Printing Long, Thin Cores for Aluminum Sand Castings
While the use of printed sand molds and cores has become increasingly common, two researchers investigated whether longer, thinner cores could hold up during the metalcasting process.
Tom Mueller, Voxeljet, Canton, Michigan, and Larry Andre, Solidiform, Fort Worth, Texas
(Click here to see the story as it appears in the November/December issue of Metal Casting Design & Purchasing.)
One of metalcasting’s major advantages, when compared to other methods of manufacturing, is its ability to produce geometrically complex components. Metalcasting allows designers to put metal only where it’s needed because liquid metal can flow into the open mold cavity. Often, cores are placed into a mold to form an interior part of a casting. These inserts, often made of sand, prevent the molten metal from filling the entire mold, so interior cavities can be included in the casting design.
In recent years, the growth of 3-D printing capabilities has become a topic of interest for every manufacturing industry. In metalcasting, the process offers the potential of producing complex molds and cores with desirable mechanical and physical properties. Printed cores and molds allow for immediate evaluation of design iterations without the additional costs and lead time associated with tooling, which speed the development of cast components for end users.
While many sand casting suppliers are experienced with printed cores and molds, little research has been done into the performance of long, thin cores. Larger cores generally are able to handle the forces exerted during the flow of metal during the casting process, but less robust designs are relatively untested.
In the research paper, “Performance of Long, Thin Printed Cores in Aluminum Castings,” two researchers evaluated the performance of printed cores that covered a range of diameters and lengths. The primary aims of the study were to see whether cores were strong enough to hold up during pouring and whether gases could be vented from the mold through internal passages in the core.
If longer and/or thinner printed cores can be integrated into castings, engineers will have more freedom in designing interior cavities in castings. The potential increased complexity offered by these 3-D printed cores could lead to reduced costs and lead times for casting customers.
Testing the Tubes
For the A356 aluminum castings, tubes with wall thicknesses of 0.06-in. were created with internal diameters ranging from 0.25 to 0.875 in. and lengths ranging from 2 to 12 in. Printed cores were used to form the internal diameters, and the cores were printed with internal vents for gases generated during casting.
To ensure conditions were consistent, all tubes of the same diameter were cast at the same time and were fed from a common feeder at the bottom of the tube. Figure 1 shows the tubes, feeder and two down sprues, which increased the feed rate so the tubes would fill before freezing off. The tapered area at the end of each tube was a riser to feed the tube as metal shrunk during solidification.
This design allowed the use of a single core for all six tubes of the same diameter. The core resembled a harp with a core print surrounding the cores, and the cores running parallel between them like strings on the harp. Figure 2 shows the CAD model of the core, which contains a passage through the base to allow the tubes to be bottom fed. The core would be extremely difficult if not impossible to create by conventional means.
To create the molds, matchplates were created for each of the tube diameters, is shown in Figure 3. The matchplate included vent lines that would connect the vents from each of the cores to the outside of the mold. Each of the mold cavities also was vented to the outside to allow complete filling of the mold.
The molds were positioned so that they were vertically parted. The pouring cups were elevated to increase the static pressure of the metal during pouring. A vacuum manifold was attached to the mold where the vent lines emerged so that a low level vacuum (supplied by a shop vacuum) could be used to assist in drawing out gases generated during filling.
On the first casting, the pouring cup was positioned 2 ft. above the mold, which filled quickly and metal flowed up and out of the vents. On successive casts, the head height was reduced until it was found that a head height of 3 in. sufficed.
In all, 12 pours were done, two for each of the six core diameters. Table 1 summarizes the 12 pours. After each pour, the casting was allowed to cool and broken out. The tubes were cut from the feeder and inspected. Each tube was visually inspected followed by X-ray and liquid penetrant inspection.
Figure 4 shows all 12 sets of tubes. Each smaller picture contains two sets of tubes, all with the same inner diameter, and two tubes of each of the six lengths of core.
Looking for Defects
The tubes were first evaluated for defects during pouring. Figure 5 shows an end view of one of each diameter of tubes. There is an obvious eccentricity of the core in the 0.25-in. tube, but the others appeared concentric. The lack of deformation in the larger diameter tubes was confirmed by X-ray inspection. The deformation of the small tube was great enough that, in longer lengths, part of the tube did not fill, as shown in Figure 6. Non-fill due to core deformation is clearly visible in the 8-, 10- and 12-in. tubes. Non-fill conditions were not found in any larger diameter tubes.
Liquid penetrant inspection revealed a number of gas defects, such as those shown in Figure 7. In some cases, a pit is obvious in the cases of larger defects. The presence of the gas defects raised two questions: What was the source of the gas? Why didn’t the core venting allow the gas to escape?
If the source of the gas was the binder in the core, such defects may be difficult to avoid in any casting. The other gas defects in the cast tubes showed two definite trends:
- Defects increased with increased pouring height.
- Smaller diameter tubes had fewer defects.
If the binder was the source of the gas, similar defect levels should be found in all castings. However, defects were more frequent in castings made with greater pouring height, which points to air entrained in the pour as the source of the gas. As the pouring height increases, the velocity of flow increases and more air is entrained in the molten metal. Consequently, more defects would be expected in those castings done with a greater pouring height. This is consistent with the results obtained.
Determining whether venting was adequate is more difficult. The diameter of the vent passage increased with the diameter of the core. As the core diameter decreased, it became more difficult to remove the unbound sand from the vent passage. It was impossible to remove the unbound sand in the smallest cores. While gases could likely flow through spaces between sand grains, the area available for flow would be drastically reduced.
The area of flow decreases as the core diameter decreases. If venting is necessary in this situation, defects should increase in cores with smaller diamteres, where nearly no venting was available. Instead, the opposite was true. Apparently, venting did not play a significant role in this situation.
The results of these tests led to two conclusions:
1. Cores with diameters of 0.375 in. or more are rigid enough to resist significant deformation during casting of aluminum when positioned vertically. Cores oriented horizontally would be subject to gravity and buoyancy forces that could result in greater deflection.
2. Gaseous emissions from a binder concentration of 1.7% is not a significant issue in casting aluminum.