The Troubleshooter: What causes parts to warp?

Almost anything you can think of causes warpage in one form or another. This time it’s the omission of a little dimple.

The part for this month’s story is a storage shelf for a refrigerator or freezer made of high-impact polystyrene (HIPS). Before doing anything else, I got on the computer and went to the free section of the IDES.com website and looked up tensile and impact values for HIPS, ABS, and PC/ABS alloys, and easily determined that HIPS was not the best material for this application.

The tensile and impact values were quite a bit lower than most of the ABS products and the ABS values were high enough to eliminate tensile and impact issues. The current pricing of ABS makes it a good substitute for HIPS these days, and the additional cost of the PC/ABS material would not be justified in this case. ABS is an easy drop-in replacement for HIPS since the shrinkage is the same and they are both amorphous materials.

Next, I addressed the warpage. The bowing or banana shape of this flat shelf could be attributed to mold water problems, molded-in stress, too fast a cycle time, or many other possibilities. I just had to figure out which one it was.

In this case we have a hot runner system involved. I generally find the hot runner flow tube, crossover runners, and drops from the crossover to the surface of the part all to be designed quite well by the companies that are in that line of business.

However, sometimes molders take these well-designed hot runner molds and try to feed them with machine nozzles that have not been drilled out to match the flow tube or bore diameter. Still, failing to size the nozzle orifice correctly or taper the gate is about all we can do to screw up a good hot runner mold.

The transition dimple

Next, I checked the gates and, sure enough, I saw a problem right away. The hot drop gate was recessed but I couldn’t see a transition dimple on the core underneath the gate. This one mistake could indeed be all it takes to mold in a lot of stress and cause warpage. The rule for hot tip gates in either a three-plate or a hot runner mold is simple: If you recess the gate, you must have a transition dimple or melt puddle under it to maintain uniform wall thickness.

The part’s nominal wall thickness was .110 inch and the recess at each of the four gates was .050 inch. This reduced the wall thickness at each gate to just .060 inch. Without a transition dimple, material flow was reduced through each of the gates, resulting in thin-to-thick flow. How much did it reduce flow? I have not calculated it but I generally tell molders and toolmakers in my seminars that this condition can easily reduce flow through the gate areas by a minimum of 25%.

How does the molding technician get the material to fill and pack the part under these conditions? Generally he raises the barrel heats and increases injection pressure, which also means a longer cycle time to keep the part from warping. Possibly the only reason for warpage on this part was that the molding technician did not slow down the cycle to accommodate the higher barrel heats. This is a common problem in the molding shops I get called into when they want to reduce rejects and speed up cycle times.

Upon further examination, I could see the hot tip gate diameters were all .050 inch and the recessed area diameters were .200 inch. I would say that is just about right—if they had just remembered to include transition dimples in their design (.050 inch deep and .200 inch in diameter). I also could see that all the gates were vestige free, which told me the hot tip gates must be tapered, as they should be; otherwise, I would see evidence of vestige at some of the gates.

Time to radius

I could see a couple of other problems, not necessarily involved in the warpage issue, but still things to bring to the moldmaker’s attention. First, with my trusty medium ballpoint pen I drew a line at the intersection of the side walls and the flat nominal wall. The result was a double line. This told me that no one bothered to put a radius along the length of this 12-inch-long wall. I consider this to be an inside corner because it is two walls of equal thickness coming together. The radius used on an inside corner should be equal to 50% of the common wall thickness. In this case, it would be a radius of .055 inch since the nominal wall thickness was .110 inch.

Looking further, I saw we had the same lack of radius condition at each of the cored-out rectangular areas on the flat section and also where the tabs attached to the nominal wall on two sides of the part. This meant we had two side walls, 16 cored-out holes, and eight tabs attached to the nominal wall, all needing a .055-inch radius. (The radius required for the tabs is on the side of the tab instead of an inside corner, but the rule is the same.)

It’s interesting how moldmakers sometimes work on the complete mold but neglect the details required for proper part design. I think I know why it happens. Moldmakers are trying to be sure all the pieces of the puzzle come together properly, such as the cavity and core, ejector pins, waterlines, and so on. They sometimes see the big picture but miss critical areas in each of the component sections.

In this case they missed gate requirements for proper flow, radiusing requirements to avoid cracking and breakage, and cycle time issues to avoid warpage—plus, who knows what else they might have missed in the waterlines or in the area of proper draw polishing to avoid knockout pin marks on the cosmetic surface of the part? So many problems, so little time.

The solution in hand

I called the molder and had my discussion with him, which he said he would pass on to the moldmakers. A couple of hours later I heard from the moldmaker. He said this mold was built like all their other molds, so why it was different?

I told him that different wall thicknesses, bigger or smaller parts, or parts molded out of different materials might have fewer problems with warpage than this particular combination of part wall thicknesses, gate design, or radius issues. I added that the older molds could certainly use these same design treatments and would produce even better parts.

He listened, discussed some of the issues, had me send him pages out of my seminar books so he could see exactly what I was suggesting, and then went off to make the changes.

A few days later he called back and wanted to buy half a dozen copies of my book, On the Road with Bob Hatch, so they could use my stories as a guide to optimizing their existing and new molds.

If you want a copy of Bob’s book, visit www.bobhatchonline.com or call (515) 326-2598.