Designing for metal injection molding (MIM)

Table of Contents

• 2.1 Introduction

• 2.2 Available materials and properties

• 2.3 Dimensional capability

• 2.4 Surface finish

• 2.5 Tooling artifacts

• 2.5.1 Parting line

• 2.5.2 Ejector pin marks

• 2.5.3 Gate locations

• 2.6 Design considerations

• 2.6.1 Flats for sintering

• 2.6.2 Wall thickness

• 2.6.3 Draft

• 2.6.4 Threads

• 2.6.5 Ribs and webs

2.1 Introduction

Metal injection molding (MIM) is a process by which powder is shaped into complex components using tooling and injection molding machines that are very similar to those used in plastic injection molding. Therefore, the component's complexity is of the same magnitude as those seen in plastic injection molding. The artifacts associated with the injection molding process (gates, ejector pins, parting line) are also similar to those seen in plastic injection molding and must be accounted for in design. However, since the MIM process requires multiple post-molding debinding and sintering steps, some design considerations such as cross-sectional thickness and geometry features require consideration.

Table 2.1 Comparison of MIM attributes with other fabrication techniques

^aFeatures this small could have distortion

As a general rule of thumb, components that are less than approximately 100 g and fit into the palm of your hand could be good candidates for MIM technology. A mean size of 15 g is typical for a MIM component; however, components in the range around 0.030 g are possible. Table 2.1 compares the MIM process with other manufacturing processes. Notice that MIM is limited to smaller part sizes, can provide thinner wall thicknesses, has excellent surface finish, and is suited for high volumes. Table 2.2 reviews the upper and lower specifications of the MIM process.

Table 2.2 Typical attributes produced by the MIM process

^aFeatures this small could have distortion

The following are some general design considerations which will be discussed in detail in this chapter.

• Avoid components over 12.5mm (0.5in.) thick. This is a function of MIM technology and alloy, for example 4140 and alloys that use carbonyl powder can have thicker wall sections than those that use gas-atomized powders that have larger particles. Also modifications to binder systems can be made to allow thicker sections to debind.

• Avoid components over 100 g in mass; however, 300 g are possible for some technologies.

• Avoid long pieces without a draft (2 degrees) to allow ejection.

• Avoid holes smaller than 0.1mm (0.0039in.) in diameter.

• Avoid walls thinner than 0.1mm (0.0039in.), although 0.030mm walls are possible in some cases.

• Maintain uniform wall thickness; thin, slender sections attached to thick sections should be avoided to enhance flow during molding, to avoid sinks and voids, and to limit distortion during sintering.

• Core out thick areas to avoid sinks, warpage, and debinding defects.

• Avoid sharp corners. The desired radius is >0.05mm (0.002in.).

• Design with a flat surface to aid in sintering-otherwise custom ceramic setters required.

• Avoid inside closed cavities-although some technologies such as a chemically or thermally removable polymer core may be used but is not common.

• Avoid internal undercuts-although a collapsible core or extractible core mentioned earlier could be used but are not common.

• Design with lettering-raised or recessed.

• Design with threads-internal and external.

2.2 Available materials and properties

MIM is available in many of the common structural materials for medical, military, hardware, electronic, and aerospace applications. If the powder is available in the appropriate size, <25μm, and the powder sinters to a sufficiently high density, without change in alloy chemistry, the material can be manufactured using the MIM process. Table 2.3 provides an overview of available materials, applications, and specific features that make these metals desirable.

Table 2.3 Overview of MIM materials, applications, and features

MIM properties are superior to most cast products and slightly inferior to wrought products. Cast and MIM components both have microstructural pores or voids as a result of the processing methods, where the cast voids can be large and localized owing to the cooling of liquid to solid and the MIM voids are typically fine and well distributed across the microstructure after sintering. The large, localized voids of the cast material result in the inferior properties, whereas the distributed nature of the fine MIM pores provides a better microstructure for enhanced properties. Hot isostatic pressing can be used to attain full density. Another attribute of the MIM process is that the final product will be annealed after the sintering operation, thus materials that show work hardening strengthening in the machined state (i.e., 316L SS and MP35N) may require some form of post-sinter operation to enhance the strength after the MIM sintering operation. Table 2.4 provides typical data for many MIM structural materials (steels and stainless steels).

Table 2.4 Typical MIM structural material properties

MIM is an attractive method for the fabrication of soft magnetic materials. The MIM operation provides a net shape component that is in the annealed condition, which is a requirement for the best magnetic response. Table 2.5 provides data for soft magnetic applications. Each of these different alloys has physical attributes which makes them ideal for different applications. The 2200 alloy has magnetic properties similar to pure iron but with a greater strength. The Fe-50Ni alloy has a high permeability and low coercive field which makes it ideal for motors, switches, and relays. The Fe-3Si shows low core loss and high electrical resistivity in both alternating current and direct current applications. The Fe-50Co alloy has a very high magnetic saturation and is ideal for high magnetic flux density applications. Finally, if a good magnetic response is needed in conjunction with good corrosion resistance, the 430L alloy would be the alloy of choice.

Table 2.5 MIM soft magnetic alloy properties

MIM is a viable technique for the production of copper. The copper made with MIM exhibits good thermal and electrical conductivity, thus MIM copper is a viable option for electrical connectors and thermal management applications. Table 2.6 provides copper data in comparison to other forming methods. Generally speaking the electrical and thermal properties of MIM product are more affected by metal contamination, such as iron, and only modestly affected by the density, provided a sintered closed pore condition is achieved (density>93%).

Table 2.6 Copper property comparison

The purpose of the controlled-expansion alloys is to insure good mating and/or sealing with other materials as the materials change temperature. Table 2.7 provides controlled-expansion alloy data for F-15 alloy. F-15 is also known as Kovar^TM and consists of 29% nickel, 17% cobalt and the balance iron. F-15 has a coefficient of thermal expansion that matches borosilicate (Pyrex) and alumina ceramics and is primarily used for hermetically sealing applications. Other MIM control expansion alloys such as Alloy 36, Alloy 42, and Alloy 48 exist and are basically iron with the percentage nickel added that matches the alloy number to adjust thermal expansion rate. Alloy 36 has a zero coefficient of thermal expansion until 10℃, Alloy 42 has low expansion until about 300℃ and has thermal expansion behavior similar to many soft glasses. Alloy 48 has a thermal expansion behavior which matches soda lead and soda lime glasses.

Table 2.7 Controlled-expansion alloys

Implantation of MIM components is a growing market where the primary alloys in use are F-75, MP35N, and titanium-based alloys. Table 2.8 provides MIM biocompatible alloy data for F-75 and MP35N. MIM titanium data exist, but are strongly dependent upon the manufacture of the product. MIM titanium and MIM titanium alloy properties are susceptible to carbon and oxygen impurities, thus, monitoring of these impurities in these alloys is paramount.

Table 2.8 Bioimplantable alloys

The last class of alloys discussed here is the tungsten-based heavy alloys, which are of interest because of their high density. These alloys find application in military, medical, cell phone, inertia balancing, and sporting goods applications. Some specific applications are inertia penetrators, cell phone vibration weights, golfing club weights, medical electrodes, and fishing and hunting weights. Table 2.9 provides tungsten heavy alloy data.

Table 2.9 Heavy alloys

2.3 Dimensional capability

MIM is a very repeatable process with variability in the range 0.2%-0.5%. This dimensional variability is associated with the amount of shrinkage and distortion that the component experiences from the time that it is molded to after it is sintered. Components shrink about 1% during the molding operation and an additional 15%-25% after sintering. Also, the ceramic fixtures that are used for component support during sintering may have variability in them, which results in variability of the components from one fixture to the next. Some extreme cases may have greater variability if the particular feature tends to distort or if the feature lies along a parting line, ejector pin blemish, or a gate blemish. If a dimension of a component needs to have high precision, that feature should be embedded in one piece of steel and not have the negative effect of gates, parting lines, and ejector pins. Also, core pins that form holes may be tunneled into the far half of the tool to prevent the butt shut-off from forming flash that would cause variability in the inner diameter in that region. In general, MIM variability is superior to investment casting and inferior to high-precision machining.

2.4 Surface finish

MIM produces remarkable surface finish. Typically, 0.8μm (32μin) Ra is achieved; however, a surface finish as smooth as 0.3-0.5μm (12-20μin) Ra is possible. The surface finish is a function of the size and chemistry of powders that are used, the sintering conditions, and on any secondary operations, i.e., bead blasting or tumbling. Sandblast and beadblast tend to increase surface roughness because of pitting, and tumbling tends to decrease surface roughness. Component surface roughness can also be affected by the surface finish on the tooling used to manufacture the components. Electrical discharge machining (EDM) pits can be translated to the finished MIMed component.

2.5 Tooling artifacts

2.5.1 Parting line

The parting line is where the two halves of the mold intersect. Typically, this can leave a witness line that is as small as 0.0003in. to as large as 0.001in. This is highly dependent on the quality of the tool. An example of a parting line is shown in Fig. 2.1. In extreme cases, flashing can occur along the parting line in worn or poorly manufactured tools. These tools can be adjusted to eliminate this flash by either grinding or adjusting a tooling feature that is holding the tool "open" during the molding operation. If the tool is worn, the parting line must be welded and the cavity remanufactured by processes like EDM, hardmilling, or grinding.

Fig. 2.1 Parting line blemish on a MIM component showing where the two halves of a tool come together.

The location of parting line is a compromise between maintaining a low tool cost and ensuring that the witness line does not interfere with the functionality or look of the component. The location of the parting line is initially defined to make the tool as simple as possible. Ideally the parting line is placed so that all features can be handled in either side of the tool without the need for slides. Slides raise the cost of tooling considerably. However, one must also consider that greatest dimensional repeatability is obtained when the feature to be measured is in one piece of steel. Envision parting line flash variability causing variability in the dimension across the component where the parting line is located. Also, the position to which the tool opens and closes is not identically the same with each cycle, owing to the presence of material on the mold face during processing. Although minor, this can result in a minor variability of ±0.0003in.

Another consideration about the parting line is related to draft angle. Typically, draft begins at the parting line for long components. This enables the part to be easily removed from the tool. Parting lines are typically along one plane to minimize cost; however, the parting line can be stepped to accommodate features that could not be molded any other way or for components that require certain surfaces to be free of any witness lines for esthetic or functional applications.

2.5.2 Ejector pin marks

Ejector pins are required to remove the green component from the tool. A sufficient quantity of these pins is required to ensure that the green component can be removed without distortion or cracking. As a natural consequence, witness marks where these ejector pins were located are evident. Fig. 2.2 shows a typical ejector pin blemish. These witness marks become more evident as the tool ages owing to the wear between the pin and the cavity where the pin resides. The size of the pin should be selected to allow the cavity where the pin fits to be opened to accommodate larger ejector pins as the tool ages. Ejector pins are typically round, since round ejector pins are available in many standard sizes and the ejector pin housing in the cavity block is most easy to EDM. Rectangular ejector bars are sometime used in special cases; however, the radius associated with the corner of the bar causes issues with fit-up and long-term integrity of the steel in thin sections, since these corners can act as stress concentrators for crack initiation in the tool. Ejector pins are located where the greatest ejection force is required, for example near bosses, cored holes, and ribs. The esthetics and functionality of the finished component should also be considered when selecting an ejector pin location.

Fig. 2.2 Typical ejector pin marks shown on a MIM component.

2.5.3 Gate locations

The gate is the location where the MIM feedstock flows into the cavity. As a consequence, a blemish will be present at this location on the finished product. A gate is typically located in the thickest section of the component and situated to ensure that uniform packing pressure is available across the component to prevent distortion during debinding and sintering. Gates are often situated so that the material flowing into the cavity impinges on a pin or another wall to prevent jetting of the molten feedstock across the cavity, which leads to surface flow defects. Other considerations for gate location are placement in nonconspicuous or secondary machining locations. Figs. 2.3-2.6 show different gate configurations. Fig. 2.3 shows a typical tab gate blemish, which is located along a parting line. Fig. 2.4 shows a tab gate blemish that is recessed to prevent any gate vestige from interfering with the functionality of the device operation. Fig. 2.5 shows a tunnel (also known as a subgate) blemish. Fig. 2.6 shows a center gate that is produced by either a three-plate tool or a hot sprue. This type of gating is used to ensure uniform packing density along the length of the component, which is a round nose cup.

Fig. 2.3 Tab gate blemish along parting line on a MIM component.

Fig. 2.4 Recessed tab gate blemish along parting line.

Fig. 2.5 Tunnel (sub)gate blemish.

Fig. 2.6 Center gate blemish located for uniform packing pressure and concentricity.

2.6 Design considerations

2.6.1 Flats for sintering

One of the key features that should be considered when designing a MIM component is how the component will be fixtured or set during the sintering operation. The MIM material is prone to distortion during the thermal debinding and sintering operation if insufficient support is provided. To eliminate distortion, a flat is designed on the part for sintering which also allows low-cost standard flat fixturing to be utilized. The fixturing is typically ceramic for most MIM materials and may have holes in it to accommodate bosses that the component may have along the flat surface.

In cases where it is not possible to design a flat into the component, a contoured fixture which follows the shape of the component can be used. This ceramic fixture is typically designed for the as-molded green state component size. Some fixturing can have features that accommodate both the green size and the sintered size. In this design, the component shrinks from the green state support into the sintered state support. In general, the greatest amount of support is required in the green state since the softening of the polymer during the thermal debinding operation is the weakest condition of the component during the entire MIM process. Alternatives to contoured ceramic fixturing include the use of "molded" supports to the actual component that can be removed using a secondary operation after sintering. Another technique is to use a cut ceramic shim that is cut to the desired height dimension of the sintered component.

2.6.2 Wall thickness

Wall thickness should be maintained as uniformly as possible to avoid warping and subsequent dimensional variability of the components during processing. Warpage can be the result of differences in cross-sectional thickness caused by variations in packing pressures during the molding operation, differences in binder removal time during the thermal debinding, and differences in thermal mass during the sintering operation. Other issues associated with large cross-sectional thicknesses are sinks, the potential for voids, and blister defects associated with difficulty in binder removal.

Wall thickness >15mm (0.6in.) should be avoided and wall thicknesses below 10mm (0.4in.) are ideal. On the lower end of the spectrum, some technologies can achieve a wall thickness of 25-50μm (0.001-0.002in.). These can be achieved over a short span, but as the span increases the likelihood of success decreases due to inability to fill or due to air entrapment. Additionally, very thin (<0.005in.) cross sections will sinter prior to the adjacent thick section, resulting in constrained sintering of the thin section and subsequent tearing and distortion of the thin section.

A one-to-one ratio on section thicknesses is desired. If this is not possible because of design constraints, one should consider coring out the thick section with webs that are of the same thickness as the thin section on the component. Different wall thickness design considerations are illustrated in Fig. 2.7. Typically, the change in thicknesses should not be <60% of the main body of the component. If a uniform wall thickness is not allowed because of design considerations, a transition of wall thickness should follow a thickness change over a distance of three times the thickness change desired, as shown in Fig. 2.8.

Fig. 2.7 Illustration of good and poor practice in wall thickness.

Fig. 2.8 Illustration of thickness transition recommendations.

2.6.3 Draft

Draft is a standard requirement for any type of injection mold to ensure easy release of components from the tool, but of particular necessity in long sections. Draft is a change in the tooling dimensions or an angle in a direction parallel to tool movement, as illustrated in Fig. 2.9. Draft should be specified to be as great as the design permits; however, MIM can have a minimum of draft since some binders can act as a lubricant. As a nominal recommendation, 0.5-2 degrees of draft should be specified. As the length of the component element becomes longer or if the surface is textured, a greater draft should be used. Often the draft is selected to be within the tolerance of the dimension that the draft affects. External dimensions require none to minimal draft since the MIM material shrinks away from the wall during the cooling stage of the molding operation.

Fig. 2.9 Illustration of both an inside and outside draft angle to allow easy component removal from the tool.

Features that the MIM material shrinks onto during molding are most often drafted to permit easy component removal from the tooling. An example would be a core pin having a draft or taper of 0.5-1 degrees with the largest diameter dimension deepest into the tool and the small diameter dimension at the end of the core pin. This allows the part to slide easily off the pin with minimum friction, as upon initial ejection, the part is freed from all surfaces. A component may have draft on the stationary half of the tool to permit easy release during mold opening and may not have any draft on the movable half of the tool; this is to ensure that the component remains on the movable half, so that the ejectors can be used to remove the component from the tool on a consistent basis and the molding machine can be run continuously in auto cycle. The mold parting line can be located to split the draft in two different directions in order to minimize the change in dimension from one side of the part to the other. Different inside diameter drafting is illustrated in Fig. 2.10; the draft meeting in the middle requires less tolerance on the dimension than the configuration where the pin is only drafted in one direction. Reverse draft is also used in some novel applications to pull the component out of mold features on the stationary half of tooling that could not otherwise be formed on the moving half of the molding machine.

Fig. 2.10 Comparison of two drafted inside diameters with an undrafted inside diameter.

2.6.4 Threads

Threads are possible to produce using MIM. Outside threads are commonly and economically produced using a parting line which runs the length of the thread, as shown in Fig. 2.11. This parting line can encompass each thread or it can have a 0.005-0.010in. flat along its length, which produces incomplete threads along the two sides next to the parting line. If the parting line is designed to produce a complete thread, flash from the parting line can interfere with thread functionality as the tool wears. When a flat is used to provide a good tool shut-off condition, as shown in Fig. 2.12, flash or vestige is prevented from interfering with the thread; however, this may lead to insufficient thread engagement strength for some applications. Alternately, a pneumatic or hydraulic actuated drive is used to rotate the threaded tool member in and out during the molding process after molding and before or during the ejection stage to form a male thread. Internal threads are produced exclusively by using an oversized core that mimics the thread. In this type of tooling, a pneumatic or hydraulic actuated drive is used to rotate the threaded core in and out as part of the molding process after molding and before or during the ejection stage. This tooling method is expensive and thus is limited to high-volume applications.

Fig. 2.11 Typical external thread produced by metal injection molding.

Fig. 2.12 Thread configuration with a flat to prevent flash from interfering with thread operation.

Thread quality is a concern when using the MIM process. Molded threads are inferior to those threads that are machined. MIM is prone to anisotropic shrinkage which may result in shrinkage differences of a few tenths across the component, which shows up as interference on tight thread tolerances. Coarse threads are more practical than fine threads for MIM. Thread engagement length should be minimized to minimize interference of the threads with the matching component due to the potential for MIM thread variability. When exterior threads are used, e, f, and g tolerance grades should be utilized. If an internal thread is designed, a G tolerance should be specified in the design. In general, internal thread pitch diameters should be on the large side of specification and external thread pitch diameters should be on the small side of specification to maintain functionality while allowing for minor anisotropic shrinkage and distortion.

2.6.5 Ribs and webs

Ribs and webs are utilized to reinforce thin sections and also as replacement for thick sections. Ribs increase the bending stiffness by increasing the moment of inertia (bending stiffness = E（Young's modulus）× I （moment of intertia）. Consequently, ribs help to strengthen the MIM component for both processing and application. Strengthening during processing helps to enhance dimensional stability and prevent warpage. The ribs also act to enhance flow along thin sections. However, if too thick, the ribs may cause sinks on the opposite flat side where the rib mates with the main body of the component, as shown in Fig. 2.13. They may also result in warpage if the wrong thickness of rib is utilized. Ideally the rib thickness should be 40%-60% the size of the section on which it is located and the height should be no more than three times the thickness of the rib. Also, these ribs should have a good radius at their base to prevent cracking.