Interface reconstruction of adhering droplets for distortion correction using glare points and deep learning
Maximilian Dreisbach, Itzel Hinojos, Jochen Kriegseis, Alexander Stroh, Sebastian Burgmann · 2025
Read Interface reconstruction of adhering droplets for distortion correction using glare points and deep learning by Maximilian Dreisbach, Itzel Hinojos, Jochen Kriegseis, Alexander Stroh, Sebastian Burgmann (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-03453.txt.
Introduction
Five engineers reconstruct the wriggling surface of droplets stuck to walls in fast air flows, in a 2025 paper. Measuring flow inside such droplets needs a correction for how the curved surface bends light, and old corrections fail at high speeds. Their trick photographs the droplets with side lights, whose glare points secretly encode the surface's depth. A neural network trained on simulated droplet movies then rebuilds the full three-dimensional shape from single images.
Experiments at several external flow speeds show the reconstruction stays sharp and accurate. With the true surface known frame by frame, the light distortion in velocity measurements can be undone even at high flow speeds. The authors present the method as a flexible metrology tool for droplet science. Applications range from inkjet printing to spray cooling and fuel injection.
Catalog record
| CREATOR | Maximilian Dreisbach, Itzel Hinojos, Jochen Kriegseis, Alexander Stroh, Sebastian Burgmann |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-03453
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Interface reconstruction of adhering droplets for distortion correction using glare points and deep learning Maximilian Dreisbach, Itzel Hinojos, Jochen Kriegseis, Alexander Stroh, Sebastian Burgmann Abstract The flow within adhering droplets subjected to external shear flows has a significant influence on the stability and eventual detachment of the droplets from the surface. Most commonly, the velocity field inside adhering droplets is measured by means of particle image velocimetry (PIV), which requires a correction step to account for the distortion caused by the refraction of light at the curved gas-liquid interface. Current methods for distortion correction based on ray tracing are limited to low external flow velocities, for which the deformation of the droplet is insignificant and axisymmetry can be assumed. However, the ray-tracing method can be extended straightforwardly to arbitrarily deformed droplet shapes if the instantaneous three-dimensional droplet interface can be acquired. In the present work, a previously introduced method for the image-based reconstruction of gas-liquid interfaces by means of deep learning is adapted to determine the instantaneous interface of adhering droplets in external shear flows. In this regard, a purposefully developed optical measurement technique based on the shadowgraphy method is employed that encodes additional three-dimensional (3D) information of the interface in the images via glare points from lateral light sources. On the basis of the images recorded in the experiments, the volumetric shape of the droplet is reconstructed by a neural network that was trained on the spatio-temporal dynamics of the gas-liquid interface from a synthetic dataset obtained by numerical simulation. The results for experiments with adhering droplets at different velocities of external flow demonstrate that the combination of the learned droplet geometry with the depth encoding through the glare points facilitates a robust and flexible reconstruction. The proposed method reconstructs the instantaneous three-dimensional interface of adhering droplets at both high resolution and spatial accuracy and thereby enables the distortion correction of PIV measurements at high external flow velocities. 0 0 footnotetext: Author to whom any correspondence should be addressed January 2025 Keywords : Two-phase flow, Adhering droplet, Shear flow, Interface reconstruction, Deep learning \ioptwocol 1 Introduction The internal flow within droplets or bubbles plays an important role in numerous different multiphase flow phenomena, such as droplet dispersing, adhering droplets subjected to external flows, the evaporation or freezing of droplets, and bubble flow among others. Particle image velocimetry (PIV) is commonly employed in the investigation of multiphase flows to reveal internal flows. Karlsson et al. [ 1 ] used PIV to visualize the flow patterns and measure velocities inside freezing droplets. The authors found that Marangoni effects, which arise due to temperature differences in and around the droplet during freezing, induce a flow inside the droplet. Kinoshita et al. [ 2 ] revealed the complex three-dimensional circulation patterns within a moving droplet in a microchannel using micro-PIV. Later, Duxenneuner et al. [ 3 ] investigated the flow inside and around a droplet during liquid-liquid dispersion using micro-PIV and revealed the existence of a flow inside the droplet in main flow direction as well as a vortex flow around the droplet. By means of PIV measurements Minor et al. [ 4 ] and later Burgmann et al. [ 5 ] have found that there is a complex interaction between the internal and external flow of adhering droplets, that is dependent on interface tension and the oscillation of both phases. Figure 1 : Visualisation of an adhering droplet subjected to an external shear flow; time-averaged flow topology inside the droplet and in the external flow (left), droplet and flow oscillation (right). Figure adapted from Burgmann et al. [ 5 ] . Adhering droplets in external flows occur in many technical applications, for instance in cleaning and drying processes [ 6 , 7 ] , oil recovery [ 8 , 9 , 10 , 11 ] , heat exchangers [ 12 , 13 , 14 ] , airfoil icing prevention [ 1 ] and in fuel cells, where a removal of the droplets is crucial for an efficient operation [ 15 , 16 , 17 , 18 ] . The flow around the droplet eventually leads to a detachment of the droplet from the solid surface at critical conditions, which occur when the external forces acting on the droplet become greater than the adhesion of the liquid droplet to the surface [ 19 , 20 , 21 , 5 ] . Several investigations have revealed an intricate interaction of the droplet with the surrounding flow. The droplet deforms [ 10 , 7 , 22 ] and additionally oscillates, when subjected to external shear flows [ 23 , 24 ] . It was shown that the contact angle hysteresis of the substrate and the droplet volume have an influence on the critical velocity for droplet detachment [ 25 , 16 , 21 ] . Previous investigations based on PIV-measurements revealed that the internal flow within the droplet is relevant to understanding the mechanism of droplet detachment. It was found that the internal flow follows the main flow direction of the external flow [ 26 ] and exhibits a clockwise rotational flow pattern [ 4 ] . Minor et al. [ 4 ] deduced that the secondary flow inside a droplet is induced by shear forces on the interface resulting from the external flow. Burgmann et al. [ 24 , 27 ] confirmed the findings of a clockwise vortex at low external flow velocities but also found an additional counterclockwise rotation at higher flow velocity that becomes dominant when approaching the critical velocity. Furthermore, the authors found an oscillation of the droplet with increasing amplitude towards higher external flow velocities. Burgmann et al. [ 5 ] proposed that the droplet detachment is associated with a self-excitation process resulting from the oscillation of the gas-liquid interface, as well as the oscillation of the inner flow structure and the flow field around the droplet. The authors demonstrated that an emerging backflow region at higher external flow velocities induces the change in the flow pattern within the droplet and that the wake flow oscillates with the same frequency as the droplet. Later, Burgmann et al. [ 28 ] found that the coupling of the internal and external flow leads to increased flow separation in comparison to rigid bodies and, consequently, a higher pressure force driving droplet detachment. Bilsing et al. [ 29 ] measured the internal three-dimensional (3D) flow field by means of microscopic particle tracking velocimetry (PTV) through a transparent substrate from below and thereby revealed that the internal flow topology consists of multiple 3D vortical structures that change drastically depending on the external flow velocity. In experiments with rigid droplet models, Zhang et al. [ 30 ] found that the external flow topology is significantly influenced by the shape of the adhering droplet. Consequently, a better understanding of the droplet deformation is required for any future investigation of the external flow. A visualization of the flow phenomena inside and around the droplet is shown in Figure 1 . The mutual dependence of wake flow, contour oscillation, and internal flow structure influences the stability of the droplet and eventually leads to the detachment of the droplet at critical conditions. While there has been significant scientific progress in