A systematic slant realizes slender injected parts

Simply adopting a high-flow resin grade and throwing it into any old machine is not the recipe for successfully molding a component with a flow- length-to-wall-thickness ratio of 100–200:1. “You must also consider how robust the material is during processing given the high temperatures typically adopted for thin-wall molding,” says Greg Tremblay, senior project engineer at SABIC Innovative Plastics (Pittsfield, MA). “We generally recommend a shot size of at least 25% of the injection unit’s capacity in order to minimize barrel residence time, and a smaller barrel also enables the high injection speed and pressure required for thin-wall molding.” Tremblay notes that an off-the-floor 550-ton press might normally be equipped with a 60-oz barrel whereas with a thin-wall machine, the barrel capacity might be 9 oz.

Similarly, tooling also needs to be of a robust design. “Cavity blocks and backup plates need to be thicker, and you need more support pillars,” says Jim Vanderwende, engineering manager for healthcare at processor GW Plastics (Bethel, VT). “We recommend preloading support pillars in the center by 3–5 thousandths of an inch (0.076–0.127 mm) to compensate for tool deflection,” says Sabic IP’s Tremblay.

Venting is also paramount given the high injection speeds involved. “A lot of people don’t polish vents but if you really want them to work, you have to get in there and shine them up,” says GW Plastics’ Vanderwende. “The land length for primary vents also needs to be no more than 45 thousandths (about 1.14 mm), otherwise gas won’t get through it.”

Mold polishing also helps with flow, as do certain specialty surface treatments. Incorporating radii at corners also makes a notable difference. “The difference between 5 thousandths (0.127 mm) and 7 thousandths (0.178 mm) can be incredible,” says Vanderwende. “Where possible, you want to minimize the number of corners the flow front has to round because this results in reduced pressure. Gate positioning therefore becomes critical.”

Larger ejector pins are also preferable for pushing out thin-wall parts, especially if there are a lot of features such as ribs or bosses, or any kind of resistance to ejection. “Ejector sleeves are best for pushing on ribs,” says SABIC Innovative Plastics’ Tremblay.

Ribs should also measure 100% of the nominal wall thickness at their base, as opposed to 60% for standard parts, to prevent them breaking off. “Fillets should also be incorporated in ribs and other areas to eliminate sharp corners and potential sources of notch propagation,” says Tremblay. This is particularly important given many thin-wall products are portable devices such as cellpones and notebook computers that need to comply with stringent drop-test requirements.

Quirky LCP
Liquid crystal polymer was a resin almost born for thin-wall molding, but the rules that need to be followed for its processing are somewhat different. “With Vectra [LCP], all the standard design rules apply—half of them to nth degree, and the other half to the opposite degree,” says Paul R. Chauvin, Vectra Engineering Manager at Ticona Engineering Polymers (Florence, KY). For example, “The material flow should be from thin to thick sections so that shear is initially induced, while wall thickness also needs to be kept constant. A variation of just 2% can change the flow direction, which means you lose pressure and shear, and any amount of hesitation will stop flow.”

Additional design hints include impinging immediately after the gate in order to form a flow front and prevent the resin jetting down the part. This can be achieved through something as simple as positioning the gate as deep as possible inside the mold and impinging onto a core pin, or using dual gates to impinge separate material flows onto each other.

Runner systems are considered part of the part when designing with LCP and should be designed so that they taper or step down at every branch in a multi-cavity tool as the resin flows towards the mold. This results in shear starting to be generated just when the flow front starts to cool as it approaches the tool. The runner system should also be devoid of sharp corners. “We don’t recommend hot runner systems for LCP as the manifolds are too large and you lose pressure,” says Chauvin.

Venting of the tool is also important. “There are never enough vents in a Vectra tool,” says Chauvin. “Every means should be considered, including utilizing inserts, ejector pins, and pins and holes used to form the product.”

Designing a part that allows incorporation of regrind is also advantageous for LCP. “The chopped reinforcing fibers in the regrind can flow into tighter spaces,” says Chauvin. And if processors are successful in thin-wall molding of LCP, they will achieve higher strength in parts (see LCP diagram on p. 42).

Software solutions
Among software tools available to assist processors in minimizing wall thickness and optimizing part strength is modeFRONTIER from ESTECO srl (Trieste, Italy: also see September MPW, p. 85). Essentially, the software acts as an interface between two or more design tools, such as flow analysis and structural analysis, and enables the user to change several design parameters such as wall thickness, injection pressure, and injection rate, and investigate the effect on “objectives” such as flexural strength and sink index. A graphic representation of the results allows the designer to identify the optimum solution (see graphic, top right).

Luca Fuligno, modeFRONTIER technical manager at software reseller and product engineering consultant EnginSoft SpA (Bergamo, Italy), says the software is relatively easy to use. “The designer should first write down on a piece of paper what parameters can be changed and then build up a workflow in modeFRONTIER based on small blocks that each describe an input variable,” he explains. “modeFRONTIER then automatically changes parameters and searches for the optimum solution.”

When wall thickness is one of the variables, the part’s CAD model has to be re-meshed for each simulation but because this is automated, the designer need not be present. “If it takes three hours to re-mesh a model, we cannot change that but the simulation can run over the weekend,” notes Fuligno. “In any case, the solution is found in just several iterations, which would be impossible to achieve manually,” says Fuligno.

Moldflow MPI software from Moldflow (Framingham, MA) has also scored success in downgaging wall thickness. Engineering consultancy firm Promold (Paris) used it to virtually experiment with three material options (polycarbonate, nylon, and PBT), five different part thicknesses ranging from 2.15-0.8 mm, and six different gating configurations for a headlamp housing where the initial design weighed 244g.

The constraints for manufacturing were that the parts must be produced on an 800-tonne molding machine and there could be no weld lines in the vicinity of where the primary headlamp bulb would be installed. The experiment was set up to use the full capacity of the designated molding machine’s 174-MPa maximum injection pressure so the wall thickness could be reduced as much as possible. The experiment resulted in the selection of the PBT material for a design with a wall thickness of only 1.2 mm that used a five-gate hot-runner system. The final weight of the part was 41% less.

The driver in this exercise was to reduce overall vehicle weight. Obviously, cutting 100g from the weight of a headlamp housing is not going to have a major effect by itself on the fuel economy of a vehicle, however with plastics approaching 10% of average vehicle weight, incorporating such best practices will guarantee that all new products developed will be optimized for mass. Taken together, they can have a sizable impact on a vehicle’s fuel efficiency.