An attempt to predict spray characteristics at early stage of the atomization process by using surface density and curvature distribution
Nowadays, numerical simulations of atomization has reached a mature state through interface capturing approaches. With these approaches, the liquid–gas flow with its complex interface morphology can be precisely described but at the price of high mesh resolution, that makes these simulations very in...
Saved in:
| Published in: | International journal of multiphase flow Vol. 147; p. 103879 |
|---|---|
| Main Authors: | , , , , , , |
| Format: | Journal Article |
| Language: | English |
| Published: |
Elsevier Ltd
01-02-2022
Elsevier |
| Subjects: | |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Nowadays, numerical simulations of atomization has reached a mature state through interface capturing approaches. With these approaches, the liquid–gas flow with its complex interface morphology can be precisely described but at the price of high mesh resolution, that makes these simulations very intensive in terms of computational cost. Practically, for most of the injection system, it is not possible to achieve a simulation up to a point where the spray formation is completed. Thus, there are several attempts based on the transformation of liquid elements such as ligaments, liquid sheet or other non-spherical to equivalent droplets. Overall, these approaches assumed that before the complete finalization of the atomization process the spray under formation carries enough information to be representative of the final spray. In this work, we have pushed this idea a step further by using global variables such as the liquid volume fraction and the surface density available in ELSA models to ensure the transports and the conservation of the main features of the spray. Then, the surface curvature distribution is analysed assuming that a part of the interface carries already curvatures that are relevant with respect to the final spray. A well-known academic test case representative of an aeronautical injection system and based on a planar prefilmer atomizer with a gas co-flow has been selected to evaluate our proposal. This configuration was studied both experimentally and numerically thanks to high-fidelity simulations. Our purpose has been to follow the already validated numerical approaches but with less computationally intensive simulations. Then, new analysis based on surface density and surface curvature distribution have been tested to recover spray characteristics and even the spray diameter distribution. It appears that these variables are meaningful even when there is no droplet yet formed, and thus they allow the description of the full atomization process from the early stage even on the liquid film. Finally, a procedure has been proposed where the spray is sufficiently atomized to reconstruct the diameter distribution from the curvature distribution. At our best knowledge, this last step is a first attempt on practical injection system. We have proposed a very simple method to detect which part of the interface carries the relevant curvature, thus obtaining a reconstructed diameter distribution that fit well with available experimental data. Although the proposed approach cannot be still considered universally applicable to any case, this preliminary assessment clearly shows its potential.
•Numerical simulation of the atomization occurring in an air blast atomizer.•Numerical determination of the amount of liquid–gas surface along that atomization process.•Numerical determination of the surface curvature distribution along the atomization process.•A new approach is proposed to obtain the drop size distribution from curvature distribution and liquid volume fraction.•This approach allows the determination of the spray size distribution at an early stage of the atomization process. |
|---|---|
| ISSN: | 0301-9322 1879-3533 |
| DOI: | 10.1016/j.ijmultiphaseflow.2021.103879 |