Evaluation of spray impact on a sphere with a two-fluid nozzle

https://doi.org/10.1016/j.jaerosci.2019.105483Get rights and content

Highlights

  • Spray-impact is combined with a two-fluid nozzle.

  • The mass of droplets in the small micrometer range was increased during spray-impact.

  • Droplet formation mechanisms such as splashing, crown formation and spreading on the sphere were observed.

Abstract

The generation of a secondary aerosol after impact, consisting of smaller droplets at a given velocity and mass flow, is relevant for various applications. Thus far, the investigations and modelling approaches on spray impact are based on extrapolation of the single-droplet impingement or empirical correlations. The validity of the models presented is limited to the given experimental setup and conditions such as initial droplet size, velocity and the impact surface characteristics. The aim of this work was to empirically evaluate the spray impact of a two-fluid nozzle on a sphere. A small-scale nozzle was used, which produced a primary aerosol with a mass median diameter of about 12μm (liquid-to-gas mass flow ratio = 1, gas pressure: ΔpG = 5 bar). After impact on a sphere, a multimodal distribution was observed and a higher mass flowrate of droplets in the small micrometer range (2 and 3μm) was produced for a liquid mass flow rate in the range of 1.2–6 kg/h and an atomizing gas mass flow rate of 1–4 kg/h. For easier observation, a geometrically similar, larger nozzle was used, which produced an aerosol with a mass median diameter of about 80μm (liquid-to-gas mass flow ratio = 4, gas pressure: ΔpG = 1 bar). The measured droplet size after impact is smaller for a lower liquid-to-gas mass flow ratio and increased atomizing gas inlet pressure. Droplet formation mechanisms such as splashing, crown formation and spreading on the sphere surface were observed. A characteristic film with large variations in thickness was generated.

Introduction

The spray impact on a solid surface is relevant for various applications, such as spray coating, painting, agriculture sprays, spray cooling and medical inhalers, to name a few. Several spray impact phenomena are of interest, depending on the application (Breitenbach, Roisman, & Tropea, 2018; Moreira, Moita, & Panão, 2010; Roisman, Horvat, & Tropea, 2006). In spray coating, the deposition of the liquid on the surface is pertinent, and the generation of secondary droplets should be prevented (Roisman et al., 2006). For medical inhalers or for aerosol generation in flame spectrometers, the secondary aerosol produced with corresponding change in droplet size and mass flux after spray impingement is of interest (Sneddon, 1990; Yarin, Tropea, & Roisman, 2017). The formation of smaller droplets can also be relevant for spray drying processes. Droplets in the small micrometer range are necessary to produce submicron-sized particles, for example, to enhance the bioavailability of low water soluble drugs (Liversidge & Cundy, 1995). A detailed experimental investigation on the droplet size, velocity and mass flux near the impact surface and the influence of the process parameters and liquid properties on the generation of secondary droplets is necessary to understand the complete impact phenomenon (Moreira et al., 2010).

In recent years, the evaluation of single drop impact onto a solid dry or liquid surface has attracted much attention. Many experimental investigations have been conducted and numerous models have been developed. The outcome of single droplet impact depends on a combination of different parameters: the droplet size and velocity before impact, the impact angle, the liquid properties, the surface geometry and roughness. The presence of a liquid film may especially influence the disintegration of the impinging droplet. Depending on these conditions, possible outcomes of the single droplet impact are prompt splash, corona splash, spreading, sticking on the surface, rebound, fingering (Bakshi, Roisman, & Tropea, 2007; Banitabaei & Amirfazli, 2017; Chandra & Avedisian, 1990; Hardalupas, Taylor, & Wilkins, 1999; Josserand & Thoroddsen, 2016; Moreira et al., 2010; Rein, 1993; Rioboo, Marengo, & Tropea, 2002; Villermaux, 2007; Yarin et al., 2017). Typical values to describe the boundaries for the different impact outcomes are the Weber and Reynolds numbers, which depend on the droplet characteristics before impact.Webernumber=ρGvrel2d50,3,PA/σLReynoldsnumber=ρlvPAd50,3,PA/ηL

Here, ρl and ρG are the liquid and gas density, respectively, vrel is the relative velocity, vPA is the velocity of the droplets in the primary aerosol, d50,0,PA is the mass median droplet size of the primary aerosol, σL is the surface tension of the liquid and ηL is the viscosity of the liquid.

The impact threshold was defined in order to predict the above-mentioned outcomes after spray-impact. The disintegration of the droplet in the mechanisms of prompt splash appear, when inertial forces overcome capillary effects (Moreira et al., 2010) and can be described with the splashing parameter Kc=AOhaWeb. (Mundo, Sommerfeld, & Tropea, 1995; Stow & Hadfield, 1981). The parameters A, a and b depending on the experimental conditions. A detailed explanation and systematic review of the possible mechanisms were given by several researchers (Moreira et al., 2010; Rioboo et al., 2002; Roisman et al., 2006). Additionally, correlations to predict the size of secondary droplets after impact were developed, considering the mentioned impact conditions (Moreira et al., 2010; Yarin et al., 2017).

Compared to the single droplet impact, the mechanisms that govern multidrop impact in a spray are complex, due to the interaction of droplets with different size and velocity and the complex hydrodynamic conditions. One possibility for predicting spray characteristics after impact is the simple extrapolation of many non-interacting single droplets using an Euler/Lagrange numerical simulation (Breitenbach et al., 2018; Cossali, Santini, & Marengo, 2005; Stanton & Rutland, 1998). Nevertheless, the applicability and universality of these models on the multidrop or spray impact is highly limited, as is seen from experimental studies (Moreira et al., 2010; Tropea & Roisman, 2000). Multiple crowns or droplet interaction may occur, which influence the generation of the secondary droplets. The spray impact leads to an accumulation of droplets on the surface, forming a liquid film that in turn affects the velocity and size of the ejected droplets (Kalantari & Tropea, 2007). The average film thickness, the fluctuation and hydrodynamics of the film and the adhesion of previous droplets after impact have an enormous effect on the splash threshold, the mass flux and the formation mechanism of the secondary droplets (Kalantari & Tropea, 2007). Thus, many important physical effects impact the spray outcome and cannot be described with the known single droplet impact mechanisms (Moreira et al., 2010).

Another possibility is the experimental investigation of the spray impact which enables the formulation of empirical correlations (Roisman et al., 2006; Tropea & Roisman, 2000; Yarin et al., 2017). However, the validity of empirical correlations is limited to the used experimental conditions, such as nozzle type, initial spray characteristic and surface properties. Therefore, the empirical models are applicable to a narrow parameter range only. Experimental evidence is necessary to prove the generality of the used models (Moreira et al., 2010). So far, an attempt has been made to characterize the sparse spray impingement on an impact sphere using a single-fluid nozzle (Kalantari & Tropea, 2007; Mühlbauer, Roisman, & Tropea, 2011; Roisman et al., 2006; Weiss, 2005; Yarin et al., 2017). The droplet size distribution and velocity of the impinging droplets were analyzed above the impact sphere using the Phase-Doppler technique. The outcome of the secondary spray was mainly identified as a function of the Reynold and Weber number before impact (Kalantari & Tropea, 2007). Based on the characteristics of the primary aerosol, models for the prediction of the average film thickness, which was formed directly when the spray impinged on a surface, were introduced by many researchers (Kalantari & Tropea, 2007; Yarin et al., 2017). Various parameters, such as the distance between the orifice and the impact surface or the dimension of the impact surface can also have an influence on the secondary spray outcome, but the influence is not yet known for a broad range of conditions. Currently, a two-fluid nozzle in combination with an impact bead is used as a sample introduction system in flame atomic absorption spectrometry, in which a secondary aerosol with decreased droplet size is produced after impact (Sneddon, 1990). The use of two-fluid nozzles is of interest due to the direct dispersion of the produced secondary aerosol into a gas flow. The comparability to the droplet generation mechanisms of a single-fluid nozzle has not been investigated. It is known that the velocity of secondary droplets after spray impact is low, compared to the primary droplets. Therefore, droplets remain close to the impact surface but may be transported by the gas flow. In experimental investigations thus far, the secondary droplet characteristics were analyzed in the area between the nozzle and impact surface for different impact surface properties (flat or curved, different surface roughness). (Kalantari & Tropea, 2007; Roisman et al., 2006).

In this work, an experimental study of spray impact on a sphere is presented, where the primary aerosol is generated by a two-fluid nozzle. The secondary aerosol, which is entrained by the gas, is of interest and the influence of the high velocity gas flow on the impact is investigated. Additional parameters, including the liquid-to-gas mass flow ratio and the gas inlet pressure (typical for two-fluid atomization) and their role in the secondary aerosol generation were analyzed.

Section snippets

Materials

The atomization and spray impact experiments were conducted with deionized water. In this work the liquid properties of water at 25°C were used (Street, Watters, & Vennard, 1996, p. 7).

Capillary nozzle

The primary aerosol (PA) was produced with a two-fluid nozzle with internal mixing at two size scales (Schlinge, Mescher, & Walzel, 2012; Stratmann, 2017). The nozzle consists of seven capillaries with an inner diameter of 0.5mm in the small and 2mm in the large scale. The liquid is introduced in the mixing

Small-scale nozzle

The droplet size distribution and the mass flow of small droplets were analyzed for the secondary aerosol. Fig. 2 clearly shows the outcome of spray impact with the generation of the secondary aerosol. Due to the impact, the primary aerosol was widened. A liquid film at the impact region was observed. Additionally, due to the bar holding of the sphere, some liquid was collected and flowed downward, and larger droplets were entrained in the gas flow. The volumetric droplet size frequency of the

Conclusion

The spray impact on a sphere was analyzed with a two-fluid nozzle with two different nozzle scales, producing droplets in different size ranges. This demonstrates that the spray impact can be used to increase the number of droplets in the small micrometer range (<3 μm). With the direct dispersion of the fine droplets into a gas flow, the produced secondary aerosol can be utilized for further applications such as spray drying. However, in addition to the fine droplets, larger droplets are also

Declaration of competing interest

None.

Acknowledgement

The authors thank for the assistance of Elizabeth Ely (EIES, Lafayette, IN, USA) in preparing the manuscript.

References (29)

  • J. Breitenbach et al.

    From drop impact physics to spray cooling models: A critical review

    Experiments in Fluids

    (2018)
  • S. Chandra et al.

    The collision of a droplet with a solid surface

    Physics of Fluids A: Fluid Dynamics

    (1990)
  • G.E. Cossali et al.

    Single-drop empirical models for spray impact on solid walls: A review

    Atomization and Sprays

    (2005)
  • C. Josserand et al.

    Drop impact on a solid surface

    Annual Review of Fluid Mechanics

    (2016)
  • Cited by (1)

    View full text