Khaing, M. W., Fuh, J. Y. H., & Lu, L. (2001). Direct metal laser sintering for rapid tooling: processing and characterization of EOS parts. Journal of Materials Processing Technology, 113(1), 269-272.
1. Rapid prototyping (RP) uses computer aided design (CAD) data to directly fabricate parts without using traditional tooling.
2. Some RP methods include selective laser sintering and 3D-printing which can create complex parts with very short lead times.
3. Direct Metal Laser Sintering (DMLS) is a process that directly exposes metal powder to a laser and melts the powder into form, allowing it to create metal parts within days. A commercially available version of this has been produced by EOS GmbH in 1995 and many others moving forward.
4. The EOSINT M 250 machine uses a build chamber, control system, powder dispenser, and CO2 laser (200 W, wavelength = 10.6um, and spot size 0.3 mm). The part is built on a steel base plate coated with bronze to help bond the part to the base-plate. One mixture used is a mixture of nickel, bronze, copper- phosphide with average grain size of 30um (EOSINT M Cu 3201).
5. A couple parts of different geometry’s were made using this system, all with a layer thickness of 50um. Scanning and hatching rates were varied to change process time, and things like shrinkage compensation were also accounted for.
6. To fabricate the parts, the bronze is scanned to bond it to the base plate and then powder is deposited according to each slice data. Compressed air and a painting brush were used to remove excess powder.
7. Epoxy infiltration was done in two steps, penetration and then curing. The parts and epoxy were both heated up to 140 F and then the epoxy was added either through pouring, capillary forces, or being dipped in a resin bath. Afterwards epoxy was painted on and then the parts were cured at 320 F for 2 hrs. 8. It was found that the humidity of the environment and levelness of the base plate coating affected the adhesion of the part to the base plate.
9. Part deviations were -0.043 mm for the X- axis, -0.018 mm for the Y-axis, and 0.025 mm for the Z-axis. Specifically the diameters of cylindrical features were consistently off by a rather large amount, between 0.025 mm to 0.34 mm.
10. This deviation in circular parts was attributed to unequal shrinkage from the X and Y directions.
11. The hardness of the parts were originally between 26 to 33 on the Rockwell B scale, and the epoxied parts increased the hardness to 65-69 HR B. This could be increased if the infiltration was replaced with a low melting point metal. However the epoxied parts hardness is similar to hardened steel so this could be acceptable for parts.
12. The impact toughness for the parts was quite low, close to 4 J. This was similar in the epoxied parts. This value is considered quite low and is close to a magnesium alloy.
13. The roughness of the untreated parts were about 12-16 um and the treated parts were about 4-7 um, which is compared to a rough electrical discharge machining (EDM). Manual polishing can reduce the roughness to below 1um.
14. Even though the roughness improves with infiltration, the thermal conductivity of the parts decrease.
15. The average density of the parts was 391 lb/ft3 before infiltration and 395 lb/ft3 afterwards.
16. Without infiltration, the porosity of the parts could be from 30 to 45 percentage, which affected the strength of the part due to the voids in the microstructure.
17. An optical microscope showed the scanning lines and pores. It also showed that some larger particles did not melt due to their larger mass.
18. This DLSM process was able to produce 3D-parts with very fine details, but the parts were soft and porous.
19. Control of the powder handling, humidity, scanning rates could all improve the parts, but only to a certain point that may not be acceptable for manufacturing.
20. To improve hardness, low melting point infiltration with a silver alloy or with nickel plating (which would also improve wear resistance). However, better strength would require a new material system.