Download PDFOpen PDF in browserNumerical Analysis of the Flow Phenomena Inside the Vortex Tube with Different Turbulence ModelsEasyChair Preprint 136348 pages•Date: June 11, 2024AbstractTo accelerate the progress towards sustainability, a fundamental shift from conventional energy sources to renewable energy is necessary but it is insufficient on its own. Enhancing the efficiency of existing energy systems and optimizing energy management are equally crucial. One promising avenue for achieving these goals involves integrating previously overlooked devices, such as the vortex tube, into the current systems. Thus, there is a growing emphasis on addressing its limitations, including limited development and suboptimal efficiency, to unlock its potential in diverse applications. The current focus is primarily on computational research rather than experimental approaches due to the advantages in terms of cost and time. Nevertheless, computational studies present their own set of challenges, with two prominent hurdles being the attainment of acceptable mesh and the selection of an appropriate turbulence model. This study aims to address these challenges. An acceptable mesh has been obtained by optimization, which involves refinement of mesh at the inlet and outlet regions, followed by a comprehensive assessment of mesh independence at each stage. Furthermore, achieving a y+ value of 1 in the most important regions of the vortex tube is crucial, particularly when employing low Re turbulence models, to accurately predict boundary layer behavior. Building on the mesh studies, the performance of different turbulence models is assessed with reference data. Among the considered models, the standard k-ε turbulence model has the best performance, aligning closely with experimental results for almost the same geometric set-up. As a result, the standard k-ε turbulence model is selected for further numerical investigations. Keyphrases: CFD, Ranque-Hilsch vortex tube, Temperature separation effect, heat transfer, turbulence modelling
|