Analyzing plastics with FEA: Part 7

Finite-element analysis has many faces, as this series continues to verify. One of the most obvious uses of FEA for injection molded part design involves mold filling simulation. C-Mold's Stewart Barton and Leray Dandy believe one of the best functions for this CAE tool is to apply mold filling simulation to the gas-assist injection molding process. "Gas injection can be unpredictable," Barton reports, "and in addition, any slight change in design, material, or processing can affect gas penetration. Rather than relying on expensive trial and error, simulation can cut both time and cost from the design process." Additionally, gas assist boasts reduced cycle times with thick sectioned parts and improved structural performance. Both of these benefits, according to Barton and Dandy, can go unrealized if parts aren't designed to properly take advantage of them.

What can simulation offer? "It has the ability to quickly combine the effects of four key areas that affect the molding process-material, part design, mold design, and processing-and to predict how gas penetration changes under different combinations of these variables," Barton says. In addition, simulation offers a nearly limitless number of virtual mold trials at low cost, early in the game before any investments in tool steel have been made.

Material Matters
Resins react differently to gas injection, according to the C-Mold team. To prove its point, the team conducted two identical simulations, one using polycarbonate, another polypropylene. For the demonstration, team members used a spiral mold design often used in gas-assist experiments. The sample part was molded in PC, and is often referred to as a hollow molding because the entire part is basically a single gas channel.

In the simulations, injected gas was used mainly to pack the resin during post-filling, so cavities were filled to 100 percent shot size prior to gas injection. Comparing the two tests, shown in Figures 1 and 2, shows that polypropylene allows gas to penetrate much farther than PC. This relates well to properties of the two resins-namely, that PP compresses and shrinks more than PC.

Simulating the gas-injection process using the correct material data can help designers accurately gauge how far gas will penetrate.

By Design
Part design factors that exert a significant influence on gas penetration include gas channel size and layout. In the case of size, Barton and Dandy recommend a gas-channel-to-wall-thickness ratio between 2:1 and 4:1 for sufficient gas infusion, shown in Figure 3. If channels are nearly as thick as walls, gas can penetrate into thin-wall sections uncontrollably, a phenomenon known as gas fingering, because gas seeks the path of least resistance to reach the last areas to fill. If the channels are too large, polymer races through the channel ahead of the melt front in the thinner walls. This causes the melt to racetrack so far ahead that it can back-fill into the thin walls and cause air traps, burning, fingering, or gas blow-through.

By analyzing a plaque part with wall thickness of .100 inch, Barton and Dandy illustrate these principles. With a channel diameter of .200 inch, the channel-to-thickness ratio is 2:1. The simulation (Figure 4) shows gas fingering into thin walls.

Changing the channel diameter to .600 inch for a 6:1 ratio, the simulation shows a distinct racetracking problem prior to gas transition. After gas is added (Figure 5), fingering occurs. Barton explains that the low-pressure, last-to-fill areas are closer to the gate. As a result, the gas goes through the thin walls to reach these areas rather than going through the polymer-filled channels.

One of two final simulations (Figure 6) shows acceptable gas penetration patterns for the same plaque using a channel-to-thickness ratio of 4:1. In Figure 6, gas channel layout was changed to allow a more direct path from the gas entrance to the low-pressure areas. This makes it easier to contain the gas within the channels rather than having it migrate to thin- wall sections.

Gates and Gas Pins
Mold design for gas assist can make the difference between success or failure. Both the polymer gate and gas pin locations influence how and where gas will penetrate. Simulations can help predict the optimum sites before tooling is cut.

To illustrate, Barton and Dandy simulate the filling and gas injection stages for an automotive door handle. The part is molded in polypropylene, with varying wall thickness from .050 to .500 inch and a nominal wall of .180 inch overall.

Predicted gas penetration for the original pin and gas locations shows that gas is not reaching some of the thick sections, while it does permeate into thin walls at the far end of the part. Moving the gas pin to the opposite side (Figure 7) causes further gas penetration into thin sections. Moving both gate and gas pin (Figure 8) repositions the last-to-fill areas for acceptable gas penetration.

It's the Process
Key process variables can greatly influence gas penetration. Dandy and Barton believe that the most significant of these is polymer shot size. Most gas-assist parts are molded by first introducing a short shot, then injecting gas to push molten material outward, filling the cavity and creating a hollow section. "If the shot is too short, however, gas may penetrate into thin walls," says Dandy. "A full shot can prevent gas from reaching the ends of its channel, leaving thick sections of molten resin that require longer cooling times."

Other parameters such as melt and mold temperature, filling time, gas pressure, and gas hold times can also be critical. In all cases, a molding simulation can help designers find the optimum levels without resorting to a traditional trial-and-error approach.