Binder Jet Additive Manufacturing of Copper

Abstract

This research aimed to optimize the process parameters of small copper parts printed with binder jetting. The optimisation was distinguished by dividing the parameters in print and post-process parameters. Binder jetting is an additive manufacturing technique that combines layers of the material in powder form with a binder. After the printing process the printed samples need curing, debinding and sintering. The binder acts as a temporary glue to hold the shape and is later removed during the debinding. The sintering fuses the powder particles together, below the melting temperature. With binder jetting it is the aim to achieve the high density parts, as there is no compression of powder during the printing process. This often leads to low density parts with binder jetting. The two problems that needed to be solved, were: how can the bleeding issues be stopped and at what temperature, time and environment can the printed copper be sintered. Multiple print sessions were done to: first test the known print parameters, second optimize the print parameters and third, print with optimized print parameters. ThermoGravimetric Analysis (TGA) was conducted to determine the correct debinding and sintering temperature of the copper samples. The bleeding issues could be solved by adjusting the saturation level. The saturation level was reduced from 104 % to 72 %. The lower percentage gave 0.5 mm dimensional accurate samples and no bleeding issues. Compression testing was done to compare the mechanical behaviour to that of conventionally manufactured copper. However the discs used for compression testing needed to be bigger in size than the cubes used for optimizing the sintering process. This lead to not fully sintered discs used for the compression testing. Meaning the achieved compressive strength and elasticity modulus were 1 % of conventionally manufactured copper. Three working sinter sessions were compared to see which gave the highest achieved density. Sintering 10 hours in argon at 1093 ℃ or in vacuum at 1080 ℃ both achieved the same high densities. Shell printed cubes performed better with an average density of 73 ± 8 % in argon and 74 ± 7 in vacuum. The samples placed on the top level inside the sintering oven achieved higher densities than in lower levels. Two shell printed cubes in argon achieved a density of 83 ± 0.5 % with a volumetric shrinkage of 40 ± 1 % and two shell printed cubes in vacuum achieved a density of 82 % with a volumetric shrinkage of 40 %. Solid cubes reached lower densities, with 67 ± 10 % in argon and 71 ± 7 % in vacuum. As with the shell printed cubes, the solid cubes have a high deviation due to the placement of the samples on different levels. The solid cubes had less volumetric shrinkage and rather high deviation due to the same reason, with 25 ± 16 % in argon and 35 ± 11 % in vacuum. The density is measured with a caliper and with Archimedes which gave quite different results. Densities measured with Archimedes were up to 22 % higher due to the porosity of the samples. All the measurements done by other copper binder jetted parts are done with Archimedes. Thus it seems the 74 % and 82 % is low when compared to 91 % density achieved in literature, but it is not known how accurate these measurements are.