Liberty Powder Metals (LPM), part of GFG Alliance’s Liberty Steel Group, has identified a process configuration which increases the yield for atomised Additive Manufacturing powders.
The DANDY 2 project, funded by Innovate UK, successfully linked the modelling of the atomisation process via computation fluid dynamics (CFD) and experimental modelling (EM) with the plant operations to increase the yield of Additive Manufacturing powders with a particle size distribution of 15-45 μm, by over 40% via an iterative design optimisation of the step change liquid atomiser nozzle and powder production process changes. This involved collaborative R&D with the Materials Processing Institute, Middlesborough, North Yorkshire, UK.
During the year-long project, the CFD simulated gas flows, orifice plate, new nozzle designs, gas input temperature link to atomisation velocity, gas input pressure link to aspiration pressure (cross-validated by plant measurement), gasket thickness changed both atomisation velocity and aspiration pressure. EM provided confidence in machine security, enabled further modelling on gasket thickness and changes in nozzle design. CFD combined with the EM on nozzle design development, including high speed videoing at 115,000 frames per second, providing cross-validation between model flows and against nozzle performance.
LPM completed 193 atomiser melts of 316L and Alloy718, with individual trials on nozzle design and material, orifice plate, gasket thickness, gas input temperature, gas input pressure (both dynamic and steady state). Each parameter has influenced Additive Manufacturing, which, when combined have produced a cumulative AM yield, increase to 57%. Machine learning tools were applied to this dataset by Swansea University and Materials Processing Institute independently and whilst the dataset was small in terms of ‘big data’, these tools have confirmed the findings from the plant trials and identified various parameters to investigate further.
Powder from twenty-four melts were fully characterised to assess flowability and morphological characteristics. These powders were also used to build mechanical test specimens by Additive Manufacturing. It was found that the powder from the high yield trials was good quality, easy to manufacture with and provided good mechanical properties.
From the modelling, plant trials, machine learning and powder characterisation, a new configuration was identified to increase the yield for LPM AM powder production.
The project, funded by Innovate UK, also examined the use of audiometry and vibration monitoring tools on the atomiser to give real time feedback and control. LPM intends to continue development of this technology.