Combining Powder Characterization Methods and Numerical Simulations to Optimize Printing Parameters


Presentation Stage 2

Powder flowability and packing fraction are critical parameters for additive manufacturing techniques involving powders. In order to obtain thin and homogeneous layers, a compromise between grain size and flowability has to be found. Unfortunately, when the grain size decreases, the cohesiveness increases and the flowability decreases. Too often, both the powder spreadability assessment and the optimization of the printing parameters are costly empiric processes. This presentation will evaluate an original method associating powder flow characterization instruments and numerical simulations to select the powders and to optimize printing parameters like recoater speed and layer thickness. The powder characterization allows first to assess the spreadability of the powder in the printer and secondly to calibrate the simulation parameters. Then the recoating process is simulator with the calibrated simulations to predict the behavior of the powder for different printing parameters and different recoater geometry. Finally, the results are validated by testing the powder in a printer equipped with an in-situ powder layer homogeneity tester based on image analysis. The powder characterization is performed with the association of three measurement techniques:

1. An improved version of the very classical Hall flow meter (GranuFlow)

2. A powder Rheology analyzer using the rotating drum geometry (GranuDrum)

3. A packing fraction dynamics analyzer which is an improved tapped density measurement device (GranuPack).

To investigate the powder behavior inside a printer, the researchers used numerical model based on the Discrete Element Method (DEM), using the software LIGGGHTS. The bulk behavior of granular materials is governed by particle-scale parameters, which can hardly be measured directly. This session will, therefore, present a workflow, closely connected to the aforementioned measurements, that uses an optimization approach to obtain the required set of DEM model parameters. Finally, the spreading tests are performed on an SLM 250 equipped with an LCD camera. A picture is taken after each spreading step. Further tests were also performed on an SLM 280. During these tests, several cycles passing through different speed ranges were repeated several times in order to be able to work on a minimum of statistical data.

The presentation is the result of collaboration between the University of Liège, the Belgian research center Sirris and two companies active in additive manufacturing (GranuTools and DCS Computing).

Associate Professor
University of Liege