The present article reports an experimental analysis of the mechanisms of secondary atomization which occur at the impact of individual droplets onto heated targets. The experiments follow those reported in a previous article (Moreira et al. 2007) and encompass the use of different liquids and impact conditions. An image analysis system is combined with a phase Doppler interferometer to measure extended size distributions, which cover the full range of diameters generated at all heat transfer regimes. The results evidence that disintegration mechanisms depend on the heat transfer regimes; therefore, a universal relation cannot be devised for the outcome of droplet impact. Analysis shows that droplets impacting within the nucleate-boiling regime break-up by a thermal-induced mechanism associated with the vapour pressure at bubble nucleation sites, combined with liquid surface tension. On the other hand, within the film-boiling regime, disintegration is associated with radial disruption of the rim at the early instants after impact, as in non-heated targets, and with the rupture of the ligaments of the cellular structures. Functional relations available at the literature, mostly developed for impacts onto non-heated surfaces, are well fitted to the experimental results obtained within the film-boiling regime, since the break-up mechanisms are qualitatively similar. On the other hand, such relations cannot predict the secondary atomization occurring within the nucleate-boiling regime, as the break-up mechanisms within this regime have significantly different characteristics. In this context, the present article recognizes the relevance of the relations devised for ‘cold impacts’, to fit the size of secondary droplets within the film-boiling regime, as the correlation formulated here has a similar form: SMD/D 0 = f(We, Re) ~ A 1 We −0.6 Re −0.23 and proposes a new correlation for impacts within the nucleate-boiling regime: SMD/D 0 = f(We, Re, Ja) ~ A 2 We −0.14 Re −011 Ja −03. These correlations are observed to hold for impacts onto rough surfaces with dimensionless roughness R a/D 0 smaller than 2E-3, but not for larger roughness amplitudes, for which the data are quite scattered.
Read full abstract