Laser powder bed fusion (LPBF) is a rapidly growing additive manufacturing process that is based on shaping a part by stacking successive layers of metal powder. With LPBF, each layer of the part is recreated using a high-powered laser that selectively fuses the spread particles of the powder bed; then a scraper evenly spreads a new layer of powder on top of the previously deposited one to create the next layer and so on until the entire part is made.
The performance of the LPBF process, in addition to the printing parameters and the characteristics of the used material, is highly dependent on the properties of the spread powder bed, especially in terms of density and uniformity. The properties of the powder bed result from the interaction between the intrinsic properties of the powder particles, the recoating mechanism and the spreading parameters used. Therefore, an accurate assessment of the powder characteristics and a deep understanding of its mechanical behaviour during spreading becomes important. Such an evaluation involves in particular the characterization of the rheological properties of the powder and their impact on the performance of the LPBF process.
This thesis presents, in the form of three scientific papers, a proposed approach aiming at understanding the link between the intrinsic characteristics of the powder, its rheological properties and the performance of the LPBF process, and at presenting a new powder characterization technique able to evaluate metrics of interest relevant to powder bed based additive manufacturing processes.
First, an exhaustive study was conducted on three lots of Ti-6Al-4V powders selected in order to study the impact of particles’ sphericity (plasma atomized powder vs. gas atomized powder) and particle size distribution (powder with more fine particles vs. powder with less fine particles) on several properties of interest, namely, the rheological performances of the powders before vs. the geometrical and mechanical properties of the parts fabricated with these powders by LPBF.
Using rheological indices measurable with an FT4 rheometer (dynamic flow resistance, cohesion, compressibility, permeability...), a figure of merit was proposed to quantify the overall suitability of a given powder to the LPBF process. Then, the same powder lots were used to fabricate a series of test samples to study the impact of the selected powders’ characteristics on the properties of the printed parts (density, surface finish, fabrication limits, static mechanical properties), and this, for two layer thicknesses, two post-treatments and two fabrication orientations. To further investigate this, a fatigue performance study of samples produced with two of the three powder lots previously selected, i.e. plasma atomized powder vs. gas atomized powder, was conducted.
All in all, the relatively small differences between the morphologies of the three powder lots proved to be very significant in terms of their rheological performance (15% on average), surface roughness (a14%) and dimensional accuracy (a24%) of the printed samples. However, as far as the mechanical properties of the printed samples are concerned, this difference became small for the static properties (a7% for the ultimate tensile strength, a4%, for the yield strength and a3%, for the elongation at break) and finally faded, for the fatigue properties (a500 MPa, N=106). This decrease in the impact of powder morphology in the advanced stage of the process can be explained by the robustness of the LPBF process to small morphological variations between powder lots. However, microtomography defect analysis demonstrated the presence of aligned pores in samples fabricated with one of the powders indicating the use of sub-optimal printing parameters. Hence the importance of optimizing the manufacturing parameters for the different powder lots even if the difference in particle morphologies seems minimal. This optimization can be done with simulation software of the LPBF process melt pool. Knowing that the melt pool characteristics depend not only on the laser parameters (power, scanning speed, hatch spacing...), but also on the laser-material interactions and, consequently, on the properties of the powder bed (density, uniformity, surface roughness), it is important to be able to quantify the latter. Nevertheless, conventional powder characterization techniques are not in line with the metrics representative of the properties of a spread powder bed.
In order to address the needs expressed above and given that the flow properties of powders depend greatly on the conditions imposed by the application, a test bench reproducing the spreading system of a typical industrial machine is developed and presented in this work. This test bench allows the operator to control the spreading parameters, including the increments of the different platforms, the spreading speed and the powder deposition mechanism. The test bench also allows to evaluate the density of the spread powder bed with the help of load cells installed under the spreading platform. On the other hand, an imaging system installed above the spreading platform allows to evaluate the surface profile and thus the uniformity and the variation of the surface profile of the powder bed. The spreading forces are also evaluated by measuring the torque variations at the actuator of the spreading mechanism. A case study is presented to show the capabilities of the developed testing apparatus. The closed architecture of commercial machines, the extreme conditions in a manufacturing chamber, the large volume of powder required and the high costs associated with printing greatly limit studies focusing on the rheological properties of powders, hence the interest in having access to an instrumented tool that can easily test different spreading configurations by evaluating several metrics of interest in order to optimize the spreading operation and decrease the selection and quality control time of powders intended for the FLLP process.
| Date | 21 Aug 2023 |
|---|
| Original language | French |
|---|
| Awarding Institution | - École de technologie supérieure
|
|---|
| Supervisor | Vladimir Brailovski (Supervisor) |
|---|