Comparison of Numerical Simulations to Experiments for Atomization in a Jet Nebulizer |
| |
Authors: | Nicolas Lelong Laurent Vecellio Yann Sommer de Gélicourt Christian Tanguy Patrice Diot Alexandra Junqua-Moullet |
| |
Affiliation: | 1. DTF Aerodrug, Faculté de Médecine, Université François Rabelais, Tours, France.; 2. Centre d’Etudes des Pathologies Respiratoires INSERM U1100/EA 6305, Faculté de Médecine, Université François Rabelais, Tours, France.; 3. ANSYS France, Montigny le Bretonneux, France.; 4. Dantec Dynamics, Nozay, France.; 5. CEA Le Ripault, Monts, France.; University Paul Sabatier, France, |
| |
Abstract: | The development of jet nebulizers for medical purposes is an important challenge of aerosol therapy. The performance of a nebulizer is characterized by its output rate of droplets with a diameter under 5 µm. However the optimization of this parameter through experiments has reached a plateau. The purpose of this study is to design a numerical model simulating the nebulization process and to compare it with experimental data. Such a model could provide a better understanding of the atomization process and the parameters influencing the nebulizer output. A model based on the Updraft nebulizer (Hudson) was designed with ANSYS Workbench. Boundary conditions were set with experimental data then transient 3D calculations were run on a 4 µm mesh with ANSYS Fluent. Two air flow rate (2 L/min and 8 L/min, limits of the operating range) were considered to account for different turbulence regimes. Numerical and experimental results were compared according to phenomenology and droplet size. The behavior of the liquid was compared to images acquired through shadowgraphy with a CCD Camera. Three experimental methods, laser diffractometry, phase Doppler anemometry (PDA) and shadowgraphy were used to characterize the droplet size distributions. Camera images showed similar patterns as numerical results. Droplet sizes obtained numerically are overestimated in relation to PDA and diffractometry, which only consider spherical droplets. However, at both flow rates, size distributions extracted from numerical image processing were similar to distributions obtained from shadowgraphy image processing. The simulation then provides a good understanding and prediction of the phenomena involved in the fragmentation of droplets over 10 µm. The laws of dynamics apply to droplets down to 1 µm, so we can assume the continuity of the distribution and extrapolate the results for droplets between 1 and 10 µm. So, this model could help predicting nebulizer output with defined geometrical and physical parameters. |
| |
Keywords: | |
|
|