The Roller Skate (Part 2): A Change in Plastic Material & Manufacturing Method is Troublesome

Apply for Access to 84,000 Datasheets • Share this page

A memorable mistake reveals there is more than one way to correct a processing defect.

In the previous article (The Roller Skate, Part 1: Handling a Delicate Situation With Data), we reviewed stress cracking around the molded-in steel insert in a roller skate undercarriage. Once this obvious defect was eliminated, our attention was drawn to two equally troublesome, but subtler, problems.

The 10.3-inch-long part had warped 0.270 inch along its length. ThAhrinis part was so stiff that during assembly it deformed the sole of the leather skate shoe.

There were also problems with the two teardrop-shaped bosses near the ends of the part. What looked like coring in these large bosses was actually voids or bubbles. These voids interfered with drilling the skate wheels’ mounting holes.

The cardinal rule for the design of all injection molded parts is to maintain a uniform wall thickness. The roller skate undercarriage violated that important rule. The wall between the teardrop-shaped bosses measured 0.200-0.370 inch. The thickness of the wall surrounding the metal inset was 0.270 inch. The area under the heel measured 0.325 inch and the section under the toes was 0.395 inch thick. The two teardrop-shaped bosses were 0.825 inch high and 0.875 inch across.

Locating the gate in a thin section of this part caused a pressure drop, making it difficult to pack out the thicker areas; it also meant that material in the thin portion solidified quickly, preventing flow from reaching the other end.

Defining the problem

The problem with varying wall thicknesses is that thicker sections take longer to cool and that they shrink more than the thinner sections. This nonuniform shrinkage created sink marks, high levels of molded-in stress, and warpage along the length of this part.

Examination of these parts under polarized light confirmed that they contained a high level of molded-in stress. These stresses were, however, less than they could have been. This was due to most of the changes in wall thickness being gradually blended together; abrupt changes in wall thickness would have resulted in even higher levels of stress.

Injection molding is a melt-flow process. Injection molds must provide a flow path from the nozzle of the injection cylinder through the gate and into the cavity. The design of the entire flow path is important, but the size and location of the gate is critical. As a general rule, the thickness of a gate should be 50% of the part’s thickness. The rectangular gate on the undercarriage was 0.187 inch wide and 0.125 inch thick—too thin for this part. This resulted in the gate solidifying and stopping any additional flow into the cavity long before the thicker sections had cooled and stopped shrinking.

It is equally important to gate a part in its thickest section. This allows the melt to flow from the thicker into the thinner walls, which is ideal for injection molding. The gate on the roller skate undercarriage was located a little forward of the teardrop-shaped boss under the heel, where the wall was only 0.200 inch thick. In this instance, the part was gated in the thinnest section, which is the exact reverse of the recommended location.

With this gate location, the melt had to pass through the thin section in order to flow into the thicker walls in the heel and the toe areas. As the melt flowed through the thin section, there was a pressure drop. This reduced pressure made it difficult to pack out the thicker sections.

Another problem with this gate location was that the thinner wall, just like the too-thin gate, cooled and solidified relatively quickly. This prevented melt from continuing to flow into, and compensate for additional shrinkage in, the thicker section at both ends of the part. This was especially troublesome in the case of the two very thick teardrop-shaped wheel mounting bosses. The inability to continue to flow melt into the bosses and the reduced packing pressure accounted for the voids.

These problems, plus the stress cracks around the metal insert, prevented the lightweight roller skate product from being introduced in time for the Christmas gift-buying market. The best approach to resolving these problems would have been to start over. This would allow the part to be redesigned to produce a more uniform wall thickness. A new mold would allow an improved melt-flow-path design.

However, the OEM was not interested in a redesigned part and a new mold. It was only interested in a compromise between what was best and what was good enough. A suggestion to increase the coring out from the back side in the thick heel and toe areas was rejected for fear of a loss in strength. A small amount of coring out was allowable in the two teardrop-shaped bosses. The rest of the design had to remain the way it was.

Corrective action

The two-cavity mold for this part was a relatively simple two-plate design with ejector pins and edge gating. It was fairly easy to relocate the edge gate to the 0.395-inch-thick section near the metal insert. This location was chosen as the thickest region with the greatest mass of plastic material. At the same time, the gate thickness was increased from 0.125 to 0.187 inch. The thicker gate took longer to cool and solidify. That allowed more melt to flow into the cavity to compensate for shrinkage in the thick sections.

This modified gating arrangement did a better job of molding the thick section at the toe end of the part, but it didn’t do anything for the thick sections at the heel end of the part.

Fortunately, this part had a triangular-shaped reinforcing rib extending between the two teardrop-shaped bosses. A solid section at the tip of this rib measured 0.300 by 0.370 inch. This thick rib became an internal runner that rapidly conveyed melt into the heel end of the part.

In this new cavity-filling sequence, the melt flowed from the relocated gate through the thick section, into the teardrop-shaped boss, and around the metal insert. Once these areas were partially filled, the injection pressure increased and the melt raced through the reinforcing rib into the teardrop-shaped boss at the heel end of the part. The melt then spread out to fill the rest of that end of the part. Simultaneously, melt flowed down from the tip of the reinforcing rib to fill out the thin sections between the teardrop-shaped bosses. This flow path arrangement allowed for high-pressure packing of all of the thick sections as they cooled and shrunk.

The relocated gate eliminated the voids in the teardrop-shaped bosses. The end-to-end warpage was reduced to an acceptable 0.045 inch. Parts from the modified mold enjoyed six or seven years of commercial success before being obsoleted by inline skates in the mid-1990s. I blamed all of this part’s problems on the OEM’s lack of experience. This was only partially true. I found out later that this exact part had been successfully produced for years as a diecast zinc component. In this case, the company simply changed the sprue bushing and added the threaded mold inserts to use the same tool to mold polycarbonate parts that were lower in cost and lighter in weight. This was a new experience for me and that is what makes this a memorable mistake.

The moral of this story is that each plastic material and process combination has its own set of part design guidelines. There is nothing wrong with specifying a never-before-used material or process; however, anyone doing that should consult the trade literature or someone with experience in that field.