13 th Annual Conference on Liquid Atomization and Spray System (ILASS 97), 9-11 Juli 1997, Firenze, Italien 1 Droplet Impingement on a Solid Surface Covered by a Liquid Film C. Ghielmetti, M. Marengo, T. Steigleder, C. Tropea Lehrstuhl für Strömungsmechanik/ Universität Erlangen-Nürnberg/ Germany ABSTRACT In many technical processes, liquid sprays impinge on a solid surface, producing a liquid film on the impacted surface and a secondary atomization in the form of smaller secondary droplets detaching from the wetted surface. One of the main purposes of the present work, it is to obtain an empirical model to describe the splashed flux and the velocity and diameter probability distribution functions of the secondary droplets as a function of the impact parameters (film thickness, impacting droplet velocity and diameter, liquid properties). In this work, an experimental set-up to obtain such a result is presented and a first application of the empirical model to a polydispersed spray is given. 1. INTRODUCTION In this work, the impingement of droplets on a solid wetted surface is considered. This phenomenon is significant for many industrial and natural processes: direct injection of diesel engines, painting of surfaces, coating of substrates, icing on airplane wings, cooling processes, cavitation, and erosion phenomena are some examples of the range of applications for droplet impingement studies. The secondary atomisation regime (the production of droplets from a splash process) may be characterized by the impact parameters, including primary droplet velocity u p and diameter d p , liquid viscosity μ , surface tension σ, density ρ, the gravitational acceleration g and the film thickness h. Using dimensional analysis, it is possible to determine the dimensionless parameters that describe the splashing phenomenon: the Weber number We ud p p = ρ σ 2 , the Ohnesorge number Oh d p = μ ρ σ , the dimensionless film thickness δ = h/d p , the Bond number Bo gh = ρ σ 2 , and the dimensionless characteristic time τ = u p t/d p . Previous studies concerning droplet impact on a solid surface can be roughly subdivided into three groups: impingement on a high temperature surface (higher than the Leidenfrost temperature for the given impact parameters), impingement on a cold dry surface and impingement on a cold wetted surface. The first group is not considered in the following review. Walzel [1], Stow and Hadfield [2], and Mundo et al. [3] recently investigated the impact of single droplets on dry surfaces. The splashing/deposition limit is described in terms of the number K = We Oh -0.4 , whereby only the wall-normal velocity component is used in computing the Weber number. If a primary droplet has a K value greater that the critical value of K relative to the impact surface, splashing occurs. The critical value of K (K cr ) depends on the impacted surface conditions only. For a dry surface, K cr is a function of surface roughness, wettability [4], temperature, surface elasticity and possible some further higher order effects. For a wetted surface, the critical constant depends mainly on the film thickness. Coghe et al. [5] examined the splash due to the Rayleigh break-up from the jets on the crown (the only secondary atomization process for the case of Oh > 0.005) and found that for δ < 1, the critical K can be written as K cr,1 = 2100 + 5880 δ 1.44 . For low viscosity liquids, such as water, a prompt splash was observed, i.e. some secondary droplets had originated from the lamella already at a very early phase of crown formation. For this phenomenon, the critical value of K for the impact of water droplets has been found to be closer to: K cr,2 = 2074 + 870 δ 0.23 [6]. For values of δ approaching zero, the first coefficient of the correlation for K cr will describe entirely the process and vary due to the effect of the solid surface characteristics. When δ is significantly greater than one (deep pool), an asymptotic critical value is reached [7]. In the following experimental study the two correlations given above will be tested using an independent experiment. In fact the agreement is not especially good indicating that further parameters may be of significance, for instance impact angle, impact frequency, waviness of the impacted liquid surface. Mundo et al. [3] showed that, for a dry surface, the splashing/deposition limit depends only on the component normal to the surface, at least for impact angles α > 60°. Actually, as we observed in a preliminary experiment, when the impact angle is less than a critical value ( α cr < 20°), the droplets rebound leaving some of their mass on the surface. Also the waviness of the liquid film should have an effect, increasing the perturbations in the first phase of impact and probably causing the splashing limit to become lower. With regard to the impact frequency, it is possible to link the impact frequency to the thickness of the deposited film and, hence, as a first approximation, to the number δ [5]. Nevertheless some uncertainty will also remain in the actual measurement of the splash limit, since the transition from deposition to splashing impact is not sudden and to provide a quantitative limit, for instance in terms of K,