Follow these tips for injection molding to get better parts faster.
It wasn’t so long ago that most injection molds had to be cut manually by machinists who worked directly from drawings. Then CAD and CNC came along, enabling milling work from digital input. Proto Labs added another layer of automation to mold design and CNC tool-path generation. However you get from model to mold, the accuracy of the original design determines the quality of the finished product.
All parts begin as 3D CAD files, but some files work better than others, and, depending on your mold manufacturer, some don’t work at all. Some files can be adapted, but following a few simple guidelines will speed up the process and help you get exactly what you want without unnecessary delays.
The more common file formats injection molding vendors can use include:
• SolidWorks Native (.sldprt)
• Pro/ENGINEER Native (.prt)
• IGES (.igs): Initial Graphic Exchange Standard
• STEP (.stp): Standard for the Exchange of Product model data
• ACIS (.sat): Andy, Charles, Ian’s System (no kidding!)
• Parasolid (.x_b or .x_t)
• AutoDesk (.ipt and .dwg, 3D only)
One common file format that may not work with injection molding is STL. This format is designed for stereolithography and, though many CAD packages offer it as an output option, it does not contain data that are precise enough for injection molding. Similarly, 2D files, wireframe models or .dxf (Drawing Interchange Files) do not contain all the information needed for the injection molding process.
Tips for efficient injection molding
If you must rework a part design, it is better to undo whatever needs changing than to patch it. For example, if you create a hole that you later decide you don’t need, plugging the hole is not the same as deleting the feature and recreating it. Patching can create internal surfaces, which can be confusing as there is no way to be certain if the internal surfaces are errors, parts of an assembly, or a garbled model.
You can, however, “join” separate parts to create a single part, as long as you do so within the software.
Set export tolerances as high as possible—1/1,000th is good; 1/10,000th is better. This ensures maximum accuracy in your final part. For a variety of reasons, saved files may occasionally appear incomplete. This can often be resolved by resaving the file in a different format.
Some features not only don’t need to be drafted, they work better if they aren’t, especially if you are using the Proto Labs process. The features in question are typically screw holes used to connect plastic parts—front and back halves of a plastic shell, for example—with thread-forming screws. The holes are formed by posts in the mold called “cores” (see Figure 1).
Protomold can produce high aspect-ratio small diameter holes using steel core pins in the mold. For example, with a part with a ¾ in. deep, 1/8 in. diameter hole (see Figure 2), you include that feature in your 3D CAD model. Protomold’s proprietary software will design the mold with a cylindrical steel core pin for forming the hole (see Figure 3).
This innovation changes two things for you. First, you can have mold parts with deeper, narrower holes. Second, you don’t have to draft those features. A steel pin is strong enough to handle the stress of ejection and its surface is smooth enough to release cleanly from the part without draft.
The size of the hole in your part will be determined by the size of the thread-forming screw you’ll use for assembly. The hole itself will be slightly larger than the minor diameter of the fastener—the diameter of the shaft at the root of the threads. Typically, the manufacturer will specify a diameter for the pilot hole in their screw specs. Finally, some screws will be specified for particular plastic resins, so if you change your resin during prototyping, make sure you’re still looking at the right type of fastener.
The No.1 rule
According to Dave Nyseth, customer service engineer, Proto Labs, there are four Number One rules; “Ignore any one of them and it can stop you in your tracks.” Thus,
1a) Maintain uniform wall thickness
1b) Maintain appropriate draft
1c) Understand the resins you plan to use
1d) Understand the molding process
Uniform wall thickness
Uneven wall thickness is an open invitation to problems. Depending on gate placement, it can lead to incomplete mold filling if resin has to pass through a thin area to reach a thick one. Because resin shrinks as it cools, thick areas may shrink more than thin ones, which can lead to warp in the finished part. So, if walls are to be identical (or at least similar) in thickness, what should that thickness be?
If it is too thin, parts won’t be structurally sound, but if they are too thick they may shrink enough to cause unsightly, potentially risky surface sink. Also, because dissolved gases are released as resin cools, thick walls can develop bubbles at or below the surface, weakening the part. The ideal thickness of a wall will depend on its function and on the resin used.
If a feature needs to extend above or below the rest of the part surface, it need not be thicker than the adjacent areas. Instead, it can be designed as a cored-out feature rather than a solid one.
Sometimes when an undrafted part is ejected from a mold, the sliding surfaces deform rather than hold their shape. Proper draft ensures that the part surface and mold surface will draw apart instead of being dragged across one another during ejection. The required degree of draft to avoid damage depends on a variety of factors including height, location, and surface texture of the feature.
Draft is almost always required for surfaces that are parallel to the direction of mold opening. In parts with cam-driven side actions, draft is also required for surfaces parallel to the direction of cam action. And shutoffs—surfaces where mold faces meet—that are parallel to the direction of mold or cam opening require draft as well.
While the characteristics of various resins differ across too many dimensions to discuss in detail here, there are important issues that can affect the molding of your part:
• Mechanical properties such as strength can be an issue; stronger resins may require less material to meet your requirements.
• Shrinkage varies among resins and can definitely affect moldability. This can be of special concern with filled resins, which shrink unevenly depending on the direction of resin flow.
• Viscosity, and the ability to fill small features, also varies among resins.
Making choices at the resin buffet
Choosing a resin for a plastic part can be as important (and challenging) as designing the part itself. You can pore over information in books or on the web, but you may not know for certain that the resin you’ve chosen is the right one for the job until you make and test some prototypes.
Of course there are times when the choice of resin is obvious, when there is a special requirement that overshadows all others and only one product will do. On the other hand, there are applications in which the requirements aren’t very challenging and any of a number of resins would meet the need. More often, however, the choice of resin is a matter of balancing competing demands and finding just the right combination of cost, cosmetics, moldability, and performance. In such cases there may be several serious contenders when you’re ready to start making prototypes.
Statistics on a resin data sheet, which you can find at such sites as www.matweb.com, www.ides.com, and others, can help narrow the choices, as will knowing how many pieces you will likely produce. The time and energy invested in finding the right resin product can bring substantial returns.
Instead of trying resins sequentially, though, have several versions of your prototype made at the same time and test them side-by-side. This can be done either by machining or molding prototypes.
If you want to compare resins with significantly different shrink coefficients, you can have prototypes machined from solid resin. This should allow you to at least reduce the resin choices to a set with similar shrink coefficients.
If you’ve narrowed down your resins to a set with similar shrink coefficients or if your tolerances are not particularly tight, you can injection mold your prototypes. Keep in mind that each mold is manufactured with a shrink coefficient. So, if you need to use a different resin and its shrink coefficient is different, a new mold may have to be cut through CNC machining.
The sequential use of CNC machining followed by injection molding will help you fully explore your resin options while avoiding excessive tooling costs, and the additional production time for molding parts in more than one resin will be negligible—probably just a matter of hours. The information you’ll get can be invaluable.
For a small incremental cost for several prototypes, you can answer questions such as: how does your part look in apple red versus lemon yellow? How opaque is each resin at a particular point in the design? How does it stand up to a three-foot fall, a sharp rap with a ball-peen hammer, or an hour in a closed car on a sunny afternoon? How much glass fill does it really need to meet your strength goals? Will it warp or sink?
With parts in several resins you can get right down to serious comparative testing and stay on schedule while you let your customers take a look at them, check their mechanical properties, evaluate their fit with other parts, test them in different environments, and so on.
Pushing your parts around
Ejector pins are the ‘bouncers’ of the injection-molding world. They apply a force to eject a part from the mold, and in some cases can leave marks. Pins are located in the B-side mold half, the side in which the part will stay when the mold opens. Once the mold is opened, the pins extend into the mold cavity, push the part out, and then retract, allowing the mold to close and be refilled.
Ejector pin placement depends on a number of factors. Obviously the shape of the part is one. Factors like draft and texture of sidewalls and depth of walls and ribs can increase the likelihood that areas of the part will cling to the mold.
Resin choice can affect pin placement or size. Some resins are ‘stickier,’ requiring more force for release from the mold. Softer resins may also require the use of more or wider pins to spread force and prevent puncturing or marring of the cooled plastic.
To be effective, the pins need a flat ‘pad’ to push against, and the surface of the pad must be perpendicular to the direction of pin movement. If the part surface at that location is textured, the smooth surface of the pad will be apparent. And if the surface of the part is not parallel to the flat end of the ejector pin, the cosmetic impact will be even more obvious.
In a traditional steel production tool it may be possible to machine the end of the pin to match the contour of a part surface that is not perpendicular to the direction in which the pin moves, producing a contoured pin.
Because it is in a different plane than the part surface, the pad may be raised slightly above the part surface at one edge or recessed slightly below the part surface at one edge.
Configuring a pad that is slightly recessed into the part surface is the default configuration for pins on contoured surfaces.
A post gate produces an extreme example of a raised ejector pad (see Figure 5). In cases in which an edge gate cannot be used, resin is injected through an extension of an ejector pin channel. When the part has cooled, the ejector pin pushes against the resulting post and, in the process, clips off the runner. The post is typically removed from the finished part in a secondary operation.
In most cases, ejector pads (or the vestiges left by their removal) are on the non-cosmetic sides of parts. In some cases, however, this may not be possible. Take for example the case of a clip formed using a pass-through core (see Figure 6). In this case, because the clip increases the surface area of that side of the part, the ‘clip-side’ part surface will adhere more strongly to its mold half. This will make that mold half the B-side. The clip would normally be on the cosmetic side of the part, but its presence requires that ejector pads also be on that side of the part.
All of the above cases assume that there are surfaces against which pins can push to eject parts from the mold. There are, however, some designs in which there are no such surfaces.
Take, for example, a grate, in which all that faces into the B-side mold half are the tops of ribs. If the rib edges do not provide enough surface area for the pins to push against, you would need to add some bosses to act as ejector pads. MPF
This material was excerpted from the Design Tips white paper by Proto Labs.