Hostname: page-component-7b7567c8c-fvjsw Total loading time: 0 Render date: 2026-11-16T11:19:58.795Z Has data issue: false hasContentIssue false

Droplet size distribution in a swirl airstream using in-line holography technique

Published online by Cambridge University Press:  06 January 2023

Someshwar Sanjay Ade
Affiliation:
Center for Interdisciplinary Program, Indian Institute of Technology Hyderabad, Kandi, Sangareddy 502 284, Telangana, India
Pavan Kumar Kirar
Affiliation:
Department of Chemical Engineering, Indian Institute of Technology Hyderabad, Kandi, Sangareddy 502 284, Telangana, India
Lakshmana Dora Chandrala*
Affiliation:
Department of Mechanical and Aerospace Engineering, Indian Institute of Technology Hyderabad, Kandi, Sangareddy 502 284, Telangana, India
Kirti Chandra Sahu*
Affiliation:
Department of Chemical Engineering, Indian Institute of Technology Hyderabad, Kandi, Sangareddy 502 284, Telangana, India
*
†Email addresses for correspondence: lchandrala@mae.iith.ac.in, ksahu@che.iith.ac.in
†Email addresses for correspondence: lchandrala@mae.iith.ac.in, ksahu@che.iith.ac.in

Abstract

We investigate the morphology and size distribution of satellite droplets resulting from the interaction of a freely falling water droplet with a swirling airstream of different strengths by employing shadowgraphy and deep-learning-based digital in-line holography techniques. We found that the droplet exhibits vibrational, retracting bag and normal breakup phenomena for the no swirl, low and high swirl strengths for the same aerodynamic field. In the high-swirl scenario, the disintegrations of the nodes, rim and bag-film contribute to the number mean diameter, resulting in smaller satellite droplets. In contrast, in the low-swirl case, the breakup of the rim and nodes only contributes to the size distribution, resulting in larger droplets. The temporal variation of the Sauter mean diameter reveals that for a given aerodynamic force, a high swirl strength produces more surface area and surface energy than a low swirl strength. The theoretical prediction of the number-mean probability density of tiny satellite droplets under swirl conditions agrees with experimental data. However, for the low swirl, the predictions differ from the experimental results, particularly due to the presence of large satellite droplets. Our results reveal that the volume-weighted droplet size distribution exhibits two (bi-modal) and three (multi-model) peaks for low and high swirl strengths, respectively. The analytical model that takes into account various mechanisms, such as the nodes, rim and bag breakups, accurately predicts the shape and characteristic sizes of each mode for the case of high swirl strength.

Information

Type
JFM Papers
Copyright
© The Author(s), 2023. Published by Cambridge University Press
Figure 0

Figure 1.Schematic of the experimental set-up (top view) equipped with a shadowgraphy and digital in-line holography systems to investigate the size distribution of satellite droplets due to the fragmentation of a freely falling water droplet under an imposed swirl airstream. The inset depicts the back and side views of a swirler. Here, FP and MFC represent the focal plane associated with the high-speed camera 2 and mass flow controller, respectively. A picture of our experimental set-up is presented as figure S1 in the supplementary material available at https://doi.org/10.1017/jfm.2022.1028.

Figure 1

Figure 2.Various steps of the digital in-line holography technique employed in the present study.

Figure 2

Figure 3.A demonstration of different steps to obtain the droplet size distribution after breakup using the in-line holography technique. (a–c) Temporal evolution of the bag formation and breakup obtained using the shadowgraphy technique. The dimensionless coordinates of the droplet ($x/D_{0}$, $y/D_{0}$, $z/D_{0}$) are mentioned at the bottom of the shadowgraphy images at different instants. Outlet of the nozzle and the tip of the dispensing needle are indicated in the left and top of the shadowgraphy images. (d) Image at $\tau =0.67$ is considered to reconstruct the hologram at different planes in the $x$ direction (panel (e)). Here, ${\textit {We}}=14.53$, $Sw = 0$ and $(x_d/D_o,y_d/D_o,z_d/D_o) = (0.0, 0.27, 0.95)$. An animation showing the different holography reconstructed planes at $\tau =0.67$ is provided as supplementary movie 1.

Figure 3

Table 1.Fluid properties of water and ethanol used in the present study and considered by Guildenbecher et al. (2017), respectively.

Figure 4

Figure 4.Comparison of the temporal variation of normalized mean diameter $(d_{10}/d_0)$ obtained from our experiments with that of Guildenbecher et al. (2017). In the present study, a water droplet of $d_0 = 3.09$ mm (${\textit {We}} = 14.53$) is used, while Guildenbecher et al. (2017) considered an ethanol droplet of $d_0 = 2.54$ mm (${\textit {We}} = 13.8$). In our experiments, the error bar represents the standard deviation obtained from the three repetitions. The uncertainty bar for the data of Guildenbecher et al. (2017) corresponds to the standard deviation from forty-four realizations.

Figure 5

Figure 5.Effect of the swirl number, $Sw$, on the fragmentation process of a water droplet (obtained from shadowgraphy) for ${\textit {We}} = 12.1$. (a,d,g,j,m,p) $Sw = 0$ (vibrational breakup), (b,e,h,k,n,q) $Sw = 0.47$ (retracting bag breakup) and (c,f,i,l,o,r) $Sw = 0.82$ (normal bag breakup). The value of $\tau$ is indicated in each panel. The dimensionless scale-bar is shown in panel (a). The dimensionless location $(x_d/D_o,y_d/D_o,z_d/D_o)$ of the dispensing needle for $Sw=0$, $0.47$ and $0.82$ are $(0.0,0.01,0.80)$, $(0.12,0.01,0.80)$ and $(0.12,0.01,0.80)$, respectively. To show the migration due to the airstream, the dimensionless position of the droplet in the flow direction, $y/D_0$, is mentioned at the bottom of each panel. Kirar et al. (2022) have shown similar breakup modes for an ethanol droplet. The droplet breakup phenomena for $Sw=0$, 0.47 and 0.82 are provided as supplementary movies 2–4, respectively.

Figure 6

Figure 6.Schematic of the core-annular flow field of the high shear and wake zones created due to the swirling effect. The droplets during their interaction with the swirl airstream for low and high swirl strengths are also depicted.

Figure 7

Figure 7.Temporal evolution of the droplet size distribution for $Sw = 0.47$ and ${\textit {We}} = 12.1$. Panels (a,c,f,i,l) and (b,d,g,j,m) represent the shadowgraphy images and recorded holograms obtained using the in-line holography. The values of the dimensionless time, $\tau$, are mentioned on the left-hand side of panels (a,c,f,i,l) measured from the instant at the onset of breakup. The scale-bar is shown in panel (a). Panels (e,h,k,n) depict the histograms of the droplet size distribution (the droplet counts, $N$ versus the droplet diameter, $d$) at different instants after the breakup. The dimensionless location of the dispensing needle is at $(x_d/D_o,y_d/D_o,z_d/D_o) = (0.12, 0.01, 0.80)$. The dimensionless position of the droplet in the flow direction, $y/D_0$, is mentioned at the top-right corner of each panel of the shadowgraphy images.

Figure 8

Figure 8.Temporal evolution of the droplet size distribution for $Sw = 0.82$ and ${\textit {We}} = 12.1$. Panels (a,c,f,i,l) and (b,d,g,j,m) represent the shadowgraphy and the holography (recorded hologram) images, respectively. The values of the dimensionless time, $\tau$ are mentioned on the left-hand side of (a,c,f,i,l) measured from the instant at the onset of breakup. The scale-bar is shown in panel (a). Panels (e,h,k,n) depicts the histogram of the droplet size distribution (the droplet counts, $N$, versus the droplet diameter, $d$) at different instants after breakup. The dimensionless location of the dispensing needle is $(x_d/D_o,y_d/D_o,z_d/D_o) = (0.12, 0.01, 0.80)$. The dimensionless position of the droplet in the flow direction, $y/D_0$, is mentioned at the top-right corner of each panel of the shadowgraphy images.

Figure 9

Figure 9.Temporal variation of normalized mean diameters (a) $(d_{10}/d_0)$ and (b) $(d_{32}/d_0)$ for different swirl numbers at ${\textit {We}} = 12.1$.

Figure 10

Figure 10.Comparison of the variation of $p(\chi )$ versus $\chi (d/\langle d \rangle )$ obtained from our experiments with the theoretical prediction from (3.4) (Villermaux & Bossa 2009) for (a) $Sw = 0.47$ at $\tau = 0.75$ and (b) $Sw = 0.82$ at $\tau = 0.66$. The rest of the parameters are the same as those in figure 7.

Figure 11

Figure 11.Volume-weighted density of all fragments at $\tau =0.75$ for $Sw=0.47$ and ${\textit {We}}=12.1$. This depicts a bi-modal distribution.

Figure 12

Figure 12.Comparison of the volume-weighted density obtained from our experiments with the theoretical prediction (Jackiw & Ashgriz 2022) at $\tau =0.66$ for $Sw=0.82$ and ${\textit {We}}=12.1$. Note the multi-modal distribution is likely due to fragments from the bag, the rim and the node.

Ade et al. Supplementary Movie 1

The animation depicts different holography reconstructed planes at different depths. This corresponds to figure 3.

Download Ade et al. Supplementary Movie 1(Video)
Video 4.9 MB

Ade et al. Supplementary Movie 2

The vibrational breakup mode in Figure 5 (Sw = 0).

Download Ade et al. Supplementary Movie 2(Video)
Video 10.4 MB

Ade et al. Supplementary Movie 3

The retracting bag breakup mode in Figure 5 (Sw = 0:47).

Download Ade et al. Supplementary Movie 3(Video)
Video 8 MB

Ade et al. Supplementary Movie 4

The bag breakup mode in Figure 5 (Sw = 0:82).

Download Ade et al. Supplementary Movie 4(Video)
Video 7.8 MB
Supplementary material: PDF

Ade et al. Supplementary Material

Ade et al. Supplementary Material

Download Ade et al. Supplementary Material(PDF)
PDF 1.5 MB