everyone looking at the mold halves, we could see that this mold didn’t have a single vent. A round part that is center gated should have perimeter venting all the way around the parting line. The depth of the vent for a part like this should be about .001 inch. Next, go out from the parting line with a .060-inch land, and drop into a .040-inch-deep channel to a racetrack that can then be vented to atmosphere. To make it self-cleaning we should draw polish the vent lip to an A1 finish.
Why did the moldmaker skip the venting? He said that because of deadline restrictions he thought he would let the molder sample the mold first and then put in vents where needed. Perhaps he was just going to vent near the burns on the parts. As a rule, I vent as much of the parting line as I can, and if I have a runner I vent the runner first.
Getting a Satisfactory Part
After our examination, the moldmaker changed the sprue bushing from cold to heated, drilled out the nozzle, sized the heated sprue bushing orifice, relieved the orifice land at a 45° angle, vented the mold, and eliminated the waterline quick disconnects and jumpers. Once these changes were completed, it was time to put the mold back into the press.
The molder lowered the barrel melt temperatures and started shooting plastic. Immediately some differences were visible. The part surface wasn’t as glossy as before, so we warmed up the mold from 120F to 180F, which helped quite a bit. The injection pressure, which was 1200 psi before, was now about 800 psi, and the holding pressure, which was 600 psi before, was raised to 900 psi. My suggestion to raise the holding pressure was based on my hunch that the part was previously being underpacked.
The results were not amazing, but we reduced part warpage from the previous .080 inch to a little less than .020 inch. With a little more tweaking, the molder was able to get it down to .012 inch—still not at zero tolerance, but satisfactory for the customer. The customer agreed to the ±.020-inch flatness called out on the print and not to hold the molder to the just-barely-achievable .012-inch variation.
Did the moldfilling analysis help much here? Probably not, but it got us to a point where we could make selective changes and mold a part that satisfied the needs of the customer. Did the moldmaker learn anything? He probably learned quite a few things from the problem part that he will carry with him for the rest of his life. Did I learn anything new? Not really. We still had to go back to my tried-and-true optimizing techniques that always seem to work. I still don’t think moldfilling analysis is that good yet, but it is getting better every year.
You May Also Like